Thank you

for purchasing this e-book.

To receive special offers and news
about our latest products,
sign up below.

Sign Up
Or visit LWW.com
 HANDBOOK OF
CLINICAL
AUDIOLOGY
SEVENTH EDITION

XXXDBNCPEJBNFECMPHTQPUDPN
]
#FTU
.FEJDBM
#PPLT
]
$IZ
:POH

]
$SFEJU
4503.
HANDBOOK OF
CLINICAL
AUDIOLOGY
SEVENTH EDITION
EDITOR-IN-CHIEF EDITORS

JACK KATZ, Ph.D. MARSHALL CHASIN, Au.D.

Director Director of Auditory Research
Auditory Processing Service Musician’s Clinics of Canada
Prairie Village, Kansas Toronto, Ontario, Canada
and Research Professor
University of Kansas Medical Center KRISTINA ENGLISH, Ph.D.
Kansas City, Kansas and Professor Emeritus Professor and Interim School Director
University at Buffalo School of Speech Pathology and Audiology
State University of New York University of Akron/NOAC
Buffalo, New York Akron, Ohio

LINDA J. HOOD, Ph.D.

Professor
Department of Hearing and Speech Sciences
Vanderbilt Bill Wilkerson Center
Vanderbilt University
Nashville, Tennessee, USA
Honorary Professor
University of Queensland
Brisbane, Australia

KIM L. TILLERY, Ph.D.

Professor and Chair
Department of Communication Disorders & Sciences
State University of New York at Fredonia
Fredonia, New York
Acquisitions Editor: Michael Nobel
Product Development Editor: Linda G. Francis
Marketing Manager: Leah Thomson
Editorial Assistant: Tish Rogers
Production Project Manager: Priscilla Crater
Design Coordinator: Stephen Druding
Illustration Coordinator: Jennifer Clements
Manufacturing Coordinator: Margie Orzech
Prepress Vendor: Aptara, Inc.

7th edition

Copyright © 2015 Wolters Kluwer Health

Copyright © 2009, 2001 Lippincott Williams & Wilkins. Copyright © 1993, 1985, 1978, 1972 Williams & Wilkins.
Two Commerce Square
2001 Market Street
Philadelphia, PA 19103 USA
LWW.com

Not authorised for sale in United States, Canada, Australia, New Zealand, Puerto Rico, or U.S. Virgin Islands.

All rights reserved. This book is protected by copyright. No part of this book may be reproduced or transmit-
ted in any form or by any means, including as photocopies or scanned-in or other electronic copies, or utilized
by any information storage and retrieval system without written permission from the copyright owner, except
for brief quotations embodied in critical articles and reviews. Materials appearing in this book prepared by
individuals as part of their official duties as U.S. government employees
are not covered by the above-mentioned copyright. To request permission, please contact Wolters
Kluwer Health at Two Commerce Square, 2001 Market Street, Philadelphia, PA 19103, via email at
[email protected], or via our website at lww.com (products and services).

987654321

Printed in China

Library of Congress Cataloging-in-Publication Data

Handbook of clinical audiology / editor-in-chief, Jack Katz ; editors, Marshall Chasin, Kristina English,
Linda J. Hood, Kim L. Tillery. – Seventh edition.
p. ; cm.
Includes bibliographical references.
ISBN 978-1-4511-9163-9
I. Katz, Jack, editor. II. Chasin, Marshall, editor. III. English, Kristina M., 1951- editor. IV. Hood, Linda J.,
editor. V. Tillery, Kim L., editor.
[DNLM: 1. Hearing Disorders. 2. Hearing–physiology. WV 270]
RF291
617.8–dc23
2014014240

Care has been taken to confirm the accuracy of the information presented and to describe generally accepted
practices. However, the authors, editors, and publisher are not responsible for errors or omissions or for any
consequences from application of the information in this book and make no warranty, expressed or implied,
with respect to the currency, completeness, or accuracy of the contents of the publication. Application of this
information in a particular situation remains the professional responsibility of the practitioner; the clinical
treatments described and recommended may not be considered absolute and universal recommendations.
The authors, editors, and publisher have exerted every effort to ensure that drug selection and dosage set
forth in this text are in accordance with the current recommendations and practice at the time of publication.
However, in view of ongoing research, changes in government regulations, and the constant flow of informa-
tion relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug
for any change in indications and dosage and for added warnings and precautions. This is particularly impor-
tant when the recommended agent is a new or infrequently employed drug.
Some drugs and medical devices presented in this publication have Food and Drug Administration (FDA)
clearance for limited use in restricted research settings. It is the responsibility of the health care provider to
ascertain the FDA status of each drug or device planned for use in his or her clinical practice.

LWW.com
 CO N T R I B U TO R S

DANIEL ARTHUR ABRAMS, Ph.D. MARSHALL CHASIN, Au.D.

Research Associate, Department of Psychology and Director of Research
Behavioral Sciences Musicians’ Clinics of Canada
Stanford University Toronto, Ontario, Canada
Palo Alto, California
LAUREL A. CHRISTENSEN, Ph.D.
ANGELA LOUCKS ALEXANDER, Au.D. Chief Audiology Officer
Director, Taupo Audiology and Auditory Processing Network Vice President, Research and Development
Taupo, New Zealand GN ReSound Group
Glenview, Illinois
EDOARDO ARSLAN, M.D.*
Department of Neuroscience, University of Padova, Padova, Italy JOHN GREER CLARK, Ph.D.
Service of Audiology and Phoniatrics, Treviso Regional Hospital, Associate Professor, Department of Communication
Piazza Ospedale,Treviso, Italy Sciences and Disorders
University of Cincinnati
A.U. BANKAITIS, Ph.D.
Cincinnati, Ohio
Vice President, Oaktree Products, Inc.
President, Clark Audiology, LLC
St. Louis, Missouri
Middletown, Ohio
JANE A. BARAN, Ph.D.
CHRISTOPHER GRAY CLINARD, Ph.D.
Professor and Chair
Assistant Professor
Department of Communication Disorders
Department of Communication Science & Disorders
University of Massachusetts Amherst
James Madison University
Amherst, Massachusetts
Harrisonburg, Virginia
DOUGLAS L. BECK, Au.D.
CLAUDIA BARROS COELHO, M.D., Ph.D.
Director of Professional Relations, Oticon, Inc.
Research Scientist
Somerset, New Jersey
Department of Otolaryngology
Web Content Editor, American Academy of Audiology
University of Iowa
Reston, Virginia
Iowa City, Iowa
LINDSAY BONDURANT, Ph.D.
WILLIAM COLE, B.a.Sc., P.Eng.
Assistant Professor of Audiology, Communication
President
Sciences and Disorders
Audioscan Division of Etymonic Design, Inc.
Illinois State University
Dorchester and
Normal, Illinois
Adjunct Associate Professor
CARMEN BREWER, Ph.D. School of Communication Science and Disorders
Chief Research Audiologist, Audiology Unit, Otolaryngology Western University
Branch London, Ontario, Canada
National Institute on Deafness and other Communication
BARBARA CONE, Ph.D.
Disorders
Professor Speech, Language and Hearing Sciences
National Institutes of Health
The University of Arizona
Bethesda, Maryland
Tucson, Arizona
ROBERT BURKARD, Ph.D.
ALLAN O. DIEFENDORF, Ph.D.
Professor and Chair Department of Rehabilitation Science
Professor, Department of Otolaryngology, Head and Neck
University at Buffalo
Surgery
State University of New York
Indiana University School of Medicine
Buffalo, New York
Director, Audiology and Speech/Language Pathology
ANTHONY T. CACACE, Ph.D. Indiana University Health
Professor Indianapolis, Indiana
Communication Sciences and Disorders
ANDREW DIMITRIJEVIC, Ph.D.
Wayne State University
Assistant Professor
Detroit, Michigan
Communication Sciences Research Center
Cincinnati Children’s Hospital
Department of Otolaryngology, University of Cincinnati
*Deceased Cincinnati, Ohio

v
vi Contributors

RACHEL N. DINGLE, Ph.D. SAMANTHA GUSTAFSON, Au.D.

Student, School of Communication Sciences and Disorders Ph.D. Student, Department of Hearing and Speech Sciences
Western University Vanderbilt University
London, Ontario, Canada Nashville, Tennessee
MANUEL DON, Ph.D. TROY HALE, Au.D.
Head, Electrophysiology Department, Scientist III (retired) House Assistant Professor, Audiology
Research Institute Los Angeles, California AT Still University
Director, AFA Balance and Hearing Institute
M. PATRICK FEENEY, Ph.D.
Mesa, Arizona
Professor, Department of Otolaryngology, Head and
Neck Surgery MELANIE HERZFELD, Au.D.
Oregon Health and Science University Practice Owner Hearing and Tinnitus Center
Director, Veterans Affairs National Center for Rehabilitative Woodbury, New York
Auditory Research
THERESA HNATH-CHISOLM, Ph.D.
Portland Veterans Affairs Medical Center
Professor and Chair, Communication Sciences and Disorders
Portland, Oregon
University of South Florida
JEANANE FERRE, Ph.D. Tampa, Florida
Adjunct Faculty
LINDA J. HOOD, Ph.D.
Communication Sciences & Disorders
Professor
Northwestern University
Department of Hearing and Speech Sciences
Evanston, and
Vanderbilt Bill Wilkerson Center
Audiologist, Central Auditory Evaluation and Treatment
Vanderbilt University
Oak Park, Illinois
Nashville, Tennessee, USA
TRACY S. FITZGERALD, Ph.D. Honorary Professor
Staff Scientist/Director, Mouse Auditory Testing Core Facility University of Queensland
National Institute on Deafness and Other Communication Brisbane, Australia
Disorders
LISA L. HUNTER, Ph.D.
National Institutes of Health
Associate Professor
Bethesda, Maryland
Department of Otolaryngology and Communication
BRIAN J. FLIGOR, Sc.D. Sciences and Disorders
Chief Audiology Officer University of Cincinnati
Lantos Technologies, Inc. Scientific Director, Department of Audiology
Wakefield, Massachusetts Cincinnati Children’s Hospital Medical Center
Cincinnati, Ohio
RICHARD E. GANS, Ph.D.
Founder & CEO ANDREW B. JOHN, Ph.D.
The American Institute of Balance (ABI) Assistant Professor
Largo, Florida Department of Communication Sciences and Disorders
College of Allied Health
DOUGLAS B. GARRISON, Au.D.
University of Oklahoma Health Sciences Center
Director, Duke Vestibular Lab
Oklahoma City, Oklahoma
Department of Otolaryngology—Head and Neck Surgery
Duke University Health System ANDREW B. JOHN, Ph.D.
Durham, North Carolina Assistant Professor, Communication Sciences and Disorders
University of Oklahoma Health Sciences Center
JENNIFER E. GONZALEZ, B.A.Au.D./Ph.D.
Oklahoma City, Oklahoma
Candidate, Department of Speech, Language and Hearing
Sciences CHERYL DeCONDE JOHNSON, Ed.D.
University of Connecticut Private Consulting Practice
Storrs, Connecticut The ADVantage
Auditory-Deaf Education Consulting
JENNIFER GROTH, M.A.
Leadville, Colorado
Director, Audiology Communications Research and
Developement HYUNG JIN JUN, M.D., Ph.D.
GN ReSound Group Department of Otolaryngology-Head and Neck Surgery
Glenview, Illinois Guro Hospital, Korea University College of Medicine
Seoul, South Korea
 Contributors vii

JACK KATZ, Ph.D. RICK NEITZEL, Ph.D.

Director Assistant Professor, Department of Environmental
Auditory Processing Service Health Sciences
Prairie Village, Kansas University of Michigan
and Research Professor Ann Arbor, Michigan
University of Kansas Medical Center
PEGGY NELSON, Ph.D.
Kansas City, Kansas and Professor Emeritus
Professor, Department of Speech-Language-Hearing Sciences
University at Buffalo
University of Minnesota
State University of New York
Minneapolis, Minnesota
Buffalo, New York
WILLIAM NOBLE, Ph.D.
WILLIAM JOSEPH KEITH, Ph.D.
Psychology
Director, SoundSkills Auditory Processing Clinic
School of Behavioural, Cognitive and Social Sciences
Auckland, New Zealand
University of New England
PAUL KILENY, Ph.D. Armidale, Australia
Professor and Academic Program Director, Audiology
TABITHA PARENT-BUCK, Au.D.
Otolaryngology, Head-and-Neck Surgery
Chair, Audiology Department
University of Michigan
AT Still University
Ann Arbor, Michigan
Audiologist, AFA Balance and Hearing Institute
KELLY KING, Ph.D. Mesa, Arizona
Research Audiologist
DENNIS P. PHILLIPS, Ph.D.
Audiology Unit, Otolaryngology Branch National Institute on
Professor, Department of Psychology and Neuroscience
Deafness and Other Communication Disorders
Dalhousie University
National Institutes of Health
Halifax, Nova Scotia, Canada
Bethesda, Maryland
ERIN G. PIKER, Ph.D.
NINA KRAUS, Ph.D.
Assistant Professor, Department of Surgery-Division of
Hugh Knowles Professor, Communication Sciences
Otolaryngology
and Disorders
Duke University
Northwestern University
Durham, North Carolina
Evanston, Illinois
BETH A. PRIEVE, Ph.D.
BRIAN KREISMAN, M.D., Ph.D
Professor, Communication Sciences and Disorders
Department of Speech Pathology and Audiology
Syracuse University
Calvin College
Syracuse, New York
Grand Rapids, Michigan
EVELING ROJAS RONCANCIO, M.D.
FREDERICK N. MARTIN, Ph.D.
Department of Otolaryngology
Lillie Hage Jamail Centennial Professor Emeritus
University of Iowa
Department of Communication Sciences and Disorders
Iowa City, Iowa
The University of Texas at Austin
Austin, Texas CHRIS SANFORD, Ph.D.
Assistant Professor
RACHEL McARDLE, Ph.D.
Communication Sciences and Disorders
Associate Professor, Communication Sciences and Disorders
Idaho State University
University of South Florida
Pocatello, Idaho
Tampa, Florida
Chief, Audiology and Speech Pathology ROSAMARIA SANTARELLI, Ph.D., M.D.
Bay Pines Veterans Affairs Healthcare System Department of Neuroscience
Bay Pines, Florida University of Padova
Padova, and
JOSEPH J. MONTANO, Ed.D.
Deputy Director of Audiology and Phoniatrics
Associate Professor of Audiology
Treviso Regional Hospital
Department of Otolaryngology
Treviso, Italy
Weill Cornell Medical College
New York, New York KIM SUZETTE SCHAIRER, Ph.D.
Adjunct Faculty, Department of Audiology and
FRANK E. MUSIEK, Ph.D.
Speech-Language Pathology
Professor, Speech Language and Hearing Sciences
East Tennessee State University
University of Connecticut
Johnson City
Storrs, Connecticut
Audiologist, Department of Audiology
James H. Quillen Veterans Affairs Medical Center
Mountain Home, Tennessee
viii Contributors

ROBERT S. SCHLAUCH, Ph.D. RICHARD S. TYLER, Ph.D.

Professor Professor, Department of Otolaryngology
Department of Speech-Language-Hearing Sciences University of Iowa
University of Minnesota Iowa City, Iowa
Minneapolis, MN
KRISTIN M. UHLER, Ph.D.
SUSAN SCOLLIE, Ph.D. Assistant Professor
Associate Professor School of Communication School of Medicine
Sciences and Disorders Department of Otolaryngology
Western University, University of Colorado Denver
London, Ontario, Canada
MICHAEL VALENTE, Ph.D.
JOSEPH SMALDINO, Ph.D. Director of Adult Audiology
Professor, Communication Sciences and Disorders Department of Otolaryngology
Illinois State University Washington University School of Medicine
Normal, Illinois St. Louis, Missouri
JENNIFER L. SMART, Ph.D. MAUREEN VALENTE, Ph.D.
Associate Professor, Audiology, Speech-Language Pathology Director of Audiology Studies
and Deaf Studies Program in Audiology and Communication Sciences
Towson University Associate Professor
Towson, Maryland Department of Otolaryngology
Washington University School of Medicine
CARRIE SPANGLER, Au.D.
St. Louis, Missouri
Educational Audiology Clinician
School of Speech Language Pathology and Audiology BARBARA E. WEINSTEIN, Ph.D.
The University of Akron Professor and Founding Executive Officer
Akron, Ohio Health Sciences Doctoral Programs, Au.D. Program
Graduate Center, City University of New York
JAMES R. STEIGER, Ph.D.
New York, New York
Professor
School of Speech-Language Pathology and Audiology KARL R. WHITE, Ph.D.
The University of Akron; Northeast Ohio AuD Consortium Director, National Center for Hearing Assessment and
Aakron, Ohio Management
Emma Eccles Jones Endowed Chair in Early Childhood Education
DE WET SWANEPOEL, Ph.D.
Professor of Psychology
Professor, Speech-Language Pathology and Audiology
Utah State University
University of Pretoria
Logan, Utah
Pretoria, South Africa
Adjunct Professor, Ear Sciences Centre, School of Surgery LAURA ANN WILBER, Ph.D.
The University of Western Australia Professor Emeritus, Communication Sciences and Disorders
Perth, Australia Northwestern University
Evanston, Illinois
ANNE MARIE THARPE, Ph.D.
Professor and Chair, Department of Hearing and WILLIAM S. YACULLO, Ph.D.
Speech Sciences Professor, Communication Disorders
Vanderbilt University Governors State University
Nashville, Tennessee University Park, Illinois
KIM L. TILLERY, Ph.D. CHRISTINE YOSHINAGA-ITANO, Ph.D.
Professor and Chair Professor, Department of Speech, Language and Hearing Sciences
Department of Communication Disorders & Sciences University of Colorado, Boulder
State University of New York at Fredonia Boulder, Colorado
Fredonia, New York
TERESA A. ZWOLAN, Ph.D.
HENRY P. TRAHAN, Au.D. Professor, Otolaryngology
Assistant Professor, Audiology University of Michigan
AT Still University Ann Arbor, Michigan
Mesa, Arizona
KELLY TREMBLAY, Ph.D.
Professor, Speech and Hearing Sciences
University of Washington
 D E D I C AT I O N

The Seventh Edition of the Handbook of Clinical Audiol- tists (both of whom he considered as audiologists) to learn
ogy is Dedicated to Raymond Carhart (192?–1975), who is about the hearing process and develop ways to help persons
recognized as the “Father of Audiology.” He talked about the living with hearing loss.
shared responsibility of the clinician and the hearing scien-

ix
 FO R EWO R D

It is fitting that this book is dedicated to Dr. Raymond Four of those dimensions were (1) sensitivity (how faintly
Carhart. He stated in a 1976 interview that he conceived of can one hear); (2) clarity of sound (speech or otherwise)
an audiologist “as someone who has a prime commitment in quiet; (3) clarity in noise; and (4) tolerance (how loud
to learning about hearing and its processes as well as a com- can sound be without becoming a problem). Two of these
mitment to understanding and coping with its problems.” dimensions (sensitivity and clarity), which were detailed by
He talked about the shared responsibility of the clinician Carhart in a 1951 paper, became the basis of Plomp’s (1978)
and the hearing scientists (both of whom he considered as two-component model of hearing loss involving audibility
audiologists) to learn about the hearing process and ways and distortion. Carhart said there were many more dimen-
to help the persons with hearing impairment. The seventh sions to hearing, but those four should always be measured
edition of Handbook of Clinical Audiology book strives to do when fitting hearing aids and working with patients with
that, as have the previous editions. hearing loss.
Carhart has been referred to as the “Father of Audiol- Although Carhart worked as a speech scientist, a clini-
ogy”—or sometimes the “Grandfather of Audiology.” Per- cian (in speech and in hearing), a researcher in speech and
haps it would be most appropriate to call him the “Grand especially in hearing, his primary contribution is probably as
Father of Audiology.” a teacher and educator. An educator can be described as one
Although he came to the field somewhat indirectly, his who conveys learning in which the knowledge, skills, and
contributions were enormous. habits of a group of people are transferred from one genera-
Dr. Carhart was born in Mexico City. He received his tion to the next through teaching, training, or research, and
Bachelor’s degree from Dakota Wesleyan University in 1932 that certainly describes Dr. Carhart.
in speech pathology and psychology; his Master’s and Ph.D. In his capacity as an educator, Carhart directed some 35
degrees from Northwestern in 1934 and 1936, respectively, dissertations, beginning in 1946 with a study by John Keys
in Speech Pathology, Experimental Phonetics, and Psychol- entitled “Comparative Threshold Acuity of Monaural and
ogy. He was an instructor in speech reeducation at North- Binaural Hearing for Pure Tone and Speech as Exhibited by
western from 1936 to 1940 and then an assistant, and associ- Normal and Hard of Hearing.” Although his primary inter-
ate professor in 1943 in speech science. est was in speech and speech understanding, the disserta-
Although Carhart initially worked in speech science, he tions he directed covered a range of hearing problems from
was asked to replace C.C. Bunch following Bunch’s untimely difference limens, to effects of surgery and specific diseases
death in June, 1942. Carhart then began to teach Bunch’s on hearing, to auditory fatigue, loudness and many more
courses in hearing and became so interested in the problems topic areas. In addition, as an educator he taught some of
that, as he said, “I’ve been working with them ever since.” the leaders in the field of audiology like James Jerger, Don-
In 1943, Carhart joined the Medical Administrative ald Dirks, Cornelius Goetzinger, Jack Willeford, and many
Corps, U S Army, as a captain, he was assigned to DeShon more. Many of those went on to teach, and to educate other
Hospital in Butler, Pennsylvania as Director of the Acous- students in audiology programs at our most prestigious
tic Clinic and as Acoustic Physicist where he was asked to universities.
develop a program for veterans who had lost their hearing In 1949, he directed the dissertation of Miriam Pauls
during the war. In that capacity he contacted the scien- Hardy, who may have been the first female to graduate with
tists at the Psycho-Acoustic Laboratory (PAL) at Harvard, a Ph.D. in audiology. Unlike some of the professors of the
who, among other things, had come up with word lists that time, Dr. Carhart did not discriminate on the basis of gen-
might be used in evaluating a person’s ability to understand der. He believed that it was the mind—not the gender—that
speech. He also developed a fairly comprehensive rehabilita- was important. He did, however, believe that one should do
tion program that involved selecting and fitting hearing aids the work, not just talk the talk. He set an example in that one
(which were not so complex or elaborate as they are today), often found him in his office or laboratory in the evening
and teaching the soldiers and veterans how to use them. and on weekends.
When Carhart returned to Northwestern in 1946, he His early research interests at Northwestern were in con-
convinced the dean to establish an academic program in ductive hearing loss (the “Carhart notch,” which can be an
Audiology, which was the name that Dr. Norton Canfield indicator of possible otosclerosis was named for him), includ-
chose for the department at DeShon. He became the first ing a method of checking the accuracy of bone conduction
professor of Audiology at Northwestern. measurements before there was an artificial mastoid, let alone
Carhart later said (in class if not in print) that sound— an ANSI standard. He was interested in masking (forward,
and especially speech—was comprised of many dimensions. backward, and perceptual, which we now call informational

xi
xii Foreword

masking), and did much to enhance our understanding of the We dedicate this seventh edition to Dr. Carhart, because
way speech is processed by the damaged ear. like him the Handbook of Clinical Audiology has educated so
Before there were computers in our clinics and most many leaders of our field and has inspired countless audiol-
academic research laboratories, he developed a key-sort sys- ogists throughout the world. This edition of the Handbook
tem as a way of classifying audiograms so that one could will provide a broad perspective of the field of audiology
better interpret the puretone audiogram. by nearly 100 contributing experts in the field. They offer
Finally, Carhart believed that audiology was more their knowledge, wisdom, and enthusiasm to help another
than a clinical field and that the clinicians who practiced it generation of audiologists to fulfill their mission.
should continue to explore and research the ways in which
we hear and how to improve the lives of those who do not
hear normally. REFERENCES
Raymond Carhart died at his desk in October 1975, Carhart R. (1951). Basic principles of speech audiometry. Acta
leaving behind a legacy to the academic discipline of audiol- Otolaryngol. 40:62–71.
ogy, the numerous leaders in the field of audiology whom he Plomp R. (1978). Auditory handicap of hearing impairment
had educated, and the greater understanding of audiologic and the limited benefit of hearing aids. J Acoust Soc Am.
assessment and hearing aid rehabilitation for which he is 63:533–549.
known as the father—or Grand Father of Audiology. Laura Ann Wilber
 P R E FAC E

For more than 40 years, the Handbook of Clinical Audiology

(HOCA) has maintained an important role in the education SECTIONS, CHAPTERS, AND
of graduate students in audiology, both in North America CONTRIBUTORS
and throughout the world. It also serves as a useful reference
The strength of HOCA has always been the knowledge and
for audiologists, otologists, and speech–language patholo-
expertise of the contributors in the many aspects of audiol-
gists who wish to have a comprehensive and practical guide
ogy. They have both clinical and research credentials in the
to the current practices in audiology.
topics they write about and most are also professors who
Each edition of the HOCA has been an update of the
are proficient in communicating with students. Audiologists
previous one, but we have also striven to make the newest
looking down the list of contributors will recognize famil-
edition better than the one that came before. For this edition,
iar and highly respected colleagues. They have contributed
there are four highly skilled and knowledgeable editors plus
much to the field in the past and now contribute again by
one senior editor. We have worked together to select highly
providing important and readable materials for both col-
qualified contributors on topics that are both core and cur-
leagues and students. We have made every effort to provide
rent for students and professionals in audiology. Online case
up-to-date, accurate, and clinically applicable information.
studies and references have been added to this edition to
Each of the four main editors of this book has a dis-
enable the reader to go beyond the basic scope of this book.
tinguished record of teaching, research, writing, and clinical
work. Each one took responsibility for significant portions
of the book. Linda Hood helped to edit the Sixth Edition
THE FOREWORDS and decided to go “another round.” Her chapters deal pri-
In the previous edition of the Handbook the foreword was marily with physiological methods for audiologic diagnosis.
written by Moe Bergman, a distinguished gentleman with Marshall Chasin, our first Canadian editor, edited the chap-
many years of audiology behind him. Moe Bergman, Ed.D., ters dealing with amplification and other technical aspects.
was in the very first group of audiologists who began this dis- Kim Tillery edited the chapters dealing with central audi-
cipline more than 70 years ago. Starting prior to World War tory processing disorders and shared in the final editing of
II and for decades following, Dr. Bergman was a clinician, all chapters. Kristina English edited the basic chapters and
administrator, professor, researcher, and writer, and after he those dealing with re/habilitation. Jack Katz reviewed all of
retired from Hunter College in New York City, he went to the chapters and saw to the overall manuscript issues.
Israel to establish audiology as a profession there. He is con- The Handbook is divided into four sections. There are
sidered as the Father of Audiology in Israel. For many years, eight chapters dealing with Introduction, Basic Tests, and
Dr. Bergman has continued to be active as an advisor and an Principles. A chapter that summarizes diagnostic audiol-
officer in international professional organizations. His clar- ogy and brings together the various contributions has been
ity about the events and developments so many years ago added in this edition. Other top-notch audiologists wrote
(see Bergman, 2002) makes him a treasured link to our roots. on Puretone Air Conduction, Bone Conduction, and Speech
This edition is dedicated to Raymond Carhart, “The Audiometry, as well as Masking and Case History.
Father of Audiology.” We are delighted to have the book’s The second section is made up of 14 chapters dealing
dedication and foreword discussing Dr. Carhart; written with Physiologic Principles and Measures. This section of
by Laura Ann Wilber a former student of his and a distin- the book includes auditory measures starting with the con-
guished audiologist in her own right. Dr. Carhart was her ductive mechanism up to the brain and vestibular measures
dissertation advisor and she worked with him and Tom Till- that assess from the inner ear to the brain. Some chapters
man to develop what later became the NU-6 word recogni- include specialty areas such as intraoperative monitoring
tion test. When Laura Wilber earned her Ph.D. there were and therapy for vestibular disorders. Some of the most pro-
few women who were educated at that level in audiology nounced advances in recent years have been made in these
and many people felt that it was a male profession. So Dr. areas.
Carhart’s acceptance of her and clearing a path for her was The third section is devoted to a wide variety of Special
especially important. Populations. It contains 14 chapters beginning with New-
It is worth noting that Laura Wilber has contributed to born Hearing Screening, Assessment of Hearing Loss in
each edition of the Handbook since the first edition in 1972. Children and Educational Audiology and ends with Hearing
She herself was honored by the editors of the HOCA in the Loss in the Elderly, Tinnitus/Hyperacusis, and Tele-Audiol-
previous edition by dedicating the book to her and to three ogy. Four chapters deal with Central Auditory Processing
other audiologists. Disorders and Central Auditory Functions.

xiii
xiv Preface

The final section, Management of Hearing Disorders, is SENSORY/NEURAL

made up of 10 chapters. Five of the chapters deal with hear-
ing aids and cochlear implants, two focus on management, On the one hand, while there is good reason to use puretone
and two more are on Room Acoustics and Assistive Tech- as a compound word, on the other hand, it would be benefi-
nologies as well as Building a Successful Audiologic Practice. cial for the term sensorineural to be separated into sensory
In addition, for the first time we will have a chapter dealing or neural using a slash as we often use for “or” (i.e., sensory/
with infection control in audiology practice that was written neural). This makes it clear that the test or result it did not
by A.U. Bankaitis. This important topic relates to all aspects distinguish sensory from neural. From the term sensorineu-
of audiology that deal with clinical patients and/or research ral it is often not clear what is intended as many profession-
subjects. als assume that sensorineural means sensory. This problem
Sadly, during the writing of the chapter on electroco- has led to important confusions and errors that can be easily
chleography with Rosamaria Santarelli, contributor Edoardo remedied by the use of sensory/neural if it is unclear which
Arslan passed away. He was both her co-author and mentor. is indicated (e.g., with no air-bone gap for puretone thresh-
olds we do not know if it is sensory, neural, or both). If the
specific region is identified (e.g., present Otoacoustic Emis-
NEW FEATURES sions but absent Middle Latency Response) we indicate that
Six new chapters have been introduced in the seventh edi- it was specifically neural or retrocochlear. If it is both we
tion. They are Diagnostic Audiology, The Dizzy Patient and state “both sensory and neural” or just “sensory and neural.”
Vestibular Rehabilitation, Newborn Hearing Screening,
Genetic Hearing Loss, Tele-Audiology, and Infection Con- EPILOGUE
trol. At the close of every chapter is a new section called
We are pleased that the Handbook of Clinical Audiology
Food for Thought, which encourages readers to interact
(HOCA) is used widely by audiologists around the world.
more deeply with the text.
Interestingly when the HOCA first came out in 1972, we were
living in the Turkish Republic. There the word hoca means a
religious leader or a revered teacher. While HOCA is certainly
In this edition of this Handbook we have added supplemen- not a religious leader, we do hope it will be a revered teacher
tal materials (e.g., extended references and case studies) on for the many students and colleagues that read this book.
thePoint companion website at http://thepoint.lww.com/
Katz7e. ACKNOWLEDGMENTS
We would like to thank the editors of Wolters Kluwer, espe-
TERMINOLOGY cially Linda Francis and Meredith Brittain, not only for their
The following is an explanation of some of the spelling con- fine editing of this book but also for advising us and keep-
ventions used in the HOCA and briefly why we chose them. ing us to our deadlines. We would like to mention the fol-
Further discussion may be found in Chapter 1. lowing colleagues and other individuals who also helped us
in many ways to make the Handbook as high quality a text
and reference book that we could. They are Mark Chertoff,
Compound Words Robin Gashler, Jay Hall, Amy Lane, Larry Medwetsky, Mar-
In clinical audiology, as well as in English generally, com- cello Peppi, Lynden Ronsh, and Natalie Turek. In addition,
pound words (two words written as one) are common. Com- I appreciate the tremendous support from my family and
pound words are simplifications of words that are frequently would like to highlight those who made special contribu-
used together. For example, brain and stem are combined in tions to the completion of this edition. They are Eric Kas-
the term auditory brainstem response. When two words are eff, Lainie Kaseff, Mark Katz, and Miriam Kaseff. Another
frequently used together to express a certain meaning, in time, member of my family deserves very special mention. My
they may be connected by a hyphen and eventually joined wife Irma Laufer Katz has been heavily involved as reader,
together into a single word (base ball, base-ball, baseball). advisor, and organizer of most of my projects over the past
58 years. For the Handbook she was also the secretary who
PURETONE kept track of the 46 chapters as they came and went from
authors to editors and back again in the various stages of
The terms pure tone and pure-tone are constantly used in completion. We must certainly mention the many authors
audiology with or without a hyphen. This has encouraged us who contributed their knowledge and skills to make this
to combine them into a compound word, puretone. By choos- Handbook an important contribution to the education and
ing a single word it eliminates the inconstancies that we see practice of those in the field of Audiology. To all of them
when they are used or misused with or without a hyphen. and, of course, to my fellow editors my thanks and gratitude.
 CO N T E N T S

Contributors v 14 Auditory Brainstem Response: Estimation

Dedication ix of Hearing Sensitivity . . . . . . . . . . . . . . . . . . . . .249
Foreword xi Linda J. Hood
Preface xiii
15 Auditory Steady-State Response . . . . . . . . . . .267
Andrew Dimitrijevic and Barbara Cone
SECTION I:
16 Intraoperative Neurophysiological
BASIC TESTS AND PROCEDURES 1 Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .295
Paul R. Kileny and Bruce M. Edwards
1 A Brief Introduction to Clinical Audiology
and This Handbook . . . . . . . . . . . . . . . . . . . . . . . . . 3 17 Middle-Latency Auditory-Evoked
Jack Katz Potentials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .315
Anthony T. Cacace and Dennis J. McFarland
2 Calibration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Laura Ann Wilber and Robert Burkard 18 Cortical Auditory-Evoked Potentials. . . . . . . .337
Kelly Tremblay and Christopher Clinard
3 Puretone Evaluation . . . . . . . . . . . . . . . . . . . . . . .29
Robert S. Schlauch and Peggy Nelson 19 Otoacoustic Emissions . . . . . . . . . . . . . . . . . . . .357
Beth Prieve and Tracy Fitzgerald
4 Bone Conduction Evaluation . . . . . . . . . . . . . . .49
James R. Steiger 20 Clinical Neurophysiology of the
5 Speech Audiometry . . . . . . . . . . . . . . . . . . . . . . . .61 Vestibular System. . . . . . . . . . . . . . . . . . . . . . . . .381
Rachel McArdle and Theresa Hnath-Chisolm Erin G. Piker and Douglas B. Garrison

6 Clinical Masking. . . . . . . . . . . . . . . . . . . . . . . . . . .77 21 Evaluation of the Patient with Dizziness

William S. Yacullo and Balance Disorders . . . . . . . . . . . . . . . . . . . .399
Troy Hale, Henry Trahan, and Tabitha Parent-Buck
7 Case History . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113
Douglas L. Beck 22 Vestibular Rehabilitative Therapy . . . . . . . . . .425
Richard Gans
8 Diagnostic Audiology . . . . . . . . . . . . . . . . . . . . .119
Brian M. Kreisman, Jennifer L. Smart, and Andrew B. John
SECTION III:

SECTION II:
SPECIAL POPULATIONS 435
PHYSIOLOGICAL PRINCIPLES 23 Newborn Hearing Screening . . . . . . . . . . . . . . .437
AND MEASURES 135 Karl R. White

24 Assessment of Hearing Loss in Children . . . .459

9 Tympanometry and Wideband Acoustic Allan O. Diefendorf
Immittance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .137
Lisa L. Hunter and Chris A. Sanford 25 Genetic Hearing Loss . . . . . . . . . . . . . . . . . . . . .477
Carmen Brewer and Kelly King
10 Acoustic Stapedius Reflex Measurements. . . .165
M. Patrick Feeney and Kim S. Schairer 26 Educational Audiology . . . . . . . . . . . . . . . . . . . .501
Cheryl DeConde Johnson and Carrie Spangler
11 Introduction to Auditory Evoked
Potentials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .187 27 Central Auditory Processing: A Functional
Robert Burkard and Manuel Don Perspective from Neuroscience. . . . . . . . . . . . .513
Dennis P. Phillips and Rachel N. Dingle
12 Electrocochleography . . . . . . . . . . . . . . . . . . . . .207
Rosamaria Santarelli and Edoardo Arslan 28 Auditory Pathway Representations of
Speech Sounds in Humans. . . . . . . . . . . . . . . . .527
13 Auditory Brainstem Response: Differential Daniel A. Abrams and Nina Kraus
Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .231
Frank E. Musiek, Jennifer E. Gonzalez, and Jane A. Baran

xv
xvi Contents

29 Central Auditory Processing Evaluation: 39 Troubleshooting and Testing

A Test Battery Approach. . . . . . . . . . . . . . . . . . .545 Hearing Aids . . . . . . . . . . . . . . . . . . . . . . . . . . . . .727
Kim L. Tillery William Cole and Marshall Chasin

30 Central Auditory Processing Disorder: 40 Hearing Aid Fitting for Children: Selection,
Therapy and Management . . . . . . . . . . . . . . . .561 Fitting, Verification, and Validation. . . . . . . . .759
Jack Katz, Jeanane Ferre, William Keith, and Angela Loucks Susan Scollie
Alexander
41 Hearing Aid Fitting for Adults: Selection,
31 Individuals with Multiple Disabilities . . . . . .583 Fitting, Verification, and Validation. . . . . . . . .777
Anne Marie Tharpe and Samantha Gustafson Michael Valente and Maureen Valente
32 Noise Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . .595 42 Building and Growing an Audiologic
Brian Fligor, Marshall Chasin, and Rick Neitzel Practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .805
33 Nonorganic Hearing Loss. . . . . . . . . . . . . . . . . .617 Melanie Herzfeld
Frederick N. Martin and John Greer Clark 43 Implantable Hearing Devices . . . . . . . . . . . . . .817
34 Hearing Loss in the Elderly: A New Look Teresa A. Zwolan
at an Old Problem . . . . . . . . . . . . . . . . . . . . . . . .631 44 Intervention, Education, and Therapy
Barbara E. Weinstein for Children with Hearing Loss . . . . . . . . . . . .835
35 Tinnitus and Hyperacusis . . . . . . . . . . . . . . . . .647 Christine Yoshinaga-Itano and Kristin M. Uhler
Richard S. Tyler, William Noble, Claudia Coelho, Eveling Rojas 45 Audiologic Rehabilitation . . . . . . . . . . . . . . . . .849
Roncancio, and Hyung Jin Jun Joseph Montano
36 Tele-audiology . . . . . . . . . . . . . . . . . . . . . . . . . . .659 46 Infection Control . . . . . . . . . . . . . . . . . . . . . . . . .861
De Wet Swanepoel A.U. Bankaitis

SECTION IV: SECTION V:

MANAGEMENT OF HEARING APPENDICES 869
DISORDERS 673
37 Room Acoustics and Auditory Rehabilitation
Author Index 895
Technology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .675
Joseph Smaldino, Brian Kreisman, Andrew John, and Lindsay Subject Index 907
Bondurant

38 Hearing Aid Technology . . . . . . . . . . . . . . . . . . .703

Jennifer Groth and Laurel A. Christensen
 S EC T I O N I

Basic Tests and

Procedures
 C H A P T ER 1

A Brief Introduction to Clinical

Audiology and This Handbook
Jack Katz

Audiology is the study of hearing and hearing disorders, a

field devoted to helping those with auditory and vestibular
AUDIOLOGY FROM 1940s
dysfunctions. This work may involve evaluation, re/habili- TO TODAY
tation, counseling, education, research, and/or screening/ As mentioned above, the field of audiology was founded
prevention. during WWII. Prior to that time hearing testing was carried
Audiology combines aspects of science and art with out using tuning forks and whispered speech by medical
techniques that are based on both basic and clinical doctors, although some puretone audiometers that pro-
research. We use sophisticated equipment to provide preci- vided repeatable stimuli were also in use. The combined
sion in determining the type and extent of the problems. efforts of the different disciplines fostered the variety of
But audiology is also an art. It involves the ability to per- procedures we have to address the problems caused by hear-
form the various tasks precisely and to provide informa- ing impairment. Bone-conduction testing and speech audi-
tion and maximum support to the individuals affected ometry were soon added to the clinical tools. Aspects such
and their families. Because of these intellectually and emo- as lip reading/speech reading, auditory training, and coun-
tionally gratifying aspects, it makes audiology an exciting seling were borrowed from deaf education, psychology, and
career. speech-language pathology. An important adjunct for the
In my more than 50 years in this field, audiology service members was the fitting of hearing aids which were
has continued to be interesting and rewarding work. quite limited by today’s standards. Nevertheless for years
It is a comparatively new field that emerged in the after- after the war these veterans were still using and benefiting
math of World War II (WWII) to aid service members from the amplification and training that they had received
who suffered hearing impairments. It brought together from those early audiologists when the profession was in its
speech-language pathologists, deaf educators, psycholo- infancy.
gists, and ear, nose, and throat (ENT) physicians. This After leaving military service, the early audiologists
interdisciplinary cooperation was responsible for the began to train others at colleges and universities. Audiolo-
excellent services that were provided to the injured gists began to research the clinical problems that they faced
military personnel. At the same time these multidisci- and many of these approaches and solutions are still in use
plinary activities helped to lay the groundwork for the today. These procedures also led the way to important inno-
field of audiology. Indeed this interdisciplinary aspect of vations. Because it was clear that we did not have enough
the field of audiology remains one of its great strengths diagnostic information to accurately measure and catego-
even today. Initially, audiologic work was carried out in rize hearing disorders, early on, there was a heavy emphasis
military hospitals and then spread to universities and on developing new diagnostic procedures. For a number of
university clinics, afterward to hospitals and community years the area of diagnosis was the primary focus in audio-
clinics. logic research and practice.
Presently there are about 12,000 members of the When audiologists began dispensing hearing aids, this
American Academy of Audiology and approximately 2,000 caused an expansion of attention, from just evaluation and
members of the International Society of Audiology. Also identification of hearing loss to include providing means
memberships continue to grow in local, state, and national of managing hearing difficulties and therapy to address
associations around the world. Audiology has several peer- the communication problems. Hearing aid fitting was also
reviewed journals and other publications, both printed and a major impetus for audiologists to go into private prac-
digital, that report on research and clinical developments. tice. At the same time there were major breakthroughs in
The field of audiology is constantly expanding its horizons physiological measurements. This began with what we now
and developing deeper understandings of both normal and refer to as cortical responses, but after a few years, earlier
abnormal processes. responses were identified from the auditory nerve and even

3
4 SECTION I • Basic Tests and Procedures

the cochlea. The field of audiology has expanded to include Evanston, Illinois went on to contribute significantly to the
the assessment of more complex functions at all levels of the field of audiology in their own right.
peripheral and central auditory nervous system. Immittance
measurements enabled audiologists to assess mechanical
properties of the auditory system of the outer and middle
ABOUT THIS HANDBOOK
ears, as well as middle-ear muscle responses that rely on The first edition of the Handbook of Clinical Audiology was
auditory nerve and brainstem activity. Specialties such as published in 1972 and subsequent editions have served sev-
auditory processing disorders, educational audiology, ves- eral generations of audiologists in the United States and
tibular function, and interoperative monitoring have added increasingly throughout the world. It is used widely as both
to the breadth and depth of the field. a text and reference book by students and professionals in
The growing sophistication and understanding of audi- various fields.
tory functions and development can be seen in the lowering Currently, for this edition, we have five editors who
of the target ages for various services. In the mid-1950s it have diverse areas of specialization in clinical, research, and
was taught that we should wait until deaf children are per- teaching aspects for which they are responsible. To broaden
haps 10 years of age before testing them and presumably our horizons and to be as inclusive as possible, in this edi-
provide amplification after that. Given our current state of tion we have tried to include more international voices and
knowledge, in retrospect, this seems absurd and counterpro- procedures.
ductive. At that time we did not understand that develop- We have squeezed in as much information in 1,000
mental problems should be identified at the earliest possible pages as we could. The more than 90 contributors are highly
time. Otherwise, the person could miss critical periods and regarded audiologists who also have clinical, research, and
lose plasticity, as well as fall further behind with reduced teaching experience. This makes the chapters authoritative,
learning and more acquired misconceptions. Now, neona- well organized, and geared for sharing our knowledge in a
tal hearing screening is widespread and we strive to begin field that we love. We have always considered readability an
habilitation by 6 months of age. In fact, in the past, one important feature of this book and especially now that it is
audiologist was ridiculed when she advocated that audiolo- used by many people whose first language is not English.
gists fit hearing aids for children as young as 1 year of age. The 46 chapters are divided into four sections. Section I
Once we realized the critical importance of the early years deals with basic tests and procedures that are used by most
for later development, early identification and assessment audiologists for most of the people with whom they work.
procedures, as well as training procedures were targeted and This involves puretone air and bone conduction, as well as
developed. standard speech audiometry. Calibration and case history
As the field of audiology expanded so did the academic chapters are also important components for any audiologic
demands on the practitioners. Initially, a bachelor’s degree assessment. The Diagnostic Audiology chapter helps the
was required to practice and then a master’s degree was reader to combine all of the previous information into a
the entry level along with basic clinical certification. As in coherent diagnosis.
the past a Ph.D. was generally desired for university teach- Section II introduces the various physiological and
ing and research. In more recent years (in the United States) electrophysiological procedures used by audiologists at this
the Doctorate of Audiology (Au.D.) degree was introduced time. These include immittance measures that primarily
to provide even broader clinical teaching and training expe- reveal the status of the middle ear. Electrocochleography and
riences. Also, higher levels of competency and certification Otoacoustic Emissions provide detailed information about
are generally required today to practice audiology. Students the responses from the cochlea, the end organ of hearing.
interested in a career that includes independent research Five chapters in this section discuss the electrophysiological
continue to pursue a Ph.D. in audiology, hearing science, responses from the auditory nerve, brainstem, and various
or related areas. Now many of the top university programs areas of the brain. The chapter on intraoperative monitor-
in audiology have both Ph.D. and Au.D. trained professors ing describes the analysis of the auditory system during sur-
to provide the student the best of both worlds. We also see gery that informs the surgeons about the status and pos-
combined Au.D./Ph.D. programs that offer students excel- sible adverse effects of their manipulations of the auditory
lent ground for both clinical and research endeavors. system. The final three chapters in this section deal with the
We owe a debt of gratitude to those early audiologists vestibular system. They begin with the study of vestibular
who helped to form this vibrant and vital health profession. neurophysiology and end with vestibular rehabilitation.
Although we cannot mention the many important con- Section III is called Special Populations. This recog-
tributors, it is perhaps appropriate to mention Raymond nizes that certain groups often require modifications in
Carhart (1912 to 1975) who is generally recognized as “The audiometric procedures or accommodations. Evaluation of
Father of Audiology.” He was an important contributor to young children offers a special challenge to the audiologist
the developing field of audiology and an excellent teacher. because they do not have the auditory or cognitive devel-
Many of his students from Northwestern University in opment needed for some of the tests, and it is sometimes
 CHAPTER 1 • A Brief Introduction to Clinical Audiology and This Handbook 5

difficult to have them perform in the expected fashion. This known syndrome. Since audiologists are often the first
chapter describes ways to obtain the desired results. Hear- professionals to suspect a genetic basis for a hearing loss,
ing, screening, and educational audiology generally involve it is important to have current information available as
the work carried out in schools with those who have nor- well as the knowledge of resources.
mal hearing as well as those with auditory impairments. e. Audiology Tele-practice follows the global trend to pro-
This section also includes chapters that deal with those who vide appropriate services at a distance from the profes-
have multiple disabilities, hereditary hearing loss, and the sional. Using a range of communication technologies
elderly. Other special groups are those with noise-induced and appropriate training of para-professionals, audiolo-
hearing loss, those who have tinnitus, and individuals who gists can treat individuals in remote places who might
have “nonorganic” hearing loss. Four of the other chapters otherwise not receive care. Tele-practice also provides
involve auditory processing disorders, which include the convenience to patients who live relatively close by, but
bases of central auditory problems, diagnostic procedures, nonetheless find it challenging to visit the clinic for rou-
and subsequent remediation. This section concludes with tine problems. By making oneself available using tele-
a chapter on tele-practice in which audiologists can work technology, the audiologist helps patients conserve their
with people at far distances via communication systems. physical energy, time, and travel expenses, while keeping
Section IV deals with the management of hearing dis- abreast of the patient’s challenges as they develop.
orders. It begins with acoustical environments and tech- f. The topic of Infection Control relates to every aspect of
nologies that are used to aid the hard-of-hearing person in audiology, because it is important not to harm the peo-
reducing the influence of noise and other factors that can ple whom we are here to help. Infection control is part
compromise communication. This is followed by four chap- of every aspect of our work and for this reason it is the
ters related to various aspects of hearing aids and hearing first of many chapters, in the future, that will be available
aid fittings. Another chapter, which deals with implantable from the Point on internet.
hearing devices, is a rapidly expanding area. The chapter
discusses cochlear implants and other devices that are sur- Other New Features in This
gically imbedded into the person with a hearing loss. Two
other chapters deal with management of those with hearing Handbook
problems in the classroom and with rehabilitation of adults. In this edition of the Handbook we have reduced the num-
There is also a chapter in this section that advises audiolo- ber of references provided in each chapter, but there are
gists on how to start an effective audiologic practice. extensive lists of references for the interested students, pro-
fessors, and researchers on the Point. In this way the reader
New Chapters in This Edition is not encumbered with reading through or skipping over
many references when trying to understand the concepts
a. Diagnostic Audiology serves an important purpose in and to remember the facts in this book. At the same time
bringing together the information from the basic evalua- there are thousands of references organized by chapters
tion procedures in this book to form an audiologic inter- online for those who are interested in research or for greater
pretation and an understanding of the patient’s needs. depth on the topics covered in this book.
This chapter will also discuss some procedures that are Another new feature is the thought questions at the
not covered in the preceding chapters, as well as men- end of each chapter. They will ask how and what you
tioning what our tests do not tell us. would do in dealing with, or solving, problems associated
b. The Dizzy Patient and Vestibular Rehabilitation chapter with the information in the chapter. This is not another
is an extension of a former Handbook chapter, apply- hoop to jump through but a valuable exercise. The stu-
ing diagnostic information to enable appropriate treat- dent must take what they have learned from the chapter
ment decisions for patients with vestibular problems. It and combine it with their other knowledge to figure out
will describe audiologic procedures designed to relieve a good solution to a problem/question. In this way they
patient’s symptoms, as well as the role of physical therapy take what was on page and internalize it, while it is fresh
and the necessity of collaboration among healthcare pro- in their minds, and put the information to a practical use.
fessionals. This will help you to internalize the information and make
c. Hearing Screening discusses newborn hearing screen- the material your own.
ings, school screenings, and other screening procedures
using universal approaches and targeted population
approaches. The specific procedures, their value, and
Terminology
outcomes of screening programs will be discussed. Most of the terms used in this edition are standard in
d. Hereditary Hearing Loss describes much-needed infor- the field at this time. However, when a change is made it
mation for audiologists related to genetic aspects of should be for a worthwhile purpose and not one that creates
hearing loss that may be nonsyndromic or part of a important problems. For example, this writer was pleased to
6 SECTION I • Basic Tests and Procedures

see a recent change back to a previous term. What was once For this edition we will combine both the Jacobson and
called Central Auditory Processing was changed to Auditory Northern and the Martin and Clark approached as this
Processing and recently was changed back to the clearer and seems to be better than sensory-neural and avoids the prob-
more specific Central Auditory Processing again (American lems that ‘sensorineural’ has caused.
Academy of Audiology, 2010).
PURETONE
SENSORY/NEURAL The reader might infer that the writer does not like com-
A conductive loss is a mechanical impairment of hearing, pound words (two words that are combined to form
associated with the outer and/or middle ears. For many years a composite of the two, e.g., flashlight, textbook). We
a nonconductive loss had been called a “nerve loss.” After rarely combine opposites (e.g., dogcat, daynight, or even
WWII it was changed to “sensory-neural loss” when ENT sensorineural). But when two words are frequently spo-
doctors and audiologists were then able to separate sensory ken together (e.g., base and ball) often the first step is to
(cochlear) from neural (acoustic nerve or brainstem) dis- hyphenate them (base-ball) and when people get used to
orders. For example, cochlear problems (such as Meniere’s this expression, they are often combined and made a com-
disease) were demonstrated by a rapid growth of loudness pound word (baseball).
when a sound was presented above the person’s threshold The term “pure tone” is shown one or more times on
of hearing. On the other hand with retrocochlear losses every audiogram and appears in almost every report and
(e.g., auditory nerve or brainstem) there was no accelerated is a very common type of audiometer (but in that case it
growth of loudness with sounds above the neural hearing might be hyphenated because it is followed by a noun, e.g.,
level (as with a person who had an auditory nerve tumor). pure-tone audiometer). Because (1) we have to explain this
However, after a number of years the term sensory-neural to students and often have to decide if it needs a hyphen
was changed to “sensorineural.” There was little reaction to when we are writing, and (2) it is surely time to graduate
this minor change. from pure-tone to puretone, this change seems appropriate.
I was shocked, however, to receive an angry phone call In this case there is no compelling reason for doing so (as it
from a doctor who claimed that I made a mistake which would be in the case of sensorineural) but it seems that it is
caused him to delay surgery for his patient’s auditory nerve time for “pure” and “tone” to be officially married and to be
tumor. From a review of my report it was abundantly clear a compound word forever more.
that the patient had retrocochlear characteristics that are
consistent with an “eighth nerve or brainstem involvement” ESPECIALLY FOR STUDENTS—
and not cochlear involvement. How could that have been
misinterpreted? The physician only had read up to the first
SOME SUGGESTIONS
test result, that puretone testing showed a “sensorineural As a student, it is most helpful to educate yourself broadly
loss in the right ear.” On seeing the term “sensorineural” he in your profession and related subjects. You may benefit
incorrectly concluded that it was a cochlear problem and not from speech, psychology, and many other courses as much
a very dangerous auditory nerve tumor. He did not know as from some of your audiology courses. The ability to take
that the term sensorineural could represent two importantly a broader view is certainly an advantage no matter how you
different types of hearing loss. Puretone thresholds distin- plan to practice audiology.
guish conductive from both sensory and neural disorders. When you have a choice in taking your first job, it is well
Later on similar mistakes, with the term sensorineural, to take one that covers a wider area of professional activity
were made by knowledgeable audiologists in two separate over one that is narrow. You may find that an area that pre-
publications. This convinced me that the term sensorineural viously did not seem too interesting is one that you realize
can create serious problems that should be less problematic is very interesting or gratifying. Also, if you have a broad
with the original term sensory-neural. experience you can qualify for more opportunities later on.
Since the second edition of the Handbook we have used As you get deeper into your areas of major interest you
the term sensory-neural to avoid the errors caused by senso- will necessarily reduce how broadly you can practice. But
rineural (Katz, 1978). If those who coined the term sensori- having a prior background or learning can help you in what
neural originally did not try to combine two auditory com- you are doing and perhaps provide variety in your profes-
ponents that we try hard to distinguish from one another, sional activities. Later on, if you have specialized in one
it is likely that fewer problems would have occurred. Other area then an exciting and enriching aspect is to carry out
authors have recognized the problem with the term senso- research to improve your success or simply to obtain a better
rineural. Jacobson and Northern (1991) suggest using just understanding. One way to repay your profession for train-
sensory or neural, when it is clearly one or the other. Martin ing you is to supervise students in your external practicum
and Clark (2012) avoid the confusion by using the term site. Mentoring students and sharing what you have learned
sensory/neural which is also a good way to clarify the term. can be most rewarding, but in addition you may learn some
 CHAPTER 1 • A Brief Introduction to Clinical Audiology and This Handbook 7

new concepts from the students that you may have missed the main reason for choosing each of them for your
or learn from having to answer their questions. department.
It is our pleasure to provide you with this book full of 3. If you were the editor of Handbook of Clinical Audiology
knowledge that was written by dozens of audiologists who and could only add one chapter to this edition, based on
have enjoyed sharing with you their hundreds of years of what you know or imagine, which of the six new chapters
experience in this wonderful field. Finally, as professionals (see above) would you choose and why?
we should be committed to helping those we serve. We also
need to follow the rules. Of course, in addition your work
needs to provide you with the necessities of life. Despite
KEY REFERENCES
these constraints, to a great extent, your profession is pretty A full list of references for this chapter can be
much what you make of it. found at http://thePoint.lww.com. Below are the key refer-
ences for this chapter.
American Academy of Audiology. (2010) Guidelines for diagnosis,
FOOD FOR THOUGHT treatment and management with children and adults with cen-
1. What personal characteristics and experiences do you tral auditory processing disorders. Available online at: http://
www.audiology.org/resources/documentlibrary/documents/
have that you think will be helpful to you as an audiolo-
CAPDGuidelines 8–2010.pdf.
gist? Jacobson J, Northern J. (1991) Diagnostic Audiology. Austin, TX:
2. You are the Director of an Audiology Department at Pro-Ed; p 8.
a medical center. There is a need to establish guide- Katz J. (1978) Clinical audiology. In: Katz J, ed. Handbook of Clini-
lines for the audiologists to provide a degree of consis- cal Audiology. Baltimore, MD: Williams & Wilkins Co.; p 5.
tency (e.g., in reports). You have seen “sensorineural” Martin F, Clark JG. (2012) Introduction to Audiology. Boston:
spelled like that and also as sensory/neural. Consider Pearson; p 446.
 C H A P T ER 2

Calibration

Laura Ann Wilber and Robert Burkard

out this chapter, we will refer to various standards. In the

WHY CALIBRATE? United States, we (mostly) rely on standards that have been
In some ways, calibration can be compared to exercising. We approved by the American National Standards Institute
know it is good for us, but some of us would prefer not to (ANSI). Nonetheless, we will also refer standards written and
participate. However, unlike exercising, if one does not cali- approved by the International Electrotechnical Commission
brate, it hurts others (our clients) more than it does us. For (IEC) and the International Organization for Standardiza-
years, many clinicians felt that calibration was something tion (ISO). Since these standards do not have the status of
that researchers did but that such procedures were not nec- law, it is important to understand how, and perhaps why,
essary in the clinic. Today, that basic attitude has changed they are developed: Standards are developed so that manu-
dramatically. The Occupational Safety and Health (OSHA) facturers of equipment (from all countries) and users of the
regulations (1983) require that audiometric equipment be equipment are all on the same page. According to its website
regularly checked. Some state regulations for hearing aid dis- (http://www.ansi.org/about_ansi/overview/overview.aspx?
pensers and/or for audiologists also require that equipment menuid=1), ANSI is “a private, nonprofit organization
calibration (and records of calibration) be maintained. Fur- (501(c) 3) that administers and coordinates the U.S. vol-
thermore, many state health departments concerned with untary standardization and conformity assessment system.”
school screening also insist on having calibration checked Its “mission is to enhance both the global competitiveness
on a routine basis. Thus, we must calibrate if we are to meet of U.S. business and the U.S. quality of life by promoting
the current regulations, and we should calibrate to make and facilitating voluntary consensus standards and confor-
sure our results are within specified tolerances. mity assessment systems, and safeguarding their integrity”
Initial audiometric calibration provided by the man- (ANSI, 2004). Some values (e.g., the “0” hearing level [HL])
ufacturer is insufficient to guarantee that the audiometer have both international and national approval. In most
will function correctly over time. Although modern digi- cases, ANSI standards and ISO and IEC standards are tech-
tal audiometers are less likely to arrive out of calibration nically very similar (in current jargon, this is called harmo-
and are less likely to develop problems later than the older nization). Harmonization of ANSI and international stan-
vacuum tube machines, even brand new audiometers that dards enhances commercial interchange between nations.
have just arrived from the factory, as well as audiometers If, for example, the ANSI audiometer standard was radi-
that were in perfect calibration when they were new, can cally different from the IEC standard, manufacturers would
show variations in sound level, frequency, distortion, to have to build instruments solely for the American market
name a few. Problems are often related to the transduc- and solely for the European or World market. In a relatively
ers (earphones, bone vibrators, loudspeakers), but the small-volume industry (such as audiometric instrumenta-
electronic components can also lead to the audiometer tion), this would be impractical at best.
failing to remain in calibration. It is the responsibility of All standards are reviewed periodically. If they are reaf-
the user (i.e., the audiologist) to either check its calibra- firmed (and not changed), then the standard will read, for
tion personally or to arrange for regular calibration of the example, ANSI S3.39-1987 (R2012). This means the stan-
equipment by an outside service. The audiologist who has dard was approved in 1987 and was most recently reaffirmed
demonstrated that the clinic equipment is “in calibration” in 2012. If the standard is revised, then the date changes (e.g.,
can then feel confident in reporting the obtained results. ANSI S3.6-2010, which was previously ANSI S3.6-2004). An
Calibration checks can determine if an audiometer meets announcement is made when the standard is going to be
appropriate standards and also whether the instrument has voted on so that interested parties can obtain a copy and
changed over time. comment to the person or persons who will be voting. For
The purpose of this chapter is to tell the audiologist or example, audiologists might contact the American Speech-
student how to check audiometers to determine if they meet Language-Hearing Association (ASHA) or the American
current national (or international) standards. Through- Academy of Audiology (AAA), both of which are voting

9
10 SECTION I • Basic Tests and Procedures

members. This is the basic procedure for development and equipment, if daily listening checks are strictly enforced,
approval of standards. For more information on the stan- transducers should be verified at least annually, unless there
dards process, the reader is referred to Melnick (1973) and is reason to suspect that the output has changed. If daily
Wilber (2004). There are three primary sources of funding listening checks are not strictly enforced more complete
for the production of standards in acoustics: financial sup- checks might be necessary. In addition to regularly sched-
port from Acoustical Society of America (ASA), fees paid by uled checks, audiometers should be tested whenever the
the voting members of an Accredited Standards Committee, clinician notices anything unusual in their performance.
and income from the sales of standards. Through your pur- Sometimes test results themselves reveal the need for
chase of standards, you are supporting the efforts of those an immediate calibration check (e.g., when the same air–
professionals who donate their time and effort to develop bone gap is obtained for two successive patients). It is always
and maintain ANSI standards. Contact information of the better to check the audiometer first rather than assume the
secretariat of ANSI S1, S2, S3, and S12 is: problem lies with the client or clinician. A quick biologic
check (described later) can always be performed. If this con-
Acoustical Society of America
firms the probability of an equipment problem, then a more
ASA Secretariat
elaborate electroacoustic check should be carried out.
35 Pinelawn Road, Suite 114E
If the audiologist discovers that the frequency or time
Melville, NY 11747-3177
components of the audiometer are out of calibration, then
E-mail: [email protected]
in most instances the manufacturer or a local representa-
tive should be contacted for immediate repair and/or proper
calibration of the instrument. However, if there is a stable
PARAMETERS OF CALIBRATION deviation in output level at a given frequency, calibration
The first step in learning how to check calibration should corrections can be made by adjusting the trim pots (potenti-
always be to read the appropriate manual(s) that accom- ometers) on the audiometer, by using the audiometer’s self-
pany the audiometric equipment that you have purchased. calibrating mechanism, or by posting a note on the front of
Additional resources include electronic parts, stores that the audiometer indicating the corrections. If paper correc-
often have basic manuals on test equipment, ASHA and tions must be used, then the adjustment in decibels (plus
ASA. A number of books have also discussed procedures or minus) that should be made at the various frequencies
for acoustic measurements and equipments that might be should be shown for each transducer. Note that if the SPL
used in such measurements (Beranek, 1988; Decker and output is too high (e.g., by 5 dB), then you must increase
Carrel, 2004; Silverman, 1999). The United States Govern- their audiometric threshold (e.g., by 5 dB HL). Most mod-
ment Printing Office is also a good source of information ern audiometers provide some sort of internal (typically
on basic test procedures. The specific parameters that must software based) calibration system for earphones, and many
be checked in an audiometer are outlined in standards pro- also provide this for bone conduction or sound field. If one
vided by the ANSI and the IEC. See Table 2.1 for a listing of plans to use bone vibrators for both mastoid and frontal
standards relevant to calibration of audiometric equipment. bone testing or two sets of earphones with the same audi-
It is beyond the scope of this chapter to discuss each area of ometer (e.g., supra-aural earphones and insert receivers), it
calibration in detail. For the readers who intend to perform is probably advisable to use “paper corrections,” rather than
their own calibration of audiometric equipment, they need trying to adjust trim pots between each transducer’s use. If
to purchase copies of the latest standards to verify the exact frequent level adjustments are required, it is probably wise
parameters to be checked and their permissible variability. to check with a qualified technician.
To better understand the procedures for checking calibra-
tion, one must first understand the parameters that need to
be checked, as well as the equipment used to perform these
INSTRUMENTATION
calibration checks. For puretone and speech audiometers, As mentioned earlier, the calibration of an audiometer
the three parameters are (1) frequency, (2) level (sound pres- requires the use of various pieces of electroacoustic and
sure level [SPL] or force level or [FL]), and (3) time. These electronic instrumentation. Most, if not all, graduate audi-
parameters apply whether one is using a portable audiom- ology programs will have the instrumentation needed to
eter, a standard diagnostic audiometer, or a computer-based at least evaluate whether the audiometer meets the refer-
audiometric system. ence equivalent threshold sound pressure level (RETSPL),
Some organizations, such as ASHA and OSHA, specify frequency, linearity, and distortion standards specified in
time intervals at which calibration checks should be made. ANSI S3.6 Specification for Audiometers. In this section, we
With current solid-state electronic circuitry, frequency, and will review the use of several basic instruments, including
time, parameters should be checked when the audiometer sound level meter (SLM), multimeter, frequency counter,
is first acquired and at yearly intervals thereafter. Older oscilloscope, and digital spectrum analyzer. More details on
equipment should be checked at least biannually. For newer acoustics and instrumentation can be found in numerous
 CHAPTER 2 • Calibration 11

TA B L E 2 .1

ANSI, IEC, and ISO Standards for Audiometers and Audiometric Testing
Number Title
ANSI S3.1-1999 (R 2008) Maximum Permissible Ambient Noise for Audiometric Test Rooms
ANSI S3.2-2009 Method for Measuring the Intelligibility of Speech Over Communication Systems
ANSI S3.6-2010 Specification for Audiometers
ANSI S3.7-1995 (R 2008) Coupler Calibration of Earphones, Method for
ANSI S3.13-1987 (R 2012) Mechanical Coupler for Measurement of Bone Vibrators
ANSI S3.20-1995 (R 2008) Bioacoustical Terminology
ANSI S3.21-2004 (R 2009) Method for Manual Pure-Tone Threshold Audiometry
ANSI S3.25-2009 Occluded Ear Simulator
ANSI S3.36-2012 Specification for a Manikin for Simulated In Situ Airborne Acoustic Measurements
ANSI S3.39-1987 (R 2012) Specifications for Instruments to Measure Aural Acoustic Impedance and Admittance
(Aural Acoustic Immittance)
ANSI S1.4-1983 (R 2006) Specifications for Sound Level Meters
IEC 60318-1:2009 Electroacoustics: Simulators of Human Head and Ear. Part 1—Ear Simulator for the
Calibration of Supra-aural and Circumaural Earphones
IEC 60318-4:2010 Electroacoustics: Simulators of the Human Head and Ear. Part 4—Occluded-Ear Simula-
tor for the Measurement of Earphones Coupled to the Ear by Means of Ear Inserts
IEC 60318-6:2007 Electroacoustics: Simulators of Human Head and Ear. Part 6—Mechanical coupler for
the measurement on bone vibrators
IEC 60645-3:2007 Electroacoustics: Audiometric equipment. Part 3—Auditory Test Signals of Short Dura-
tion for Audiometric and Neuro-otological Purposes
IEC 60645-5:2004 Electroacoustics: Audiometric Equipment. Part 5—Instruments for the Measurement of
Aural Acoustic Impedance/Admittance
IEC 60645-6:2009 Electroacoustics: Audiometric Equipment. Part 6—Instruments for the Measurement of
Otoacoustic Emissions
IEC 60645-7:2009 Electroacoustics: Audiometric Equipment. Part 7: Instruments for the Measurement of
Auditory Brainstem Responses
IEC 60645-6:2009 Electroacoustics: Audiometric Equipment. Part 6: Instruments for the Measurement of
Otoacoustic Emissions
IEC 60645-7:2009 Electroacoustics: Audiometric Equipment. Part 7: Instruments for the Measurement of
Auditory Brainstem Responses
ISO 8253-1:2010 Acoustics: Audiometric Test Methods. Part 1: Basic Pure-Tone and Bone Conduction
Threshold Audiometry
ISO 389-1:l998 Acoustics: Reference Zero for the Calibration of Audiometric Equipment. Part 1: Reference
Equivalent Threshold Sound Pressure Levels for Pure Tones and Supra-aural Earphones
ISO 389-2:l994 Acoustics: Reference Zero for the Calibration of Audiometric Equipment. Part 2: Reference
Equivalent Threshold Sound Pressure Levels for Pure Tones and Insert Earphones
ISO 389-3:l994 Acoustics: Reference Zero for the Calibration of Audiometric Equipment. Part 3: Reference
Equivalent Threshold Force Levels for Pure Tones and Bone Vibrators
ISO 389-4:l994 Acoustics: Reference Zero for the Calibration of Audiometric Equipment. Part 3: Reference
Equivalent Levels for Narrow-Band Masking Noise
ISO 389-5:2006 Acoustics: Reference Zero for the Calibration of Audiometric Equipment. Part 5: Reference
Equivalent Threshold Sound Pressure Levels for Pure Tones in the Frequency Range
8 kHz to 16 kHz
ISO 389-6:2007 Acoustics: Reference Zero for the Calibration of Audiometric Equipment. Part 6: Reference
Threshold of Hearing for Test Signals of Short Duration
ISO 389-7:2005 Acoustics: Reference Zero for the Calibration of Audiometric Equipment: Part 7: Reference
Threshold of Hearing under Free-Field and Diffuse-Field Listening Conditions
ISO 389-8:2004 Acoustics: Reference Zero for the Calibration of Audiometric Equipment. Part 8: Reference
Equivalent Threshold Sound Pressure Levels for Pure Tones and Circumaural Earphones
ANSI, American National Standards Institute; ASHA, American Speech-Language-Hearing Association; IEC, International Electrotechnical
Commission; ISO, International Organization for Standardization.
NOTE: All ANSI, ISO, and IEC Standards referred to in this chapter are listed in this table.
12 SECTION I • Basic Tests and Procedures

texts (e.g., Decker and Carrell, 2004; Harris, 1998; Rosen correct this problem. If the frequency counter is an option
and Howell, 1991; Speaks, 1996). in a multimeter, there is often no adjustable trigger level,
and the signal level must be changed to correctly trigger the
counter function.
Multimeter
The term “multimeter” indicates that this device can be
used to make multiple measurements. In most cases, a
Sound Level Meter
multimeter will allow one to make measurements of volt- The SLM is actually multiple instrumentation components
age, current, and resistance. Each of these measurements provided in a single instrument. You can combine separate
is made differently, and we will limit our discussion herein instruments into a usable device when an SLM is not avail-
to making voltage measurements. To measure voltage, we able. At a minimum, for checking the calibration of RET-
must make the measurement in parallel to (across) the SPL (i.e., 0 dB HL values on the audiometer), you need an
device of interest. For example, if we are interested in atten- acoustic calibrator, an appropriate coupler (2 cc and/or 6
uator linearity, we want to place the leads of the multime- cc), a microphone, and the SLM. SLMs used for checking
ter across the earphone leads, with the earphone plugged the calibration of audiometers should be Type 1, as should
into the audiometer output. We can replace the earphone microphones used for such calibrations. The most com-
with an equivalent impedance (in most cases, a 10-, 50-, or monly used Type 1 microphone is a condenser microphone.
300-ohm resistor for ER-3 A, TDH-39, TDH-49, or TDH-50 Condenser microphones come in four standard sizes (refer-
earphones). Simply unplugging the earphones and plug- ring to their diameter): 1/8″, 1/4″, 1/2″, and 1″. For calibra-
ging in the multimeter will likely produce inaccurate tion of a supra-aural earphone, a 1″ microphone is specified
results, because this approach in most cases will change the in ANSI S3.6-2010 (because of its sensitivity—see the state-
load impedance of the audiometer output. It is important ment that follows). In general, the smaller the microphone
to purchase a true root mean square (RMS) multimeter is, the higher its upper frequency cutoff and the less its sen-
for accurate RMS voltage readings. It is important to set sitivity. Sensitivity is a measure of its efficiency transferring
the meter to AC, or alternating current (vs. DC, or direct sound pressure into voltage and is commonly reported as
current), voltage. The meter is most accurate when set to millivolts per pascal, or in dB re: 1 V/Pa. Many condenser
the lowest voltage range possible. In most cases, the volt- microphones require a DC polarization voltage of 200 V.
age range is set in powers of 10, where the listed voltage is Some condenser microphones are prepolarized and hence
the maximum voltage possible for that voltage range. When do not require an externally applied polarization voltage.
this maximum voltage is exceeded, an overload is indicated Microphones also come as pressure microphones (to be
(see multimeter manual for the overload indicator for your used in a coupler), free-field microphones (to be used in
multimeter). You adjust the multimeter range until you sound field recordings such as when measuring the ambi-
have the most sensitive range (lowest maximum voltage) ent noise in the sound booth), or random-incidence micro-
where the output is NOT overloaded. phones (for measures in, e.g., reverberant environments).
More detailed information about microphones and SLMs
can be found in Johnson et al. (1998) and Yeager and Marsh
Frequency Counter (1998). It is important that your SLM and microphone be
This might be a stand-alone device, or it might be an option compatible (i.e., provide the correct polarization voltage),
on your multimeter. In the case of a stand-alone device, a fre- or equipment damage and/or incorrect SPL measures may
quency counter will often have a trigger adjust (the voltage result.
level and direction: positive-going or negative-going) that The SLM also contains amplifiers (whose gain is
determines when an event is triggered. The frequency coun- changed when you change the SPL range), time-weighting
ter combines an event counter with an accurate clock. The circuits (for fast, slow, and possibly impulse and peak time
ratio of events (i.e., cycles) divided by the time elapsed gives weightings), various filter settings (e.g., dBA, dBC, and
you the frequency (in hertz). Thus, if 20 events are measured octave and/or third-octave band filters), as well as a display
in 10 ms (one-hundredth of a second), then the cycles per function (this could be a volume unit (VU) meter, an LED
second (or hertz) = 20 cycles/0.01 s = 2,000 cycles/s (Hz). If indicator, and/or a digital readout). The gain of an amplifier
the counter does not trigger (no events counted), you need in the SLM must be adjusted to account for the sensitivity of
to reduce the trigger level or turn up the signal (e.g., increase each microphone. For example, a 1″ microphone might have
the dB HL on the audiometer dial). If the frequency counter a sensitivity of 50 mV/Pa, whereas a 1/4″ microphone might
reads a number substantially larger than expected, then it have a sensitivity of 1 mV/Pa. If the SLM were adjusted
is possible that the trigger level is set too low (or the signal appropriately for the 1/4″ microphone, then when 1 Pa
presented is set too high) and that multiple triggers per cycle of pressure was presented to the microphone diaphragm,
are occurring. In this case, turning the signal level down or the SLM would read 94 dB SPL [20 log(1 Pa/0.0002 Pa)
increasing the trigger level of the frequency counter should = 94 dB SPL]. If we replaced the 1/4″ microphone with the
 CHAPTER 2 • Calibration 13

1″ microphone but did not change the SLM amplifier gain, allow the “freezing” of the signal on the oscilloscope. To
the 1″ microphone would read 128 dB SPL [94 dB SPL + measure, for example, peak-to-peak voltage, one counts the
20 log(50 mV/1 mV)]. How, then, do we calibrate the SLM number of vertical divisions (usually a division is a centi-
so that it displays the correct SPL? In most instances, we meter) extending from the positive to the negative extremes
would use a device that presents a known SPL to the dia- and multiplies this number of divisions by the voltage per
phragm of the microphone. Two types of calibration devices division to obtain the peak-to-peak voltage. It should be
are commercially available for this purpose: pistonphones noted that measurements made on an analog oscilloscope
and acoustic calibrators. The former produces sound by a are assumed to have an error of 5% or more.
mechanical piston, whereas the latter uses an electrical oscil-
lator and a transducer to produce the tone. Each calibrator
produces a specified SPL at a specified frequency, and this
Spectrum Analyzer
calibrator should be periodically sent back to the manu- Numerous devices can be used to provide a frequency-
facturer to assure it remains within specified tolerances of domain representation of a signal (including the octave
frequency and SPL. These calibrators can accommodate a or third-octave band filters available on many SLMs). In
variety of microphone sizes by inserting nesting adapters. this section, we will limit our discussion to instruments
Using an acoustic calibrator is very simple: turn on the SLM, referred to as digital spectrum analyzers. These instruments
place the calibrator snugly over the microphone, and turn may be stand-alone hardware devices or might be part of
on the calibrator. Making sure that the frequency response a computer-based hardware/software application. These
of the SLM is wideband (flat, or dBC if flat weighting is not devices convert an analog input signal to digital format by
available), adjust the gain of the SLM (by trimming a cali- use of an analog-to-digital converter. It is important that
bration potentiometer using a screwdriver or via software) the reader understand that if the sampling rate used dur-
until the specified output of the calibrator (e.g., 114 dB SPL) ing analog-to-digital conversion is too slow, it can cause the
is displayed on the SLM. generation of “false frequencies” in a process called aliasing.
Once the SLM is calibrated, you must remove the Many spectrum analyzers preclude aliasing by judicious use
acoustic calibrator (or pistonphone) and place an appropri- of a low-pass filter (called an antialiasing filter). It should
ate coupler over the microphone: a 2-cc coupler for insert also be noted that not all possible signal amplitudes can be
earphones (e.g., Etymotic ER3 A earphones) or a 6-cc cou- encoded following analog-to-digital conversion, but sig-
pler for supra-aural earphones (such as TDH-39, TDH-49, nal level is rounded off (“quantized”) and that the magni-
or TDH-50 earphones). ANSI S3.6-2010 has RETSPL values tude of possible quantization error is related to the voltage
for both insert and supra-aural earphone for several 6-cc range and the resolution (related to the number of bits) of
(National Bureau of Standards [NBS] 9-A, IEC 318) and the analog-to-digital converter. The time-domain signal is
2-cc (HA-1, HA-2, occluded ear simulator) couplers. digitized over a limited time period, called the time window
or the time epoch. Once the signal is digitized into a time
epoch, it is converted into the frequency domain by Fourier
Oscilloscope transformation. (See Rosen and Howell, 1991 for a more
The oscilloscope, in its most common display mode, pres- complete explanation of aliasing, antialiasing, quantizing,
ents voltage as a function of time. Oscilloscopes come in and digitization.) The fast Fourier transform (FFT) is one
analog and digital types. In the analog oscilloscope, the out- of many algorithms that have been developed to convert a
put of an electron gun transiently illuminates the screen of time-domain (voltage over time) signal into a frequency-
a cathode ray tube. Freezing the display on the oscilloscope domain (amplitude across frequency) signal. Another term
screen involves repeated triggering of the oscilloscope on a for the frequency-domain representation is the spectrum.
fixed phase of the stimulus. Specialized analog oscilloscopes In addition to the possibility of quantization errors and
that can freeze a display for prolonged periods of time are aliasing, you must be aware that signal processing prior to
called storage oscilloscopes. A digital oscilloscope is simi- Fourier transformation can have an influence on the results.
lar to an analog oscilloscope, except that instead of electron Because of some underlying assumptions about the peri-
guns and a cathode ray tube, the signal is recorded by an odic nature of the discretely sampled signal, the spectrum of
analog-to-digital converter and displayed on a flat panel dis- the signal is distorted unless an integer number of cycles of
play. Digital oscilloscopes often have features that are not all frequencies is contained in the time epoch over which
typically available on analog oscilloscopes (e.g., storage of the signal is digitized. To prevent the distortion (often called
waveforms, cursor functions, and summary statistics such leakage) that occurs when a noninteger number of cycles
as peak-to-peak and RMS voltage calculations). Simple is contained in the time epoch, the digitized time epoch
amplitude and voltage measurements are easily performed can be shaped. This shaping multiplies the signal by val-
on a signal using an oscilloscope. Manipulations of the time ues at or near zero, near the beginning and end of the time
base (in time per division) and amplitude (in volts per divi- window and weights them at or near 1, near the middle of
sion), as well as the appropriate adjustment of the trigger, the time window. One popular windowing function is the
14 SECTION I • Basic Tests and Procedures

Hanning window. A given windowing function trades ampli- Biologic Check

tude uncertainty for frequency resolution. Once the data are
converted to the frequency domain, the amplitude of a given After the audiometer has been installed, plugged in, turned
Fourier coefficient (e.g., frequency) can be determined on, and allowed to warm up, the operator should listen to
using a cursoring function. It should be noted that Fourier the signal at different dial settings through each transducer
transformation produces multiple discrete harmonically (earphone, loudspeaker, and bone vibrator). With a little
related (i.e., integer multiples) spectral components. The practice, one can hear basic faults in the equipment. A vague
lowest frequency (fundamental frequency) and, hence, complaint to the audiometer technician or distributor that
the frequency interval between components are related to it “sounds funny” is as futile as telling an auto-repair person
the recorded time-domain signal. If the time-domain signal the same thing. However, a specific description of the sound
is, for example, 200 ms (0.2 s), then the lowest frequency is and when it occurs can help determine the source of the
1/0.2 s, or 5 Hz. The longer the time window is, the better trouble. If the technicians are given a detailed description
the spectral resolution. of the problem, then the fault may be found more quickly,
without wasting their time and your money.
Much information on the source of the problem may
BASIC EQUIPMENT also be obtained by inspecting the audiometer. Following
The basic calibration equipment for checking output levels are some areas of potential malfunction that the audiologist
of an audiometer should include (1) a voltmeter or mul- should check periodically (normally on a daily basis):
timeter; (2) condenser microphones (both pressure and
free-field types); (3) acoustic calibrator; (4) a 6-cc coupler 1. Check the power, attenuator, earphone, and vibrator
(NBS 9-A or IEC 318); (5) a 2-cc coupler (ANSI HA-1 or cords for signs of wear or cracking. Listen to the tone
HA-2 or IEC occluded ear simulator); (6) a 500-g weight; through the transducer at a comfortable level while
(7) a mechanical coupler for bone vibrator measurements twisting and jiggling the cords. A defective cord will usu-
(artificial mastoid); and (8) an SLM (or equivalent). When ally produce static or will cause the tone to be intermit-
purchasing any of the above components, it is wise to check tent. Tightening the earphone screws and/or resoldering
with others who use similar types of equipment to find the the phone plug connections might fix the problem. If this
best specific brands available locally. does not alleviate the problem, it is wise to replace the
Other equipment such as a digital oscilloscope, fre- cord.
quency counter, and/or a spectrum analyzer will also prove 2. If the audiometer has dials, check for loose dials or for
to be invaluable in checking the acoustic parameters of dials that are out of alignment. If such faults exist, the
audiometers. In many instances, this equipment can be dial readings will be inaccurate. Defective dials should
shared by more than one facility. If one has only one or a few be repaired immediately (sometimes this just requires
audiometers, a service contract is most sensible. If one has tightening the set screws that hold the dial to the audi-
a substantial number of pieces of audiometric test equip- ometer), and the audiometer should be recalibrated to
ment, an SLM (with appropriate couplers, microphone(s), determine outputs at the “new” dial settings. Check to
and acoustic calibrator) and a multimeter should be pur- see that incremental changes are correctly reflected in
chased and used. If the accuracy of the audiometer is ques- the readout.
tioned, it necessitates shutting down the equipment or 3. The audiologist should listen for audible mechani-
retesting patients at a later date. This translates into time cal transients through the earphone when the dials or
and financial loss, not to mention more serious conse- switches are manipulated. The ANSI S3.6-2010 standard
quences in surgical or medicolegal cases. In a busy practice, (section 5.4.4) suggests that two normal-hearing listen-
such a loss would surely be equivalent to the cost of one or ers should listen at a distance of 1 m from the audiometer
more pieces of electronic test equipment that would prevent with the earphones in place but disconnected and with a
this problem. This of course assumes that someone working proper load resistance (coinciding with the impedance
in that practice setting is competent to check the calibration of the earphone at 1,000 Hz) across the circuit while
of the audiometric equipment. manipulating the presenter/interrupter switch, and so
on, to make sure that there are no audible signals that
would inform the subject to the presence of the test sig-
CHECKING THE CALIBRATION OF nal. A mechanical transient can often be detected more
PURETONE AUDIOMETERS easily by listening than through the use of electronic
equipment.
Basic Signal 4. To determine if electronic transients are audible, it is wise
As soon as one obtains a new audiometer, the manual to listen to the output both at a moderate hearing level
should be read and, if any calibration instructions are pro- (e.g., 60 dB) and below the threshold of hearing. Elec-
vided, they should be followed. tronic transients will show up on an oscilloscope as an
 CHAPTER 2 • Calibration 15

irregularity when the problem switch or dial is manip- caused by faulty external wiring between the examiner’s
ulated. The danger of an audible transient, whether booth and that of the test subject or within the audiom-
mechanical or electronic, is that the patient may respond eter itself. Cross-talk must be corrected before any testing
to the transient rather than the stimulus tone. Sometimes is carried out.
an antistatic or contact-cleaner spray can alleviate the 7. The clinician should listen to the signal while the attenu-
problem of electronic transients. ation dial is changed from maximum to minimum levels.
5. The audiologist should listen for hum or static with the For instance, a tone may be present at 20 dB HL on the
hearing level dial at a high value, both when a stimulus dial, whereas no tone is present at 15 dB HL on the dial.
signal is present and when it is absent. One should not In some cases, the tone stays at the same hearing level
hear static or hum at levels below 60 dB HL on the dial. from 20 dB HL to −10 dB HL on the dial. These problems
6. “Cross-talk” may occur between earphones, that is, the are easily detected by listening to the audiometer.
signal that is sent to one earphone may be heard in the 8. Finally, the threshold of the clinician (or a person with
contralateral earphone. Such a problem could greatly known hearing thresholds) should be checked with the
affect the audiometric thresholds obtained on that audi- earphones and bone vibrators to make sure that the
ometer, especially for cases with unilateral hearing loss. outputs are approximately correct. If the levels are not
Cross-talk may be detected by unplugging one earphone, within 10 dB of the previous threshold values, the output
sending a signal to that phone, and listening to the other levels should be checked electronically.
earphone. As before, when removing the earphone, a
proper resistive load must be put in its place. The signal Aside from these gross problems, which can be detected
at a suprathreshold dial setting (e.g., 70 dB HL) should by looking or listening (see Figure 2.1 for an example of a
not be heard in the opposite earphone when a signal form that may be used to aid the clinician in carrying out
is presented in the normal manner. Cross-talk may be the listening check), the precise accuracy of the output levels

Audiometer serial #

Date:

Time:

Checked by:

Earphone cords
Power cord
Attenuator cord
Hum
Dials
Frequency
Attenuation
Intensity right phone
Intensity left phone
Tone interrupter
Tone pulse rate
Cross-talk
FIGURE 2.1 Form for biologic
Acoustic radiation check of audiometer. (Reprinted
Bone vibrator(s) from Wilber L. (1972) Calibration:
pure tone, speech and noise
Loudspeakers
signals. In: Katz J, ed. Handbook
Other comments of Clinical Audiology. 1st ed.
Baltimore, MD: The Williams &
Wilkins Company; pp 11–35, with
the permission of Lippincott
Williams & Wilkins.)
16 SECTION I • Basic Tests and Procedures

must be evaluated when the audiometer is first purchased or storage oscilloscope. When gating the signal on, rise time
and at regular intervals thereafter. Frequency, output level, is the length of time it takes for the signal to increase from
linearity of attenuation, and percentage of harmonic dis- −20 to −1 dB (10% to 90%) of its final steady-state value.
tortion should all be checked electronically, in addition to The fall time is the length of time between −1 and −20 dB
the biologic check. Section 5.4 of ANSI S3.6-2010 describes (90% to 10%) relative to its steady-state value. This is usu-
various checks for unwanted sound from the transducer or ally checked at a hearing level of 60 dB HL or less. ANSI
audiometer. S3.6-2010 specifies a rise time as well as a fall time of not less
than 20 ms and not more than 200 ms. A detailed descrip-
tion of the rise and fall characteristics is given in section
Frequency Check 7.5.3 of ANSI S3.6-2010.
The frequency output from the audiometer should be
checked by using an electronic frequency counter. This
instrument will tell the exact frequency of the output signal. Linearity Check
Quite accurate frequency counters are often included in a Attenuator linearity (the hearing level dial) may be checked
digital multimeter. The electrical output from the audiome- electrically, directly from the audiometer, or acoustically
ter may be routed directly to the instrument (i.e., unplug the through its transducer (earphone or bone vibrator). If mea-
earphone, then plug in the frequency counter input to the surements are to be made electrically, the earphone should
audiometer output) because the frequency is determined by remain in the circuit and the voltage should be measured
an oscillator in the audiometer rather than the transducer. in parallel to the earphone, or a dummy load that approxi-
By using an electronic frequency counter, one can easily mates the earphone impedance should replace the trans-
determine if the output from the audiometer corresponds ducer. To check linearity, the audiometer should be turned
to the nominal frequency. The standard for audiometers to its maximum output and then attenuated in 5-dB steps
allows a tolerance of ±1% of the indicated frequency value until the output can no longer be read. Each attenuator on
for Type 1 and 2 audiometers; ±2% for Type 3 and 4 audi- the audiometer should be checked separately. To meet the
ometers; and ±3% for Type 5 audiometers. For example, if ANSI S3.6-2010 standard, the attenuator should be lin-
the audiometer dial reads 1,000 Hz, then the actual output ear within 0.3 of the interval step or by 1 dB, whichever is
must be between 990 and 1,010 Hz for a standard diagnostic smaller. That is, if you change the level in 5-dB steps, the
(Type 1) audiometer. audiometer must attenuate between 4 and 6 dB per step.
Frequency should be checked on initial receipt of the If the attenuation step is 2 dB, then the reading should be
audiometer and at yearly intervals thereafter. Neverthe- between 1.4 and 2.6 dB per step (0.3 × 2 dB = 0.6 dB, which
less, it is appropriate to listen to the audiometer each day to is less than 1 dB). As noted in section 7.2 (and section 7.3.3)
judge whether the frequencies are maintaining reasonably of ANSI S3.6-2010, the SPL or FL of earphones, speakers, or
good accuracy. bone vibrators can vary by no more than ±3 dB from 125 to
5,000 Hz and no more than ±5 dB at 6,000 Hz and above, at
Harmonic Distortion Check any dB HL dial setting.
Attenuator linearity should be checked annually. If a
Linearity measurements may also help detect distortion “fixed loss pad” (i.e., a device that automatically changes
in a transducer or in the audiometer itself. Distortion may the signal level by a set amount, e.g., 20 dB) is present
appear as a lack of linear attenuation, especially at high in the audiometer, its attenuation must also be checked. If
output levels (90 dB HL and above). Harmonic distortion the audiometer attenuates in 1- or 2-dB steps, then these
must be checked through the transducer itself. Excessive smaller attenuation steps should be checked if they are used
harmonic distortion is rarely caused by the audiometer but clinically.
often arises in the various transducers. The maximum per-
missible total harmonic distortion in the current standard
(ANSI S3.6-2010) is 2.5% for earphones and 5.5% for bone EARPHONE LEVEL CALIBRATION
vibrators. The standard also shows the maximum permis-
sible distortion for the second, third, fourth, and higher
Real Ear Methods
harmonics, as well as the subharmonics, across audiometric There are two basic approaches for the calibration of ear-
frequency. phones. One is the “real ear” method and the other is the
“artificial ear” or coupler method. With the original real ear
method, one simply tested the hearing of a group of normal-
Rise–Fall Time hearing persons, averaged the results, and checked to see
The rise–fall time of the tone is a basic parameter of the that the average hearing of this group was at zero on the dial
audiometer, which may be checked by taking the output for each frequency. Although this is theoretically feasible
directly from the audiometer and routing it into a digital with a large population sample, it is not a recommended
 CHAPTER 2 • Calibration 17

procedure. ANSI S3.6-2010, Appendix D, describes probe is described in ANSI S3.25-2009, but RETSPLs are not given
tube, loudness balance, and threshold procedures that may for supra-aural or insert receivers using the Zwislocki cou-
be used for this purpose. Clearly, these procedures are pos- pler or the manikin.
sible but quite unwieldy. For audiometers, this approach is When checking the audiometer earphone output, the
technically incorrect because the ISO 389-1:1998 reference supra-aural earphone is placed on the coupler and a 500-g
(which is also used in ANSI S3.6-2010) is not tied to normal weight is placed on top of it. If using an SLM (rather than
hearing per se, but simply refers to an arbitrarily accepted a microphone preamplifier), the output is read in dB SPL,
SPL (i.e., the RETSPL or FL). If the audiologist wishes to use where SPL = 20 log10 P/Pref (where P is the observed sound
a new earphone (that is not listed in ANSI S3.6-2010 Stan- pressure and Pref = 20 µPa). After the earphone is placed on
dard, its appendix, or any subsequent revision), a real ear the coupler, a low-frequency tone (125 or 250 Hz) is intro-
procedure might be the only way to check calibration, but if duced and the earphone is reseated on the coupler until
generally accepted earphones are used, it is much easier and the highest SPL value is read. This helps assure optimal
more efficient to use an artificial ear/coupler method. earphone placement on the coupler. The output from the
earphone is then compared to the expected values at each
frequency. The standard SPL values that are used are given
Artificial Ear (Coupler) Methods in (1) ISO 389-1:1998, often referred to as ISO-1964 because
The most commonly used procedure today is that of the of its initial publication date, and (2) ANSI S3.6-2010. These
“artificial ear,” which consists of a condenser microphone values evolved through a “round robin” in which several
and a 6-cc coupler (for supra-aural earphones) or 2-cc cou- earphones were measured on various couplers at a group of
pler (for insert earphones). The 6-cc coupler was originally laboratories throughout the world (Weissler, 1968).
chosen because it was thought that the enclosed volume was The current ANSI standard includes RETSPLs for the
approximately the same as the volume under a supra-aural TDH-type earphones, as well as insert earphones. It also
earphone for a human ear (Corliss and Burkhard, l953). provides values for both the IEC and NBS couplers for
However, since volume displacement is only one com- supra-aural earphones and values for insert phones using an
ponent of acoustic impedance, it cannot be assumed that occluded ear simulator, HA-1 or HA-2 coupler. Figure 2.2
the coupler actually represents a human ear. Burkhard and shows an audiometer earphone calibration worksheet,
Corliss (1954) pointed out that the impedance characteris- which contains the expected values at each frequency with
tics of a 6-cc coupler probably simulates the impedance of TDH-39 or TDH-49 (or TDH-50) earphones in Telephonics
the human ear over only a small part of the frequency range. type 51 cushions on an NBS 9-A coupler and insert receivers
Because the 6-cc coupler does not replicate the impedance using an HA-1–type coupler. ANSI S3.6-2010 allows a toler-
of the human ear, it cannot be considered a true artificial ance from the listed values of ±3 dB from 125 to 5,000 Hz
ear. Subsequent work by Cox (1986), Hawkins et al. (1990), and ±5 dB at 6,000 Hz and higher.
Killion (1978), and Zwislocki (1970, 1971) has quantified The supra-aural output measurements referred to above
the differences between real ear and coupler values. In an are only valid when a supra-aural–type earphone cushion
attempt to solve this problem, the IEC 318 coupler was (which touches the pinna) such as the Telephonics 51 is
developed. However, there is still some disagreement as to used and not when a circumaural cushion (which encircles
the accuracy of this ear simulator (formerly called an arti- the pinna) is used. ANSI S3.6-2010 provides RETSPL val-
ficial ear) because its impedance characteristics are also not ues for several circumaural earphones (Sennheiser HDA200
exactly those of a real human ear. However, it is clearly more and Koss HV/1 A) with an IEC 60318-2 coupler and a type
accurate than the present NBS 9-A coupler. 1 adapter (Sennheiser earphone) or type 2 adapter (Koss
In addition to the problem of acoustic impedance char- earphone). When the output of the audiometer through the
acteristics, the NBS 9-A coupler is known to have a natu- earphone has been established, it is compared to the appro-
ral resonance at 6,000 Hz (Rudmose, 1964). This interferes priate standard to determine whether it is in calibration or
with the measurement of the output of an audiometer ear- not. If possible, the audiometer trim pots (or by software
phone around that frequency. Other coupler problems are adjustments in newer digital audiometers) should be used to
its size, its shape, and the hard walls that permit the possibil- bring the audiometer into calibration. However, when this is
ity of standing waves at frequencies above 6,000 Hz. Despite not possible or when different earphones will be used with
these difficulties, the NBS 9-A coupler remains the accepted the same audiometer, and when corrections are less than
device (by ANSI S3.6-2010) for measuring the acoustic out- 15 dB, a calibration correction card may be placed on the
put from the audiometer through a supra-aural earphone. A audiometer showing the discrepancy from the established
coupler developed by Zwislocki (1970, 1971, 1980) appears norm. It should be noted that if the output of the audiome-
to very closely approximate the acoustic impedance of the ter is, for example, 10 dB too low, then the dB HL correction
human ear. It is used in KEMAR (a manikin that has a pinna sheet must be decreased by 10 dB. Such corrections must then
and an ear canal, as well as a coupler and microphone) be taken into consideration when an audiogram is plotted.
(Burkhard, 1978; Burkhard and Sachs, 1975). This manikin If an audiometer is off by more than 15 dB at any frequency
18 SECTION I • Basic Tests and Procedures

AUDIOMETER EARPHONE CALIBRATION SHEET

Audiometer: S# Earphone: Channel: Room:

Calibrated by: Date: Equipment:

FREQUENCY: 125 250 500 750 1000 1500 2000 3000 4000 6000 8000

1. SPL*

2. Audiometer dial
setting
3. Nominal ref. SPL
(Line 1 – Line 2)
4. Equipment and mike
correction
5. Corrected ref. SPL
(Line 3 – Line 4)
6a. TDH – 49/50 47.5 26.5 13.5 8.5 7.5 7.5 11.0 9.5 10.5 13.5 13.0
earphones**
TDH - 39 45.0 25.5 11.5 8.0 7.0 6.5 9.0 10.0 9.5 15.5 13.0
6b. ER 3-A 26.5 14.5 6.0 2.0 0.0 0.0 2.5 2.5 0.0 -2.5 -3.5
earphones***
7. Calibration error
(Line 5 – Line 6)
8. Corrections @
* SPL = sound pressure level in dB re: 20 µPA
** TDH-49/50 values from ANSI S3.6-1996, p. 18 (see standard for coupler and cushions)
*** ER3-A values from ANSI S3.6-1996, p. 20 using HA-1–type coupler (see standard for different coupler values)
@ Correction – Rounded to the neatest 5 dB; – = audiometer weak, make threshold better
+ = audiometer weak, make threshold better
FIGURE 2.2 Earphone calibration worksheet. (Reprinted from Wilber L. (1972) Calibration: pure tone,
speech and noise signals. In: Katz J, ed. Handbook of Clinical Audiology. 1st ed. Baltimore, MD: The
Williams & Wilkins Company; pp 11–35, with the permission of Lippincott Williams & Wilkins.)

or by 10 dB at three or more frequencies, it is advisable to tem is in proper calibration, bone-conduction corrections

have the audiometer put into calibration by the audiometer for the audiometer can be determined by using the differ-
manufacturer or their representative. If the audiometer is ence obtained between air- and bone-conduction thresh-
new, it should meet ANSI S3.6-2010 tolerances. With cur- olds. This procedure makes a few assumptions that are not
rent digital audiometers, deviations in desired output are always met. For example, it presupposes that true thresholds
usually due to the transducer rather than the audiometer, so can be obtained for all the normal-hearing subjects using
sometimes it is easier to bring the audiometer into calibra- the given audiometer. Because (1) many audiometers do not
tion by replacing the offending transducer(s). go below 0 dB HL and (2) the ambient noise in test booths
often does not allow assessment below 0 dB HL, it is not
always possible to determine the true threshold. To avoid
BONE VIBRATOR CALIBRATION these problems, Roach and Carhart (1956) suggested using
individuals with pure sensory/neural losses for subjects in
Real Ear Procedures the real ear procedure. Such an approach eliminates the
Checking the calibration of a bone vibrator presents a differ- problems of ambient noise and lack of audiometric sensi-
ent problem than that of an earphone. Whereas earphones tivity, thus increasing the probability that one will obtain
can be checked easily using a microphone as a pickup, bone “true” thresholds. However, it can be problematic to find
vibrators cannot. The original technique for checking bone a group of subjects with “pure sensory/neural” losses (i.e.,
vibrator calibration was a real ear procedure (American those who have no conductive component) and who have
Medical Association, 1951), which was somewhat differ- thresholds that do not extend beyond the bone-conduction
ent than that used for earphones. The method assumes that limits of the audiometer. A more basic problem with real
air- and bone-conduction thresholds are equivalent. If 6 to ear bone vibrator calibration is the supposition that air- and
10 normal-hearing subjects are tested for both air and bone bone-conduction thresholds are equivalent in the absence
conduction with an audiometer whose air-conduction sys- of conductive pathology. Although this is certainly true, on
 CHAPTER 2 • Calibration 19

average, for a large group of people, it cannot be expected to compact disc (CD) input of the speech circuit of the audi-
be true for any individual or for small groups (Studebaker, ometer. The input level should be adjusted so that the moni-
1967; Wilber and Goodhill, 1967). toring VU meter on the face of the audiometer reflects the
appropriate level, usually 0 dB. The output from the trans-
ducer is then measured. For most speech stimuli used for
Artificial Mastoid Procedure audiologic purposes, there is a 1,000 Hz tone on the tape or
The preferred procedure for calibrating bone vibrators CD (or in other digital forms) that has an RMS voltage that
involves the use of a mechanical coupler, often referred to is similar to the RMS voltage of the speech stimuli. Details
as an artificial mastoid. Artificial mastoids were proposed concerning the calibration of the speech circuit of an audi-
as early as 1939 by Hawley (1939). However, it was not ometer are given in section 6.2 of ANSI S3.6-2010.
until Weiss (1960) developed his artificial mastoid that they ANSI S3.6-2010 states that the speech output for the
became commercially available. Just as replication of the 1,000-Hz tone at 0 dB HL should be 12.5 dB above the
acoustic impedance of the human ear is difficult with a cou- RETSPL for the earphone at 1,000 Hz. Bone vibrators
pler, replication of the mechanical impedance of the head should be calibrated separately. All subsequent speech test-
is difficult with an artificial mastoid. Because no commer- ing must be carried out with the monitoring meter peaking
cially available artificial mastoid met the mechanical imped- at the same point as during the calibration check. If, for
ance requirements of the ANSI (S3.13-1972) or IEC (IEC example, one prefers −3 dB on the meter rather than 0 dB,
60373:1971) standards, both the ANSI and IEC standards then calibration of the 1,000-Hz tone must be peaked at
were revised to conform more closely to an artificial mas- −3 dB, or an appropriate correction must be made in
toid that is available (ANSI S3.13-1987; IEC 60318-6:2007). reporting measurements.
ANSI S3.6-2010 gives threshold values in reference equiva- The required flatness of the frequency response of the
lent threshold force levels (RETFLs) that are appropriate for speech audiometer circuit is defined as ±3 dB for the fre-
a bone vibrator such as the B-71 or B-72, or one meeting quencies of 250 to 4,000 Hz and from 0 to −10 dB between
the physical requirements described in section 9.4.3 of ANSI 125 and 250 Hz and ±5 dB between 4,000 and 6,000 Hz.
S3.6-2010. The ISO standard (ISO 389-3:1994) gives one set ANSI S3.6-2010 gives specific requirements for checking
of values that are to be used for all bone vibrators having the microphone circuit as well as the other speech input cir-
the circular tip described in the ANSI and IEC documents. cuits. If the puretone and speech audiometers are separate
These values are also used in the ANSI standard. It is impor- machines, then the speech audiometer must also be checked
tant to recognize that both the ANSI and the ISO values for cross-talk, internal noise, and attenuator linearity as
are based on unoccluded ears using contralateral masking. described earlier. More specific information on calibra-
Thus, the values presuppose that masking will be used in the tion of the speech circuit may be found in section 6.2.10 of
contralateral ear when obtaining threshold. One can use the ANSI S3.6-2010.
same type of worksheet for bone as for air—substituting
the appropriate RETFL values. In both earphone and bone
vibrator calibration, it is important to check distortion as
MONITORING METER
well as overall level through the transducer. Distortion may Monitoring (or VU) meters are indicators of signal level and
be measured directly with software integrated into the SLM are found on the face of most audiometers. The monitoring
or by routing the output of the artificial mastoid and SLM meter is calibrated relative to the input signal that it moni-
to a spectrum analyzer. As mentioned earlier, allowable dis- tors and should not be interpreted as yielding any absolute
tortion values for bone vibrators are more lenient than for values such as 0 dB SPL. On a speech audiometer, the meter
earphones. This is because bone vibrators have more distor- is used to monitor the speech signal or to aid the audiolo-
tion than earphones. In addition to the earlier mentioned gist in adjusting the input calibration tone that precedes the
physical measurement procedures, the importance of just recorded speech materials. The specifications for the meters
listening to the audiometer output through the bone vibra- may be found in section 6.2.10 of ANSI S3.6-2010. In gen-
tor cannot be overstated. The normal ear (with audiologist eral, it is important that the meter be stable, that there is
attached) should be able to perceive gross attenuation and minimal undershoot or overshoot of the needle indicator
distortion problems. The electroacoustic procedures, how- relative to the actual signal, and that any amplitude change
ever, serve to quantify the problems that the human ear can is accurately represented on the meter. The audiologist may
only identify subjectively. check the meter and its entire accompanying input system
as described below.
A puretone should be fed from an oscillator through an
SPEECH AUDIOMETERS electronic switch to the input of the audiometer. The tone
Because running speech fluctuates in SPL (as well as fre- should be monitored by a voltmeter or an oscilloscope. By
quency content) over time, the preferred method is to intro- activating the electronic switch to produce a rapidly inter-
duce a puretone (1,000 Hz) into the microphone, tape, or rupted signal, one can watch the meter to ascertain whether
20 SECTION I • Basic Tests and Procedures

there is any overshoot or undershoot relative to the signal

in its steady state. One must also check the response time
CALIBRATION OF ANCILLARY
of the needle on the VU meter. A computer-generated or EQUIPMENT
tape-recorded tone may be used to ensure that the needle
reaches its 99% state deflection in 350 ± 10 ms. In addition,
Masking Generator
the overshoot should be no more than 1.5%. One can insert ANSI S3.6-2010 defines white noise, weighted random noise
a linear attenuator in the line between the oscillator and for masking of speech, and narrowband noise. Instead of
the audiometer input, one may reduce the output from the HL, masking noise is discussed in terms of effective mask-
oscillator and the audiometer input, or one may reduce the ing (dB EM), meaning that, for example, a 20-dB EM noise
output from the oscillator by a known amount (as moni- is that noise level that perceptually masks a 20-dB HL sig-
tored by a voltmeter or oscilloscope). The change in input nal. The bandwidths for narrow bands of noise are specified
should be accurately reflected by a corresponding change on by frequency with RETSPL corrections for third-octave and
the monitoring meter. half-octave measurements. Cutoff values are given in the
standard (see table 4 of the standard). When checking the
bandwidth of the narrowband noise, it is necessary to have a
SOUND FIELD TESTING frequency analyzer or spectrum analyzer (or a computer
ANSI S3.6-2010 describes the primary characteristics of program that allows one to produce a Fourier analysis of
sound field testing in section 9.5. This includes the test the noise) to determine if the noise bandwidths from the
room, frequency response, method for describing the audiometer conform to specifications. The same transducer
level of the speech signal, and the location of the speakers. that will be used when delivering the masking sound should
Table 9 of the standard also gives specific RETSPL values be used to make final calibration measurements. However,
for band-limited stimuli (frequency-modulated tones or because the characteristics of various transducers are quite
narrow bands of noise) for binaural and monaural lis- different from one another, it is sensible to first do an elec-
tening. An ASHA working group prepared a tutorial for tronic check directly from the audiometer to verify that
sound field testing that discusses some of the problems any variation from the bandwidth is due to the transducer
of setting up the test procedure (ASHA, 1991). Charac- rather than the electrical output of the audiometer.
teristics of the frequency-modulated signals are given in The masking sound should be checked periodically
section 6.1.3 of ANSI S3.6-2010. In addition, the char- through the transducers used to present it. The examiner
acteristics of narrowband noise levels are presented in should be careful to use a signal that is high enough in level
table 4 of the standard. The level for speech in sound field to avoid interference by ambient room noise (generally
should be comparable to the corrected free-field response about 80 dB HL). In the case of narrowband noise, the SPL
for earphones. values measured should be within ±3 dB of the RETSPLs for
When calibrating stimuli is present in the sound field, the geometric center frequency and corrected appropriately
it is important to place some sort of marker (such as a ring for masker bandwidth. If broadband white noise (noise that
suspended from the ceiling) at the place where the subject’s has equal level across frequency) is the only masking signal
head will be. A free-field microphone should be placed on the audiometer, one need only check the output through
so that the diaphragm is facing toward the direction of the earphone with a linear setting (no filter) on the SLM.
the plane-propagated wave (called frontal incidence). If a The overall output and attenuation characteristics should
pressure microphone is used, the microphone diaphragm be checked in the same basic manner as described for pur-
should be placed facing at a right angle to the direction of etones using an appropriate coupler.
the plane-propagated wave (called grazing incidence). In When making noise measurements, the characteristics
either case, the microphone should be placed at the place of the measuring equipment are critical. Since noise is not a
where the subject’s head will be during testing. There “clean” (i.e., uniform and unvarying) signal, it is highly sus-
should be nothing between the speaker and the calibration ceptible to errors of overshoot and undershoot on a meter
equipment. and to damping on a graphic level recorder. A spectrum
The amplifier hum or internal noise of the loudspeaker analyzer that is capable of frequency-domain averaging and
system should be checked. This may be done by adjusting with storage capabilities is optimal for checking calibra-
the attenuator dial to some high setting (between 80 and tion of noise. Unfortunately, most clinics do not have such
90 dB HL) and then measuring the output from the loud- sophisticated equipment.
speaker when no signal is present. That is, everything is in
normal position for testing except that there is no signal COMPACT DISC AND TAPE
presented to the speaker. The equipment noise (in SPL)
should be at least 50 dB below the dial setting (in HL; i.e., if
PLAYERS
the dial reads 80 dB HL, then the equipment noise should CD or tape players that are used in a clinic for reproducing
be <30 dB SPL). speech signals, filtered environmental sounds, or other test
 CHAPTER 2 • Calibration 21

stimuli should be checked electroacoustically at least once available, one can record across the terminals of the tim-
every 12 months. However, if the CD or tape player is in reg- ing mechanism inside the audiometer. It is difficult to check
ular use, weekly maintenance should be carried out (such as the pulse speed on a graphic level recorder because of pen
cleaning and demagnetizing the heads for tape players). The damping, but it is possible to check it on a digital or storage
instruction manuals normally outline the procedures to be oscilloscope. It is not difficult to estimate whether there is
used with the particular tape or CD player. If not, any good roughly a 50% duty cycle (on half the time and off half the
audio equipment dealer can explain the procedure. In addi- time), but it is quite difficult to judge whether the signal is on
tion, the frequency response and time characteristics of the for 200 ms versus 210 ms. The characteristics of the pulsed
tape player should be checked. tone are described in section 7.5.4 of ANSI S3.6-2010.
At present, there are no standards for tape players used If both pulsed and continuous signals are used, it is
with audiometers. However, the frequency response and important to check the relative level of the pulsed and
time characteristics of the tape player may be checked by continuous signals. If they are not equal, this should be
using a standard commercial tape recording of puretones of corrected. The relative levels can be compared by observ-
various frequencies. If you do not have access to such a tape, ing the envelope of the waveform on an oscilloscope or by
it is possible to make one by introducing puretones from an recording the output with a graphic level recorder if there is
audio oscillator into the machine, recording them, and play- no damping problem. The attenuation rate and pulse rate
ing them back. This enables the operator to check both the should be checked annually unless there is a reason to sus-
record and playback sections of the tape recorder. Unfortu- pect a problem earlier.
nately, if both the record and playback are equally reduced Computerized audiometers are becoming commer-
(or increased) in frequency, the output will appear at the cially available. It should be noted that computerized audi-
nominal frequency. The output from the oscillator should ometers, just like manual audiometers, must meet all of the
be monitored with a voltmeter to make certain that a con- technical specifications included in ANSI S3.6-2010.
stant voltage signal is used. Distortion of the puretone from
the tape player should also be checked. If none of this is pos- AUDITORY-EVOKED POTENTIAL
sible, the speed of the tape player can be checked grossly by
marking a tape and then, after timing a segment as it goes
INSTRUMENTS
across the tape head, measuring to see how many inches There is an IEC standard for auditory test signals of short
passed over the heads per second. Also, if the machine is duration for audiometric and neuro-otologic purposes (IEC
badly out of calibration, it will be audible as a pitch change 60645-3:2007). There is also an IEC standard for auditory
in the recorded speech (higher if too fast, lower if too slow). brainstem response instruments (IEC 60645-7:2009 Electroa-
coustics—Audiometric Equipment—Part 7: Instruments
AUTOMATIC (AND COMPUTERIZED) for the measurement of auditory brainstem responses). ISO
AUDIOMETERS 389-6:2007 reports RETSPLs for clicks and standard (2–1–2
cycle) tonebursts. There is currently no ANSI standard that
A calibration check of automatic (or Bekesy) audiometers provides RETSPL for clicks and tonebursts.
begins with frequency, level, cross-talk, and other aspects The basic parameters of the acoustic signals used for
described for manual puretone audiometers. In addition, auditory-evoked potentials (AEPs) are the same as for con-
the attenuation rate and interruption rate for pulsed signals ventional audiometry. One must check output level, fre-
should be checked. ANSI S3.6-2010 requires that a rate of quency, and time. When calibrating acoustic transients from
change of 2.5 dB/s be provided for Type 1, 2, and 3 audiom- an AEP instrument, the instrumentation used to calibrate
eters. Permissible rates for all types of audiometers are given an audiometer may be inappropriate. It is especially impor-
in the ANSI S3.6-2010 standard. As in manual audiometers, tant to check the output from the AEP unit acoustically as
the permissible variance in level per step is 1 dB or 0.3 of the well as electrically. It is easy to display the electrical output
indicated step size, whichever is smaller. The attenuation rate from the AEP unit on an oscilloscope, but to analyze that
may be measured quite easily with a stopwatch. After starting display, one needs to repeat it very rapidly or, preferably,
the motor, a pen marking on the chart is started at the same use a digital or storage oscilloscope. Determination of the
instant as a stopwatch is started. One reads the chart to deter- acoustic level of these acoustic transients requires an SLM
mine how far the signal was attenuated (or increased) during that can record true peak SPL (pSPL) or that allows rout-
the measured time interval. By dividing the duration (in sec- ing the output to an oscilloscope to determine pSPL or peak
onds) into the decibel change in level, one can find the deci- equivalent SPL (peSPL).
bel per second attenuation rate. The audiometer should be
checked for signals both increasing and decreasing in level.
To check the pulsed stimulus duration, one may go
Calibration of Acoustic Transients
from the “scope sync” output on the back of the audiom- Acoustic transients must be calibrated utilizing specialized
eter (if such exists) to an electronic counter, or if that is not instrumentation and procedures. For clinical procedures,
22 SECTION I • Basic Tests and Procedures

it is prudent to use earphones that can be coupled to either

a 6- or a 2-cc coupler. TDH-39s, TDH-49s, and TDH-50s
Vp-p
housed in MX 41/AR or Telephonics type 51 cushions can Vp
be coupled to a 6-cc (NBS 9-A) coupler. Etymotic insert
earphones (ER-1, ER-2, ER-3 A) can be coupled to a 2-cc
coupler. At the base of a 2- or a 6-cc coupler is (typically)
a space designed to house a 1″ condenser pressure micro-
phone, but for some couplers, they are designed to house a
½″ microphone. The microphone output can be routed to
either an SLM or to a conditioning amplifier that provides
Vp
the polarization voltage for the condenser microphone
and (in some instruments) provides voltage amplification
(gain). The measurement of the SPL of an acoustic tran-
sient with an SLM is complicated by the time constants
used for the SPL measurement. Most SLMs have at least two
Vp-p
exponential-time–weighted averaging modes: fast and slow.
Fast exponential-time–weighted averaging has a measure-
ment time constant of 125 ms, whereas slow exponential-
time–weighted averaging has a measurement time constant FIGURE 2.3 The procedure for obtaining peSPL (both
of 1,000 ms (Yeager and Marsh, 1998). In either case, clicks baseline-to-peak and peak-to-peak measures) is shown.
or tonebursts have durations that are much shorter than the (Reprinted from Burkard R, Secor C. (2002) Overview of
time constant of even the fast exponential-time weighting, auditory evoked potentials. In: Katz J, ed. Handbook of
and you will underestimate the true SPL of the toneburst Clinical Audiology. 5th ed. Baltimore, MD: Lippincott
using fast (or worse yet slow) exponential-time–weighted Williams & Wilkins; pp 233–248, with the permission of
averaging. There are several solutions to this measurement Lippincott Williams & Wilkins.)
problem. First, if you can turn the toneburst on for several
seconds, you can record the level of the tone in the fast or
slow exponential-time–weighted averaging mode. If you presented and the baseline-to-peak voltage is measured on
measure over three time constants (375 ms in fast, 3,000 ms the oscilloscope. Making sure not to change the SPL range
in slow), the recorded value will closely approximate the on the SLM, a tonal stimulus is routed through the earphone,
true exponential-time–weighted SPL of the stimulus. and the level of the tone is adjusted until the baseline-to-
This is one method to obtain what is commonly referred peak voltage on the oscilloscope is identical to that measured
to as the peSPL of the toneburst. This approach will not for the click. The reading on the SLM is the baseline-to-peak
work for a click stimulus, as increasing the duration of the peSPL of the click.
electrical pulse will alter the spectrum of the stimulus. A The second method of determining peSPL involves
second approach for determining the level of an acoustic measuring the peak-to-peak voltage of the click (or other
transient is to purchase an SLM that records the largest transient) on the oscilloscope, and adjusting the sine wave
instantaneous pressure (the “peak” pressure) and “holds” until its peak-to-peak voltage is equal to the peak-to-peak
this value in the display until the meter is reset. This peak- voltage of the click. The SPL value displayed on the SLM is
hold measurement may vary with the specific SLM, as the recorded as the peak-to-peak peSPL of the click (or other
measurement interval over which this “peak” is evaluated transient). The baseline-to-peak peSPL can never be less
(the time constant) varies with the particular SLM. It is than the peak-to-peak peSPL. If the voltages of the positive
desirable to use a meter with a pSPL time constant of sev- and negative phases of a click (or other transient) are equal,
eral tens of microseconds, or less. A third approach is to then the baseline-to-peak and peak-to-peak peSPL values
use an oscilloscope and an SLM that has an analog (AC) will be numerically equal. If the click is critically damped
output. This type of output enables you to route the ana- and shows a voltage deflection in only one direction (posi-
log voltage output of the microphone to the oscilloscope. tive or negative), then the baseline-to-peak peSPL will be
This technique can be used with any transient, including 6 dB larger than the peak-to-peak peSPL. As the baseline-
a click, and this approach is another method to determine to-peak peSPL and the peak-to-peak peSPL values can differ
the peSPL. Figure 2.3 diagrams two procedures for deter- by as much as 6 dB, it is imperative that the measurement
mining click peSPL. technique used when reporting peSPL is reported.
In the first procedure, referred to as the baseline-to-peak A fourth approach to measuring the level of a transient
peSPL procedure, the click (in this case) or other transient is is to eliminate the SLM and to use a coupler, microphone,
routed through the earphone/coupler/microphone/SLM to microphone conditioning amplifier and oscilloscope. The
the oscilloscope. The click (or other transient) stimulus is click (or other transient) stimulus is presented, and the peak
 CHAPTER 2 • Calibration 23

voltage on the oscilloscope is measured. The microphone is microphone conditioning/amplifier or SLM is routed to a
calibrated by using a sound source with a known SPL (i.e., spectrum analyzer or to an analog-to-digital converter to a
a pistonphone or acoustic calibrator), or the microphone computer that is programmed to do a Fourier analysis of
sensitivity curve can be used to convert microphone out- the signal. In each case, a display of the spectrum of each
put voltage to input sound pressure (and ultimately to SPL). signal type can be obtained. NOTE: Most clinics will likely
The acoustic calibrator is coupled to the microphone, and not have the instrumentation or perhaps the expertise to do
the voltage out of the microphone conditioning amplifier the calibration of transient stimuli. These calibration guide-
is measured. In this way, a given voltage is produced when a lines will thus be most useful to the manufacturers and to
specified SPL is present at the microphone diaphragm. For the technician who periodically calibrates the AEP instru-
example, an acoustic calibrator produces 114 dB SPL, and 1 mentation.
V is measured at the conditioning amplifier output. Then, Many AEPs are elicited by presentation of brief acous-
a 2-cc coupler is placed on the microphone. An Etymotic tic transients. The two most commonly used transients
ER-3 A insert microphone is then mated to the coupler. A are clicks and tonebursts. A click is produced by exciting
sine wave is presented through the earphone, and 50 mv is a transducer with a brief-duration electrical pulse. For use
measured. The SPL is with human subjects, a common click duration is 100 µs.
In a pulse with duration d, there are spectral zeroes (energy
20 log (50 mv/1,000 mv) + 114 dB SPL = 88 dB SPL
at that frequency dips toward zero) at frequencies that are
The first part of the formula estimates the dB re: 114 dB integer multiples of 1/d. Thus, for an electrical pulse with a
SPL, which is −26 dB; 114 is added to convert from “dB re: duration of 100 µs, the first spectral zero occurs at 10,000 Hz,
114 dB SPL” to “dB SPL.” with spectral zeroes occurring at 20,000, 30,000, 40,000 Hz,
Using the sensitivity curve of the microphone, the peak and so on. The acoustic representation of this 100-µs pulse,
voltage on the oscilloscope is converted to peak pressure as recorded through a TDH-50 earphone in a 6-cc coupler, is
(e.g., in Pascals). The sensitivity of a microphone relates the shown in Figure 2.4. It is important for the reader to under-
voltage out of the microphone to the sound pressure at the stand that the click has energy over a wide range of frequen-
microphone diaphragm, in for example, millivolts per pas- cies, and thus is a broadband stimulus. It is also important
cal. Then the pressure is converted to SPL. For example, a to understand that the spectrum of a transient (indeed, of
peak voltage of 100 mV is measured. The microphone sen- any signal) can and is influenced by not only the earphone,
sitivity is 50 mV/Pa. If 50 mV represents 1 Pa, then 100 mV but also the coupler (and perhaps the microphone) used in
represents 2 Pa. Converting to pSPL, the calibration process.
Clicks are often the stimuli of choice when using the
pSPL = 20 log (2 Pa/0.00002 Pa) = 20 log 10,000
auditory brainstem response (ABR) for hearing screening,
= ∼100 dB pSPL
site-of-lesion testing, and intraoperative monitoring. In
This value should correspond to the pSPL produced by contrast, when interested in obtaining an electrophysiologi-
an SLM in peak-hold mode, although if the time constant of cal estimate of the behavioral audiogram, then a broadband
the SLM is too long (say 100 µs), then the SLM may produce stimulus is not optimal. Several approaches have been used
a lower value. For a given stimulus, the baseline-to-peak to elicit responses from a limited cochlear region. The vari-
peSPL value should be 3.01 dB less than the true pSPL value. ous approaches can be broken down into two strategies:
This is because the peSPL value is actually referenced to a (1) using stimuli with narrow spectra and/or (2) using
RMS measure of a sine wave, rather than a peak measure. masking procedures to eliminate the contribution of spe-
To obtain the true pSPL value, 3.01 dB must be added to cific regions of the cochlea. The latter approach (the use of
the peSPL (using the baseline-to-peak procedure), because masking procedures) goes beyond the scope of the present
the crest factor (ratio of peak to RMS) of a true sine wave is chapter, but is considered later in this book (Chapter 11:
1.414, or 3.01 dB (3.01 dB = 20 log 1.414). Introduction to auditory-evoked potentials).
In addition to the overall level, it is important to One approach to generating a limited-spectrum stimu-
determine the frequency characteristics of the signal (i.e., lus is to present a sine wave for a brief duration. ANSI S3.6-
its spectrum) as it is played through the transducer. The 2010, which reviews technical specifications for audiom-
acoustic spectrum of a signal is not necessarily identical to eters, states that for audiometric purposes a tone must be
the spectrum of the electrical signal. This is because each presented for a duration of not less than 200 ms, and have
system has its own transfer function (i.e., filtering charac- a rise time and fall time ranging between 25 and 100 ms.
teristics), and the acoustic spectrum of the stimulus will Figure 2.5 (upper panel) shows a toneburst with a carrier
be affected by both the earphone and the coupler used to frequency of 2,000 Hz. The time required for the toneburst
mate the earphone to the SLM microphone. The spectrum envelope to increase from zero to its maximal amplitude
of the signal can be measured by routing the acoustic signal is termed its rise time. The time that the toneburst enve-
through a coupler, condenser microphone, and microphone lope remains at this maximal amplitude is called its pla-
conditioning/preamplifier or SLM. Finally, the output of the teau time. The time required for the toneburst envelope to
24 SECTION I • Basic Tests and Procedures

A Done 0s 2.646 Vpk

10
Vpk

2
Vpk/div

–10
Vpk 0 s 3.891 ms
Time 1 real part No avg
FIGURE 2.4 Time-domain
waveform (upper panel) and B Done 256 Hz –7.787 dBVpk
spectrum (lower panel) of an 0
dBVpk
electrical pulse with a dura-
tion of 100 µs. (Reprinted from
Burkard R, Secor C. (2002)
Overview of auditory evoked 6
potentials. In: Katz J, ed. Hand- dB/div
book of Clinical Audiology. 5th
ed. Baltimore, MD: Lippincott
Williams & Wilkins; pp 233–248, –60
with the permission of Lippincott dBVpk 256 Hz 25.6 kHz
FFT 1 Log mag uniform Vec avg 64
Williams & Wilkins.)

decrease from its maximal amplitude to zero amplitude is frequency (i.e., the frequency of the sine wave). For the
called its fall time. Figure 2.5 (lower panel) shows the ampli- toneburst, there is significant energy over a range of fre-
tude spectrum (i.e., frequency-domain representation– quencies centered at 2,000 Hz. This spread of energy to fre-
amplitude across frequency) of this toneburst. A tone that quencies above and below the carrier frequency is referred
is infinitely long in duration only has energy at the carrier to as spectral splatter. Acoustic transients are used to elicit

A Done 0s 4.826 mVpk

250
mVpk

50
mVpk/div

–250
FIGURE 2.5 Upper panel: The mVpk 0 s 15.59 ms
Time 1 real part No avg
time-domain representation of
a toneburst. Lower panel: The B Done 64 Hz –99.354 dBVpk
frequency-domain representa- –20
dBVpk
tion of the same toneburst.
(Reprinted from Burkard R,
Secor C. (2002) Overview of
auditory evoked potentials. In: 10
Katz J, ed. Handbook of Clinical dB/div
Audiology. 5th ed. Baltimore,
MD: Lippincott Williams &
Wilkins; pp 233–248, with the –120
permission of Lippincott dBVpk 64 Hz 12.8 kHz
FFT 1 Log mag uniform Vec avg 64
Williams & Wilkins.)
 CHAPTER 2 • Calibration 25

AEPs, because many of these AEPs, such as the ABR, are fundamental frequency level when measured in an HA-1–
onset responses. Onset responses are elicited by the first few type coupler. The probe signal shall not exceed 90 dB SPL
milliseconds of the stimulus onset, and hence are primar- as measured in that coupler, in an attempt to minimize the
ily affected by the stimulus rise time. When using toneburst possibility that the probe signal will elicit an acoustic reflex.
stimuli, it is intended to serve as a compromise between an The range of acoustic admittance and acoustic impedance
audiometric tone (long duration, long rise/fall times, and values that should be measurable varies by instrument type.
a very narrow spectrum) and a click (short duration, very The accuracy of the acoustic immittance measurements
fast rise/fall times, and very broad spectrum). To add even should be within 5% of the indicated value or ±10−9 cm3/Pa
more complexity to this topic, there are many unique gat- (0.1 acoustic mmho), whichever is greater. The accuracy of
ing functions that can be used to shape the onset and offset the acoustic immittance measurement can be determined
of the toneburst, including linear, Blackman, and Hanning by connecting the probe to the test cavities and checking
(cosine2) functions. As mentioned previously, methods for the accuracy of the output at specified temperatures and
the calibration of acoustic transients can be found in IEC ambient barometric pressures. A procedure for checking the
60645-3:2007, and RETSPLs for clicks and select toneburst temporal characteristics of the acoustic immittance instru-
stimuli can be found in ISO 389-6:2007. ment is described by Popelka and Dubno (1978) and by
Lilly (1984).
OTOACOUSTIC EMISSION Air pressure may be measured by connecting the probe
to a “U tube” manometer and then determining the water
DEVICES displacement as the immittance device air pressure dial is
There are currently no ANSI standards for otoacoustic rotated. If the SI unit of decapascals (daPa) is used, then
emission (OAE) devices. IEC 60645-6 is an international an appropriate manometer or pressure gauge must also be
standard that can be used to specify the stimuli used to used. The air pressure should not differ from that stated on
obtain OAEs. If using a click to obtain transient-evoked oto- the device (i.e., 200 daPa) by more than ±10 daPa or ±15%
acoustic emissions (TEOAEs), reporting the level in pSPL of the reading, whichever is greater. The standard states that
or peSPL is appropriate. For distortion product otoacoustic the air pressure should be measured in cavities with vol-
emissions (DPOAEs), verification that the primary signals umes of 0.5 to 2 cm3.
(as measured in an appropriate coupler/microphone/SLM Finally, the reflex-activating system should be
and frequency counter or spectrum analyzer) are close to checked. In checking the activation of a contralateral or
the levels and frequencies as specified by the OAE device is ipsilateral reflex, normally an insert receiver will be used
desirable. Because the DPOAE response is an intermodula- that may be measured on a standard HA-1 coupler. The
tion distortion product (typically the cubic difference tone), frequency of the activator can be measured electrically
it is critical to measure the amplitude of the distortion at this directly from the acoustic immittance device. In this
frequency in a hard-walled cavity to know when measured case, one uses a frequency counter as described earlier
distortion in fact represents distortion in the instrumenta- for checking the frequency of puretones in audiometers.
tion itself, rather than representing nonlinearities generated Frequency should be ±3% of the stated value, and har-
by the inner ear. monic distortion should be less than 3% at specified fre-
quencies for earphones and 5% or less for the probe tube
transducer or insert receiver. Noise bands should also be
ACOUSTIC IMMITTANCE DEVICES checked if they are to be used as activating stimuli. Broad-
The standard for acoustic immittance (impedance/admit- band noises should be uniform within ±5 dB for the range
tance) devices is ANSI S3.39-1987. Note that there is also between 250 and 6,000 Hz for supra-aural earphones.
an IEC standard, IEC 60645-5:2004, for measurement of This can be checked by sending the output through the
aural acoustic impedance/admittance. ANSI S3.39-1987 transducer connected to a coupler, a microphone, and a
describes four types of units for measuring acoustic immit- graphic level recorder or spectrum analyzer. The SPL of
tance (listed simply as Types 1, 2, 3, and 4). The specific tonal activators should be within ±3 dB of the stated value
minimum mandatory requirements are given for Types for frequencies from 250 to 4,000 Hz and within ±5 dB for
1, 2, and 3. There are no minimum requirements for the frequencies of 6,000 to 8,000 Hz and for noise. The rise
Type 4 device. Types 1, 2, and 3 must have at least a 226-Hz and fall times should be the same as those described for
probe signal, a pneumatic system (manual or automatic), audiometers and may be measured in the same way. Daily
and a way of measuring static acoustic immittance, tympa- listening checks as well as periodic tests of one or two per-
nometry, and the acoustic reflex. Thus, to check the acous- sons with known acoustic immittance to check tympano-
tic immittance device, one may begin by using a frequency grams and acoustic reflex thresholds should be performed
counter to determine the frequency of the probe signal(s). to catch any gross problems.
The frequency should be within 3% of the nominal value. In summary, acoustic immittance devices should be
The total harmonic distortion shall not exceed 5% of the checked as carefully as one’s puretone audiometer. Failure
26 SECTION I • Basic Tests and Procedures

to do so can lead to variability in measurement, which may FOOD FOR THOUGHT

invalidate the immittance measurement.
1. Do you think it is more important for standards to tell us
(a) how to optimally characterize our stimuli or (b) how
TEST ROOM STANDARDS to be consistent in how we characterize our stimuli? What
It is insufficient to limit the periodic calibration checks to are your reasons? If you believe that they are equally
the audiometric equipment. The environment in which important, please state your reasons for that point of
the test is to be carried out must also be evaluated. ANSI view.
S3.1-1999 provides criteria for permissible ambient noise 2. Describe the advantages and disadvantages of having a
during audiometric testing. Section 11 of ISO 8253-1:2010 formal electroacoustic characterization of your audio-
also specifies appropriate ambient noise levels. The ambi- metric equipment every 3 months versus annually.
ent level in the test room is checked by using an SLM that 3. State why you agree or disagree with the following state-
is sensitive enough to allow testing to levels as low as 8 dB ment: All Au.D. students should learn (both in lecture
SPL. Many modern SLMs can measure to levels of 5 dB and in hands-on laboratories) how to calibrate all audio-
SPL or less. One should place the SLM (preferably using metric equipment that they use in the clinic.
a free-field microphone) in the place where the subject is
to be seated. The doors of the test room should be closed
when making the measurements. If one plans to use moni- REFERENCES
tored live voice testing, the ambient levels in the examin- American Medical Association. (1951) Specifications of the Council
er’s room should also be checked. However, there are no of Physical Medicine and Rehabilitation of the American Medical
standards concerning acceptable noise levels in the exam- Association. J Am Med Assoc. 146, 255–257.
iner’s room. ANSI S3.1-1999 provides acceptable ambient American National Standards Institute. (2004) About ANSI over-
noise values for threshold estimation at 0 dB HL, for one- view. Available at: http://www.ansi.org/about_ansi/overview/
and third-octave bandwidths, for use with supra-aural and overview.aspx?menuid=1 (accessed July 13, 2013).
insert earphones, and for free-field (or bone conduction) American Speech-Language-Hearing Association. (1991) Sound
testing. These values vary with the range of audiometric Field Measurement Tutorial. Rockville, MD: American Speech-
Language-Hearing Association.
frequencies investigated. If the level reported by ANSI
Beranek LL. (1988) Acoustical Measurements. New York, NY:
S3.1-1999 is exceeded, then the minimum dB HL value
American Institute of Physics.
that can be recorded is increased from 0 dB HL. This is Burkhard MD. (1978) Manikin Measurements–Conference Proceed-
(more or less) a linear function, so if the accepted ambient ings. Elk Grove Village, IL: Industrial Research Products.
noise level in a given band is exceeded by 5 dB, then the Burkhard MD, Corliss ELR. (1954) The response of earphones in
minimum dB HL value that you can measure is increased ears and couplers. J Acoust Soc Am. 26, 679–685.
to 5 dB HL. Burkhard MD, Sachs RM. (1975) Anthropometric manikin for
acoustic research. J Acoustic Soc Am. 58, 214–222.
Corliss ELR, Burkhard MD. (1953) A probe tube method for the
CONCLUSIONS transfer of threshold standard between audiometer earphones.
This chapter has emphasized that the audiologist must, on a J Acoust Soc Am. 25, 990–993.
Cox R. (1986) NBS-9 A coupler-to-eardrum transformation:
daily basis, listen to the output of the equipment. There are
TDH-39 and TDH-49 earphones. J Acoust Soc Am. 79, 120–123.
many problems that can be detected by a trained human ear. Decker TN, Carrell TD. (2004) Instrumentation: An Introduction
However, the listener is simply not good enough to check for Students in the Speech and Hearing Sciences. Mahwah, NJ:
the auditory equipment with the precision that is needed Lawrence Erlbaum Associates.
to ensure that it is working properly. Thus, it has also been Harris C. (1998) Handbook of Acoustical Measurements and
stressed that, to determine the precise characteristics of the Noise Control. 3rd ed. Woodbury, NY: Acoustical Society of
equipment, routine electroacoustic checks must be carried America.
out. Even when there are no current standards (i.e., there Hawkins DB, Cooper WA, Thompson DJ. (1990) Comparisons
is, at the time this chapter is being written, no ANSI stan- among SPLs in real ears, 2 cm3 and 6 cm3 couplers. J Am Acad
dard specifying the RETSPL of transients), one should at Audiol. 1, 154–161.
least check the stability of one’s equipment. Because the Hawley MS. (1939) An artificial mastoid for audiophone measure-
ments. Bell Lab Rec. 18, 73–75.
test results that one obtains are no more accurate than the
Johnson D, Marsh A, Harris C. (1998) Acoustic measurement
equipment on which they are performed, both clinical and instruments. In: Harris C, ed. Handbook of Acoustical Measure-
calibration equipment must be chosen and maintained with ments and Noise Control. 3rd ed. Woodbury, NY: Acoustical
care. The ultimate responsibility for the accuracy of the test Society of America; pp 5.1–5.21.
results lies with the audiologist. Therefore, the audiologist Killion MD. (1978) Revised estimate of minimum audible
must make sure that the equipment is working properly by pressure: where is the “missing 6 dB?” J Acoust Soc Am. 63,
carrying out routine calibration checks. 1501–1508.
 CHAPTER 2 • Calibration 27

Lilly DJ. (1984) Evaluation of the response time of acoustic-immit- Weiss E. (1960) An air-damped artificial mastoid. J Acoust Soc Am.
tance instruments. In: Silman S, ed. The Acoustic Reflex. New 32, 1582–1588.
York, NY: Academic Press. Weissler P. (1968) International standard reference zero for audi-
Melnick W. (1973) What is the American National Standards Insti- ometers. J Acoust Soc Am. 44, 264–275.
tute? ASHA. 10, 418–421. Wilber LA. (2004) What Are Standards and Why Do I care? Semi-
Occupational Safety and Health Administration. (1983) Occupa- nars in Hearing—Current Topics in Audiology: A Tribute to Tom
tional Noise Exposure, Hearing Conservation Amendment. Rule Tillman. Stuttgart, Germany: Thieme; pp 81–92.
and Proposed Regulation. Washington, DC: Federal Register, Wilber LA, Goodhill V. (1967) Real ear versus “artificial mastoid”
United States Government Printing Office. methods of calibration of bone-conduction vibrators. J Speech
Popelka GR, Dubno JR. (1978) Comments on the acoustic- Hear Res. 10, 405–416.
reflex response for bone-conducted signals. Acta Otolaryngol Yeager D, Marsh A. (1998) Sound levels and their measurement. In:
(Stockh). 86, 64–70. Harris C, ed. Handbook of Acoustical Measurements and Noise
Roach R, Carhart R. (1956) A clinical method for calibrating the Control. 3rd ed. Woodbury, NY: Acoustical Society of America;
bone-conduction audiometer. Arch Otolaryngol. 63, 270–278. pp 11.1–11.18.
Rosen S, Howell P. (1991) Signals and Systems for Speech and Hear- Zwislocki JJ. (1970) An Acoustic Coupler for Earphone Calibration.
ing. London: Academic Press. Rep. LSC-S-7, Lab Sensory Commun. Syracuse, NY: Syracuse
Rudmose W. (1964) Concerning the problem of calibrating TDH-39 University.
earphones at 6 kHz with a 9 A coupler. J Acoust Soc Am. 36, 1049. Zwislocki JJ. (1971) An Ear-Like Coupler for Earphone Calibration.
Silverman FH. (1999) Fundamentals of Electronics for Speech- Rep. LSC-S-9, Lab Sensory Commun. Syracuse, NY: Syracuse
Language Pathologists and Audiologists. New York, NY: Allyn University.
and Bacon. Zwislocki JJ. (1980) An ear simulator for acoustic measurements.
Speaks C. (1996) Introduction to Sound. Acoustics for the Hearing and Rationale, principles, and limitations. In: Studebaker G, Hoch-
Speech Sciences. 2nd ed. San Diego, CA: Singular Publishing. berg I, eds. Acoustical Factors Affecting Hearing Aid Perfor-
Studebaker G. (1967) Interest variability and the air-bone gap. mance. Baltimore, MD: University Park Press.
J Speech Hear Disord. 32, 82–86.
 C H A P T ER 3

Puretone Evaluation

Robert S. Schlauch and Peggy Nelson

to 6,000 Hz) that is important for speech understanding

INTRODUCTION (French and Steinberg, 1947).
Most people who attend primary school in the United States Puretone amplitude or level is usually quantified in
and in other industrialized nations experience puretone* decibels. Decibels (dB) represent the logarithm of a ratio
testing firsthand as a method to screen for hearing loss. of two values; the term is meaningless without a reference.
Puretone threshold testing is seen in films, such as Woody Two commonly used decibel scales are sound pressure level
Allen’s award-winning movie Hannah and Her Sisters or the (SPL) and hearing level (HL). The reference level for dB SPL
film Wind Talkers. These casual experiences with audiology is 20 µPa, a pressure value. This reference value for SPL was
may give lay people the false impression that audiology is a selected to correspond to the faintest pressure that is audible
narrow profession. in the frequency region where hearing is most sensitive. The
Most audiologists would likely agree that puretone (PT) frequency is not specified in the reference level for dB SPL;
thresholds represent a key component of the assessment all sounds expressed in units of dB SPL share the same refer-
battery. Proper administration and interpretation of PT ence of 20 µPa. The SPL scale is frequently used in audiology
threshold tests require considerable knowledge, as it is not to compare the level of speech or other sounds at different
always simple and straightforward. The goal of this chapter frequencies. Such comparisons are critical for prescribing
is to introduce readers to the complexity of PT threshold and evaluating hearing aids. HL, a second decibel scale, is
testing, as well as to provide clinicians with a reference for used to plot an audiogram, the accepted clinical representa-
clinical applications. tion of puretone thresholds as a function of frequency. The
reference for dB HL is the median threshold for a particular
frequency for young adults with no history of ear problems.
WHAT ARE PURETONES AND Unlike dB SPL, the zero reference level for dB HL varies with
HOW ARE THEY SPECIFIED? frequency, because humans have more sensitive hearing at
some frequencies than others. Because the reference is nor-
PT thresholds represent the lowest level of response to a tonal mal human hearing, thresholds that deviate from 0 dB HL at
stimulus. Puretones are the simplest of sounds described by any frequency show how much one’s hearing deviates from
their frequency, amplitude, phase, and duration. The most this normal value.
important of these characteristics for puretone audiometry Figure 3.1 illustrates thresholds displayed in dB SPL
are frequency and amplitude (or intensity level). and dB HL. The left panel shows hearing thresholds plot-
Puretone frequency is perceived as pitch, the charac- ted in dB SPL as a function of frequency. Thresholds plot-
teristic of sound that determines its position on a musical ted in this way constitute a minimum audibility curve. The
scale. Young people with normal hearing are able to perceive right panel shows a conventional audiogram plotted in dB
frequencies between 20 and 20,000 Hz. Human hearing is HL. Note that on the dB SPL scale, larger decibel values
more sensitive (better) in the range of frequencies between are plotted higher on the graph. By contrast, larger values
500 and 8,000 Hz than it is at either extreme of the audible in dB HL are plotted lower on the audiogram. To illustrate
range of frequencies. Conventional puretone audiometry the relationship between dB SPL and dB HL, the reference
typically assesses thresholds for frequencies between 250 (or values for 0 dB HL (average normal hearing) for a specific
125) and 8,000 Hz. The frequency range for conventional earphone are plotted in dB SPL as a solid line. Illustrated
audiometry is very similar to the range of frequencies (100 with a dashed line on these same two figures are the thresh-
olds for a person with a high-frequency hearing loss. Note
in the figure on the left that the separation between the solid
*The use of the compound noun “puretone” is the editor’s choice for line and the dashed line represents values for dB HL on the
consistency purposes. audiogram.

29
 30 SECTION I • Basic Tests and Procedures

Minimal audibility curve Audiogram

120 0

100
Sound pressure level (dB re: 20 µP)

20

Hearing loss (dB HL)

80
40

60

60
40

80
20

0 100
0.125 0.25 0.5 1.0 2.0 4.0 8.0 0.125 0.25 0.5 1.0 2.0 4.0 8.0
Frequency (kHz) Frequency (kHz)
FIGURE 3.1 Thresholds in dB sound pressure level (SPL; left panel) and dB hearing level (HL; right
panel) as a function of frequency. The solid line represents average normal hearing; the dashed line
represents a person’s threshold who has a high-frequency hearing loss.

frequencies at the apical end. Damage to sensory cells of the

WHY PURETONE THRESHOLDS? cochlea at a specific place along the basilar membrane can
The reader might be wondering why audiologists use pure- result in a loss of hearing that corresponds to the frequen-
tones at specific frequencies when the most meaningful cies coded by that place. For this reason, PT threshold tests
stimulus is speech. Two important reasons are that PT provide details that would otherwise remain unknown if a
thresholds provide information about the type of hearing broadband stimulus such as speech were used.
loss, as well as quantify frequency-specific threshold eleva- In addition to providing audiologists with critical diag-
tions that result from damage to the auditory system. nostic information about the amount and type of loss, PT
PT thresholds provide quantification of amount of loss thresholds find applications (1) for estimating the degree of
due to problems with the outer and middle ear (the conduc- handicap, (2) as a baseline measure for hearing conserva-
tive system) separately from the cochlea and the auditory tion programs, (3) for monitoring changes in hearing fol-
nerve (the sensory/neural system). This distinction helps in lowing treatment or progression of a disease process, (4) for
the diagnosis and guides audiologists and physicians with screening for hearing loss, (5) for determining candidacy
important details for providing treatment strategies. for a hearing aid or a cochlear implant, and (6) for select-
Damage to the auditory system often results in a loss of ing the frequency-gain characteristics of a hearing aid. PT
sensitivity that is frequency specific. For instance, changes thresholds also provide a reference level for presentation of
in the stiffness and mass properties of the middle ear affect suprathreshold speech testing and for the meaningful inter-
the relative amount of loss in the low and high frequen- pretation of other audiologic tests, such as evoked otoacous-
cies (Johanson, 1948). For air-conduction thresholds, an tic emissions and acoustic reflex thresholds. PT thresholds
increase in stiffness results in a greater low-frequency loss, are also used to assess the functional attenuation of hearing
whereas an increase in mass results in a greater loss in the protection devices.
high frequencies. Thresholds for puretones (or other nar-
rowband sounds) also provide us with diagnostic informa-
tion about the integrity of different channels in the sensory/
TUNING FORK TESTS
neural pathway. The auditory system is organized tonotopi- A struck tuning fork produces a sustained puretone that
cally (i.e., a frequency-to-place mapping) from the cochlea decays in level over time. Unlike an audiometer, tuning
to the cortex. The tonotopic organization of the cochlea is forks cannot present a calibrated signal level to a listener’s
a result of the frequency tuning of the basilar membrane, ear. Despite this shortcoming, tuning fork tests provide
with high frequencies represented at the basal end and low qualitative information that can help determine whether a
 CHAPTER 3 • Puretone Evaluation 31

hearing loss is conductive or sensory/neural. Tuning fork falsely identified as having conductive losses with the Rinne
tests are promoted by some as an important supplement to test, this test misses many people with significant conductive
puretone audiometry. In a recently published book, otolo- losses (Browning, 1987), including 50% of losses that have
gists are advised to include tuning fork tests as an integral 20-dB air–bone gaps. The Weber test fares equally poorly.
part of the physical examination for conductive hearing loss Browning (1987) reports that a majority of children with
(Torres and Backous, 2010). conductive losses give inappropriate responses on the Weber
The two best known tuning fork tests are the Weber test. From these and other studies, one must conclude that
and Rinne. Judgments about the type of hearing loss are tuning fork tests are not a replacement or even a supplement
made by comparing the pattern of results on both tests. to audiometry. Audiometry is capable of identifying nearly
Air conduction (AC) is tested by holding the tuning fork at 100% of air–bone gaps, as small as 15 dB.
the opening of the ear canal, and bone conduction (BC) is
tested by placing the tuning fork on the mastoid process (the
bony area behind the pinna) or on the forehead or incisors
PURETONE AUDIOMETRY
(British Society of Audiology, 1987). For the Weber test, Audiometers are used to make quantitative measures of AC
a client judges whether sound is perceived in one or both and BC PT thresholds. AC thresholds assess the entire audi-
ears when the tuning fork is placed on the forehead. For tory pathway and are usually measured using earphones.
the Rinne test, the client judges whether sound is louder When sound is delivered by an earphone, the hearing sen-
when presented by AC or by BC. Ideally, conductive hear- sitivity can be assessed in each ear separately. BC thresholds
ing losses produce a pattern of responses that is uniquely are measured by placing a vibrator on the skull, with each
different from the one for sensory/neural hearing losses. In ear assessed separately, usually by applying masking noise
the Weber, the sound is lateralized to the poorer ear with a to the nontest ear. The goal of BC testing is to stimulate the
conductive loss and to the better ear for a sensory/neural cochlea directly, thus bypassing the outer and middle ears.
loss. In the Rinne, the sound is louder by BC in a conductive A comparison of AC and BC thresholds provides separate
loss and by AC with a sensory/neural loss. estimates of the status of the conductive and sensory/neural
Some recommend tuning fork tests to check the valid- systems. If thresholds are elevated equally for sounds pre-
ity of audiograms (Gabbard and Uhler, 2005) or to confirm sented by AC and BC, then the outer and middle ear are not
the audiogram before conducting ear surgery (Sheehy et al., contributing to a hearing loss. By contrast, if thresholds are
1971). However, it is important to recognize that tuning poorer by AC than by BC, then the source of at least some of
fork tests administered to people with known conductive the loss is the outer or middle ear. Figure 3.2 illustrates the
losses have shown that these procedures are often inaccurate AC and BC pathways and how hearing thresholds are typi-
(Browning, 1987; Snyder, 1989). Although only about 5% cally affected by damage to these structures. See Chapter 4
of people with normal hearing or sensory/neural losses are for a complete review of BC assessment.

Conductive mechanism Sensory/neural mechanism

Outer ear Middle ear Cochlea VIIIth nerve to brain

Normal hearing
Air conduction
(AC)

Bone conduction (BC)

Conductive
hearing loss
AC impaired

BC normal
Sensory/neural
hearing loss
AC normal

BC impaired
Mixed
hearing loss
AC impaired

BC impaired FIGURE 3.2 Conductive and sensory/

= potential source neural pathways. (Adapted from Martin
of hearing loss
(1994))
32 SECTION I • Basic Tests and Procedures

Equipment
AUDIOMETERS
Puretones are generated within an audiometer. Audiometers
have the ability to select tonal frequency and intensity level
and to route tones to the left or right earphone. All audi-
ometers also have an interrupter switch that presents the
stimulus to the examinee. The American National Standards
Institute (ANSI) Specification for Audiometers (ANSI, 2010)
describes four types of audiometers, with Type 1 having the
most features and Type 4 having the fewest features. A Type 1
audiometer is a full-featured diagnostic audiometer. A Type
1 audiometer has earphones, bone vibrator, loud speakers,
masking noise, and other features. A Type 4 audiometer is FIGURE 3.4 Etymotic model ER-3A insert earphones.
simply a screening device with earphones, but none of the
other special features.
due to their ease of calibration and the lack of other types
Type 1 (full-featured, diagnostic audiometer) has the
of commercially available earphones. In the past few years,
ability to assess puretone AC thresholds for frequencies
insert earphones and circumaural earphones have become
ranging from 125 to 8,000 Hz and BC thresholds for fre-
available and provide some useful applications for puretone
quencies ranging from 250 to 6,000 Hz. If an audiometer
assessment.
has extended high-frequency capability, air-conduction
Insert earphones are coupled to the ear by placing a
thresholds can be extended to 16,000 Hz. Maximum output
probe tip, typically a foam plug, into the ear canal. The com-
levels for AC testing are as high as 120 dB HL for frequen-
mercially available model that has a standardized calibra-
cies where hearing thresholds are most sensitive. By contrast,
tion method for audiology is the Etymotic model ER-3A,
distortion produced by bone oscillators at high intensities
which is illustrated in Figure 3.4. These earphones have
limits maximum output levels for BC thresholds to values
gained popularity in the past few years because they offer
nearly 50 dB lower than those for AC thresholds for the same
distinct advantages over supra-aural earphones. One major
frequency.
advantage is that insert earphones yield higher levels of
interaural attenuation than supra-aural earphones (Killion
TRANSDUCERS and Villchur, 1989). Interaural attenuation represents the
decibel reduction of a sound as it crosses the head from the
Earphones
test ear to the nontest ear. The average increase in interau-
Earphones are generally used to test puretone AC thresholds. ral attenuation is roughly 20 dB. This reduces the need for
A pair of supra-aural earphones is illustrated in Figure 3.3. masking the nontest ear and decreases the number of mask-
For decades, supra-aural earphones, ones in which the cush- ing dilemmas, situations for which thresholds cannot be
ion rests on the pinna, were the only choice for clinical audi- assessed, because the presentation level of the masking noise
ology. The popularity of supra-aural phones was mainly is possibly too high. (See Chapter 6 for a comprehensive
review of masking.) Another important advantage of insert
earphones over supra-aural earphones is lower test–retest
variability for thresholds obtained at 6 and 8 kHz; variability
for other frequencies is comparable. Given that thresholds
for 6 and 8 kHz are important for documenting changes in
hearing due to noise exposure and for identifying acoustic
tumors, lower variability should increase the diagnostic pre-
cision. A third advantage that insert earphones offer is elim-
ination of collapsed ear canals (Killion and Villchur, 1989).
In about 4% of clients, supra-aural earphones cause the ear
canal to narrow or be closed off entirely when the cushion
presses against the pinna, collapsing the ear canal (Lynne,
1969), resulting in false hearing thresholds, usually in the
high frequencies (Figure 3.5) (Ventry et al., 1961). Because
insert earphones keep the ear canal open, collapsed canals
FIGURE 3.3 Telephonics model TDH-49, an example of are eliminated. A fourth advantage of insert earphones is
supra-aural earphones. that they can be easily used with infants and toddlers who
 CHAPTER 3 • Puretone Evaluation 33

Occlusion Symbol
0 80.5%
91.6%
94.4%
20 98.3%
dB hearing level (re: ANSI 2010)

99.2%
100%
40

60

FIGURE 3.5 Air-conduction

80 (AC) thresholds (in dB hearing
level [HL]) for different percen-
tages of ear canal occlusion.
100 One hundred percent indicates
that the ear canal is completely
occluded. Deviations from 0 dB
120 HL represent the loss due to
250 500 1,000 2,000 4,000 8,000 occlusion. (Adapted from
Frequency (Hz) Chandler (1964))

cannot or will not tolerate supra-aural earphones. A fifth earphone type (Voss et al., 2000; Voss and Herman, 2005).
advantage of insert earphones is the option of conducting For these cases, measuring the SPL at the eardrum to
middle-ear testing and otoacoustic emission testing with- specify the level presented to an individual patient would
out changing the earphones; some recently introduced improve the accuracy of hearing thresholds. The probe-
diagnostic instruments use this approach. Although insert tube microphones necessary for these types of measures
earphones offer a hygienic advantage over supra-aural ear- already exist, and hopefully, this technology will become
phones, because the foam tips that are placed into a client’s routinely available for use in diagnostic audiometers (see
ear canal are disposable, the replacement cost of those tips Scheperle et al., 2011 for a discussion of calibration in
is prohibitive for many applications. In addition to higher the ear canal).
costs, insert earphones also yield errant thresholds in persons
with eardrum perforations, including pressure-equalization Speakers
tubes (Voss et al., 2000). (See Figure 3.12 for additional AC thresholds can be measured using speakers as the trans-
information about perforations.) Insert earphones also have ducer. Thresholds so obtained are known as sound-field
maximum output levels that are lower than those produced thresholds. Sound-field thresholds are unable to provide
by supra-aural earphones for some frequencies. Because of ear-specific sensitivity estimates. In cases of unilateral
these differences, many diagnostic clinics keep both ear- hearing losses, the listener’s better ear determines thresh-
phone types on hand and switch between them depending old. This limitation and others dealing with control over
on the application. stimulus level greatly limit clinical applications involving
Circumaural earphones, a third type, have cushions sound-field thresholds. Applications for sound-field thresh-
that encircle the pinna. ANSI (2010) describes reference olds are screening infant hearing or demonstrating to the
equivalent threshold SPL values (SPL values corresponding parents their child’s hearing ability. Sound-field thresholds
to 0 dB HL) for Sennheiser model HDA200 and Koss model may also be desirable for a person wearing a hearing aid or
HV/1A earphones. These earphones and the Etymotic ER-2 cochlear implant.
insert earphones are the only ones in the current standard In sound-field threshold measures, the orientation of
that have reference values covering the extended high fre- the listener to the speaker has a large effect on stimulus level
quencies (8 to 20 kHz). presented at the eardrum. A person’s head and torso as well
Current standards for earphone calibration specify as the external ear (e.g., pinna, ear canal, concha) affect
the level based on measures obtained with the earphone sound levels (Shaw, 1974). Differences in SPL at the eardrum
attached to an acoustic coupler or artificial ear. These coup- are substantial for speaker locations at different distances
lers are designed to approximate the ear canal volume of an and different angles relative to the listener. For this reason,
average person. Given that some clients (e.g., infants) have sound-field calibration takes into consideration these fac-
very small or very large ear canals (e.g., some postsurgi- tors. A mark is usually made on the ceiling (or floor) of the
cal clients and persons with perforated eardrums), coupler room to indicate the location of the listener during testing.
measures may produce erroneous results, regardless of the Even at the desired location, stimulus level at the eardrum
34 SECTION I • Basic Tests and Procedures

for some frequencies can vary as much as 20 dB or more air–bone gaps and bone–air gaps are equipment miscalibra-
by simply having the listener move his or her head (Shaw, tion, test–retest variability, and individual differences in
1974). Calibration assumes the listener will always be fac- anatomy that cause thresholds to deviate from the group-
ing the same direction relative to the sound source (ANSI, mean data used to derive normative values for relating AC
2010). Furniture and other persons in the sound field also and BC thresholds.
affect the stimulus level at a listener’s eardrum (Morgan For threshold measurements bone vibrators are typi-
et al., 1979). All of these factors add to the challenge of cally placed behind the pinna on the mastoid process or
obtaining accurate sound-field thresholds. on the forehead. Although forehead placement produces
Another important consideration in sound-field slightly lower intrasubject and intersubject threshold dif-
threshold measures is the stimulus type. Thresholds cor- ferences (Dirks, 1994), placement on the mastoid process
responding to different frequencies are desired for plotting is preferred by 92% of audiologists (Martin et al., 1998).
an audiogram, but puretones can exhibit large differences Mastoid placement is preferred mainly because it produces
in level at different positions in a testing suite as a result between 8 and 14 dB lower thresholds than forehead place-
of standing waves. Standing waves occur when direct sound ment for the same power applied to the vibrator, depend-
from the speaker interacts with reflections, resulting in ing on the frequency (ANSI, 2010). The median difference
regions of cancellation and summation. Differences in is 11.5 dB. Given that the maximum output limits for bone
stimulus level due to standing waves are minimized by using vibrators with mastoid placement are as much as 50 dB
narrowband noise or frequency-modulated (FM) tones as lower than that for AC thresholds, forehead placement
the stimulus (Morgan et al., 1979). FM tones, also known yields an even larger difference. The inability to measure
as warbled tones, are tones that vary in frequency over a BC thresholds for higher levels means that a comparison of
range that is within a few percent of the nominal frequency. AC and BC thresholds is ambiguous in some cases. That is,
This variation occurs several times per second. Under ear- when BC thresholds indicate no response at the limits of the
phones, thresholds obtained with these narrowband stimuli equipment (e.g., 70 dB HL) and AC thresholds are poorer
are nearly identical to thresholds obtained with puretones, than the levels where no response was obtained (e.g., 100 dB
with some exceptions in persons with steeply sloping HL), the audiologist cannot establish from these thresholds
hearing loss configurations. FM tones and narrowband whether the loss is purely sensory/neural or whether it has a
noise are the preferred stimuli for sound-field threshold conductive component.
measures.

Bone Vibrators Test Environment

A bone vibrator is a transducer that is designed to apply Hearing tests ideally are performed in specially constructed
force to the skull when placed in contact with the head. sound-treated chambers with very low background noise. A
Puretone BC thresholds are measured with a bone vibrator sound-treated room is not a sound-proof room. High-level
like the one illustrated in Figure 3.6. A separation of 15 dB external sounds can penetrate the walls of a sound-treated
or more between masked AC and BC thresholds, with BC room and may interfere with test results. Because test tones
thresholds being lower than AC thresholds, is often evidence near threshold can be easily masked by extraneous, exter-
of a conductive hearing loss. Other possible explanations for nal noise, test chambers have strict guidelines for maximum
permissible ambient noise levels. Low background noise
levels are particularly important for BC testing, when the
ears remain uncovered. When testing is done in a room that
meets the ANSI guidelines, the audiogram reflects that by
citing ANSI S3.1 (1999), the standard governing permissible
ambient noise levels. Table 3.1 shows the minimum levels of
ambient noise measured in octave bands encompassing the
test frequency that enable valid hearing threshold measure-
ments at 0 dB HL.
At times, audiologists must estimate hearing thresh-
olds in rooms that do not meet the guidelines for minimal
ambient noise. Some patients in hospital rooms or nursing
homes must be tested at bedside. In those cases, test results
should be clearly marked so that others know the conditions
under which the test was done. When possible, these bedside
tests should be performed using insert earphones, which
FIGURE 3.6 A clinical bone-conduction vibrator provide a greater amount of attenuation in low frequencies
(Radioear Model B-72). where ambient noise is typically more of a problem. In these
 CHAPTER 3 • Puretone Evaluation 35

(American Speech-Language-Hearing Association [ASHA],

TA B L E 3 .1
2005). The goal of the guideline is to minimize differences
Maximum Permissible Ambient Noise across clinics by standardizing procedures. The commit-
Levels for Puretone Threshold Testing tee that drafted this consensus document understood that
its recommendations represent general guidelines and
Max dB SPL Max dB SPL that clinical populations may require variations of the
Octave Band Center with Ears with Ears procedure.
Frequency (Hz) Covered Uncovered
125 39 35 Instructions
250 25 21 Puretone audiometry begins with instructing the indi-
500 21 16 vidual being tested. The instructions are a critical part of
1,000 26 13 the puretone test, because thresholds measured using this
2,000 34 14 clinical procedure are biased by the willingness of a per-
4,000 37 11 son to respond. Some listeners wait for a tone to be distinct
8,000 37 14 before they respond, which leads to higher thresholds than
Adapted from American National Standards Institute. (1999) Maxi- for someone who responds whenever they hear any sound
mum Permissible Ambient Noise for Audiometric Test Rooms. ANSI that could be the tone. This bias is controlled in the instruc-
S3.1–1999. New York, NY: American National Standards Institute, tions by informing listeners to respond any time they hear
Inc. Octave band levels cannot exceed the tabled values to mea- the tone no matter how faint it may be. A study by Marshall
sure valid thresholds at 0 dB HL or lower.
and Jesteadt (1986) shows that response bias controlled for
in this manner plays only a small role (a few dB at most) in
environments, BC testing, particularly in the low frequen- PT thresholds obtained using the ASHA guideline. Marshall
cies, may not be valid. and Jesteadt (1986) also reported that the response bias of
elderly listeners was not different than that of a group of
Measuring Puretone Thresholds younger persons. Before the study by these authors, it was
believed that elderly persons might adopt an extremely con-
Psychophysics is the field of study that relates the physical servative response criterion, resulting in artificially elevated
world with perception. PT thresholds are an example of a thresholds compared to those of younger persons.
psychophysical measure relating the physical characteristics According to the ASHA (2005) guideline, the instruc-
of a tone to a behavioral threshold. tions should also include the response task (e.g., raise your
A psychophysical procedure describes the specific hand or finger, or press a button), the need to respond when
method used to obtain behavioral thresholds. The most the tone begins and to stop responding when it ends, and
common one used in puretone audiometry is a modified that the two ears are tested separately. Although not in the
method of limits. In the method of limits, the tester has ASHA guideline, instructions asking the examinee to indi-
control over the stimulus. A threshold search begins with cate which ear the sound is heard in may be useful. This is
the presentation of a tone at a particular frequency and especially important in cases of unilateral or asymmetrical
intensity that is often specified by the procedure. After each hearing losses where cross-hearing is possible.
presentation of the tone (or a short sequence of pulsed The examiner should present the instructions prior
tones), the tester judges whether or not the listener heard it to placement of earphones. Earphones attenuate external
based upon the listener’s response or lack of response. Each sounds making speech understanding more difficult, par-
response determines the subsequent dB-level presentation. ticularly for persons with hearing loss. Listeners should also
If a tone on a given presentation is not heard, the tone level be queried after the instructions are presented to determine
is raised. If a tone is heard, the level is lowered. The rules if they understood what was said. Sample instructions are
of the psychophysical procedure govern the amount of given below:
the level change following each response, when to stop the
You are going to hear a series of beeps, first, in one
threshold search, and the definition of threshold. The pro-
ear and then in the other ear. Respond to the beeps by
cedure, which is described in detail in subsequent sections,
pressing the button [switch] when one comes on and
may be modified slightly based on the clinical population
release it as soon as it goes off. Some of the beeps will
(e.g., the age of the listener).
be very faint, so listen carefully and respond each time
you hear one. Do you have any questions?
COOPERATIVE LISTENERS AGE 5 YEARS
TO ADULT (EARPHONES) Earphone Placement
Guidelines for Manual Pure Tone Audiometry is a publication The earphones should be placed by the examiner. For
that describes a uniform method for measuring thresholds convenience, earphones are color coded; red and blue
36 SECTION I • Basic Tests and Procedures

correspond to the right and left ears, respectively. Prior to duration tone is recommended for manual presentation
placement of earphones, clients are asked to remove jewelry (ASHA, 2005). The duration is determined by the amount
such as earrings and glasses if they will interfere with the of time the interrupter switch is held down. Pulsed tones
placement of the earphone. This is particularly relevant for are achieved by selecting this option on the audiometer’s
supra-aural earphones. front panel. If pulsed tones are selected, then the audio-
For circumaural and supra-aural earphones, the dia- meter alternately presents the tone followed by a short silent
phragm of the earphone should be centered over the ear interval (typically 225 ms on followed by 225 ms off) for as
canal. The examiner should view each ear while the phone long as the interrupter switch is depressed. The minimum
is being placed. Immediately after placement, the headband duration for a single pulse of the tone is critical. Numerous
is tightened enough to make the earphone perpendicular to psychoacoustic studies have shown that tonal durations
the floor when the examinee is sitting upright. between roughly 200 ms and 1 second or more yield nearly
The first step in placement of insert earphones is to identical thresholds (Watson and Gengel, 1969). By con-
attach a spring-loaded clip that holds the transducer in trast, the same studies show that durations less than 200 ms
place to the examinee’s clothing. The clip can be attached to result in higher thresholds. For this reason, audiometers are
clothing near the shoulder (or behind a child’s neck) to keep designed to have a nominal pulse duration of 225 ms (ANSI,
the plug from being pulled out of the ear. In some newer 2010). Pulsed and manually presented tones presented from
implementations that combine middle-ear and otoacoustic audiometers that maintain tonal durations between 200 ms
emission measurements, the earphone is attached to a head- and 2 seconds yield nearly identical thresholds, as the psy-
band. For both types of support, the audiologist compresses choacoustic studies suggest. However, pulsed tones are pre-
the foam plug and inserts it into the ear canal so that its ferred for two reasons. Most patients prefer pulsed tones
outer edge lines up with the tragus. (Burk and Wiley, 2004), and pulsed tones also reduce the
number of presentations required to find threshold in per-
Placement of the Bone-conduction Vibrator sons with cochlear hearing loss who have tinnitus (Mineau
Although some recommend forehead placement (Dirks, and Schlauch, 1997). Apparently, pulsed tones help patients
1994), typically audiologists place the BC oscillator on the to distinguish the puretone signal from the continuous
most prominent part of the mastoid process. While hold- or slowly fluctuating noises generated from within their
ing the oscillator against the mastoid process with one hand, auditory system (tinnitus), thereby reducing false-positive
the headband is fit over the head to hold the oscillator in responses. False-positive responses can lengthen test time
place using the other hand. The oscillator surface should (Mineau and Schlauch, 1997), which is costly to an audiology
be set directly against the skin, not touching the pinna, and practice.
with no hair or as little hair as possible between the oscilla- Thresholds typically are obtained using a modified
tor and the skin. Some audiologists play a continuous low- Hughson–Westlake down-up procedure, which is a specific
frequency tone while moving the oscillator slightly side to implementation of a method-of-limits procedure (Carhart
side, asking the listener to report the location at which the and Jerger, 1959; Hughson and Westlake, 1944). The exam-
tone is the strongest. iner begins the threshold-finding procedure by presenting
a tone at 30 dB HL (ASHA, 2005). If the listener responds,
Audiometric Procedure for Threshold Measurement the level of the tone is decreased in 10-dB steps until the lis-
The ASHA Guideline (2005) recommends starting a thresh- tener no longer responds. If the listener does not respond to
old search from either well below threshold or using a supra- this initial 30-dB tone, the examiner raises the tone in 20-dB
threshold tone that familiarizes the participant with the steps until a response is obtained. After every response to a
stimulus. Most clinicians prefer the familiarization method. tone, the level of the tone is decreased in 10-dB steps until
For the familiarization approach, testing usually begins at there is no response. For subsequent presentations when
1,000 Hz at 30 dB HL unless prior knowledge of the exami- there is no response, the examiner raises the level of the tone
nee’s hearing suggests otherwise (ASHA, 2005). At 1,000 Hz, in 5-dB steps until a response is obtained. Following this
an examinee is more likely to have residual hearing than at a “down-10/up-5” rule, the tester continues until the thresh-
higher frequency, and test–retest reliability is excellent. Test- old is bracketed a few times, and a threshold estimate is
ing begins with an examinee’s self-reported better ear. If the obtained. ASHA (2005) recommends that threshold should
examinee believes both ears are identical, testing begins by correspond to the level at which responses were obtained
convention with the right ear. The better ear is tested first for two ascending runs, which is what most clinicians based
to provide a reference to know whether masking needs to their thresholds on even when the ASHA (1978)Guideline
be delivered to obtain a valid estimate of threshold for the recommended that thresholds be based on three ascending
poorer ear. runs. Research based on computer simulations of clinical
Tonal duration is an important factor in a puretone procedures (Marshall and Hanna, 1989) supports the cli-
test. On most audiometers, the option exists to select nician’s position and that of the ASHA (2005) guideline.
either pulsed or manual presentation. A 1- to 2-second The computer simulations of thresholds based on three
 CHAPTER 3 • Puretone Evaluation 37

ascending runs showed only a minimal reduction of the vari- TESTING CHILDREN YOUNGER THAN AGE
ability when compared to thresholds based on two ascend- 5 YEARS AND PERSONS WITH SPECIAL NEEDS
ing runs. Listeners who produce inconsistent responses are
an exception, and for these listeners, additional measure- For most children younger than age 5 years, audiologists
ments can be made to confirm the threshold estimate. have special procedures that they employ to measure PT
After a threshold is measured at 1,000 Hz, the next thresholds. Some of these same procedures are also appro-
frequencies that are examined depend on the goal, but the priate for persons older than 5 years who have cognitive def-
higher frequencies are typically tested prior to the lower icits. Chapter 24 on pediatric hearing assessment describes
frequencies. For diagnostic audiometry, thresholds are mea- these procedures and their interpretation.
sured at octave intervals between 250 and 8,000 Hz, along
with 3,000 and 6,000 Hz. Intra-octave thresholds between
500 and 2,000 Hz should be measured when thresholds dif-
Audiometric Interpretation
fer by 20 dB or more between two adjacent octaves. ASHA PT thresholds are displayed in tabular or graphical for-
(2005) also recommends that 1,000 Hz be tested twice as a mats. The tabular format is useful for recording the results
reliability check. Refer to the ASHA (2005) guidelines for of serial monitoring of thresholds, as in a hearing conser-
specifics about the recommended protocol and Chapter 6 vation program, but in many applications, thresholds are
for details about the use of masking noise to eliminate the plotted on an audiogram. ASHA (1990), in a publication
participation of the nontest ear. Masking noise is needed entitled Guidelines for Audiometric Symbols, suggests a stan-
whenever the threshold difference between ears is equal to dardized form for the audiogram. Although other formats
or exceeds the lowest possible values for interaural attenu- for plotting audiograms are acceptable, it is helpful to use
ation. For BC testing, masking is needed to verify results a standardized format for ease of interpretation across clin-
anytime an air–bone gap in the test ear of greater than ics. The audiogram consistent with that recommended in
10 dB is observed. For AC testing, masking is needed when the ASHA guidelines (1990) is shown in Figure 3.7 along
the difference between the AC threshold in the test ear with recommended symbols. This audiogram only covers
and the BC threshold of the nontest ear is greater than or the conventional frequencies. Thresholds for extended high
equal to 40 dB for supra-aural earphones, and considerably frequencies are plotted often in units of dB SPL, because
more for insert earphones, especially in the low frequencies average extended high-frequency thresholds vary over a
(Killion and Villchur, 1989). Specific recommendations for wide range with the age of the listener, making dB SPL a
insert earphones cannot be made until a study with a larger better reference than dB HL for comparing thresholds to
sample size is completed. norms for listeners of different ages. Conversion between

Key to symbols
RE LE RE LE
0 (no resp)

AC

20 AC
Masked

BC
Mastoid
dB hearing level

40
BC
Mastoid
masked
BC
60 Forehead
masked

Ear not specified

80 (resp) (no resp)

BC
Forehead

100 Sound field

BC
Mastoid

120
250 500 1,000 2,000 4,000 8,000
Frequency (Hz)
FIGURE 3.7 Recommended audiogram and symbols (ASHA, 1990) with a sensory/neural
hearing loss. RE and LE represent the right ear and left ear, respectively. The word “response”
is abbreviated “resp.”
38 SECTION I • Basic Tests and Procedures

TA B L E 3 . 2
0
Classification of Degree of Hearing Loss
Calculated from the Average of Thresholds 20
for 500, 1,000, and 2,000 Hza
Northern Jerger and 40

dB hearing level
Degree of and Downs Goodman Jerger
Loss (2002) (1965) (1980)
60
None <16 <26 <21
Slight 16–25
80
Mild 26–30 26–40 21–40
Moderate 30–50 41–55 41–60 Right Left
Moderately 56–70 100 Air
Bone
severe
Severe 51–70 71–90 61–80
120
Profound >70 >90 >80 250 500 1,000 2,000 4,000 8,000
a
Although all three references cited differ in the value accepted Frequency (Hz)
as a profound loss, a loss of 90 dB HL or more is widely accepted FIGURE 3.8 A bilateral conductive hearing loss. The
as representing a qualitative as well as a quantitative boundary
plotted values represent the average loss reported by
between hearing and deafness.
Fria et al. (1985) in a group of children with otitis media.

units of dB SPL and dB HL can be accomplished for three successful at achieving this goal, because (1) handicap is
different earphone models by consulting reference levels dependent on many factors related to an individual’s needs
published in ANSI (2010). and abilities, (2) only some of the speech frequencies are
Audiograms are often classified by categories based on assessed using this three-frequency average (speech frequen-
the degree of hearing loss. A number of authors have pub- cies range from 125 to 6,000 Hz), and (3) identical amounts
lished systems for classifying hearing loss based on the aver- of hearing loss sometimes result in large differences in the
age AC thresholds for three frequencies. The frequencies ability to understand speech and, as a consequence, the
used for this purpose are usually 500, 1,000, and 2,000 Hz, degree of disability associated with the loss. Despite these
often referred to as the three-frequency puretone average limitations, many audiologists use these categories routinely
(PTA). Table 3.2 shows the categories for the degree of loss to summarize the amount of loss in different frequency
based on this PTA for three different authors (Goodman, regions of an audiogram when describing results to other
1965; Jerger and Jerger, 1980; Northern and Downs, 2002). professionals or to a client during counseling.
The first category is normal hearing. Note that none of the Another factor in audiometric classification is the
three authors agree on the upper limit for normal, which type of hearing loss. The type of hearing loss is determined
ranges from 15 to 25 dB HL. Northern and Downs (2002) by comparing the amount of hearing loss for AC and BC
suggest using 15 dB HL as the upper limit for normal hear- thresholds at the same frequency. A sensory/neural hear-
ing for the three-frequency PTA for children between 2 ing loss has an equal amount of loss for AC and BC thresh-
and 18 years of age and a higher limit for adults. A 15 dB olds (as shown in Figure 3.7). By contrast, a conductive loss
HL upper limit for normal hearing may produce a signifi- has lower BC thresholds than AC thresholds (as shown in
cant number of false positives when applied to thresholds Figure 3.8). Conductive-loss magnitude is described by the
for individual audiometric frequencies, even in children decibel difference between AC and BC thresholds. This dif-
(Schlauch and Carney, 2012). Regardless of the value used ference is known as the air–bone gap, a value that has a max-
as an upper limit for normal hearing, keep in mind that an imum of about 65 dB† (Rosowski and Relkin, 2001). Due
ear-related medical problem can still exist even though all to test–retest differences, an air–bone gap needs to exceed
thresholds fall within the defined normal range. For exam- 10 dB before it is considered significant. A mixed hearing
ple, the presence of a significant air–bone gap might indi- loss shows a conductive component and a sensory/neural
cate the presence of middle-ear pathology even though all component. In other words, a mixed loss has an air–bone
AC thresholds fall within normal limits.
The original intent of classification system for severity
of loss based on a three-frequency PTA was to express, in †
Physiologic models suggest that the maximum air–bone gap occurs
a general way, the degree of handicap associated with the when there is an intact tympanic membrane and a disarticulated ossic-
magnitude of the loss. These categories are only somewhat ular chain (Rosowski and Relkin, 2001).
 CHAPTER 3 • Puretone Evaluation 39

TA B LE 3 . 3
0 Right Left
Air Criteria for Classifying Audiometric
Bone
20 Configurations
Term Description
dB hearing level

40 Flat <5 dB rise or fall per octave

Gradually falling 5–12 dB increase per octave
60 Sharply falling 15–20 dB increase per octave
Precipitously Flat or gradually sloping, then
falling threshold increasing at 25 dB
80
or more per octave
Rising >5 dB decrease in threshold per
100 octave
Peaked or saucer 20 dB or greater loss at the
extreme frequencies, but not
120
250 500 1,000 2,000 4,000 8,000 at the mid frequencies
Frequency (Hz) Trough 20 dB or greater loss in the mid
frequencies (1,000–2,000 Hz),
FIGURE 3.9 A mixed hearing loss. but not at the extreme fre-
quencies (500 or 4,000 Hz)
gap, and the thresholds for BC fall outside the range of Notched 20 dB or greater loss at one
normal hearing (Figure 3.9). frequency with complete or
Yet another way that audiograms are described is by near-complete recovery at
the hearing-loss configuration. The configuration takes adjacent octave frequencies
into account the shape of the hearing loss. A description of Modified from Carhart R. (1945) An improved method of classifying
the configuration of the loss helps in summarizing the loss audiograms. Laryngoscope. 5, 1–15 and Lloyd LL, Kaplan H. (1978)
to patients and to other professionals and often provides Audiometric Interpretation: A Manual for Basic Audiometry. Baltimore,
insight into the etiology or cause of the loss. Some typical MD: University Park Press.
shapes and the criteria used to describe them are shown in
Table 3.3.
An audiogram is summarized verbally by the degree, thresholds). The source of this variability is a combination
type, and configuration of the hearing loss for both ears. If a of variations in the person’s decision process, physiologic or
person has normal thresholds in one ear and a hearing loss bodily noise, a shift in the response criterion, and differences
in the other ear, this is known as a unilateral hearing loss. in transducer placement. It is assumed that the equipment is
A loss in both ears is described as a bilateral hearing loss. calibrated correctly for successive tests and that the standard
Bilateral losses are described as symmetric (nearly equal is not in error (Margolis et al., 2013).
thresholds in both ears) or asymmetric. The inherent variability of PT thresholds poses a prob-
lem for audiologists who are faced with making clinical deci-
sions based on these responses. Audiologists frequently need
Some Limitations of Puretone to assess whether hearing has changed significantly since the
Testing last test, whether hearing is significantly better in one ear
than the other, and whether an air–bone gap is significant.
TEST–RETEST RELIABILITY A good place to begin with understanding test–retest
PT thresholds are not entirely precise. Consider a coopera- variability is to consider the standard deviation (SD) of
tive adult whose AC thresholds are measured twice at octave test–retest differences at a single frequency. When a 5-dB
intervals between 250 and 8,000 Hz. For these two measures, SD is assumed, threshold differences on retest of 15 dB or
assume too that the earphones are removed and replaced more are rarely expected if only a single threshold measure-
between tests. For this situation, the probability of obtaining ment is retested. By contrast, when complete audiograms
identical thresholds at each frequency is small. This is due to are assessed, the likelihood of obtaining a large threshold
test–retest variability. Test–retest variability is also responsible difference at one frequency on retest increases. For exam-
for BC thresholds not always lining up with AC thresholds ple, 15 dB or greater differences on retest are expected only
in persons with pure sensory/neural losses. As reported by 1.24% of the time when the threshold for a single frequency
Studebaker (1967), test–retest variability causes false air–bone is assessed. When thresholds for six frequencies are assessed
gaps and false bone–air gaps (BC thresholds poorer than AC in each ear (octave intervals between 0.25 and 8 kHz), 14%
40 SECTION I • Basic Tests and Procedures

of the persons tested would be expected to have at least one

threshold differing by 15 dB or more (Schlauch and Carney, 0
Air
2007). Thus, differences of 15 dB or more in these appli- Bone
cations would be much more commonplace than those 20
predicted by the SD of inter-test differences for a single
frequency. 40

dB hearing level
Several methods have been proposed to assess the sig-
nificance of threshold differences on retest for complete 60
audiograms (Schlauch and Carney, 2007). These methods
usually require that thresholds for more than one frequency 80
contribute to the decision process, although some accept a
large change for a single frequency, such as 20 dB or more, 100
as a significant difference. One of these methods defines a
significant threshold shift by a minimal change in a PTA.
120
For instance, the Occupational Safety and Health Adminis-
tration (1983) defines a notable threshold shift (in their ter-
140
minology, a standard threshold shift) as a 10-dB or greater 250 500 1,000 2,000 4,000 8,000
change in the PTA based on thresholds for 2, 3, and 4 kHz Frequency (Hz)
in either ear. These frequencies were selected because they
include those that are susceptible to damage by occupa- FIGURE 3.10 Mean vibrotactile thresholds for bone
conduction (dashed line) and air conduction (solid line).
tional noise and have stable test–retest reliability. A second
The range of responses is indicated by the shaded
commonly used approach requires threshold differences to region. (Adapted from Boothroyd and Cawkwell (1970)).
occur at adjacent frequencies. One rule that is applicable
to many situations defines a significant threshold shift as
one for which two adjacent thresholds differ by 10 dB or relatively low vibrotactile thresholds are observed for BC
more on retest. This criterion has been applied widely in at 250 and 500 Hz, a false air–bone gap is likely to occur
audiometric studies and is sometimes combined with other in persons with significant sensory/neural losses at these
criteria to arrive at a decision (ASHA, 1994). A third approach frequencies.
recommends repeating threshold measurements during a Boothroyd and Cawkwell (1970) recommend asking the
single session to improve audiometric reliability (National client if they “feel” the stimulus or “hear” the stimulus as a
Institute for Occupational Safety and Health, 1998). This means to differentiate between these two outcomes. Persons
method is paired with a rule or rules defining the crite- with experience with auditory sensations can usually make
rion for a significant threshold shift. The notable difference this distinction.
between this method and the others described earlier is that The values for vibrotactile thresholds illustrated in
the criterion defining a threshold shift must be repeatable to Figure 3.10 are based on only nine listeners. A more detailed
be accepted as significant. study needs to be conducted to specify these ranges more
The examples in this section on the variability of PT precisely for the transducers in current use.
thresholds have assumed a fixed SD of test–retest differences
of ±5 dB for all audiometric frequencies. Although 5 dB is a BONE-CONDUCTION THRESHOLDS:
reasonable average value for many situations, studies show NOT A PURE ESTIMATE OF
that the SD varies with type of earphone, the time between
tests, and even with audiometric frequency (Schlauch and
SENSORY/NEURAL RESERVE
Carney, 2007). The goal of BC testing is to obtain an estimate of sensory/
neural reserve, but BC thresholds sometimes are influenced
by the physiologic properties of the external, middle, and
VIBROTACTILE THRESHOLDS inner ears. The BC vibrator sets the skull into vibration,
In persons with significant hearing losses, sound vibra- which stimulates the cochlea, but this does not happen in
tions produced by earphones and bone vibrators may be isolation. When the skull is vibrated, the middle-ear ossicles
perceived through the sense of touch. Such thresholds are are also set into motion, and this inertial response of the
known as vibrotactile thresholds. ossicular chain contributes to BC thresholds. Changes in
Figure 3.10 illustrates the range of levels found to yield the external and middle ear can modify the contribution
vibrotactile thresholds for a supra-aural earphone and a of the inertial response, which may result in significant
bone vibrator. A threshold occurring within the range of changes in BC thresholds (Dirks, 1994).
possible vibrotactile thresholds is ambiguous; it could be A classic example of a middle-ear problem that influ-
a hearing threshold or a vibrotactile threshold. Because ences BC thresholds is otosclerosis. Otosclerosis frequently
 CHAPTER 3 • Puretone Evaluation 41

causes the footplate of the stapes to become ankylosed or of “no trial” silence to confirm that their responses are, in
fixed in the oval window. This disease process and some fact, responses to test tones.
other types of conductive losses (e.g., glue ear) (Kumar et al., In rare cases, patients have tinnitus resulting from blood
2003) reduce the normal inertial response of the ossicles flowing nearby auditory structures. Blood flowing through
to BC hearing. The result is poorer thresholds that form a a vein or artery sometimes produces masking noise or
depressed region of BC hearing known as Carhart’s notch “bruit” that can elevate thresholds for low-frequency tones
(Carhart, 1950). This notch, which typically shows poorer (Champlin et al., 1990). On the audiogram, this form of
BC thresholds between 500 and 4,000 Hz with a maximum tinnitus may produce an apparent sensory/neural loss. The
usually at 2,000 Hz of 15 dB, disappears following successful loss occurs because the tinnitus masks AC and BC thresh-
middle-ear surgery. The finding that BC thresholds improve olds. Bruit, a recordable form of tinnitus resulting from
following middle-ear surgery is strong evidence that these vibrations in the head or neck, is documented by audiolo-
poorer BC thresholds observed in stapes immobilization gists by measuring sound levels in the ear canal (Champlin
are due to a middle-ear phenomenon rather than a change et al., 1990). This problem is treatable when the problem
in the integrity of the cochlea. is caused by a vein. In a case study reported by Champlin
A frequently observed example of middle-ear problems et al. (1990), the patient received some reduction in tinnitus
affecting BC thresholds occurs in persons with otitis media loudness before surgery by applying pressure to her neck.
with effusion. In this group, falsely enhanced BC thresh- Surgical ligation of the vein responsible for the tinnitus was
olds in the low frequencies (1,000 Hz and below) are seen shown to be an effective treatment. Surgery reduced tinnitus
often. The magnitude of the enhancement can be as much loudness, SPLs of the bruit measured in the ear canal were
as 25 dB (Snyder, 1989). Upon resolution of the middle-ear lower, and the audiogram showed significantly improved
problem, these previously enhanced BC thresholds become thresholds.
poorer and return to their premorbid values.
Similarly, enhancement in BC thresholds occurs for
low frequencies with occlusion of the external ear canal by
PSEUDOHYPACUSIS
a supra-aural ear phone. This low-frequency BC enhance- Pseudohypacusis, also known as functional hearing loss and
ment, known as the occlusion effect, must be considered nonorganic hearing loss, is the name applied to intra-test
when occluding the nontest ear to present masking noise and inter-test inconsistencies that cannot be explained by
during BC testing. However, when the masking noise is medical examinations or a known physiologic condition
presented using an insert earphone with the foam plug (Ventry and Chaiklin, 1965). Most persons who present
inserted deeply into the ear canal, the amount of the low- with this condition are feigning a hearing loss for monetary
frequency enhancement is smaller than it is when supra- or psychological gain, but a very small percentage of per-
aural earphones are used to deliver the masking noise (Dean sons have subconscious motivations related to psychologi-
and Martin, 2000). Further, apparent enhancement of BC cal problems (see Chapter 33).
thresholds can occur in cases of superior canal dehiscence Persons presenting with pseudohypacusis are often
(see Chapter 4). identified from inconsistencies in their responses to the
puretones. In addition to general poor reliability during
threshold searches, there is a tendency for the threshold
Special Populations to become poorer as more presentations are made (Green,
1978). Methods of identifying the pseudohypacusis by com-
TINNITUS paring PT thresholds with other measures and the use of
Many people who come for hearing testing experience tinni- special tests are covered in Chapter 33.
tus, the sensation of hearing internal sounds when no sound
is present (see Chapter 35). Tinnitus can interfere with the
AUDITORY NEUROPATHY
perception of test tones, which can lead to a large number
of false-positive responses, and false-positive responses can Auditory neuropathy (or auditory dys-synchrony) is a con-
produce an inaccurate (too sensitive) threshold estimation. dition that may account for 11% of hearing losses found in
Some listeners simply require additional instruction and children at risk for hearing loss (Rance et al., 1999). Infor-
encouragement to wait until they are more certain they have mation about this disorder may be found in Chapters 13
heard a test tone. In some cases, the audiologist can pres- and 19. Many of these children appear to be severely hard of
ent a clearly audible tone at the test frequency to remind hearing because of very poor speech recognition; however,
the listener of the test tone. For more intractable cases, the PT thresholds do not follow any specific pattern. Puretone
examiner can present a series of pulsed tones and ask the hearing thresholds for these children range from minimal to
listener to count the number of tones. It is important with profound losses. Individuals with auditory neuropathy clas-
listeners who are giving false-positive responses to avoid a sically show very inconsistent audiometric responses during
fixed presentation rhythm and to provide irregular intervals a test and between tests.
 42 SECTION I • Basic Tests and Procedures

on the outside of the auditory nerve code high frequen-

0 cies, the hearing loss is associated with the high frequencies
(Schlauch et al., 1995). Studies have shown that a screening
20 test that compares the average threshold difference between
ears for 1, 2, 4, and 8 kHz is most effective (Schlauch et al.,
1995). Threshold differences between ears for this PTA that
dB hearing level

40 exceed 15 dB or 20 dB maximize identification of persons

with these tumors while minimizing false-positive diagnoses
60 of persons with cochlear losses. The pass–fail criterion
Male age Symbol (e.g., requiring a 20-dB difference between ears) may dif-
30
fer depending on the money available for follow-up tests.
80 40
50 A pass–fail criterion requiring 15-dB or greater differences
60 between ears identifies more tumors than one requir-
100 70 ing 20-dB or larger differences, but the smaller difference
80 also yields more false-positive responses. False-positive
responses (in this case, persons with cochlear losses iden-
120 tified incorrectly as having tumors) place a burden on the
250 500 1,000 2,000 4,000 8,000
Frequency (Hz) healthcare system, because follow-up tests such as MRI or
auditory-evoked potentials are expensive.
FIGURE 3.11 Average audiograms for adult males for The effectiveness of a screening test based on the thresh-
different decades of life. Data from National Institute on
old asymmetries between ears is dependent on the clinical
Deafness and Other Communication Disorders (2005).
population. This test was found to be ineffective in a Veterans
Administration hospital where many patients are males who
have presbycusis and noise-induced hearing loss (NIHL)
AGING (Schlauch et al., 1995). By contrast, preliminary data from
Presbycusis is a term that describes the gradual loss of hear- young women with normal hearing in their better ear sug-
ing sensitivity that occurs in most individuals as they grow gest that true-positive rates and false-positive rates for this
older. Studies suggest (Schuknecht, 1974; Dubno et al., test are comparable to those for auditory brainstem response
2013) that several different types of damage can occur to the (Schlauch et al., 1995). It should also be noted that a small
auditory system because of aging. Hearing loss due to aging percentage of persons (<3%) with acoustic tumors have no
typically causes a gently sloping, high-frequency sensory/ hearing loss or hearing threshold asymmetry (Magdziarz
neural hearing loss that tends to be slightly greater in men et al., 2000).
than in women. Figure 3.11 shows the average amount of
threshold elevation expected based on aging in men who
have had limited exposure to intense sounds. Even among
MÉNIÈRE’S DISEASE
this select group of participants, large individual differences Ménière’s disease is diagnosed based on the symptoms of
are often observed. sensory/neural hearing loss, vertigo, tinnitus, and aural full-
ness (Committee on Hearing and Equilibrium, 1995) as
well as the exclusion of other known diseases. Adding to the
ACOUSTIC TUMORS
diagnostic challenge, the four symptoms do not occur all at
An acoustic tumor (acoustic neuroma/neurinoma or ves- once, and some of them may occur only during the inter-
tibular schwannoma) is a rare disorder. Once identified, mittent attacks that characterize this disease. It takes, on
these tumors are usually removed surgically, because they average, 1 year after the first symptom occurs before all of
can compress the brainstem and threaten life. Early diag- the symptoms are experienced by a person stricken with this
nosis and removal lessen the risk of complications during disease. Ménière’s disease rarely occurs before age 20 and
surgery and increase the opportunity to preserve hearing if is most likely to begin between the fourth and sixth decades
that approach is pursued. (Pfaltz and Matefi, 1981).
Magnetic resonance imaging (MRI) is the definitive test Ménière’s disease usually begins as a unilateral sensory/
for acoustic tumors. Unfortunately, it is expensive and only neural hearing loss, but the frequency of bilateral involve-
becomes cost effective when a screening test is used to assess ment increases with disease duration (Stahle and Klockhoff,
which patients should receive an MRI. Puretone audiometry 1986). Although audiometric configuration is not too help-
should be considered as part of that screening procedure. ful in diagnosing Ménière’s disease, a peaked audiogram
When the auditory nerve is compressed by the tumor, it (described in Table 3.3) is most common (roughly 60%
often, but not always (Magdziarz et al., 2000), results in a of involved ears), and a rising audiogram is also seen quite
unilateral or asymmetrical hearing loss. Because the fibers frequently, especially in the earliest stages of the disease.
 CHAPTER 3 • Puretone Evaluation 43

However, the peaked audiogram is also seen in 13% of ears ments that might preserve hearing. Ototoxic drugs typically
with acoustic tumors (Ries et al., 1998). cause reduction in high-frequency hearing before having
any adverse effect on hearing for the speech range. For this
reason, extended high-frequency hearing testing is recom-
NOISE-INDUCED HEARING LOSS AND
mended for ototoxic monitoring test protocols. Several
ACOUSTIC TRAUMA studies have demonstrated the effectiveness of early identifi-
Exposure to intense sound levels can cause permanent or cation of ototoxic hearing loss by monitoring thresholds for
temporary hearing loss due to hair cell damage. When a nar- frequencies higher than 8,000 Hz (Fausti et al., 1992). How-
rowband sound is presented at a level high enough to result ever, for ototoxic drugs that selectively damage inner hair
in damage, a loss occurs at a frequency roughly one-half cells in the cochlea (e.g., carboplatin), PT thresholds may
octave above the frequency of exposure (Henderson and be unaffected even though extensive damage has occurred
Hamernik, 1995). Most people who are exposed to damag- (Lobaranis et al., 2013).
ing noise levels in their work or recreational endeavors are
exposed to broadband sounds, but their losses, especially OTITIS MEDIA
during early stages of NIHL, are characterized by a “notch”
(a drop in hearing) on the audiogram. The greatest hear- Young children are susceptible to temporary, recurring
ing loss typically occurs in the region of 3,000 to 6,000 Hz. middle-ear inflammations (otitis media) that are often
The susceptibility of these frequencies is a result of sound accompanied by fluid in the middle ear (effusion). Otitis
amplification by the external ear (Gerhardt et al., 1987). The media, often referred to as a middle-ear “infection,” may
amplification is mainly a result of the ear canal resonance, be viral or bacterial but is most often serous (noninfected
which increases the level of sound by 20 dB or more. Tem- fluid). Otitis media is the most common medical diagnosis
porary hearing loss is referred to as temporary threshold for children, accounting for 6 million office visits in 1990 for
shift (TTS), and permanent changes are referred to as per- children between the ages of 5 and 15 years (Stoll and Fink,
manent threshold shifts (PTS). 1996). Adults, too, may have otitis media with effusion,
The greater variability of thresholds at 6 and 8 kHz than although the prevalence decreases significantly with age
at other frequencies makes small noise notches associated (Fria et al., 1985). During the active infection, often lasting
with early NIHL difficult to identify. Some frequently used a month or more, a patient’s hearing loss may fluctuate,
rules for quantifying noise notches can produce high false- usually varying between 0 and 40 dB. The average degree
positive rates when decisions are based on a single audio- of hearing loss is approximately 25 dB. Figure 3.8, which
gram (Schlauch and Carney, 2011). Averaging multiple was used earlier in this chapter to illustrate an audiogram
audiograms improves diagnostic accuracy as does clearing for a conductive loss, shows an audiogram derived from the
the ear canals of all earwax, which can result in the appear- average thresholds from a group of children diagnosed with
ance of a high-frequency loss (Jin et al., 2013; Schlauch and otitis media.
Carney, 2012). NIHL can be slowly progressive, as listen-
ers are exposed to high sound levels over months and years TYMPANIC MEMBRANE PERFORATIONS
(Ward et al., 2000), or it can rapidly change, such as noise
trauma after a sudden explosion or impulsive sound (Kerr Tympanic membrane perforations are caused by trauma,
and Byrne, 1975; Orchik et al., 1987). The shooting of a rifle disease, or surgery. The diameter and location of perfora-
can result in a greater loss in the ear closest to the muzzle tion and the involvement of other middle-ear structures
of the gun. In right-handed persons, the left ear is exposed determine the amount of conductive hearing loss, if any.
directly to the muzzle, and the right ear is protected from For instance, a myringotomy and the placement of pressure-
the direct blast by the head. New evidence (Kujawa and equalization tubes represent a physician-induced perforation
Liberman, 2009) suggests that PT thresholds may return to that results in a minimal air–bone gap in successful surgeries.
near normal following noise exposure, whereas functional The measurement of AC thresholds in the presence of
auditory abilities may remain compromised due to the tympanic membrane perforations requires special consider-
noise exposure. (See Chapter 32 for NIHL.) ation. Figure 3.12 shows an audiogram obtained in a single
session from a school-age child who has a tympanic mem-
brane perforation in the left ear and a pressure-equalization
OTOTOXICITY tube in the right ear. Thresholds were measured twice in
Regular monitoring of PT thresholds is particularly impor- each ear, once with supra-aural earphones and again with
tant for patients who take drugs known to be ototoxic. For insert earphones. Note that the low-frequency thresholds
example, certain powerful antibiotics and cancer-fighting obtained from insert earphones were as much as 15 to 25 dB
drugs are known to cause cochlear and vestibular damage in poorer than the ones obtained with supra-aural earphones.
many patients. Monitoring hearing sensitivity during treat- This outcome is typical and is predicted because insert
ment could allow a physician to consider alternative treat- earphones are more susceptible to calibration problems
 44 SECTION I • Basic Tests and Procedures

inconsistent puretone responses or who may present with

0 pseudohypacusis (Schlauch et al., 1996).
Suprathreshold word-recognition performance is
20
assessed in most clinical settings by scoring a client’s ability
to repeat back a list of monosyllabic words. WRSs provide
a valid estimate of speech understanding ability (Wilson
dB hearing level

40 and Margolis, 1983) and quantification of the distortion, if

any, caused by sensory/neural hearing loss. WRSs are corre-
60 lated with puretone audiometric thresholds in persons with
cochlear losses (Pavlovic et al., 1986), but individuals’ scores
vary considerably depending on the type of damage to the
80 auditory system. If the words are presented at a level high
Left Right
TDH–49 enough to make the speech sounds audible (overcoming the
100 Insert attenuation caused by the loss), persons with mild cochlear
hearing loss are expected to have high WRSs, and those with
severe to profound losses are likely to have fairly low scores.
120
250 500 1,000 2,000 4,000 8,000 Dubno et al. (1995) and Yellin et al. (1989) have published
Frequency (Hz) tables relating WRSs and the average of PT thresholds for
500, 1,000, and 2,000 Hz for groups of persons with typical
FIGURE 3.12 Audiograms obtained with two types cochlear losses. WRSs that are abnormally low for a given
of earphones from the same child who had bilateral
PTA are associated with a variety of conditions including
perforations.
an acoustic tumor, multiple sclerosis, Ménière’s disease,
auditory neuropathy, and cochlear dead regions (Moore,
in the presence of perforations than are supra-aural ear- 2004), to name a few. When there are dead regions (areas
phones when calibration is based on coupler rather than of missing inner hair cells in the cochlea), PT thresholds
real ear measurements (Voss et al., 2000). The thresholds may appear artificially better than expected because of the
obtained using the supra-aural earphones are more accu- spread of energy along the cochlea. Healthier cochlear cells
rate in this instance and in any situations in which the effec- adjacent to the missing cells will elicit a response to pure-
tive volume of the ear canal is significantly larger than is tones presented at the dead region frequency.
typical.
Automated Audiometry
Relation between Puretone
Clinical researchers automated the measurement of routine
Thresholds and Speech Measures hearing thresholds to increase clinical efficiency (Rudmose,
PT thresholds are often compared with speech audiomet- 1963). Devices were developed for this purpose, and several
ric test results. The two most common comparisons are machines were manufactured and sold commercially. Some
with speech reception thresholds (SRT) and suprathreshold of these automated audiometers had the ability to vary
word-recognition scores (WRSs). (See Chapter 5 for a com- intensity and frequency during a hearing test.
prehensive review of speech audiometry.) The Bekesy audiometer is an automated audiometer
SRTs obtained using spondaic words (or spondees) that was a common piece of equipment in major clinical
agree well with PT thresholds for low frequencies. Spond- and research settings in the 1960s. In its routine application,
ees are easily recognized; listeners only need to recognize AC thresholds were assessed for interrupted tones and sus-
the vowels to identify these words correctly. Because of the tained tones for frequencies ranging from 100 to 10 kHz.
importance of the vowels at low intensities, spondee thresh- Frequencies were swept through the range over time, typi-
olds are found to agree closely with the average of PT thresh- cally at a rate of one octave per minute. The examinee
olds for 500 and 1,000 Hz (Carhart and Porter, 1971). In controlled the level of the sound by depressing a handheld
the event of a rising audiogram, better agreement between switch for as long as he or she heard a tone and released it
the spondee and PT thresholds is the average for 1,000 and when none was heard. The resulting brackets around thresh-
2,000 Hz. Spondee thresholds and a two-frequency PTA, as old were recorded on an audiogram. Patterns of responses
noted earlier, nearly always agree within ±10 dB in coop- for sustained tones and interrupted tones were found to
erative examinees. This agreement makes the threshold distinguish between different etiologies of hearing loss (see
for spondaic words an excellent check on the validity and Chapter 33 on pseudohypacusis). In recent years, the use of
reliability of the audiogram. This comparison is important Bekesy audiometry has decreased in medical settings, but it
for most children. It is also a valuable tool for assessing still has important applications in research, the military, and
the reliability of PT thresholds in adults who demonstrate in hearing conservation programs.
 CHAPTER 3 • Puretone Evaluation 45

Within the past few years, a new generation of automated testing, BC thresholds are measured. Although it would be
audiometers has been developed (Margolis et al., 2010). useful to have puretone BC thresholds prior to AC thresh-
The new automated audiometers are capable of measuring olds to know how much masking noise can be presented
masked AC and BC thresholds, as well as WRSs, with only a safely, this advantage is outweighed by the inconvenience
single placement of the earphones and BC oscillator. Bekesy of having to enter the booth multiple times to reposition
audiometry is still used along with some other automated the BC vibrator and earphones. Valid, masked AC thresh-
methods (Laroche and Hetu, 1997), including ones that olds can be obtained successfully from most clients before
implement the threshold-finding procedure used in manual obtaining BC thresholds.
puretone audiometry (Margolis et al., 2010). Computer- A few clinics begin with immittance testing, which usu-
based rules control the presentation of stimuli, examinee ally includes a tympanogram and acoustic reflex thresholds.
responses, and the plotting of thresholds. The goal is to If the case history does not indicate a middle-ear problem
automate threshold collection for routine cases, which will and these tests of middle-ear function are normal, then BC
free audiologists to perform more complex measures or to thresholds may not be performed, and the loss, if present, is
work with difficult-to-test populations. assumed to be a sensory/neural loss. A possible risk of this
strategy is that, in rare instances, persons with middle-ear
Calibration problems have normal immittance measures. In this situa-
tion, a conductive loss would be missed. This approach also
Clinical data require accurate stimulus specification, or the adds the expense of immittance testing for each client. Stud-
results are meaningless. When most persons think of cali- ies should be done using an evidence-based practice model
bration of audiometers, the obvious examples include the to determine whether the assessment of middle-ear status of
accuracy of puretone frequency and level. However, pure- each client using immittance or wideband reflectance (see
tone calibration involves much more, including an assess- Chapter 9) is justified. Another time-saving strategy might be
ment of attenuator linearity, harmonic distortion, rise and to measure BC thresholds at two frequencies, a low and a high
fall times, and more. Consult ANSI (2010) and Chapter 2 frequency, and if an air–bone gap is not observed, BC thresh-
on calibration in this book to learn more about this topic. olds are not measured for other frequencies. A low frequency,
such as 500 Hz, would assess stiffness-related middle-ear
Puretone Thresholds and the pathologies. A high frequency, such as 4,000 Hz, would iden-
tify mass-related middle-ear pathologies and collapsed canals.
Audiologic Test Battery Since this method requires placement of the BC vibrator, the
PT thresholds are measured on nearly everyone entering a amount of time actually saved would be limited.
diagnostic audiology clinic, but the test sequence and the Despite the observed variability, it seems that it is possible
extent of the measurements often differ across clinics. Most for audiologists to obtain important diagnostic information
of these differences in protocol are implemented to save test- about the degree, type, and configuration of hearing losses
ing time, which contributes to the cost of running a clinic. using a variety of valid, evidence-based puretone audiometric
ASHA’s guide to manual PT threshold audiometry (2005) methods. Although at first glance, the puretone test proce-
makes no recommendation concerning the puretone test dure may appear elementary, it is clear that well-informed test
sequence. In 2000, the Joint Audiology Committee on Prac- procedures using appropriate and calibrated test equipment
tice Algorithms and Standards recommended an algorithm provide a necessary part of the complete audiologic test
that listed puretone AC testing (with appropriate masking battery and form the basis for clinical decision making.
applied) followed by puretone BC testing with appropriate
masking. They acknowledged that the assessment process
may vary “based on patient need and the assessment setting.”
FOOD FOR THOUGHT
Furthermore, they stated that “decision-making . . . occurs(s) 1. Given the known test–retest variability of PT thresholds,
throughout this process.” what is the threat to the quality of a hearing conservation
Based on informal surveys of clinicians in a variety of program if the tester chooses not to make multiple esti-
settings, it seems that there is considerable variability in mates of baseline audiograms?
test protocols among clinics. In many clinics, BC thresh- 2. Think about examples of auditory pathology where the
olds are not usually obtained from persons with normal AC PT threshold might be misleading, when it might not
thresholds (near 0 dB HL) unless the case history or risk reflect the full nature of the underlying cochlear injury.
of middle-ear problems suggests otherwise. BC threshold 3. Consider several cases where the ear canal volume (the
testing is also omitted in some clinics for returning patients volume under the earphone) might significantly affect
with pure sensory/neural losses if their AC thresholds match the resulting PT threshold. Consider the client’s total vol-
those of the prior visit. A common alternative test sequence ume of the outer ear, perforations of the eardrum, and
is to begin with puretone AC thresholds followed by supra- possible occlusions in the ear canal. What effect will these
threshold word-recognition testing. After word-recognition have on the resulting thresholds?
46 SECTION I • Basic Tests and Procedures

KEY REFERENCES Gerhardt KJ, Rodriguez GP, Hepler EL, Moul ML. (1987) Ear canal
volume and variability in patterns of temporary threshold
A full list of references for this chapter can be shifts. Ear Hear. 8, 316–321.
Goodman A. (1965) Reference zero levels for pure tone audiom-
found at http://thePoint.lww.com. Below are the key refer-
eter. Am Speech Hear Assoc. 7, 262–263.
ences for this chapter.
Henderson D, Hamernik RP. (1995) Occupational Medicine: State
American National Standards Institute. (1999) Maximum Permissi- of the Art Reviews, Biologic Bases of Noise-Induced Hearing Loss.
ble Ambient Noise for Audiometric Test Rooms. ANSI S3.1–1999. Philadelphia, PA: Hanley & Belfus, Inc.
New York, NY: American National Standards Institute, Inc. Hughson W, Westlake H. (1944) Manual for program outline for
American National Standards Institute. (2010) Specifications rehabilitation of aural casualties both military and civilian.
for Audiometers. ANSI S3.6–2010. New York, NY: American Trans Am Acad Ophthalmol Otolaryngol. (suppl 48), 1–15.
National Standards Institute, Inc. Jerger J, Jerger S. (1980) Measurement of hearing in adults. In:
American Speech-Language Hearing Association. (1990) Guide- Paperella MM, Shumrick DA, eds. Otolaryngology. 2nd ed.
lines for audiometric symbols. ASHA. 32, 25–30. Philadelphia, PA: W.B. Saunders.
American Speech-Language-Hearing Association. (1994) Guide- Jin SH, Nelson PB, Schlauch RS, Carney E. (2013) Hearing conser-
lines for the audiologic management of individuals receiving vation program for marching band members: Risk for noise-
cochleotoxic drug therapy. ASHA. 36, 11–19. induced hearing loss? Am J Audiol. 22, 26–39.
American Speech-Language-Hearing Association. (2005) Guide- Joint Audiology Committee on Practice Algorithms and Stan-
lines for manual pure-tone threshold audiometry. Available dards. (2000) Clinical practice guidelines and statements.
online at: www.asha.org/policy/html/GL2005-00014.html. Audiol Today. Special Issue.
Boothroyd A, Cawkwell S. (1970) Vibrotactile thresholds in pure Killion MC, Villchur E. (1989) Comments on “Earphones in audi-
tone audiometry. Acta Otolaryngol. 69, 381–387. ometry.” J Acoust Soc Am. 85, 1775–1778.
Browning GG. (1987) Is there still a role for tuning-fork tests? Br Kujawa SG, Liberman MC. (2009) Adding insult to injury: cochlear
J Audiol. 21, 161–163. nerve degeneration after “temporary” noise-induced hearing
Carhart R. (1945) An improved method of classifying audiograms. loss. J Neurosci. 29 (45), 14077–14085.
Laryngoscope. 5, 1–15. Kumar M, Maheshwar S, Mahendran A, Oluwasamni A, Clayton
Carhart R. (1950) Clinical application of bone conduction. Arch MI. (2003) Could the presence of a Carhart notch predict
Otolaryngol. 51, 789–807. the presence of glue at myringotomy? Clin Otolaryngol. 28,
Carhart R, Jerger J. (1959) Preferred method for clinical determina- 183–186.
tion of pure-tone thresholds. J Speech Hear Disord. 24, 330–345. Laroche C, Hetu R. (1997) A study of the reliability of automatic
Carhart R, Porter LS. (1971) Audiometric configuration and audiometry by the frequency scanning method (AUDIOSCAN).
prediction of threshold for spondees. J Speech Hear Res. 14, Audiology. 36, 1–18.
486–495. Lloyd LL, Kaplan H. (1978) Audiometric Interpretation: A Manual
Champlin CA, Muller SP, Mitchell SA. (1990) Acoustic measure- for Basic Audiometry. Baltimore, MD: University Park Press.
ments of objective tinnitus. J Speech Hear Res. 33, 816–821. Lobarinas E, Salvi R, Ding D. (2013) Insensitivity of the audio-
Chandler JR. (1964) Partial occlusion of the external auditory gram to carboplatin induced inner hair cell loss in chinchil-
meatus: Its effect upon air and bone conduction hearing acuity. las. Hear Res. 302, 113–120. http://dx.doi.org/10.1016/j.jeares.
Laryngoscope. 74, 22–54. 2013.03.012
Committee on Hearing and Equilibrium. (1995) Committee on Magdziarz DD, Wiet RJ, Dinces EA, Adamiec LC. (2000) Normal
Hearing and Equilibrium guidelines for the diagnosis and eval- audiologic presentations in patients with acoustic neuroma:
uation of Ménière’s disease. Otolaryngol Head Neck Surg. 113, an evaluation using strict audiologic parameters. Otolaryngol
181–185. Head Neck Surg. 122, 157–162.
Dean MS, Martin FN. (2000) Insert earphone depth and the Margolis RH, Eikelbloom RH, Johnson C, Ginter SM, Swanepoel
occlusion effect. Am J Audiol. 9, 131–134. DW, Moore, BCJ. (2013) False air-bone gaps at 4 kHz in listen-
Dirks D. (1994) Bone-conduction thresholds testing. In: Katz J, ers with normal hearing and sensorineural hearing loss. Int J
ed. Handbook of Clinical Audiology. 4th ed. Baltimore, MD: Audiol. 52, 526–532. Early Online: 1–7.
William & Wilkins; pp 132–146. Margolis RH, Glasberg BR, Creeke S, Moore BCJ. (2010) AMTAS®:
Dubno JR, Eckert MA, Lee FS, Matthews LJ, Schmiedt RA. (2013) Automated method for testing auditory sensitivity: Validation
Classifying human audiometric phenotypes of age-related studies. Int J Audiol. 49 (3), 185–194.
hearing loss from animal models. J Assoc Res Otolaryngol. 14, Marshall L, Hanna TE. (1989) Evaluation of stopping rules for
687–701. audiological ascending test procedures using computer simu-
Dubno JR, Lee F, Klein A, Matthews L, Lam CF. (1995) Confidence lations. J Speech Hear Res. 32, 265–273.
limits for maximum word-recognition scores. J Speech Hear Marshall L, Jesteadt W. (1986) Comparison of pure-tone audibil-
Res. 38, 490–502. ity thresholds obtained with audiological and two-interval
Fausti SA, Henry JA, Schaffer HI, Olson DJ, Frey RH, McDonald forced-choice procedures. J Speech Hear Res. 29, 82–91.
WJ. (1992) High frequency audiometric monitoring for early Martin FN. (1994) Introduction to Audiology. 5th ed. Boston, MA:
detection of ototoxicity. J Infect Dis. 165, 1026–1032. Allyn and Bacon.
Fria TJ, Cantekin EI, Eichler JA. (1985) Hearing acuity of chil- Mineau SM, Schlauch RS. (1997) Threshold measurement for
dren with otitis media with effusion. Arch Otolaryngol. 111, patients with tinnitus: Pulsed or continuous tones. Am J
10–16. Audiol. 6, 52–56.
 CHAPTER 3 • Puretone Evaluation 47

Moore BC. (2004) Dead regions in the cochlea: Conceptual foun- Schlauch RS, Carney E. (2011) Are false-positive rates leading to an
dations, diagnosis, and clinical applications. Ear Hear. 25 (2), overestimation of noise-induced hearing loss? J Speech Lang
98–116. Hear Res. 54 (2), 679–692.
Morgan DE, Dirks DD, Bower DR. (1979) Suggested threshold Schlauch RS, Carney E. (2012) The challenge of detecting mini-
sound pressure levels for frequency-modulated (warble) tones mal hearing loss in audiometric surveys. Am J Audiol 21 (1),
in the sound field. J Speech Hear Disord. 44, 37–54. 106–119.
National Institute for Occupational Safety and Health. (1998) Cri- Schlauch RS, Levine S, Li Y, Haines S. (1995) Evaluating hearing
teria for a Recommended Standard: Occupational Noise Expo- threshold differences between ears as a screen for acoustic neu-
sure: Revised Criteria. Cincinnati, OH: National Institute for roma. J Speech Hear Res. 38, 1168–1175.
Occupational Safety and Health, US Department of Health Schuknecht HF. (1974) Pathology of the Ear. Cambridge, MA:
and Human Services Report; pp 98–126. Harvard University Press.
National Institute on Deafness and Other Communication Dis- Shaw EAG. (1974) The external ear. In: Kleidel WD, Neff WD,
orders. (2005) Presbycusis. Available online at: http://www. eds. Handbook of Sensory Physiology. Berlin: Springer;
nidcd.nih.gov/health/hearing/presbycusis.asp pp 455–490.
Northern JL, Downs MP. (2002) Hearing in Children. 5th ed. New Snyder JM. (1989) Audiometric correlations in otology. In: Cum-
York, NY: Lippincott Williams & Wilkins. mings CW, Fredrickson JM, Harker LS, et al., eds. Otolaryngology
Occupational Safety and Health Administration. (1983) Occu- Head and Neck Surgery: Update. St Louis, MO: Mosby.
pational noise exposure: hearing conservation amendment. Stahle J, Klockhoff I. (1986) Diagnostic procedures, differential
Occupational Safety and Health Administration, 29 CFR 1910. diagnosis, and general conclusions. In: Pfaltz CR, ed. Con-
95; 48 Federal Register, 9738–9785. troversial Aspects of Ménière’s Disease. New York, NY: Georg
Pavlovic CV, Studebaker GA, Sherbecoe RL. (1986) An articulation Thieme.
index based procedure for predicting the speech recognition Studebaker G. (1967) Intertest variability and the air-bone gap.
performance of hearing-impaired subjects. J Acoust Soc Am. J Speech Hear Disord. 32, 82–86.
80, 50–57. Ventry IM, Chaiklin JB. (1965) Multidisciplinary study of func-
Pfaltz CR, Matefi L. (1981) Ménière’s disease – or syndrome? tional hearing loss. J Audiol Res. 5, 179–272.
A critical review of diagnose criteria. In: Vosteen KH, Ventry IM, Chaiklin JB, Boyle WF. (1961) Collapse of the ear canal
Schuknecht H, Pfaltz CR, et al., eds. Ménière’s Disease. New during audiometry. Arch Otolaryngol. 73, 727–731.
York, NY: Thieme. Voss SE, Herman BS. (2005) How does the sound pressure gen-
Rance G, Beer DE, Cone-Wesson B, Shepherd RK, Dowell RC, erated by circumaural, supraaural and insert earphones differ
King AM, et al. (1999) Clinical findings for a group of infants for adult and infant ears. Ear Hear. 26, 636–650.
and young children with auditory neuropathy. Ear Hear. 20 Voss SE, Rosowski JJ, Merchant SN, Thornton AR, Shera CA,
(3), 238–252. Peake WT. (2000) Middle ear pathology can affect the ear-
Ries DT, Rickert M, Schlauch RS. (1998) The peaked audiometric canal sound pressure generated by audiologic earphones. Ear
configuration in Ménière’s disease: Disease related? J Speech Hear. 21, 265–274.
Lang Hear Res. 42, 829–843. Ward WD, Royster JD, Royster LH. (2000) Auditory and nonaudi-
Rosowski JJ, Relkin EM. (2001) Introduction to analysis of middle- tory effects of noise. In: Berger EH, Royster LH, Royster JD,
ear function. In: Jahn AF, Santos-Sacchi J, eds. Physiology of the Driscoll DP, Layne M, eds. The Noise Manual. 5th ed. Fairfax,
Ear. 2nd ed. San Diego, CA: Singular. VA: AIHA Press; pp. 123–147.
Scheperle A, Goodman SS, Neely S. (2011) Further assessment of Watson CS, Gengel RW. (1969) Signal duration and signal fre-
forward pressure level for in situ calibration. J Acoust Soc Am. quency in relation to auditory sensitivity. J Acoust Soc Am. 46,
130, 3882–3892. 989–997.
Schlauch RS, Arnce KD, Olson LM, Sanchez S, Doyle TN. (1996) Wilson RH, Margolis RH. (1983) Measurements of auditory
Identification of pseudohypacusis using speech recognition thresholds for speech stimuli. In: Konkle DF, Rintelmann WF,
thresholds. Ear Hear. 17, 229–236. eds. Principles of Speech Audiometry. Baltimore, MD: University
Schlauch RS, Carney E. (2007) A multinomial model for identify- Park Press.
ing significant pure-tone threshold shifts. J Speech Hear Res.
150, 1391–1403.
 C H A P T ER 4

Bone Conduction Evaluation

James R. Steiger

components. The diagnoses reviewed include, in order of

INTRODUCTION presentation:
Puretone threshold measurements are routinely carried
• normal-hearing sensitivity,
out in audiologic evaluations. By comparing air-conducted
• CHL from outer ear disorder (cerumen and osteoma
and bone-conducted thresholds, site-of-lesion information
examples),
can be obtained. Disorders of the outer or middle ears dis-
• CHL from middle ear disorder (ossicular fixation and
rupt the flow of energy from the earphone to the inner ear.
otitis media examples),
However, much of the energy that is conducted by a bone
• MHL,
vibrator on the skull bypasses the outer and middle ears and
• SNHL from hair cell and/or neuron damage (presbycusis
stimulates the inner ear essentially unimpeded. Therefore,
example),
this discrepancy favoring the bone-conducted threshold
• SNHL and pseudoSNHL from third-window disorders
(called the air–bone gap) indicates a mechanical or conduc-
(superior semicircular canal dehiscence [SSCD] and large
tive hearing loss (CHL). On the other hand when the inner
vestibular aqueduct examples), and
ear is impaired, both pathways from the earphone and the
• pseudoSNHL from intracranial hypertension (syringohy-
bone vibrator are impacted. This bioelectrical disturbance is
dromyelia example).
referred to as a sensory/neural hearing loss (SNHL). A hear-
ing loss that has both sensory/neural and conductive ele- This chapter concludes with a review of a few impor-
ments is called a mixed hearing loss (MHL). tant technical issues including vibrotactile responses, inter-
However, bone-conducted energy does not entirely aural attenuation, mastoid and forehead bone vibrator
bypass the outer ear and middle ear. Site-of-lesion diagno- placement, and air and bone conduction threshold vari-
ses will be more accurate with an understanding that occlu- ability. Throughout, readers should be aware that the chap-
sion or disorder of the outer ear, middle ear, and/or inner ter’s main focus is on persons with fully matured anatomy.
ear components may affect bone conduction thresholds too. Infants’ bone conduction hearing may differ from adults
An understanding of bone conduction hearing is essential because of immature temporal bones, outer ears, middle
for audiologists to accurately apportion hearing loss among ears, and/or neurons (Hulecki and Small, 2011).
the possible sites of lesion and to identify the etiologies of
hearing losses. EARLY WRITINGS ON BONE
CONDUCTION HEARING
CHAPTER OVERVIEW In the 1500s Italian physicians Giovanni Filippo Ingrassia,
In this chapter, a historical context and a few basic prin- Girolamo Cardano, and Hieronymus Capivacci were
ciples in bone conduction as tested with tuning fork tech- among the earliest known writers to describe bone con-
niques are first introduced. Then highlights of the appa- duction hearing (Feldmann, 1970). For instance, Capivacci
ratus and evaluation procedures used in routine bone recognized the diagnostic significance of bone conduction
conduction evaluation today are considered. An over- hearing to, as he described it, differentially diagnose dis-
view is provided of the outer ear, middle ear, and inner orders of the tympanic membrane (what we know today
ear components of bone conduction hearing, which lead as CHL) from disorders of the cochlear nerve (what we
to the main purpose of this chapter: a review of site-of- know today as SNHL). For a test signal, Capivacci used
lesion diagnoses based on an understanding of air and the vibrations from a stringed musical instrument called a
bone conduction hearing. Of course, not all possible zither. He attached a metal rod to the zither strings, and his
diagnoses can be covered in one chapter; rather, exam- hearing-impaired listeners held the other end of the metal
ples were selected of outer ear, middle ear, and inner ear rod with their teeth. If the tone was heard by bone con-
disorders to highlight their role in the bone conduction duction, he concluded that the cochlear nerve was intact

49
50 SECTION I • Basic Tests and Procedures

and the hearing loss was caused by a tympanic membrane the phenomenon in hearing-impaired listeners, and his
disorder blocking the pathway of air-conducted sound. In work was carried on by both Bonafant and Schmalz in
contrast, if the tone was not heard by bone conduction, the 1840s. It was Schmalz who first wrote extensively on
he concluded that the listener’s hearing loss was caused by the diagnostic implications of what would become known
a cochlear nerve disorder. as the Weber tuning fork test. Listeners with unilateral
CHL hear bone-conducted tones louder in the impaired
ear because of the occlusion effect, whereas listeners with
The Rinne Tuning Fork Test unilateral SNHL hear bone-conducted tones softer in
In 1855, Heinrich Adolf Rinne (1819 to 1868) described the impaired ear because of sensory/neural disorder. The
the tuning fork test that bears his name (Feldmann, 1970). Weber test is used to this day, with bone-conducted tones
Rinne noted that the intensity of air-conducted tones was typically applied to the forehead by either tuning forks or
greater than that of bone-conducted tones, owing to the bone vibrators.
relatively lesser density of air in contrast to the greater
density of bone. Most people, including normally hear-
ing listeners and listeners with SNHL, therefore hear air-
APPARATUS
conducted tones louder than bone-conducted tones. In Bone vibrators are transducers composed of diaphragms
contrast, listeners with CHL hear bone-conducted tones encased in plastic. During bone conduction evaluation
louder than air-conducted tones for two reasons. First, the circular vibrating plastic surface of the bone vibrator
listeners with CHL have outer ear occlusions or middle is held in contact with the patient’s skull by the tension
ear disorders that attenuate air-conducted tones. And of a metal band. The American National Standards Insti-
second, outer or middle ear disorder can effectively trap tute (ANSI, 2004) specifies bone vibrator surface diam-
bone-conducted tones that would otherwise radiate out eter, metal band tension, and the output characteristics
of the ear canal; thus occlusions effectively intensify of audiometers and transducers including bone vibra-
bone-conducted tones. This is the so-called occlusion tors. As stated above, the clinical apparatus is calibrated
effect. so that patients’ air and bone conduction thresholds
Rinne’s procedure was straightforward; listeners held may be compared to the same 0 dB HL normal-hearing
the tuning fork with their teeth (dense bone conduction threshold referent. Figure 4.1 shows a common B-71
transmission) allowing vibrations to attenuate until no lon- bone vibrator and also that same bone vibrator properly
ger audible. Then the still vibrating tuning fork was moved placed on the mastoid of a KEMAR manikin. Audiomet-
in front of the ear canal (less dense air conduction trans- ric evaluations are conducted in test rooms compliant
mission). If the tuning fork was audible by air conduction, with ANSI S3.1-1999 (ANSI, 2003). The standard speci-
the listener had normal hearing or an SNHL. If not audible fies maximum permissible ambient noise levels allow-
by air conduction, the listener had a CHL. Today the bone able for audiometric threshold testing as low as 0 dB HL,
conduction tuning fork placement is more likely to be on including when the ears are uncovered as during bone
the mastoid rather than the teeth. conduction threshold evaluations. A full discussion of
Audiologists to this day diagnose site of lesion by com- calibration can be found in Chapter 2 of this book.
paring the air and bone conduction thresholds of their
patients. Audiometers, however, are calibrated so that 0 dB
HL is a normal threshold referent for both air and bone
EVALUATION PROCEDURES
conduction; air and bone conduction thresholds are there- Common practice is to begin with the bone vibrator on the
fore similar for normally hearing listeners as well as listen- better hearing ear mastoid or the right mastoid if a better
ers with SNHL. For listeners with CHL, bone conduction hearing ear is not known or suspected. Evaluation guide-
thresholds are better than air conduction thresholds for the lines were published by the American Speech-Language-
reasons stated above. Hearing Association (ASHA) (2005), including a protocol
for tone presentation, patient response modes, and the defi-
nition of threshold. Bone conduction thresholds should be
Weber Tuning Fork Test tested at several frequencies and traditionally in this order:
In 1827, both the German physician C.T. Tourtual and the 1,000, 2,000, 3,000, and 4,000 Hz, a retest of 1,000, 500, and
English physicist Charles Wheatstone described lateraliza- 250 Hz (ASHA, 1990). Because bone-conducted signals
tion of bone-conducted stimuli to an occluded ear due to may reach either mastoid at similar intensities, contralateral
the occlusion effect (Feldmann, 1970). Wheatstone expe- ear masking may be necessary to obtain ear-specific bone
rienced the occlusion effect when manually occluding an conduction thresholds. A full discussion of threshold evalu-
ear while listening to a vibrating tuning fork in contact ation, masking procedures, and the symbols used for record-
with his skull. Tourtual used as his stimulus a pocket watch ing thresholds on an audiogram can be found in Chapters 3
held in his mouth. In the 1800s Heinrich Weber observed and 6 of this book.
 CHAPTER 4 • Bone Conduction Evaluation 51

FIGURE 4.1 B-71 bone vibrator and also that same bone
vibrator properly placed on the mastoid of a KEMAR manikin.
(From Vento B, Durrant JD. (2009) Assessing bone conduc-
tion thresholds in clinical practice. In: Katz J, Medwetsky L,
Burkard R, Hood L, eds. Handbook of Clinical Audiology.
Philadelphia, PA: Lippincott Williams and Wilkins, http://lww.
B
com by permission.)

Outer, Middle, and Inner Ear Tonndorf, 1968). Vibration of the mandible may also add
to the sound wave radiation into the ear canal (Stenfelt and
Components of Bone Conduction Goode, 2005). These sound waves propagate through the
Bone-conducted stimuli cause complex skull vibrations middle ear and finally to the inner ear; thus, the complex
in several directions and with several resonances and phenomenon of bone conduction hearing includes exploi-
antiresonances (Stenfelt and Goode, 2005), the details tation of the air conduction pathway.
of which are beyond the scope of this chapter. Bone- The outer ear component may play little role in the
conducted energy ultimately arrives at the cochlea by normal unoccluded ear canal, but its role is magnified by
various transmission routes, which in turn cause basilar the occlusion effect (Stenfelt and Goode, 2005). Normally
membrane-traveling waves to propagate from the stiffer the outer ear canal acts as a high-pass filter (Tonndorf,
basilar membrane base toward the more compliant basi- 1968), that is, high-frequency energy is passed into the
lar membrane apex as occurs for air conduction (Bekesy, middle ear whereas low-frequency energy escapes through
1960). Bone conduction transmission routes can be dis- the ear canal opening. Outer ear canal occlusion traps this
cussed in terms of outer ear, middle ear, and inner ear low-frequency energy and thereby enhances bone con-
components. duction hearing up to 20 dB in the lower audiometric test
frequencies (Stenfelt and Goode, 2005).
OUTER EAR COMPONENT OF BONE
CONDUCTION MIDDLE EAR COMPONENT OF BONE
The outer ear bone conduction component arises from
CONDUCTION
vibration of the bony and especially the cartilaginous walls The middle ear component of bone conduction is the iner-
of the outer ear canal that, in turn, causes sound waves to tial lag of the ossicles (Barany, 1938). Middle ear ossicles are
radiate into the outer ear canal (Stenfelt and Goode, 2005; not directly attached to the skull, but are instead suspended
52 SECTION I • Basic Tests and Procedures

by ligaments and tendons and attached at either end to the one side of the scala media and the greater fluid volume
elastic tympanic and oval window membranes. The ossicles of the scala vestibuli, vestibule, and semicircular canals on
are free to move out of phase with skull vibrations and will the other side of the scala media (Stenfelt and Goode, 2005;
do so because of inertia, much as coffee would lag and spill Tonndorf, 1968).
from a cup moved precipitously. Middle ear ossicles vibrate Inertia of the inner ear fluids contributes to inner
relative to the skull in a like manner as during air conduc- ear bone conduction hearing, especially below 1,000 Hz
tion hearing, and thus energy is propagated into the inner (Stenfelt and Goode, 2005). Cochlear fluids and windows
ear. The middle ear component occurs mainly at and above are free to vibrate out of phase with skull vibrations and
1,500 Hz and is especially significant near 2,000 Hz, the will do so because of inertia, compared earlier to coffee
approximate resonant frequency of the middle ear (Stenfelt that will lag and spill from a moving cup. Cochlear fluid
et al., 2003). Finally, some have proposed that bone conduc- movement, in turn, displaces the basilar membrane and
tion energy radiates from the walls of the middle ear into initiates traveling waves. The spiral lamina may also be
the middle ear space and sets the tympanic membrane into flexible and lag skull vibrations, thereby contributing to
vibration, but that hypothesis has been challenged (Stenfelt bone conduction hearing, especially at higher frequencies
and Goode, 2005). (Stenfelt et al., 2003).
Finally, bone conduction energy can travel in nonosse-
ous skull contents, such as the brain, membranes, and flu-
INNER EAR COMPONENT OF BONE ids, and reach the cochlea through the cochlear and/or ves-
CONDUCTION tibular aqueducts (de Jong et al., 2011; Stenfelt and Goode,
The inner ear component of bone conduction involves sev- 2005). The nonosseous skull contents route may contribute
eral contributing factors, including cochlear compression or little to normal bone conduction hearing, but it plays a role
distortion, cochlear fluid inertia, osseous spiral lamina iner- in some inner ear disorders as discussed below.
tia, and sound pressure transmission through skull contents
(brain, membranes, and fluid). Inner ear bone conduction
has been described as resulting from alternate compressions
TRANSMISSION ROUTE MODEL
and expansions (Herzog and Krainz, 1926) or distortions Figure 4.2 illustrates the air and bone conduction transmis-
(Tonndorf, 1968) of the bony cochlear capsule. In turn, sion routes. Note the orderly air conduction route through
cochlear fluids are displaced and basilar membrane-travel- the outer, middle, and inner ear in contrast with more
ing waves are initiated. One factor making cochlear fluid dis- complex bone conduction hearing involving concomitant
placement possible is the out-of-phase and disproportion- outer, middle, and inner ear components bilaterally.
ate yielding of the round and oval cochlear windows, which
creates alternating spaces for fluid displacement. Cochlear
fluid movement, in turn, displaces the basilar membrane EXAMPLES OF DIAGNOSES
and initiates traveling waves. Also, the cochlear and vestibu-
lar aqueducts may serve as outlets for fluid displacement
Normal-hearing Sensitivity
allowing for bone conduction hearing when the oval win- Normally hearing patients are without disorder that would
dow is fixed as in otosclerosis (Stenfelt and Goode, 2005; hinder energy from reaching the inner ear by the air conduc-
Tonndorf, 1968). Fluid displacement is further enabled by tion route. Owing to calibration, air conduction thresholds
a fluid volume differential between the scala tympani on will be near 0 dB HL. Similarly, normally hearing patients

Bone
vibrator
placement
Right
mastoid
Cerebrospinal
fluid

Earphone Right Right Right Left Left Left

placement outer middle inner inner middle outer
Right ear ear ear ear ear ear ear

FIGURE 4.2 Transmission routes for air conduction, right ear example (narrow arrows) and bone con-
duction, right mastoid example (bold arrows). Note: Higher intensity air-conducted signals can activate
the bone conduction transmission route.
 CHAPTER 4 • Bone Conduction Evaluation 53

0 0

20 20

40 40
dB hearing level

dB hearing level
60 60

80 80

100 100

120 120
250 500 1,000 2,000 4,000 8,000 250 500 1,000 2,000 4,000 8,000
Frequency (Hz) Frequency (Hz)

FIGURE 4.3 Audiogram depicting normal-hearing sen- FIGURE 4.4 Audiogram depicting CHL from osteomas.
sitivity. Air conduction: presurgery thresholds represented with
filled circles and postsurgery thresholds represented with
open circles. Bone conduction: presurgery thresholds
are without disorder that would hinder any of the bone connected with a dotted line, postsurgery thresholds
conduction components. Again owing to calibration, bone not connected. (Modified from Pinsker OT. (1972) Otologi-
conduction thresholds will be near 0 dB HL and be similar cal correlates of audiology. In: Katz J, ed. Handbook of
to air conduction thresholds (±10 dB). Figure 4.3 shows Clinical Audiology. Baltimore, MD: Williams and Wilkins,
an example audiogram of a patient with normal-hearing http://lww.com by permission.)
sensitivity.

neither hair cell nor neuron damage is the cause (Hall and
CHL with Air–Bone Gaps of Outer Croutch, 2008). In such cases, bone conduction thresholds
Ear Origin underestimate sensory/neural reserve. Figure 4.4 shows
Outer ear occlusive disorder may hinder air conduction an audiogram example involving occlusive osteomas; the
energy from reaching the inner ear. In such cases, air con- presurgery audiogram shows CHL, but with bone con-
duction thresholds would be poorer than 0 dB HL, at least at duction thresholds that underestimated the true sensory/
some frequencies, to a degree dictated by the occlusive disor- neural reserve revealed by postsurgery bone conduction
der. Bone conduction thresholds, in contrast, may be unaf- thresholds. Cerumen impaction that loads (adds mass to)
fected if the outer ear bone conduction component is not the tympanic membrane can cause pseudoSNHL (Hall and
hindered. Bone conduction thresholds would be near 0 dB Croutch, 2008; Tonndorf, 1968). Anderson and Barr (1971)
HL and accurately reflect sensory/neural hearing (sensory/ reported pseudoSNHL with partial cerumen occlusion of
neural reserve or true sensory/neural capability). The result the outer ear canal, though they attributed it to earphone
is a hallmark of CHL: air conduction thresholds >10 dB HL artifact.
lower (poorer) than normal bone conduction thresholds.
The maximum air–bone gap is approximately 60 dB; higher CHL with Air–Bone Gaps of Middle
intensity air-conducted sound waves set the skull into vibra-
tion and induce bone conduction hearing thus limiting the
Ear Origin
maximum difference between air and bone conduction Middle ear disorder may hinder air conduction energy
thresholds (Bekesy, 1960). from reaching the inner ear. In such cases, air conduction
However, bone conduction thresholds are often thresholds would be poorer than 0 dB HL, at least at some
improved by occlusive disorders due to the occlusion frequencies, to a degree dictated by the middle ear disor-
effect. In such cases, bone conduction thresholds over- der. Bone conduction thresholds, in contrast, may be unaf-
estimate sensory/neural reserve. In rarer cases, outer ear fected if the middle ear bone conduction component is not
occlusive disorder may interfere with the outer or middle significantly hindered. Bone conduction thresholds would
ear bone conduction components and thus lower bone therefore be near 0 dB HL and accurately reflect sensory/
conduction thresholds, a so-called pseudoSNHL because neural reserve or be improved due to the occlusion effect.
 54 SECTION I • Basic Tests and Procedures

0 0

20 20

40 40
dB hearing level

dB hearing level
60 60

80 80

100 100

120 120
250 500 1,000 2,000 4,000 8,000 250 500 1,000 2,000 4,000 8,000
Frequency (Hz) Frequency (Hz)

FIGURE 4.5 Audiogram representing the mean CHL FIGURE 4.6 Audiogram depicting CHL from otosclero-
from a group of children. (Modified from Schlauch RS, sis. Air conduction: presurgery thresholds represented
Nelson P. (2009) Puretone evaluation. In: Katz J, with filled circles and postsurgery thresholds repre-
Medwetsky L, Burkard R, Hood L, eds. Handbook of sented with open circles. Bone conduction: presurgery
Clinical Audiology. Philadelphia, PA: Lippincott Williams thresholds connected with a dotted line, postsurgery
and Wilkins, http://lww.com by permission.) thresholds not connected. (Modified from Dirks D. (1985).
Bone-conduction testing. In: Katz J, ed. Handbook of
Clinical Audiology. Baltimore, MD: Williams and Wilkins,
The resulting air–bone gaps can be as great as approxi- http://lww.com by permission.)
mately 60 dB. Air–bone gaps can be seen with many middle
ear disorders; an example of a CHL audiogram is shown in
Figure 4.5. media, and Kumar et al. (2003) reported 2,000 Hz bone
However, middle ear disorder often hinders the mid- conduction notches in patients with glue ear. Apparently,
dle ear contribution to bone conduction and thus lowers effusion may produce stiffening or loading effects, thus
(makes poorer) bone conduction thresholds. The result is hindering the middle ear bone conduction component
a pseudoSNHL in addition to the CHL, with bone conduc- (Tonndorf, 1968). An example of CHL caused by otitis
tion thresholds underestimating sensory/neural reserve. media is shown in Figure 4.7; note the pretreatment pseu-
For example, ossicular fixation caused by otosclerosis pre- doSNHL and CHL, and the resolution of the same post-
dictably manifests as lower (poorer) than normal bone con- treatment.
duction thresholds at the approximate middle ear resonant It should be noted here that true SNHL with otitis
frequency of 2,000 Hz (Carhart, 1950). The air–bone gaps media has also been proposed. Pathogens in middle ear effu-
one expects with CHL may therefore be reduced or oblit- sion may pass through the round window and cause dam-
erated at and near 2,000 Hz. For stapedial fixation caused age to cochlear hair cells, and because of the round window
by otosclerosis, this pseudoSNHL is known as the Carhart proximity to the basal turn of the cochlea, high-frequency
notch. An audiogram example is shown in Figure 4.6; note SNHL might result (Paparella et al., 1984).
the Carhart notch and CHL presurgery and the restoration
of normal air and bone conduction thresholds postsurgery
consistent with restoration to more normal middle ear
SNHL
resonance. Neither sensory nor neural disorders hinder energy from
The middle ear bone conduction component can be reaching the inner ear by the air conduction route or by
affected by other disorders as well. For example, Dirks any of the bone conduction routes. Changes in air and bone
and Malmquist (1969) reported pseudoSNHL in addi- conduction thresholds are affected only by the damage to
tion to CHL in a case of mallear fixation. Similarly, pseu- sensory/neural structures which will lower (make poorer)
doSNHL in addition to CHL has been reported for sub- air and bone conduction thresholds similarly. Air and bone
jects with otitis media (Carhart, 1950; Hall and Croutch, conduction thresholds will therefore be similar (±10 dB).
2008). Yasan (2007) reported 1,000 Hz and in some cases Figure 4.8 shows an example audiogram of a patient with
2,000 Hz bone conduction notches in patients with otitis SNHL from presbycusis.
 CHAPTER 4 • Bone Conduction Evaluation 55

0 0

20 20

40 40
dB hearing level

dB hearing level
60 60

80 80

100 100

120 120
250 500 1,000 2,000 4,000 8,000 250 500 1,000 2,000 4,000 8,000
Frequency (Hz) Frequency (Hz)

FIGURE 4.7 Audiograms depicting CHL from otitis FIGURE 4.8 Audiogram depicting a sensory/neural
media. Air conduction: pretreatment thresholds repre- hearing loss from presbycusis. (Modified from Harrell
sented with filled circles and post-treatment thresholds RW, Dirks D. (1994) In: Katz J, ed. Handbook of Clinical
represented with open circles. Bone conduction: pre- Audiology. Philadelphia, PA: Lippincott, Williams and
treatment thresholds connected with a dotted line, post- Wilkins, http://lww.com by permission.)
treatment thresholds not connected. (Modified from Hall
CM, Croutch C. (2008) Pseudosensory-neural hearing
loss. Hear Rev. 16(1), 18–22, by permission.) typically manifesting in the lower test frequencies con-
sistent with a mathematical model analysis based on the
anatomical dimensions of the inner ear (Merchant et al.,
2007). Bone conduction thresholds, in contrast, may be
MHL with Air–Bone Gaps of Outer improved by skull content sound pressure transmissions
or Middle Ear Origin
Patients may have an MHL. An MHL audiogram will there-
fore show evidence of both the SNHL (affected air and 0
bone conduction thresholds) and the CHL (air–bone gaps
>10 dB). Bone conduction thresholds may accurately reflect
sensory/neural reserve or a pseudoSNHL may underesti- 20
mate sensory/neural reserve. An MHL audiogram example
is shown in Figure 4.9. 40
dB hearing level

Superior Semicircular Canal 60

Dehiscence with PseudoSNHL and
Air–Bone Gaps of Inner Ear Origin 80
SSCD is a thinning or absence of the temporal bone over
the membranous labyrinth of the superior semicircular 100
canal. This condition opens a third elastic membranous
inner ear window at the dehiscence, the other two win-
dows of course being the oval and round windows of the 120
250 500 1,000 2,000 4,000 8,000
cochlea (Merchant et al., 2007). The audiogram manifes- Frequency (Hz)
tation of SSCD may mimic CHL or MHL, with air–bone FIGURE 4.9 Audiogram depicting an MHL. (Modified
gaps that could approach 60 dB (Chien et al., 2012). Air from Schlauch RS, Nelson P. (2009) Puretone evaluation.
conduction thresholds may be adversely affected because In Katz J, Medwetsky L, Burkard R, Hood L, eds. Hand-
energy reaching the inner ear by the air conduction route book of Clinical Audiology. Philadelphia, PA: Lippincott
is shunted away from the cochlea through the dehiscence, Williams and Wilkins, http://lww.com by permission.)
 56 SECTION I • Basic Tests and Procedures

0 0

20 20

40 40
dB hearing level

dB hearing level
60 60

80 80

100 100

120 120
250 500 1,000 2,000 4,000 8,000 250 500 1,000 2,000 4,000 8,000
Frequency (Hz) Frequency (Hz)

FIGURE 4.10 Audiogram depicting hearing loss from FIGURE 4.11 Audiogram depicting hearing loss from a
superior semicircular canal dehiscence. large vestibular aqueduct.

through the dehiscence, that is, augmentation of the inner Intracranial Hypertension with
ear bone conduction component. The term air–bone gap
of inner ear origin has been used to describe these findings
PseudoSNHL
(Attias et al., 2012). This is a useful term as it emphasizes Intracranial hypertension with abnormal cerebrospinal
that the air–bone gaps do not reflect either a CHL or MHL. fluid flow has been associated with a number of neurologic
Figure 4.10 shows an example audiogram from a patient conditions including syringohydromyelia, Chiari malfor-
with SSCD. mations, trauma, tumors, arachnoiditis, subarachnoid hem-
orrhages, meningitis, and multiple sclerosis (Steiger et al.,
2007). Resulting audiologic symptoms may include whoosh-
Large Vestibular Aqueducts ing pulsatile tinnitus and low-frequency pseudoSNHL. The
with SNHL and Air–Bone Gaps pulsatile tinnitus may arise from circle of Willis blood flow
of Inner Ear Origin or pulsations of the walls of the dural sinuses (Rudnick and
Sismanis, 2005), which travel through the cochlear or vestib-
Large vestibular aqueducts (LVA) cause SNHL with air–
ular aqueducts to the cochlea (Marchbanks et al., 2005). The
bone gaps of inner ear origin (Attias et al., 2012; Jackler and
pseudoSNHL might be attributable to masking from the
De La Cruz, 1989). The cause of SNHL with LVA is unclear;
pulsatile tinnitus (Rudnick and Sismanis, 2005; Steiger et al.,
it may result from traumatic endolymph pressure from the
2007) or from elevated cochlear fluid pressure stiffening
endolymphatic duct and sac that damages hair cells, or by
the basilar, oval, and round window membranes (Sismanis,
endolymph electrolyte content that is harmful to the hair
1987). Stiffened cochlear membranes, in turn, may interfere
cells or stria vascularis (Campbell et al., 2011; Jackler and
with cochlear fluid motion and thus hinder the inner ear
De La Cruz, 1989; Levinson et al., 1989). The audiomet-
bone conduction component. Figure 4.12 shows an audio-
ric findings for LVA may also be influenced by the third-
gram of a patient with intracranial hypertension.
window effect similar to SSCD. Air–bone gaps of inner ear
origin are possible: Air conduction thresholds may be low-
ered (made poorer) because some energy is shunted away TECHNICAL CLINICAL CAVEATS
from the cochlea through the LVA, whereas bone conduc-
tion thresholds may be unaffected or improved by sound
Vibrotactile Responses
pressure transmissions through the skull contents to the It is possible for a patient to feel bone conductor diaphragm
LVA, (Attias et al., 2012; Merchant et al., 2007). Therefore, vibrations during bone conduction evaluation, especially at
as with SSCD, the air–bone gaps seen with LVA do not high intensities and at lower test frequencies (Nober, 1964).
reflect outer ear occlusion or middle ear disorder as in When a patient responds to stimuli felt but not heard, the
CHL or MHL. Figure 4.11 shows an example audiogram responses are called vibrotactile. Vibrotactile responses must
from a patient with LVA. not be recorded as auditory thresholds as two possible errors
 CHAPTER 4 • Bone Conduction Evaluation 57

0
Interaural Attenuation, Masking,
and the Occlusion Effect
20
Audiologists typically target a test ear for evaluation while
being vigilant for the possibility of the patient hearing in
the nontest ear. This vigilance is warranted during bone-
40
dB hearing level

conduction threshold evaluation; transducer placement on

the test ear side mastoid bone results in activation of bone
60 conduction bilaterally. The so-called cross-over from the
test ear side to the nontest ear can occur with minimal inte-
raural attenuation, ranging from 0 dB at 250 Hz to 15 dB
80 at 4,000 Hz (Studebaker, 1967). Caution compels most
audiologists to assume the worst-case scenario of 0 dB inte-
100 raural attenuation, that is, equal tone intensity at the test
and nontest ears. Bone conduction thresholds are therefore
not typically considered to be ear-specific unless sufficient
120
250 500 1,000 2,000 4,000 8,000 masking noise is delivered to the nontest ear. There are
Frequency (Hz) many variables to consider while masking, such as when to
FIGURE 4.12 Audiogram depicting pseudoSNHL from mask, masking noise type, masker intensity, and accounting
intracranial hypertension. (Adapted from Steiger JR, for the occlusion effect caused by the earphone on the non-
Saccone PA, Watson KN. (2007) Assessment of objective test ear (see Chapter 6 for details).
pulsatile tinnitus in a patient with syringohydromyelia.
J Am Acad Audiol. 18(3), 197–206.) Used with permission
of the American Academy of Audiology.
Mastoid versus Forehead Placement
During bone conduction hearing evaluation, audiologists
may place the bone vibrator on either the mastoids or
might result. First, bone conduction vibrotactile responses the foreheads of their patients. Mastoid placement is pre-
could be better than the air conduction thresholds and ferred by most audiologists (Martin et al., 1998). Perhaps
therefore might result in erroneous air–bone gaps and mis- the main advantage of mastoid placement is that the result-
diagnoses. Second, recording vibrotactile responses as bone ing bone conduction thresholds are up to 14.7 dB less than
conduction thresholds might erroneously suggest hearing in bone conduction thresholds measured with forehead trans-
patients who are deaf (Nober, 1964). Individual sensitivity ducer placement (Table 4.1). This allows for a greater test-
to vibrotactile sounds is variable (Boothroyd and Cawkwell, ing range from threshold to equipment intensity limits or
1970). Perhaps the only way to know if responses are vibro- vibrotactile sensation. Moreover, vibrations from mastoid
tactile is to ask patients; this is recommended, especially bone vibrator placement are in the same plane as middle
when bone conduction thresholds appear inconsistent with ear ossicular motion, therefore engaging the middle ear
other audiometric findings or history. bone conduction mechanism. This allows the audiologist

TABL E 4 .1

Mean Differences between Bone Conduction Thresholds

Measured with Forehead and Mastoid Bone Vibrator Placementa
Frequency in Hz
250 500 1,000 2,000 3,000 4,000
Forehead–Mastoid Corrections in dB
ANSI S3.43-1992 12 14 8.5 11.5 12 8
Frank (1982) 14.3 14.7 8.7 12 12.4 13.5
ANSI, American National Standards Institute.
a
The correction should be subtracted from the forehead thresholds to approximate mastoid
thresholds.
Source: From Vento B, Durrant JD. (2009) In: Katz J, Medwetsky L, Burkard R, Hood L, eds.
Handbook of Clinical Audiology. Philadelphia, PA: Lippincott Williams and Wilkins,
http://lww.com by permission.
 58 SECTION I • Basic Tests and Procedures

to record evidence of changes in the middle ear bone con- Threshold Accuracy and the
duction component, for example, the middle ear resonance
changes that are likely to occur with otosclerosis. Not sur-
Air–Bone Gap
prisingly, audiometers in most clinics are calibrated for Throughout this chapter significant air–bone gaps were
mastoid placement. defined as >10 dB. However, patient response variability
Forehead placement can be used if correction factors can result in underestimated or exaggerated air–bone gaps
from Table 4.1 are applied or if the audiometer is calibrated and even bone–air threshold gaps. Studebaker (1967), for
for forehead transducer placement. Audiologists who prefer example, calculated the standard deviation of air–bone gaps
forehead transducer placement should consider purchas- at 5 dB and noted that air–bone threshold gaps of ≥15 dB
ing a specifically designed bone vibrator and headband. can sometimes be seen in the absence of CHL. Similarly,
Forehead placement has advantages, including low intra- Margolis (2008) calculated a hypothetical air–bone threshold
subject and intersubject variability because of the less vari- gap distribution based on the independent variability of air
able forehead placement surface and more uniform non- and bone conduction thresholds. Significant air–bone and
pneumatized forehead bone (Dirks, 1964). Also, vibrations even bone–air threshold gaps were predicted, of course, with
from forehead placement are perpendicular to middle ear greater threshold gaps occurring less frequently. Moreover,
ossicular motion and may not engage the middle ear bone Margolis reported apparent tester bias; when testing patients
conduction mechanism as with mastoid placement. The with SNHL an expert audiologist measured more air–bone
resulting forehead bone conduction thresholds should be threshold gaps ≤5 dB than the distribution predicted. Audi-
relatively unaffected by the changes in middle ear resonance ologists should not rigidly adhere to untenable expectations
and, in cases of ossicular fixation, reflect a truer measure of regarding air–bone threshold gaps.
cochlear reserve than bone conduction thresholds obtained
during mastoid transducer placement. Figure 4.13 shows an
example audiogram with forehead and mastoid bone vibra- CONCLUSION
tor placement for a patient with ossicular fixation. Bone conduction threshold evaluation is an integral com-
ponent of the basic audiologic examination. When bone
conduction thresholds are interpreted with an understand-
ing of air and bone conduction hearing, more accurate site-
0 of-lesion and etiology diagnoses can be made. It is hoped
that with this chapter the author has informed and moti-
20
vated readers to that end.

40 FOOD FOR THOUGHT

dB hearing level

1. How might the air–bone gaps of patients with outer ear

60 occlusion differ from the air–bone gaps of patients with
middle ear disorder?
2. Why are air–bone gaps usually but not always indicative
80
of CHL?
3. Why is worsening bone conduction hearing not always
100 indicative of hair cell and/or neural disorder?

120
250 500 1,000 2,000 4,000 8,000 KEY REFERENCES
Frequency (Hz)
A full list of references for this chapter can be
FIGURE 4.13 Audiogram depicting hearing loss from
found at http://thePoint.lww.com. Below are the key refer-
ossicular (mallear) fixation. Air conduction: presurgery
ences for this chapter.
thresholds represented with filled circles and postsur-
gery thresholds represented with open circles. Bone American National Standards Institute. (2003) Maximum Permis-
conduction: presurgery forehead placement thresholds sible Ambient Noise Levels for Audiometric Test Rooms (ANSI
connected with a dotted line, presurgery mastoid place- S3.1-1999; Rev. ed.). New York, NY: Author.
ment thresholds not connected. (Modified from Dirks D. American National Standards Institute. (2004) Specifications for
(1985) Bone-conduction testing. In Katz J, ed. Hand- Audiometers (ANSI S3.6-2004). New York, NY: Author.
book of Clinical Audiology. Baltimore, MD: Williams and American Speech-Language-Hearing Association. (1990) Guide-
Wilkins, http://lww.com by permission.) lines for Audiometric Symbols. Rockville, MD: Author.
 CHAPTER 4 • Bone Conduction Evaluation 59

American Speech-Language-Hearing Association. (2005) Guide- Kumar M, Maheshwar A, Mahendran S, Oluwasamni, Clayton
lines for Manual Pure-Tone Threshold Audiometry. Rockville, MI. (2003) Could the presence of a Carhart notch predict the
MD: Author. presence of glue at myringotomy? Clin Otolaryngol. 28(3),
Anderson H, Barr B. (1971) Conductive high-tone hearing loss. 183–186.
Arch Otolaryngol. 93(6), 599–605. Levinson MJ, Parisier SC, Jacobs M, Edelstein DR. (1989) The large
Attias J, Ulanovski D, Shemesh R, Kornreich L, Nageris B, Preis M, vestibular aqueduct syndrome in children: a review of 12 cases
Peled M, Efrati M, Raveh E. (2012) Air-bone gap component and the description of a new clinical entity. Arch Otolaryngol.
of inner-ear origin in audiograms of cochlear implant candi- 115, 54–58.
dates. Otol Neurotol. 33, 512–517. Marchbanks RJ, Burge DM, Martin AM, Bateman DE, Pickard J,
Barany E. (1938) A contribution to the physiology of bone con- Brightwell AP. (2005) The relationship between intracranial
duction. Acta Otolaryngol. (suppl 26), 1–4. pressure and tympanic membrane displacement. Br J Audiol.
Bekesy G. (1960) Experiments in Hearing. New York, NY: McGraw 24(2), 123–129.
Hill Book Co. Margolis RH. (2008) The vanishing air-bone gap: audiology’s
Boothroyd A, Cawkwell S. (1970) Vibrotactile thresholds in pure dirty little secret. Audiology Online. Available online at: http://
tone audiometry. Acta Otolaryngol. 69(1–6), 381–387. www.audiologyonline.com/articles/vanishing-air-bone-gap-
Campbell AP, Adunka OF, Zhou B, Qaqish BF, Buchman CA. (2011) audiology-901&referer=www.clickfind.com.au
Large vestibular aqueduct syndrome. Laryngoscope. 121, 352– Martin FN, Champlin CA, Chambers JA. (1998) Seventh survey of
357. audiometric practices in the United States. J Am Acad Audiol.
Carhart R. (1950) Clinical application of bone conduction audi- 9(2), 95–104.
ometry. Arch Otolaryngol. 51, 798–808. Merchant SN, Rosowski JJ, McKenna MJ. (2007) Superior semi-
Chien WW, Janky K, Minor LB, Carey JP. (2012) Superior semicir- circular canal dehiscence mimicking otosclerotic hearing loss.
cular canal dehiscence size: multivariate assessment of clinical Adv Otorhinolaryngol. 65, 137–145.
impact. Otol Neurotol. 33, 810–815. Nober EH. (1964) Pseudoauditory bone conduction thresholds.
de Jong M, Perez R, Adelman C, Chordekar S, Rubin M, Kirksunov J Speech Hear Disord. 29, 469–476.
L, Sohmer H. (2011) Experimental confirmation that vibra- Paparella MM, Morizono T, Le CT, Mancini F, Sipilä P, Choo YB,
tions at soft tissue conduction sites induce hearing by way of a Lidén G, Ki CS. (1984) Sensory-neural hearing loss in otitis
new mode of auditory stimulation. J Basic Clin Physiol Phar- media. Ann Otol Rhinol Laryngol. 93, 623–629.
macol. 22(3), 55–58. Rudnick E, Sismanis A. (2005) Pulsatile tinnitus and sponta-
Dirks D. (1964) Factors related to bone conduction reliability. Arch neous cerebrospinal fluid rhinorrhea: indicators of benign
Otolaryngol. 79, 551–558. intracranial hypertension syndrome. Otol Neurotol. 26(2),
Dirks D, Malmquist C. (1969) Comparison of frontal and mastoid 166–168.
bone conduction thresholds in various conduction lesions. Sismanis A. (1987) Otologic manifestations of benign intracranial
J Speech Hear Res. 12, 725–746. hypertension syndrome. Laryngoscope. 97(8, Pt 2, suppl 42),
Feldmann H. (1970) A history of audiology: a comprehensive 1–17.
report and bibliography from the earliest beginnings to the Steiger JR, Saccone P, Watson KN. (2007) Assessment of objective
present. In: Tonndorf J, ed. Translations of the Beltone Institute pulsatile tinnitus in a patient with syringohydromyelia. J Am
for Hearing Research. Chicago, IL: The Beltone Institute for Acad Audiol. 18(3), 197–206.
Hearing Research; pp 11–111. Stenfelt S, Goode RL. (2005) Bone conducted sound: physiological
Frank T. (1982) Forehead versus mastoid threshold differences and clinical aspects. Otol Neurotol. 26, 1245–1261.
with a circular tipped vibrator. Ear Hear. 3, 91–92. Stenfelt S, Puria S, Hate N, Goode RL. (2003) Basilar membrane
Hall CM, Croutch C. (2008) Pseudosensory-neural hearing loss. and osseous spiral lamina motion in human cadavers with air
Hear Rev. 16(1), 18–22. and bone conduction stimuli. Hear Res. 181, 131–143.
Herzog H, Krainz W. (1926) Das knochenleitungsproblem. Z Hals Studebaker GA. (1967) Clinical masking of the nontest ear. J Speech
Usw Heilk. 15, 300–306. Hear Disord. 32, 360–371.
Hulecki LR, Small SA. (2011) Behavioral bone conduction thresh- Tonndorf J. (1968) A new concept of bone conduction. Arch Oto-
olds for infants with normal hearing. J Am Acad Audiol. 22, laryngol. 87, 49–54.
81–92. Yasan H. (2007) Predictive role of Carhart’s notch in pre-operative
Jackler RK, De La Cruz A. (1989) The large vestibular aqueduct assessment for middle-ear surgery. J Laryngol Oto. 121, 219–
syndrome. Laryngoscope. 99(12), 1238–1243. 221.
 C H A P T ER 5

Speech Audiometry

Rachel McArdle and Theresa Hnath-Chisolm

speech recognition score (SRS). More recently, the signal-to-

INTRODUCTION noise ratio (S/N) at which 50% correct recognition is achieved
Auditory assessment using speech stimuli has a long history in has been recommended instead of the traditional SRS (Killion
the evaluation of hearing. As early as 1804, there were scien- et al., 2004; Wilson, 2003). Before discussing measurement of
tific attempts to study hearing sensitivity for speech by assess- speech thresholds and speech recognition in quiet and noise,
ing which classes of speech sounds an individual could hear: general considerations in speech audiometry related to termi-
(1) vowels; (2) voiced consonants; or (3) voiceless consonants. nology, stimulus calibration, presentation methods, response
In 1821, Itard, who is well known for his contributions to deaf modes, and presentation levels are discussed.
education, differentiated individuals who were hard of hearing
from those who were deaf by whether a person could under- SPEECH AUDIOMETRY
stand some or none of a spoken message (Feldmann, 1970).
This early focus on hearing for speech continued through the
TERMINOLOGY
19th century, and by the mid-1920s, the first speech audiom- There are two types of threshold measures using speech stim-
eter, the Western Electric 4 A, which incorporated a phono- uli: speech detection threshold (SDT) and speech recognition
graph with recorded digit speech stimuli, was employed in threshold (SRT). SDT, as defined by the American Speech-
large-scale hearing screenings (Feldmann, 1970). Language-Hearing Association (ASHA, 1988), is an estimate
Hearing and understanding speech have unique impor- of the level at which an individual perceives speech to be
tance in our lives. For children, the ability to hear and under- present 50% of the time and should be reported in decibels
stand speech is fundamental to the development of oral lan- hearing level (dB HL). SDTs are commonly used to establish
guage. For adults, difficulty in detecting and understanding the level for awareness of speech stimuli by infants, young
speech limits the ability to participate in the communication children, or adults who cannot respond verbally or whose
interactions that are the foundation of numerous activities speech recognition ability is so poor that they are unable to
of daily living. Measures of sensitivity and understanding recognize spondaic (i.e., compound) words to obtain an SRT.
form the basis of speech audiometry. This chapter focuses on SDT is sometimes called a speech awareness threshold (SAT),
providing information that can lead to the implementation although SDT is the term preferred by ASHA (1988).
of evidence-based best practices in speech audiometry. The SRT is an estimate of the level at which an individ-
ual can repeat back spondaic words (e.g., hotdog, baseball)
50% of the time; it is most commonly reported in dB HL or
WHAT IS SPEECH AUDIOMETRY? decibels sound pressure level (dB SPL). The most common
Speech audiometry refers to procedures that use speech suprathreshold measure in quiet is the SRS or word recog-
stimuli to assess auditory function (Konkle and Rintelmann, nition score and is generally measured in percent correct at
1983). Since the classic work of Carhart (1951), speech audi- a level (dB HL) relative to either the SRT or an average of
ometry has involved the assessment of sensitivity for speech puretone thresholds. Word recognition has been referred
as well as assessment of clarity when speech is heard. These to as speech discrimination; however, discrimination infers
concepts were described by Plomp (1978), in his framework that an individual is judging between two or more specific
of hearing loss, as an audibility component (i.e., loss of sen- stimuli, which is not the task in most suprathreshold speech
sitivity) and a distortion component (i.e., loss of clarity). The recognition measures.
audibility component is quantified through assessment of
speech recognition abilities in quiet. The distortion compo- GENERAL CONSIDERATIONS FOR
nent is a reduction in the ability to understand speech, espe-
cially in a background of noise, regardless of the presentation
SPEECH AUDIOMETRY
level. Quantifying the distortion component typically involves Audiometers have to meet calibration standards set forth by
percent correct recognition at suprathreshold levels for the the American National Standards Institute (ANSI, 2004). In

61
62 SECTION I
Ş
#BTJD
5FTUT
BOE
1SPDFEVSFT

addition, recorded materials used as stimuli for speech audi-

ometry must meet the ANSI standards (ANSI, 2004, Annex B). 100
To reduce error of measurement and increase consistency
from clinic to clinic, speech measures should employ accepted 80

Percent correct recognition

calibration procedures, methods and modes of presentation,
test instructions, and response modes.
60

.FUIPE
PǨ
1SFTFOUBUJPO 40

Historically, VU meters were used for the tester to “moni-

tor” the energy of his or her voice while presenting speech 20
stimuli via the speech audiometer. The development of
analog audiotape followed by compact disc technology was 0
instrumental in facilitating standardization of word lists
used in speech audiometry (Wilson et al., 1990). ASHA 30 40 50 60 70 80 90
guidelines (1988) for speech thresholds indicate that the use dB hearing level
of recorded stimuli is preferred. The majority of audiolo- FIGURE 5.1 Psychometric functions of word recogni-
gists, however, who responded to a survey of audiometric tion performance measured in percent correct (ordinate)
practices (Martin et al., 1998), still report using monitored for a listener with hearing loss as a function of presen-
live speech to determine thresholds for speech. Of the 218 tation level (abscissa). The dashed line indicates the
audiologists who completed the survey, 94% reported using 50% point. The function to the left is the SRT function
monitored live voice test methods. whereas the function to the right is the SRS function.
We feel that it is even more important to use recorded
speech for SRSs. Digitized speech recordings improve both which is a psychological variable) based on changes of an
the intrasubject and intersubject precision of threshold and independent variable (x-axis; e.g., presentation level in HL
suprathreshold measures by providing a consistent level or SNR, which is a physical variable). Figure 5.1 is a graphic
for all test items and consistent speech patterns between display of two psychometric functions. The function to the
patients. The reliability of a given set of speech stimuli can left is an SRT function whereas the function to the right is
vary across speakers and across test time for a single speaker. an SRS function. The characteristic audiogram thought to
Hood and Poole (1980) found that a speaker had a signifi- accompany this type of performance can be seen in the lower
cant impact on the difficulty of particular monosyllabic right quadrant. Presentation level is on the x-axis (dB HL),
word lists. Similarly, Roeser and Clark (2008) found sig- whereas percent correct performance is on the y-axis. As can
nificant differences in performance when the same subjects be seen for both functions, the percent correct is low when
were tested via recorded materials and monitored live voice the level is low, and as the level is increased, the percent cor-
with the latter showing better performance. Other stud- rect increases. The dashed line in Figure 5.1 highlights the
ies have found variability in recognition performance as 50% point on the functions and indicates that an SRT was
a function of speaker–list interactions (Asher, 1958; Hirsh obtained about 40 dB HL. Also illustrated in Figure 5.1 is that
et al., 1954) such that the acoustic waveforms of two speak- the maximum point of performance (100%) was reached at
ers can cause differences in recognition performance even approximately 75 dB HL for the SRS function. As the level
when the word lists are the same. The reported contribu- is increased above 75 dB HL, no change in performance is
tion of the speaker to the recognition performance of each observed. The highest percent correct score obtained by an
listener reinforces previous reports by Kruel et al. (1969), individual is often referred to as PBmax, because historically
who stated that word lists should be thought of as a group SRSs were obtained using phonetically balanced (PB) word
of utterances and not as a written list of words because lists. Further discussion of PB word lists can be found later
speaker differences may affect a person’s performance on in this chapter under the section titled “Speech Recognition
a particular list. in Quiet.”
Because listeners with normal hearing, on average,
achieve maximal performance on a speech recognition task
1SFTFOUBUJPO
-FWFM at 30 to 40 dB sensation level (SL) re: SRT, clinicians will
often test their patients at one of these levels, assuming this
PSYCHOMETRIC FUNCTION will result in maximal performance for the listener. Assessing
Understanding the influence of presentation level on per- only a single level may provide limited diagnostic or reha-
formance is best described by psychometric functions. In bilitative information. Conversely, assessing performance at
simple terms, a function is when you measure a change in a multiple presentation levels for individuals with sensory/
dependent variable (y-axis; e.g., number or percent correct, neural hearing loss provides greater diagnostic information
 CHAPTER 5
Ş
4QFFDI
"VEJPNFUSZ 63

Conductive
c
effects because the task difficulty is too great to show subtle
100 m
Normal changes in performance, whereas scores above 80% are often
1
2
affected by ceiling effects because the task difficulty is too
80 easy to be sensitive to performance changes. For an individ-
Cochlear
ual with a steep slope, the measurements should be made in
3
small (dB) steps to obtain valid results, whereas a shallow
60
function allows for larger step sizes to obtain valid results.
When selecting test material, it is best to choose stimuli
40 that produce a steep function, which suggests the materials
are homogeneous with respect to the task (Wilson and
Margolis, 1983).
Percent correct recognition

20

0 3FTQPOTF
.PEF
The response mode for speech audiometry is generally verbal.
100 However, for SDT the response mode can be similar to that
4 Retrocochlear
e a
of puretone thresholds, where patients can push a button or
80 raise their hand when they hear the speech stimuli. A written
response is generally avoided because of the increased test
Rollover time and reliance on the patient’s ability to write and spell.
60 5 For testing children or nonverbal individuals, see Chapters 24
and 31.
40

SPEECH RECOGNITION
20
6 THRESHOLD
0 4UJNVMJ
0 20 40 60 80 100 Spondaic words are generally used for obtaining SDTs and
Presentation level (dB HL) SRTs and are recommended by ASHA (1988). Spondaic
FIGURE 5.2 Psychometric functions of word recogni- (adjective) words or spondees (noun) are two-syllable words
tion performance illustrating various types of hearing with equal stress on both syllables. Lists of spondaic words
loss can be seen in both panels as a function of percent for assessing hearing loss for speech were first developed at
correct (ordinate) and presentation level (abscissa). The the Harvard Psychoacoustic Laboratories (PAL) by Hudgins
top panel illustrates a sample psychometric function for et al. (1947). Criteria for selection of words included a high
a listener with normal hearing (open circles), conduc- level of word familiarity, phonetic dissimilarity, and homo-
tive hearing loss (curve #1), and cochlear hearing loss geneity with respect to audibility. Of the original 42 spond-
(curves #2 and #3). The bottom panel shows possible ees identified by Hudgins et al. (1947), 36 of the most famil-
psychometric functions for retrocochlear hearing loss iar were used in the development of the Central Institute
(curves #4, #5, and #6). (Adapted from Department of
for the Deaf (CID) W-1 and W-2 tests (Hirsh et al., 1952).
Veterans Affairs (1997).)
Currently, ASHA (1988) recommends the use of 15 of the
original 36 spondees used in the CID W-1 and W-2 tests
as demonstrated by the example functions drawn in for obtaining SRTs. These 15 words, shown in Table 5.1, are
Figure 5.2. In the top panel of Figure 5.2, curve #2 shows the most homogeneous with respect to audibility (Young
a function that reaches maximum performance (88%) at et al., 1982), as is the list of 20 easily pictured spondees for
80 dB HL and plateaus through 100 dB HL. In the bottom use with children (Frank, 1980).
panel of Figure 5.2, curve #4 shows a function that reaches
maximum performance (85%) at approximately 60 dB HL,
and then performance decreases as level is increased, which
3FDPNNFOEFE
435
5FTUJOH
1SPUPDPM
is depicted by a rollover in the shape of the function. The SRT measurement involves four steps: (1) instructions;
Also of importance when describing performance in (2) familiarization; (3) initial and test phase for the descend-
terms of the psychometric function is the slope of the func- ing technique; and (4) calculation of threshold. Wilson
tion. The slope of the function is typically calculated from et al. (1973) described these steps, which were subsequently
the dynamic portion of the function that ranges between set forth by ASHA (1988) as a guideline for determining
20% and 80%. Scores below 20% are often affected by floor an SRT.
64 SECTION I
Ş
#BTJD
5FTUT
BOE
1SPDFEVSFT

TA B L E 5 .1 STEP 3: DETERMINATION OF THRESHOLD

a. Initial starting level—Present one spondaic word at a
Spondaic Words Recommended
level 30 to 40 dB HL above the anticipated SRT. If a cor-
by ASHA (1988)
rect response is received, drop the level in 10-dB steps

.PTU
)PNPHFOFPVT
until an incorrect response occurs. Once an incorrect
3F
"VEJCJMJUZ
:PVOH

.PTU
&BTZ
UP
1JDUVSF
response is received, present a second spondaic word at
FU
BM

'SBOL

the same level. If the second word is repeated correctly,
drop down by 10-dB steps until two words are missed
Toothbrush Toothbrush
at the same level. Once you reach the level where two
Inkwell Hotdog
Playground Baseball
spondees are missed, increase the level by 10 dB. This is
Sidewalk Airplane the starting level.
Railroad Cupcake b. Threshold estimation—Thresholds have been esti-
Woodwork Popcorn mated using 2- or 5-dB steps since most audiometers
Baseball Bathtub are equipped with those step sizes. Previous studies
Workshop Fire truck have shown that threshold differences as a function of
Doormat Football step size are too small to be clinically significant (Wilson
Grandson Mailman et al., 1973).
Eardrum Snowman 2-dB step size—Present two spondaic words at the start-
Northwest Ice cream ing level. Drop the level by 2 dB and present two spon-
Mousetrap Sailboat daic words. An individual should get the first five out
Drawbridge Flashlight of six words correct or else the starting level needs to
Padlock Bluebird be increased by 4 to 10 dB. If at least five of the first six
Toothpaste words are correct, continue dropping the level by 2 dB
Reindeer until the individual misses five of six presentations.
Shoelace 5-dB step size—Present five spondaic words at the start-
Seesaw ing level. An individual should get the first five spondaic
words correct at the starting level. Drop the level by 5 dB
and present five spondaic words. Continue dropping the
level by 5 dB until the individual misses all five spondaic
STEP 1: INSTRUCTIONS words at the same level.
Patients need to be instructed regarding what stimuli will
be used (i.e., spondaic words from the list) and how to STEP 4: CALCULATION OF THRESHOLD
respond during the testing procedure (i.e., written or ver-
bal response). Also, it is important to make patients aware Calculation of an SRT is based on the Spearman–Kärber
that the level of the stimulus will become quite soft and to equation (Finney, 1952). An SRT is calculated by subtract-
encourage them to guess throughout the testing procedure. ing the number of words repeated correctly from the starting
level and adding a correction factor of 1 dB when using the
2-dB step size and a correction factor of 2 dB when using the
STEP 2: FAMILIARIZATION 5-dB step size. For a 5-dB step example, with a starting level of
Each patient should be familiarized with the word list to 40 dB, the patient got all five words; at 35 dB, three of the
be used during the testing procedure by listening to the list words were correct; and at 30 dB, none were correct. Eight
of test words at a level that is easily audible and repeating of the 15 words were correct. Therefore, the SRT calculation
back each word as a demonstration of word recognition. If a would be 40 − 8 = 32, + 2 for the correction, equals 34 dB HL.
patient is unable to repeat back a particular spondaic word
from the test list, then that word should be removed from
the test list. Another method of familiarization is to give the
$MJOJDBM
'VODUJPOT
PǨ
435
patient a written list of the test words to read. The most recent surveys of audiometric practices in the
Previous research has shown differences in SRT values United States reported that 99.5% (Martin et al., 1998) and
obtained with and without familiarization (Conn et al., 1975; 83% (ASHA, 2000) use SRT as part of their basic audiologic
Tillman and Jerger, 1959). Specifically, Tillman and Jerger assessment. The reasons stated for using SRT were (1) cross
(1959) found poorer SRTs of almost 5 dB HL when indi- validation for puretone thresholds; (2) measurement of
viduals were not familiarized with the test list. The ASHA communication disability; and (3) reference for supra-
guideline strongly suggests that familiarization should not threshold measures. Unfortunately, most of the historical
be eliminated from the test protocol. purposes lack scientific evidence to support routine clinical
 CHAPTER 5
Ş
4QFFDI
"VEJPNFUSZ 65

use of an SRT (Wilson and Margolis, 1983). In addition, only recognition performance in quiet, it is important to note that
58% of audiologists complete the familiarization step of the empirical data (Bilger, 1984) support that speech recognition
test protocol, and 60% do not follow the recommended performance is a single construct and performance at one
ASHA (1988) protocol but, instead, determine an SRT using level of linguistic complexity (e.g., sentences) can be predicted
a modified Hughson–Westlake procedure with two out of by performance at another level (e.g., monosyllabic words).
three criteria (Martin et al., 1998). These observations are The systematic relationship between recognition per-
of concern because the SRT is a valid and reliable proce- formances at various levels of linguistic complexity by
dure when standardized recorded materials are used with adults with acquired hearing losses was demonstrated by
a specified testing procedure. The SRT is also particularly Olsen et al. (1997). Performance for phonemes, words in
useful when assessing response reliability in an individual isolation, and words in sentences was measured for 875 lis-
who appears to be malingering (see Chapter 33). teners with sensory/neural hearing loss. They found that
the scores for words in isolation and in sentences were pre-
dictable from the phoneme recognition scores, with mean
SPEECH RECOGNITION IN QUIET prediction errors of only 6% and 12%, respectively. Thus,
The purpose of speech recognition testing in quiet is to assess for example, a person scoring 60% correct on a phoneme
how well a person can understand speech in a quiet environ- recognition task would be predicted to score 22% (±6%) for
ment when the level of the speech is loud enough to obtain the recognition of words in isolation and 42% (±12%) for
a maximum SRS (PBmax). The level necessary for a person the recognition of words in sentences.
with hearing loss to perform maximally is highly variable
from person to person and is dependent on the materials
used to obtain the SRS (Jerger and Hayes, 1977). We feel
.POPTZMMBCJD
8PSET
that it is unfortunate that, in most audiology clinics, speech Historically, word lists such as the Northwestern University
recognition testing is assessed only at one presentation level Auditory Test Number 6 (NU No. 6; Tillman and Carhart,
(Wiley et al., 1995). The majority of audiologists select a 1966), the CID Auditory Test W-22 (CID W-22; Hirsh et al.,
single presentation level 30 to 40 dB SL re: SRT, meaning 1952), and the Phonetically Balanced 50 (PB-50; Egan, 1948)
that the materials are presented 30 to 40 dB above the SRT have been used to assess word recognition performance in a
(Martin et al., 1998; Wiley et al., 1995). Kamm et al. (1983) quiet background during audiologic evaluations.
found that speech recognition testing at 40 dB SL re: SRT The initial work of Egan (1944) outlined six princi-
did not approximate maximal performance for 40% of their pal criteria that the Psychoacoustics Lab at Harvard used
25 subjects with hearing loss. Evidence suggests that evaluat- to develop the PB-50 word lists. The six criteria were (1)
ing speech recognition abilities at more than one level cap- monosyllabic structure, (2) equal average difficulty of lists,
tures a portion of the psychometric function and allows a (3) equal range of difficulty of lists, (4) equal phonetic
better estimation of performance at PBmax. A procedure sug- composition of lists, (5) representative sample of American
gested by Wilson (2005, Personal communication) suggests English, and (6) familiar words. According to Hood and
the use of at least two levels with 25 words presented at each Poole (1980), it was assumed by Egan that meeting criteria
level. For persons with normal hearing or mild hearing loss 1, 4, 5, and 6 would ensure criteria 2 and 3. Further work
as evidenced by a puretone average (PTA) of ≤35 dB HL for to revise the PB-50 word lists by Hirsh et al. (1952) and
500, 1,000, and 2,000 Hz, the first level should be 50 dB HL Tillman et al. (1963) utilized the six criteria to create the
followed by the second level of 70 dB HL. For persons with W-22 word lists and the NU No. 6 word lists, respectively.
greater hearing loss, the first level should be 10 dB greater
than their PTA of 500, 1,000, and 2,000 Hz, and the second
1#
level should be 20 dB greater than the first level. If you are
unable to raise the second level 20 dB greater than the first The initial use of monosyllabic words for speech recogni-
level because of loudness discomfort issues, raise the second tion testing is attributed to Egan (1948) who worked in the
level as high as possible over the first level. PAL at Harvard University. His original pool of 1,000 words
Several types of materials are used to assess speech rec- was divided into 20 lists of 50 words, which collectively are
ognition ability in quiet such as sentences, nonsense syllables, known as the PAL PB-50 word lists. Each list was consid-
and the most commonly used stimuli, monosyllabic words. ered to be phonetically balanced such that the 50 words that
Previous research has shown that nonsense syllables are the composed a list were a proportionally correct representa-
most difficult of the three materials mentioned above for indi- tion of the phonetic elements in English discourse.
viduals to recognize, whereas sentences are the easiest. Recog-
nition performance of monosyllabic words falls on the per-
$*%
8
formance continuum somewhere between nonsense syllables
and sentences. Although monosyllables are the most com- Hirsh et al. (1952) had five judges rate the familiarity of the
monly used stimuli in clinical settings for measuring speech 1,000 monosyllabic words selected by Egan for the PB-50
66 SECTION I
Ş
#BTJD
5FTUT
BOE
1SPDFEVSFT

word lists, and 120 of the PB-50s were selected along with 80
TA B LE 5 . 2
other words to compose the new word lists. These 200 very
common words were selected and phonetically balanced Critical Difference Ranges (95%) for
into four 50-word lists known as the CID W-22 word lists. Select Percent Correct Scores as a
The CID W-22 word lists were recorded onto magnetic tape Function of Number of Test Items
as spoken by Ira Hirsh who monitored his voice on a VU
meter stating the carrier phrase “You will say” and letting % Correct 
8PSET 
8PSET 
8PSET
each target word fall naturally at the end of the phrase. The 0 0–20 0–8 0–4
CID W-22 word lists are some of the most popular word 10 0–50 2–24
lists used by audiologists for measuring suprathreshold 20 0–60 4–44 8–36
word recognition ability in quiet. 30 10–70 14–48
40 10–80 16–64 22–58
/6
/0
 50 10–90 32–68
60 20–90 36–84 42–78
Lehiste and Peterson (1959) devised lists of CNCs (consonant– 70 30–90 52–86
syllable nucleus [vowel]–consonant) that were phonemi- 80 40–100 56–96 64–92
cally balanced versus phonetically balanced. That is, lists 90 50–100 76–98
that were developed to be phonetically balanced did not 100 80–100 92–100 96–100
take into account the position of the sound in a word and
From Thornton and Raffin (1978).
how the acoustic realization of the sound would be affected
by coarticulatory factors. Lehiste and Peterson argued that phoneme scoring (Boothroyd, 1968). In a 25-word list of
phonemic balancing could be accomplished by allowing for monosyllables, you have 25 items to score using whole-word
the frequency of occurrence of each initial consonant, vowel scoring, whereas you would have 50 to 75 possible items to
nucleus, and final consonant to be similar across CNC word score using phoneme scoring.
lists. The Lehiste and Peterson lists were condensed into
four lists of 50 words known today as the NU No. 6.
Historically, 50 words were included in each test list to
4FOUFODF
5FTUT
facilitate phonetic balancing and to allow for a simple con- Sentence-level tests were developed at Bell Laboratories
version from number correct to percent correct following (Fletcher and Steinberg, 1929) and were used during World
testing. Studies have examined the benefits of abbreviat- War II to evaluate military communication equipment
ing the number of words used per list from 50 to 25 with (Hudgins et al., 1947). Until the development of the CID
mixed results in terms of test–retest reliability (Beattie et al., Everyday Sentences (Silverman and Hirsh, 1955), no sentence
1978; Elpern, 1961). The most important work regarding test had received clinical acceptance. The CID sentences con-
this issue of half versus full lists was the examination of sist of 10 lists of 10 sentences each with 50 key words in each
speech recognition data as a binomial variable by Thornton list. Interrogative, imperative, and declarative sentences are
and Raffin (1978). As discussed in the earlier section on included. Responses can be spoken or written and are scored
psychometric functions, performance ability between 20% as the percentage of key words correctly recognized.
and 80% is the most variable, whereas performance ability is The basis for the use of sentences in the clinical assess-
least variable at either extreme of the function (Egan, 1948). ment of speech recognition abilities is that sentences pro-
The results of Thornton and Raffin (1978) support these vide a more “realistic” listening condition for everyday com-
early views on performance using the binomial distribution munication than does the use of isolated words or nonsense
to mathematically model word recognition performance. It syllables (Bess, 1983; Silverman and Hirsh, 1955). Although
indicates that the accuracy between scores for the same lis- sentences may have greater face validity than other stim-
tener depends on the number of words used per list and the uli, they also provide semantic, syntactic, and lexical clues
listener’s level of performance. In addition, Thornton and (i.e., extrinsic redundancies). Thus it is difficult to distin-
Raffin created a table of the lower and upper limits of the guish individuals who do well on a speech recognition task
95% critical differences for percentage scores as a function because they have good speech recognition skills or because
of test items. Table 5.2 shows the critical differences a retest they make good use of top-down (linguistic, cognitive) pro-
score would need to exceed to be considered statistically cessing skills. Another complication of the use of sentence
different for the original test score. As seen in Table 5.2, as materials is that, as length exceeds seven to nine syllables,
the number of items increases, the range decreases, suggest- memory constraints, particularly in the elderly, may impact
ing that as the set size increases, the variability in the scores performance (Miller, 1956). Despite these potential limita-
decreases, allowing for the detection of more subtle differ- tions, several clinically useful sentence tests have been devel-
ences in performance. One way to increase set size without oped. Because the ability to use context is preserved even in
increasing test time is to move from whole-word scoring to older adults with hearing loss, for most patient populations,
 CHAPTER 5
Ş
4QFFDI
"VEJPNFUSZ 67

sentence tests are typically too easy (ceiling effect) and, netic context. Contrasts include intonation; vowel height
therefore, fail to distinguish among levels of difficulty. How- and place; and initial and final consonant voicing, contin-
ever, they are well suited as adaptive noise procedures (see uance, and place. In addition to minimizing the effects of
“Speech Recognition in Noise” section) instead of supra- lexical context and word familiarity on performance, the
threshold quiet procedures. An exception to this trend is the use of nonsense syllables allows for detailed examination
use of sentence tests in quiet for individuals with severe-to- of phonetic errors. Despite these advantages, nonsense syl-
profound hearing losses. lables lack face validity with regard to being representative
For the profoundly impaired patient population, the of everyday speech communication.
City University of New York (CUNY) Sentences (Boothroyd Minimization of lexical context and word familiar-
et al., 1988), which consist of 72 sets of topic-related sen- ity effects, while allowing for the analysis of errors and
tences, were designed to assess the use of cochlear implants confusions, can also be accomplished through the use of
and tactile aids as supplements to speech reading. Each sen- closed-set tests using real word stimuli. Classic tests of pho-
tence in a set is about one of 12 topics: food, family, work, neme recognition include the Modified Rhyme Test (MRT;
clothes, animals, homes, sports/hobbies, weather, health, House et al., 1955; Kruel et al., 1968) and its variations (e.g.,
seasons/holidays, money, or music. Each set contains four Rhyming Minimal Contrasts Test (Griffiths, 1967) and the
statements, four questions, and four commands and one California Consonant Test (CCT; Owens and Schubert,
sentence of each length from 3 to 12 words, for a total of 1977). The MRT consists of 50 test items each with six
102 words per set. Performance is scored as the number response alternatives. Twenty-five of the items differ by
of words correct. Original recordings were on laser-video the initial consonant (i.e., bent, went, sent, tent, dent, and
disc and were presented via the Computer Assisted Speech rent), and the other 25 items differ by the final consonant
Perception Software (CASPER; Boothroyd, 1987) program. (i.e., peas, peak, peal, peace, peach, and peat). The CCT also
The CUNY Sentences are being converted to DVD format consists of 100 items but uses a four-choice, rather than a
with upgrades to the CASPER software as part of current six-choice, response format in assessing the perception of
work at the Rehabilitation Engineering Research Center 36 initial consonant items and 64 final consonant items. The
(RERC) on Hearing Enhancement at Gallaudet University perception of medial vowels as well as initial and final con-
(http://www.hearingresearch.org/). sonants was added in the University of Oklahoma Closed
Response Speech Test by Pederson and Studebaker (1972).
/POTFOTF
4ZMMBCMF
5FTUT1IPOFNF
A closed-set format is also used in the Speech Pattern
Contrast (SPAC) test (Boothroyd, 1984), which was designed
3FDPHOJUJPO
5FTUT to assess the ability to perceive both suprasegmental (i.e.,
The effects of lexical context and word familiarity on test stress and intonation) and segmental phonologically (i.e.,
performance can be minimized by the use of nonsense sylla- vowel height and place, initial and final consonant voicing,
ble and/or closed-set phoneme recognition tests. Nonsense continuance, and place) relevant distinctions. Test length of
syllables were one of the first materials used to assess speech SPAC is minimized by combining two segmental contrasts
recognition ability during the development of telephone cir- in one subset (e.g., final consonant voicing and continu-
cuits at Bell Telephone Laboratories (Fletcher and Steinberg, ance) with four items (e.g., seat-seed-cease-sees). Although
1929). However, clinical use of nonsense syllables for those the SPAC as well as other speech feature tests and NSTs are
with hearing loss did not occur until the 1970s when two not routinely used in clinical audiology, the information
carefully developed tests became available—the CUNY provided about the details of an individual’s speech percep-
Nonsense Syllable Test (CUNY-NST; Levitt and Resnick, tion ability can be quite useful when assessing the need for
1978) and the Nonsense Syllable Test (NST; Edgerton and and the benefits of hearing aids and cochlear implants for
Danhauer, 1979). The CUNY-NST is a closed-set test con- both children and adults.
sisting of seven subtests, each of which has seven to nine
consonant–vowel (CV) or vowel–consonant (VC) syllables.
The CUNY-NST assesses perception of the consonants most
SPEECH RECOGNITION IN NOISE
likely to be confused by individuals with hearing loss using The most common complaint expressed by adults with
three vowel contexts. The Edgerton–Danhauer NST is an hearing loss is the inability to understand a speaker when
open-set test consisting of 25 nonsense bisyllabic CVCV listening in an environment of background noise. In 1970,
items, allowing for assessment of the perception of 50 con- Carhart and Tillman suggested that an audiologic evaluation
sonant and 50 vowel stimuli. More recently, Boothroyd et al. should include some measure of the ability of an individual
(1988) described the Three Interval Forced Choice Test of to understand speech when in a background of speech noise.
speech pattern contrast perception (THRIFT), an NST that Prior to the revival of the directional microphone in the late
can be used with children 7 years of age or older (Hnath- 1990s, however, the information gained from a speech-in-
Chisolm et al., 1998). The THRIFT measures the perception noise task for most rehabilitative audiologists was not perti-
of nine speech pattern contrasts presented in varying pho- nent to the selection of amplification because of the fact that
 SECTION I
Ş
#BTJD
5FTUT
BOE
1SPDFEVSFT

most hearing aids were mainly selected based on gain, slope, for separation between individuals with normal hearing
and output curves. Thus in the technology-driven field of and those with hearing loss (Beattie, 1989; McArdle et al.,
audiology, speech-in-noise testing failed to gain a place in 2005b). Typically, individuals with sensory/neural hear-
the traditional audiologic evaluation. The revolution of ing loss require the signal to be 10 to 12 dB higher than
digital hearing aids and their multitude of features, such the noise to obtain 50% performance on the psychometric
as directional microphones, noise reduction strategies, and function, whereas individuals with normal hearing on aver-
digital signal processing strategies, have created an impor- age obtain 50% performance at an S/N of 2 to 6 dB. McArdle
tant reason for utilizing speech-in-noise tasks on a routine et al. (2005a, 2005b) found mean performance on the
basis when evaluating an individual with hearing loss. Words-in-Noise (WIN) test (Wilson, 2003) to be 12.5 and
For the past 40 years, researchers have observed that lis- 6 dB S/N for 383 listeners with hearing loss and 24 listeners
teners with hearing loss show a greater disadvantage when with normal hearing, respectively. Similarly, under similar
listening in a competing speech background compared with experimental conditions, Dirks et al. (1982) and Beattie
listeners with normal hearing, such that the S/N needed for (1989) who used CID W-22 word lists in noise found 50%
the listener with hearing loss is 10 to 15 dB greater than that points of 12 and 11.3 dB S/N, respectively, for listeners with
needed by listeners with normal hearing (e.g., Carhart and hearing loss.
Tillman, 1970). Plomp (1978) reported that for every 1-dB Several studies have examined the possibility of predict-
increase in signal over the competing noise, a listener with ing the ability of an individual to understand speech-in-noise
hearing loss would receive, on average, an improvement of using puretone audiograms and SRSs in quiet without suc-
3% in terms of ability to recognize the signal. Thus, a 10-dB cess (Beattie, 1989; Carhart and Tillman, 1970; Cherry, 1953;
improvement in S/N should add 30% in terms of intelligi- Dirks et al., 1982; Killion and Niquette, 2000; Plomp, 1978;
bility as measured by open-set, speech recognition tests for Wilson, 2003). The data in Figure 5.3 were compiled from
listeners with hearing loss. two studies (McArdle et al., 2005a, 2005b). In the figure, per-
The addition of background noise to a speech recog- formance on a word recognition in quiet task at 80 dB HL
nition task has been shown to improve the sensitivity and is graphed on the ordinate as a function of 50% points on
validity of the measurement (Beattie, 1989; Sperry et al., the WIN test along the abscissa. The same words spoken by
1997). In terms of improving sensitivity, the addition of the same speaker were used for both the recognition task in
multiple S/Ns increases the difficulty of the task and allows quiet and in noise. The shaded area of the figure represents

100
176 (45.5%)

107 (27.6%)
80
Percent correct recognition at 80 dB HL

60 104 (26.9%)

FIGURE 5.3 A plot of word recognition

performance in quiet in percent correct
(y-axis) versus the 50% point of recogni- 40
tion performance in multitalker babble on
the Words-in-Noise (WIN) test (x-axis).
The shaded area of the figure defines the
range of performances (10th to 90th per-
20
centiles) obtained by listeners with normal
hearing on the WIN test. The numbers
represent the number of listeners who had
word recognition scores in quiet ≥90%,
≥80%, and ≥70% correct on the words in 0
quiet. The data are combined from McArdle
et al. (2005a, 2005b). (Reprinted with per-
mission from the Journal of Rehabilitative 0 4 8 12 16 20 24
Research and Development.) 50% correct point (dB S/N)
 CHAPTER 5
Ş
4QFFDI
"VEJPNFUSZ 

the range of performance by 24 listeners with normal hear- Recognition performance is scored as the percentage of LP
ing on the WIN. and HP words correctly perceived. By providing both LP
Two main observations can be seen in the data in Fig- and HP scores, the SPIN test not only allows for the assess-
ure 5.3: (1) only 5 out of 387 listeners with hearing loss per- ment of the acoustic-phonetic components of speech, but
formed in the normal range on both the recognition task in also examines the ability of an individual to utilize linguistic
quiet and in noise; and (2) 45.5% of the 387 listeners with context.
hearing loss had word recognition scores in quiet at 80 dB In the 1980s, two additional tests designed to assess rec-
HL that were ≥90% correct. Thus, it is of interest to note ognition of everyday speech based on correct word recogni-
that although 73% of the listeners with hearing loss had tion performance in sentence length stimuli were developed.
word recognition scores in quiet ≥80%, the overwhelming The Connected Speech Test (CST; Cox et al., 1987), which
majority of these listeners displayed abnormal performance was developed as a criterion measure in studies of hearing
on a word recognition task in noise. This finding suggests aid benefit, consists of 48 passages of conversationally pro-
that speech-in-noise testing may be considered a stress test duced connected speech. Each passage is about a familiar
of auditory function (Wilson, 2013, Personal communi- topic and contains 10 sentences. Sentence length varies from
cation). In addition, it is clear that word recognition abil- 7 to 10 words, and there is a total of 25 key words in each
ity in noise is not easily predicted by word recognition in passage. Sentences are presented at an individually deter-
quiet for listeners with hearing loss other than to say that mined S/N, and performance is scored as the number of key
listeners with poor recognition ability in quiet also perform words correct.
poorly on word recognition tasks in noise. Because we are The most recent application of sentence length stimuli
unable to predict the ability of an individual to understand is in tests that are scored in terms of the decibel-to-noise
speech in a noisy background, audiologists should use the ratio required to achieve 50% correct performance. The two
tests available for quantifying the S/N needed by the listener most common tests are the Hearing in Noise Test (HINT;
to understand speech in noise. Several materials, described Nilsson et al., 1994) and the Quick Speech-in-Noise (Quick-
in the following section, have been developed to measure SIN) test (Killion et al., 2004). The two tests vary in the
speech-in-noise performance. type of sentences and type of noise used. The HINT uses
the Bamford–Kowal–Bench (BKB) Standard Sentence Lists
(Bench et al., 1979) that were compiled from the utterances
.BUFSJBMT of hearing-impaired children and contain straightforward
Initially, efforts in speech-in-noise testing were focused on vocabulary and syntax. Sentences are presented in sets of
sentence-level materials to make the task more of a real- 10 sentences, and the listener must repeat the entire sen-
world experience; however, normal everyday sentences tence correctly to receive credit. The noise used is speech-
were too easy, and further manipulation of the sentences spectrum noise that is held constant while the signal is varied
was needed to obtain the 50% correct point of perfor- to find the 50% correct point. The QuickSIN uses the Har-
mance for a listener on a speech-in-noise task. Speaks and vard Institute of Electrical and Electronics Engineers (IEEE,
Jerger (1965) developed the Synthetic Sentence Identifica- 1969) sentences, which are a collection of low-context,
tion (SSI) test to minimize the effect of contextual cues that meaningful sentences, whose phonetic balance is similar to
often made it easy to understand sentence-level materials that of English. In the QuickSIN, there are six sentences per
even in a background of noise. The stimuli are called syn- list, and each sentence contains five key words. All sentences
thetic sentences because they are not actual sentences, but are presented in multitalker babble with the five key words
rather they contain normal English phonemes and syntax in each sentence scored as correct or incorrect. Recently, the
but no semantic context. An example of a sentence is “Small BKB-SIN test (Etymotic Research, 2005) was developed for
boat with a picture has become.” The task of the listener is use with children (ages ≥5), cochlear implant patients, and
to select which one of 10 sentences displayed on a response adults for whom the QuickSIN test is too difficult.
form is perceived when presented against a competing story More recently, monosyllabic and digit materials in mul-
describing the life of Davy Crockett. The competing story titalker babble have been developed at the Auditory Research
can be presented either contralaterally or ipsilaterally. Lab of the James H. Quillen Veterans Affairs Medical Center
Another interesting sentence-level test, the Speech Per- (Wilson, 2003; Wilson and Strouse, 2002; Wilson and Weak-
ception in Noise (SPIN) test (Kalikow et al., 1977), varies ley, 2004). The word and digit materials have been shown to
the amount of semantic context that leads to the last word be sensitive to the different recognition abilities of normal-
of each sentence, which is a monosyllabic target word. The hearing and hearing-impaired adults in multitalker babble
SPIN test has eight forms of 50 sentences each that are pre- (Wilson et al., 2003; Wilson and Weakley, 2004). McArdle
sented at a fixed S/N of 8 dB. The target word in 25 of the et al. (2005b) examined the effect of material type (i.e., digits,
sentences has low predictability (LP) given the limited clues words, and sentences) on S/N loss for young listeners with
from the preceding context, and the other 25 have high normal hearing and older listeners with hearing impair-
predictability (HP) from the preceding sentence context. ment. The three speech-in-noise tests that were examined
70 SECTION I
Ş
#BTJD
5FTUT
BOE
1SPDFEVSFT

24

20

(29–40%)

50% point WIN (dB S/N)

80

12
(43–60%)

FIGURE 5.4 Bivariate plot of the 50% points 8

(in dB S/N) on the Words-in-Noise (WIN) test
(ordinate) and on the Quick Speech-in-Noise
(QuickSIN) test (abscissa). The diagonal line
represents equal performance, with the larger 4
filled symbol indicating the mean datum point.
The dashed line is the linear regression fit to
the data. The numbers in parentheses are the
number of performances above and below the 0
line of equal performances. (Reprinted with
0 4 8 12 16 20 24
permission from the Journal of Rehabilitative
Research and Development.) 50% correct point quickSIN (dB S/N)

include: (1) QuickSIN (Etymotic Research, 2001); (2) WIN the construct was described by Carhart et al. (1969) and
test (Wilson and Strouse, 2002; Wilson, 2003); and (3) digit termed perceptual masking. Energetic masking is described
triplets-in-multitalker babble (Wilson and Weakley, 2004). in the literature as peripheral masking, such that a stimulus
As expected, the younger listeners performed better than the interferes with the perception of a second stimulus making
older listeners on all three tasks. For the older listeners with the first stimulus a “masker.” Nonenergetic masking, or
hearing loss, the S/N required for 50% recognition of each informational masking, occurs when the target stimulus is
material type presented was −4, 12.4, and 11.7 dB S/N for similar to the masking stimulus, creating uncertainty for the
digits, words, and sentences, respectively. Figure 5.4 shows a listener as to whether he or she is hearing the target or the
bivariate plot of the 50% points for the older listeners with masker. Informational masking can occur at different pro-
hearing loss on both the QuickSIN (abscissa) and the WIN cessing levels (e.g., phonetic, semantic) and is greater for a
(ordinate). The diagonal line in Figure 5.4 represents equal speech masker than noise, especially when the talker is the
performance on both QuickSIN and the WIN. As can be same gender or, even worse, the same talker for both the tar-
seen, mean performance, as indicated by the bold filled cir- get and the masker (Brungart, 2001). Informational mask-
cle, is close to the diagonal line, suggesting that either the use ing has a greater effect when the masker is a single speaker
of monosyllabic words or IEEE sentences in this population versus a background of multiple talkers since once you add
provided a similar measure of performance in noise. More more than a couple of speakers the background “informa-
importantly, the performance difference at the 50% point tion” in the masker becomes hard to distinguish. Most com-
between normal-hearing listeners and hearing-impaired lis- mercially available speech-in-noise tests involve multitalker
teners was 7.6 dB for both words and sentences, suggesting babble, which decreases the effects of informational mask-
that words and sentences in a descending speech-in-noise ing but future studies in this area are warranted.
task were equally sensitive to the effects of hearing loss. For
a more in-depth discussion of the use of words or sentences CONSIDERATIONS FOR SPEECH
in speech-in-noise testing, see Wilson and McArdle (2005). AUDIOMETRY IN CHILDREN AND
A new body of literature has evolved in the area of
speech-in-noise testing focused on informational masking,
OTHER SPECIAL POPULATIONS
which is defined as nonenergetic masking that increases Speech stimuli are used for the behavioral assessment of the
threshold as a result of uncertainty (Wilson et al., 2012). auditory function of a child from birth onward. With very
Although the term informational masking is more recent, young infants, speech stimuli might be used to elicit a startle
 CHAPTER 5
Ş
4QFFDI
"VEJPNFUSZ 71

response, and as the infant develops, SDTs and SRTs can be on speech perception performance in the same way that the
obtained using a variety of behavioral techniques, such as developing linguistic abilities of a child are important to con-
visual response audiometry or play audiometry. The tech- sider in assessment. Although recorded materials are avail-
nique used will be dependent on the motor capabilities of able in languages such as Spanish (Wesilender and Hodgson,
the child. In addition to considering the motor capacity 1989), unless the audiologist speaks Spanish, errors could
of the child for responding (e.g., head turn, picture point- be made in determining correct from incorrect responses.
ing), the phonologic, receptive, and expressive language Wilson and Strouse (1999) suggest the use of a multimedia
skills of the child need to be considered during speech rec- approach similar to that used by McCullough et al. (1994)
ognition testing. For example, by the time a child can func- with nonverbal patients. Stimulus words are presented in the
tion at about a 5-year-old level, conventional SRTs can be patient’s native language, and the person responds by select-
obtained as long as the spondee words used are within the ing the perceived word from a closed set of alternatives shown
receptive vocabulary of the child (ASHA, 1988). Similarly, on a touchscreen monitor. Scoring could be done automati-
several suprathreshold pediatric speech recognition tests, cally through a software program.
such as the Word Intelligibility Picture Identification (WIPI)
test (Ross and Lerman, 1970) and the Northwestern Uni- CLINICAL FUNCTIONS OF SPEECH
versity Children’s Perception of Speech (NU-CHIPS) test
(Elliot and Katz, 1980), are comprised of words expected to
RECOGNITION MEASURES
be within the receptive vocabulary of a child. One of the historical purposes for the use of speech recog-
A variety of speech recognition tests are available for nition testing in the clinical test battery was as a diagnos-
use with children. For example, both the WIPI and NU- tic tool for determining the location of peripheral auditory
CHIPS use monosyllabic words presented in a closed-set pathology. Figure 5.2 illustrates typical psychometric func-
format. Other test paradigms allow for the assessment of the tions obtained in quiet for the recognition of monosyllabic
perception of speech feature contrasts (e.g., Imitative Test of words by listeners with normal auditory function as well
Speech Pattern Contrast Perception [IMSPAC]; Kosky and as those with conductive, sensory (cochlear), and neural
Boothroyd, 2003; Visually Reinforced Audiometry Speech (retrocochlear) hearing losses. For normal-hearing listen-
Pattern Contrast Perception [VRASPAC]; Eisenberg et al., ers, regardless of word recognition materials used, when the
2004), syllabic pattern and stress (e.g., Early Speech Percep- presentation level is about 30 dB higher than the dB level
tion [ESP] test; Moog and Geers, 1990), lexically easy versus needed for 50% performance (i.e., SRT), a score of 90% or
lexically hard words (e.g., the Lexical Neighborhood Test better can be expected. For individuals with hearing loss,
[LNT]; Kirk et al., 1995), and words in sentences presented when listening at a moderate level, scores may range any-
in quiet (e.g., BKB sentences; Bamford and Wilson, 1979) where from 100% correct to 0% correct. Because of this
and in noise (e.g., BKB-SIN test; Etymotic Research, 2005), wide dispersion of speech recognition performance across
a task which requires word and sentence recognition in both individuals with various types of hearing loss, speech rec-
quiet and noise (e.g., Pediatric Speech Intelligibility [PSI] ognition testing provides only limited diagnostic informa-
test; Jerger and Jerger, 1984). tion if testing is done at only one intensity level (see, for
In addition to children, special consideration also needs discussion, Bess, 1983; Penrod, 1994). When testing is com-
to be given to the assessment of speech perception abilities pleted at several intensity levels, however, certain patterns
in profoundly hearing-impaired adults, nonverbal patients, of performance can be expected with certain hearing losses
and multilingual patients (Wilson and Strouse, 1999). Pro- (Wilson and Strouse, 1999).
foundly hearing-impaired adults typically obtain scores of Individuals with conductive hearing loss tend to exhibit
zero on standard speech recognition tests. As a result, batter- little difficulty on speech recognition tests, with perfor-
ies such as the Minimal Auditory Capabilities (MAC) battery mance typically at 90% or better when testing is conducted
have been developed (Owens et al., 1985). Tasks included in at moderate SLs (curve #1 of Figure 5.2). A patient with a
the MAC battery involve discrimination of syllabic number, sensory/neural hearing loss will generally have poorer SRSs
noise versus voice, and statements versus questions; recog- than would a person with the same degree of hearing loss
nition of spondaic words and consonants and vowels in real due to conductive pathology. Although a very wide range
words in closed-set tasks; and more standard open-set rec- of scores are found across patients with cochlear as well as
ognition of words in isolation and sentences. retrocochlear hearing losses, SRSs tend to be poorest among
Nonverbal patients are often encountered in medical set- those with retrocochlear pathology. Although some individ-
tings where patients may have medical conditions such as lar- uals with cochlear losses will demonstrate a slight decrease in
yngectomies or cerebral vascular accidents. For these patients, recognition performance when intensity levels are increased
written responses or picture pointing tasks may be appropri- beyond the initial level needed for obtaining maximum
ate. Increases in the ethnic diversity of the US population can performance (curve #3 of Figure 5.2), marked decreases in
result in the audiologist assessing a patient who speaks little performance with increasing intensity after maximum per-
to no English. Limited knowledge of English could impact formance is achieved are typically characteristic of a neural
72 SECTION I
Ş
#BTJD
5FTUT
BOE
1SPDFEVSFT

loss (curves #4 and #5 of Figure 5.2). The phenomenon of can provide insight into the use of appropriate amplification
reduced SRSs with increasing intensity that occurs with ret- and/or cochlear implant speech processing strategies.
rocochlear pathology is referred to as the “rollover” effect In addition to testing in noise, Brandy (2002) points
(Bess et al., 1979; Dirks et al., 1977). In addition to rollover, out that audiologists can gain insight into the (re)habili-
retrocochlear pathology would be suspected in the presence tative needs of patients through recording incorrect word
of a significant discrepancy in SRSs between two ears or responses, with subsequent examination of speech feature
lower than expected performance at all presentation levels error patterns (e.g., fricatives, stops, glides). Other reha-
(curve #6 of Figure 5.2). bilitative applications of speech audiometry include the
Assessment of the central auditory system also uses use of materials that allow for the assessment of use of
measures of speech recognition performance. Tasks can be linguistic context (Flynn and Dowell, 1999) and auditory–
presented either monaurally or binaurally. Monaural tasks visual performance (Boothroyd, 1987) and for the deter-
use distorted, degraded, or low-redundancy speech stimuli mination of most comfortable and uncomfortable listen-
to reduce extrinsic redundancies. Methods of degrada- ing levels (Punch et al., 2004). Information obtained with
tion include filtering (Bocca et al., 1955), time compression a variety of materials presented in a variety of paradigms
(Wilson et al., 1994), and reverberation (Nabelek and can be useful in determining optimal device settings, start-
Robinson, 1982). Binaural tests were designed to assess the ing points for therapeutic intervention, and directions for
ability of the central auditory nervous system to integrate patient counseling.
or resynthesize the different parts of a signal that are pre-
sented to each of the two ears. For example, in the Binaural
Fusion test (Matzker, 1959), a low-pass filtered version of
FOOD FOR THOUGHT
a word is presented to one ear, whereas a high-pass filtered 1. Given the use of SRTs as a verification for puretone
version of the same word is presented to the opposite ear. A thresholds in every patient has been questioned since
normal-functioning central auditory nervous system is able 1983, what is the cost benefit of measuring SRTs in a busy
to integrate the information from each ear and respond with practice? Might the time be better spent gathering other
the correct target word. On the other hand, binaural dich- information about the patient’s auditory functioning?
otic tasks involve the presentation of different speech signals 2. Determining the presentation level for word recognition
simultaneously to both ears. The patient must repeat either testing in quiet historically has been 40 dB SL, re: SRT.
or both of the signals depending on the test used. Com- Given the evidence for this level is based on listeners with
mon clinical dichotic tests include Dichotic Digits (Kimura, normal hearing, what is most appropriate for determin-
1961), the Staggered Spondaic Word test (Katz, 1962), and ing the presentation level(s) for listeners with hearing
the Dichotic Sentence Identification test (Fifer et al., 1983). loss? Additionally in an audiologic evaluation, what are
The interpretation of performance on tests designed to assess the benefits of using 25 words at each of two or three
auditory processing abilities is beyond the scope of the pres- presentation levels versus 50 words at one presentation
ent chapter and is discussed in detail in Chapters 27 and 29. level?
In addition to diagnostic applications, speech rec- 3. Speech recognition in quiet has been performed by audi-
ognition testing has an important role in estimating the ologists since the 1950s. Given that the most common
adequacy and effectiveness of communication and in the complaint of listeners with hearing loss is their difficulty
planning and evaluation of (re)habilitative efforts, includ- communicating in noisy situations, should the stan-
ing the selection and fitting of hearing aids and cochlear dard comprehensive audiometric battery be modified to
implants. For example, many audiologists label speech rec- include speech-in-noise measures?
ognition performance for monosyllabic words presented in
quiet performance as “excellent,” “good,” “fair,” or “poor” in
an attempt to link performance to the adequacy of commu- KEY REFERENCES
nication in everyday settings. However, research designed A full list of references for this chapter can be
to demonstrate systematic relationships between recogni- found at http://thePoint.lww.com. Below are the key refer-
tion performance in quiet and actual everyday communica- ences for this chapter.
tion has been largely unsuccessful (Davis, 1948; High et al.,
1964). A better estimate of the impact of a hearing loss on American National Standards Institute (ANSI). (1996) Specifica-
tions for Audiometers. S3.6–1996. New York,NY: American
daily communication might be obtained with the use of
National Standards Institute.
speech-in-noise tests such as the WIN, QuickSIN, or HINT.
American National Standards Institute. (2004) Specifications for
As Wilson and McArdle (2005) discuss, speech-in-noise Audiometers. S3.6–2004. New York, NY: American National
testing allows for the assessment of the most common com- Standards Institute.
plaint of patients—the inability to understand speech in American Speech-Language-Hearing Association. (1988) Guide-
background noise; and thus, test results provide important lines for determining threshold level for speech. ASHA. 30,
information for use in counseling. Furthermore, test results 85–89.
 CHAPTER 5
Ş
4QFFDI
"VEJPNFUSZ 73

American Speech-Language-Hearing Association. (2000) Audiology Cox RM, Alexander GC, Gilmore C. (1987) Development of the
Survey. Rockville, MD: American Speech-Language-Hearing Connected Speech Test (CST). Ear Hear. 8, 119S–126S.
Association. Davis H. (1948) The articulation area and the social adequacy
Asher WJ. (1958) Intelligibility tests: A review of their standardization, index for hearing. Laryngoscope. 58, 761–778.
some experiments, and a new test. Speech Monogr. 25, 14–28. Department of Veterans Affairs. (1997) The Audiology Primer for
Bamford J, Wilson I. (1979) Methodological considerations and Students and Health Care Professionals. Mountain Home, TN:
practical aspects of the BKB sentence lists. In: Bench J, Bamford Veterans Affairs Medical Center.
JM, eds. Speech-Hearing Tests and the Spoken Language of Hear- Dirks D, Kamm D, Bower D, Betsworth A. (1977) Use of perfor-
ing Impaired Children. London: Academic Press; pp 148–187. mance intensity function in diagnosis. J Speech Hear Disord.
Beattie RC. (1989) Word recognition functions for the CID W-22 27, 311–322.
Test in multitalker noise for normally hearing and hearing- Dirks DD, Morgan DE, Dubno JR. (1982) A procedure for quanti-
impaired subjects. J Speech Hear Disord. 54, 20–32. fying the effects of noise on speech recognition. J Speech Hear
Beattie RC, Svihovec DA, Edgerton BJ. (1978) Comparison of Disord. 47, 114–123.
speech detection and spondee thresholds and half- versus full- Edgerton BJ, Danhauer JL. (1979) Clinical Implications of Speech
list intelligibility scores with MLV and taped presentation of Discrimination Testing Using Nonsense Stimuli. Baltimore,
NU-6. J Am Audiol Soc. 3, 267–272. MD: University Park Press.
Bench J, Kowal A, Bamford J. (1979) The BKB (Bamford-Kowal- Egan JP. (1944) Articulation Testing Methods, II. OSRD Report No.
Bench) sentence lists for partially-hearing children. Br J Audiol. 3802. Cambridge, MA: Psychoacoustic Laboratory, Harvard
13, 108–112. University.
Bess FH. (1983) Clinical assessment of speech recognition. In: Egan JP. (1948) Articulation testing methods. Laryngoscope. 58,
Konkle DF, Rintelmann WF, eds. Principles of Speech Audiom- 955–991.
etry. Baltimore, MD: University Park Press; pp 127–201. Eisenberg LS, Martinez AS, Boothroyd A. (2004) Perception of
Bess FH, Josey AF, Humes, LE. (1979). Performance intensity func- phonetic contrasts in infants. In: Miyamoto RT, ed. Cochlear
tions in cochlear and eighth nerve disorders. Am J Otolaryngol. Implants: International Congress Series 1273. Amsterdam:
1, 27–31. Elsevier; pp 364–367.
Bilger RC. (1984) Speech recognition test development. In: Elkins E, Elliot L, Katz D. (1980) Northwestern University Children’s Percep-
ed. Speech Recognition by the Hearing Impaired. ASHA Reports tion Speech (NU-CHIPS). St. Louis, MO: Auditec.
14. Rockville, MD: ASHA; pp 2–7. Elpern BS. (1961) The relative stability of half-list and full-list
Bocca E, Calaero C, Cassinari V, Migilavacca F. (1955) Testing discrimination tests. Laryngoscope. 71, 30–36.
“cortical” hearing in temporal lobe tumors. Acta Otolaryngol. Etymotic Research. (2001) QuickSINTM (Compact Disc). Elk Grove
45, 289–304. Village, IL: Etymotic Research.
Boothroyd A. (1968) Developments in speech audiometry. Sound. Etymotic Research. (2005) BKB-SINTM (Compact Disc). Elk Grove
2, 3–10. Village, IL: Etymotic Research.
Boothroyd A. (1984) Auditory perception of speech contrasts by Feldmann H. (1970) A history of audiology: A comprehensive report
subjects with sensorineural hearing loss. J Speech Hear Res. 27, and bibliography from the earliest beginnings to the present.
134–144. Transl Beltone Inst Hear Res. 22, 1–111. [Translated by J. Tonndorf
Boothroyd A. (1987) CASPER: Computer Assisted Speech Perception from Die geschichtliche entwicklung der horprufungsmethoden,
Evaluation and Training. Proceedings of the 10th Annual Confer- kuze darstellung and bibliographie von der anfongen bis zur
ence on Rehabilitation Technology. Washington, DC: Association gegenwart. In: Leicher L, Mittermaiser R, Theissing G, eds.
for the Advancement of Rehabilitation Technology. Zwanglose Abhandungen aus dem Gebeit der Hals-Nasen-Ohren-
Boothroyd A, Hnath-Chisolm T, Hanin L, Kishon-Rabin L. (1988) Heilkunde. Stuttgart, Germany: Georg Thieme Verlag; 1960.]
Voice fundamental frequency as an aid to the speechreading of Fifer RC, Jerger JF, Berlin CI, Tobey EA, Campbell JC. (1983)
sentences. Ear Hear. 9, 335–341. Development of a dichotic sentence identification test for
Brandy WT. (2002) Speech audiometry. In: Katz J, ed. Handbook of hearing-impaired adults. Ear Hear. 4, 300–305.
Clinical Audiology. 4th ed. Baltimore, MD: Lippincott Williams Finney DJ. (1952) Statistical Method in Biological Assay. London:
& Wilkins; pp 96–110. C. Griffen.
Brungart DS. (2001) Informational and energetic masking effects Fletcher H, Steinberg J. (1929) Articulation testing methods. Bell
in the perception of two simultaneous talkers. J Acoust Soc Am. Syst Techn J. 8, 806–854.
109, 1101–1109. Flynn MC, Dowell RC. (1999) Speech perception in a communi-
Carhart R. (1951) Basic principles of speech audiometry. Acta cative context: An investigation using question/answer pairs.
Otolaryngol. 40, 62–71. J Speech Lang Hear Res. 42, 540–552.
Carhart R, Tillman TW. (1970) Interaction of competing speech Frank T. (1980) Clinically significance of the relative intelligibility
signals with hearing losses. Arch Otolaryngol. 91, 273–279. of pictorially represented spondee words. Ear Hear. 1, 46–49.
Carhart R, Tillman TW, Greetis ES. (1969) Perceptual masking in Griffiths JD. (1967) Rhyming minimal contrasts: A simplified
multiple sound backgrounds. J Acoust Soc Am. 45, 694–703. diagnostic articulation test. J Acoust Soc Am. 42, 236–241.
Cherry EC. (1953) Some experiments on the recognition of speech High WS, Fairbanks G, Glorig A. (1964) Scale for self-assessment
with one and with two ears. J Acoust Soc Am. 25, 975–979. of hearing handicap. J Speech Hear Disord. 29, 215–230.
Conn MJ, Dancer J, Ventry IM. (1975) A spondee list for determin- Hirsh IJ, Davis H, Silverman SR, Reynolds EG, Eldert E, Benson
ing speech reception threshold without prior familiarization. RW. (1952) Development of materials for speech audiometry.
J Speech Hear Disord. 40, 388–396. J Speech Hear Disord. 17, 321–337.
74 SECTION I
Ş
#BTJD
5FTUT
BOE
1SPDFEVSFT

Hirsh IJ, Palva T, Goodman A. (1954) Difference limen and recruit- Matzker J. (1959) Two new methods for the assessment of central
ment. AMA Arch Otolaryngol. 60, 525–540. auditory functions in cases of brain disease. Ann Otol Rhinol
Hnath-Chisolm T, Laipply E, Boothroyd A. (1998) Age-related Laryngol. 68, 1185–1197.
changes on speech perception capacity. J Speech Hear Res. 41, McArdle R, Chisolm TH, Abrams HB, Wilson RH, Doyle PJ.
94–106. (2005a) The WHO-DAS II: measuring outcomes of hearing
Hood JD, Poole JP. (1980) Influence of the speaker and other fac- aid intervention for adults. Trends Amplif. 9, 127–143.
tors affecting speech intelligibility. Audiology. 19, 434–455. McArdle R, Wilson RH, Burks CA. (2005b) Speech recognition
House AS, Williams CE, Hecker MHL, Kryter KD. (1955) Artic- in multitalker babble using digits, words, and sentences. J Am
ulation-testing methods: Consonantal differentiation with a Acad Audiol. 16, 726–739.
closed-response set. J Acoust Soc Am. 37, 158–166. McCullough JA, Wilson RH, Birck JD, Anderson LG. (1994) A mul-
Hudgins CV, Hawkins JE Jr, Karlin JE, Stevens SS. (1947) The timedia approach for estimating speech recognition in multi-
development of recorded auditory tests for measuring hearing lingual clients. Am J Audiol. 3, 19–22.
loss for speech. Laryngoscope. 57, 57–89. Miller GA. (1956) The magical number seven, plus or minus two:
Institute of Electrical and Electronics Engineers. (1969) IEEE rec- Some limits on our capacity for processing information. Psychol
ommended practice for speech quality measurements. IEEE Rev. 63, 81–97.
Trans Audio Electroacoust. 17, 227–246. Moog JS, Geers AE. (1990) Early Speech Perception Test for Profoundly
Jerger J, Hayes D. (1977) Diagnostic speech audiometry. Arch Deaf Children. St. Louis, MO: Central Institute for the Deaf.
Otolaryngol. 103, 216–222. Nabelek A, Robinson P. (1982) Monaural and binaural speech per-
Jerger S, Jerger J. (1984) Pediatric Speech Intelligibility Test. St. Louis, ception in reverberation for listeners of various ages. J Acoust
MO: Auditec. Soc Am. 71, 1242–1248.
Kalikow DN, Stevens KN, Elliot LL. (1977) Development of a Nilsson M, Soli S, Sullivan J. (1994) Development of the Hearing in
test of speech intelligibility in noise using sentence materials Noise Test for the measurement of speech reception thresholds
with controlled word predictability. J Acoust Soc Am. 61, in quiet and in noise. J Acoust Soc Am. 95, 1085–1099.
1337–1351. Olsen WO, Van Tassell DJ, Speaks CE. (1997) Phoneme and word
Kamm CA, Morgan DE, Dirks DD. (1983) Accuracy of adaptive recognition for words in isolation and in sentences. Ear Hear.
procedure estimates of PB-max level. J Speech Hear Disord. 48, 18, 175–188.
202–209. Owens E, Kessler DT, Raggio MW, Schubert ED. (1985) Analysis and
Katz J. (1962) The use of staggered spondaic words for assessing revision of the Minimum Auditory Capabilities (MAC) battery.
the integrity of the central auditory nervous system. J Audit Ear Hear. 6, 280–290.
Res. 2, 327–337. Owens E, Schubert ED. (1977) Development of the California
Killion MC, Niquette PA. (2000) What can the pure-tone audio- Consonant Test. J Speech Hear Res. 20, 463–474.
gram tell us about a patient’s SNR loss? Hear J. 53, 46–53. Pederson OT, Studebaker GA. (1972) A new minimal-contrast
Killion MC, Niquette PA, Gudmundsen GI, Revit LJ, Banerjee closed-response-set speech test. J Audit Res. 12, 187–195.
S. (2004). Development of a quick speech-in-noise test for Penrod JP. (1994) Speech threshold and word recognition/discrim-
measuring signal-to-noise ratio loss in normal-hearing and ination testing. In: Katz J, ed. Handbook of Clinical Audiology.
hearing-impaired listeners. J Acoust Soc Am. 116, 2395–2405. 4th ed. Baltimore, MD: Williams & Wilkins; pp 147–164.
Kimura D. (1961) Some effects of temporal lobe damage on audi- Plomp R. (1978) Auditory handicap of hearing impairment and the
tory perception. Can J Psychol. 15, 157–1165. limited benefit of hearing aids. J Acoust Soc Am. 63, 533–549.
Kirk KI, Pisoni DB, Osberger MJ. (1995) Lexical effects of unspo- Punch J, Joseph A, Rakerd B. (2004) Most comfortable and uncom-
ken word recognition by pediatric cochlear implant users. Ear fortable loudness levels: six decades of research. Am J Audiol.
Hear. 16, 470–481. 13, 144–157.
Konkle DF, Rintelmann WF. (1983) Introduction to speech audi- Roeser RJ, Clark JL. (2008). Live voice speech recognition audio-
ometry. In: Konkle DF, Rindtelman WF, eds. Principles of metry – stop the madness! Audiol Today. 20,32–33.
Speech Audiometry. Baltimore, MD: University Park Press; Ross M, Lerman J. (1970) A picture identification task for hearing-
pp 1–10. impaired children. J Speech Hear Res. 13, 44–53.
Kosky C, Boothroyd A. (2003) Validation of an on-line implemen- Silverman SR, Hirsh IJ. (1955) Problems related to the use of speech
tation of the Imitative Test of Speech Pattern Contrast Percep- in clinical audiometry. Ann Otol Rhinol Laryngol. 64, 1234–1244.
tion (IMSPAC). J Am Acad Audiol. 14, 72–83. Speaks C, Jerger J. (1965) Performance-intensity characteristics of
Kruel EJ, Bell DW, Nixon JC. (1969) Factors affecting speech dis- synthetic sentences. J Speech Hear Res. 9, 305–312.
crimination test difficulty. J Speech Hear Res. 12, 281–287. Sperry JL, Wiley TL, Chial MR. (1997) Word recognition per-
Kruel EJ, Nixon JC, Kryter KD, Bell DW, Lang JS, Schubert ED. formance in various background competitors. J Am Acad
(1968) A proposed clinical test of speech discrimination. Audiol. 8, 71–80.
J Speech Hear Res. 11, 536–552. Thornton AR, Raffin MJM. (1978) Speech-discrimination scores
Lehiste I, Peterson G. (1959) Linguistic considerations in the study modeled as a binomial variable. J Speech Hear Res. 21, 507–518.
of speech intelligibility. J Acoust Soc Am. 31, 280–286. Tillman TW, Carhart R. (1966) An Expanded Test for Speech Dis-
Levitt H, Resnick S. (1978) Speech reception by the hearing crimination Utilizing CNC Monosyllabic Words. Northwestern
impaired. Scand Audiol. 6 (suppl), 107–130. University Auditory Test No. 6. Brooks Air Force Base, TX: US
Martin FN, Champlin CA, Chambers JA. (1998) Seventh survey of Air Force School of Aerospace Medicine.
audiometric practices in the United States. J Am Acad Audiol. Tillman TW, Carhart R, Wilber L. (1963) A Test for Speech Discrim-
9, 95–104. ination Composed of CNC Monosyllabic Words. Northwestern
 CHAPTER 5
Ş
4QFFDI
"VEJPNFUSZ 75

University Auditory Test No. 4. Technical Documentary Report Wilson RH, Morgan DE, Dirks DD. (1973) A proposed SRT pro-
No. SAM-TDR-62–135. Brooks Air Force Base, TX: US Air cedure and its statistical precedent. J Speech Hear Disord. 38,
Force School of Aerospace Medicine. 184–191.
Tillman TW, Jerger JF. (1959) Some factors affecting the spondee Wilson RH, Preece JP, Salamon DL, Sperry JL, Bornstein SP. (1994)
threshold in normal-hearing subjects. J Speech Hear Res. 2, Effects of time compression and time compression plus rever-
141–146. beration on the intelligibility of Northwestern University
Wesilender P, Hodgson WR. (1989) Evaluation of four Spanish Auditory Test No. 6. J Am Acad Audiol. 5, 269–277.
word recognition ability lists. Ear Hear. 10, 387–392. Wilson RH, Preece JP, Thornton AR. (1990) Clinical use of the
Wiley TL, Stoppenbach DT, Feldhake LI, Moss KA, Thordardottir compact disc in speech audiometry. ASHA. 32, 3247–3251.
ET. (1995) Audiologic practices: What is popular versus what Wilson RH, Strouse A. (1999) Auditory measures with speech sig-
is supported by evidence. Am J Audiol. 4, 26–34. nals. In: Musiek FE, Rintelmann WF, eds. Contemporary Per-
Wilson RH. (2003) Development of a speech in multitalker babble spectives in Hearing Assessment. Needham Heights, MA: Allyn
paradigm to assess word-recognition performance. J Am Acad & Bacon; pp 21–66.
Audiol. 14, 453–470. Wilson RH, Strouse A. (2002) Northwestern University Audiology
Wilson RH, Abrams HB, Pillion AL. (2003) A word-recognition Test #6 in multitalker bubble: A preliminary report. J Rehabil
task in multitalker babble using a descending presentation Res Dev. 39, 105–113.
mode from 24 dB to 0 dB in signal to babble. J Rehabil Res Dev. Wilson RH, Trivette CP, Williams DA, Watts KA. (2012). The
40, 321–328. effects of energetic and informational masking on the Words-
Wilson RH, Margolis RH. (1983) Measurements of auditory in-Noise Test (WIN). J Am Acad Audiol. 23, 522–533.
thresholds for speech stimuli. In: Konkle DF, Rintelmann WF, Wilson RH, Weakley DG. (2004) The use of digit triplets to evalu-
eds. Principles of Speech Audiometry. Baltimore, MD: Univer- ate word-recognition abilities in multitalker babble. Semin
sity Park Press; pp 79–126. Hear. 25, 93–111.
Wilson RH, McArdle R. (2005) Speech signals used to evaluate the Young L, Dudley B, Gunter MB. (1982) Thresholds and psycho-
functional status of the auditory system. J Rehabil Res Dev. 42 metric functions of the individual spondaic words. J Speech
(suppl 2), 79–94. Hear Res. 25, 586–593.
 C H A P T ER 6

Clinical Masking

William S. Yacullo

In the first edition of this text, Sanders (1972) wrote the fol- a masking stimulus must be applied to the nontest ear to
lowing introduction to his chapter on clinical masking: eliminate its participation.
Of all the clinical procedures used in auditory assess-
ment, masking is probably the most often misused
and the least understood. For many clinicians the Air-Conduction Testing
approach to masking is a haphazard hit-or-miss bit Cross hearing occurs when a stimulus presented to the test
of guesswork with no basis in any set of principles. ear “crosses over” and is perceived in the nontest ear. There
(p 111) are two parallel pathways by which sound presented through
an earphone (i.e., an air-conduction transducer) can reach
Unfortunately, this statement may still hold true today.
the nontest ear. Specifically, there are both bone-conduction
The principles of clinical masking are difficult for many
and air-conduction pathways between an air-conduction
beginning clinicians to understand. Although the clinician
signal presented at the test ear and the sound reaching the
can apply masking formulas and procedures appropriately
nontest ear cochlea (Studebaker, 1979). First, the earphone
in most clinical situations, a lack of understanding of the
can vibrate with sufficient force to cause deformations of
underlying theoretical constructs becomes evident dur-
the bones of the skull. An earphone essentially can function
ing cases where modification of a standard procedure is
as a bone vibrator at higher sound pressures. Because both
required. A lack of understanding of the underlying con-
cochleas are housed within the same skull, the outcome is
cepts of masking often leads to misuse of clinical procedures.
stimulation of the nontest ear cochlea through bone con-
Theoretical and empirical bases of masking provide a
duction. Second, sound from the test earphone can travel
strong foundation for the understanding of applied clini-
around the head to the nontest ear, enter the opposite ear
cal masking procedures. It will become evident throughout
canal, and finally reach the nontest ear cochlea through an
this chapter that there is not a single “correct” approach to
air-conduction pathway. Because the opposite ear typically
clinical masking. Any approach to clinical masking that is
is covered during air-conduction testing, sound attenuation
based on sound theoretical constructs and verified through
provided by the earphone will greatly minimize or elimi-
clinical experience is correct. One approach will not meet
nate the contribution of the air-conduction pathway to the
all clinical needs. A strong foundation in the underlying
process of cross hearing. Consequently, cross hearing dur-
concepts of clinical masking serves three purposes. First,
ing air-conduction testing is considered primarily a bone-
it allows the clinician to make correct decisions about the
conduction mechanism.
need for masking. Second, it allows the clinician to make a
Cross hearing is the result of limited interaural attenu-
well-informed decision when selecting a specific approach
ation (IA). IA refers to the “reduction of energy between
to clinical masking. Finally, it allows the clinician to apply
ears.” Generally, it represents the amount of separation or
and modify a clinical masking procedure appropriately.
the degree of isolation between ears during testing. Specifi-
cally, it is the decibel difference between the hearing level
THE NEED FOR MASKING (HL) of the signal at the test ear and the HL reaching the
nontest ear:
A major objective of the basic audiologic evaluation is
assessment of auditory function of each ear. There are situ- IA = dB HL Test Ear – dB HL Nontest Ear
ations during both air-conduction and bone-conduction
testing when this may not occur. Although a puretone or Consider the following hypothetical examples pre-
speech stimulus is being presented through a transducer to sented in Figure 6.1. You are measuring puretone air-
the test ear, the nontest ear can contribute partially or totally conduction threshold using traditional supra-aural
to the observed response. Whenever it is suspected that the earphones. A puretone signal of 90 dB HL is presented to the
nontest ear is responsive during evaluation of the test ear, test ear. Because of limited IA, a portion of the test signal can

77
78 SECTION I • Basic Tests and Procedures

contact with the skin covering the cranial skull. Insert ear-
phones are coupled to the ear by insertion into the ear canal.
Generally, IA increases as the contact area of the trans-
ducer with the skull decreases (Zwislocki, 1953). More
specifically, IA is greater for supra-aural than circumaural
earphones. Furthermore, IA is greatest for insert earphones
(Killion et al., 1985; Sklare and Denenberg, 1987), partly
because of their smaller contact area with the skull. (The
reader is referred to Killion and Villchur, 1989; Zwislocki
A et al., 1988, for a review of advantages and disadvantages of
earphones in audiometry.) Because supra-aural and insert
earphones are most typically used during audiologic testing,
they will be the focus of this discussion.
There are different approaches to measuring IA for air-
conducted sound (e.g., “masking” method, “compensation”
method, method of “best beats”; the reader is referred to
Zwislocki, 1953, for discussion). The most direct approach,
however, involves measurement of transcranial thresholds
(Berrett, 1973). Specifically, IA is measured by obtaining
unmasked air-conduction (AC) thresholds in subjects with
B unilateral, profound sensory/neural hearing loss and then
FIGURE 6.1 Interaural attenuation (IA) is calculated as calculating the threshold difference between the normal and
the difference between the hearing level (HL) of the sig- impaired ears:
nal at the test ear and the HL reaching the nontest ear
cochlea. A puretone signal of 90 dB HL is being presented IA = Unmasked AC Impaired Ear – Unmasked AC Normal Ear
to the test ear through traditional supra-aural earphones.
Example A: If IA is 40 dB, then 50 dB HL is reaching the For example, if unmasked air-conduction thresholds are
nontest ear cochlea. Example B: If IA is 80 dB, then 10 obtained at 60 dB HL in the impaired ear and 0 dB HL in
dB HL is reaching the nontest ear cochlea. (From Yacullo the normal ear, then IA is calculated as 60 dB:
WS. (1996) Clinical Masking Procedures. 1st ed. Boston,
MA: Allyn & Bacon, © 1996, p 3. Adapted by permission of IA = 60 dB HL – 0 dB HL
Pearson Education, Inc., Upper Saddle River, NJ.) = 60dB
There is the assumption that air- and bone-conduction
thresholds are equal (i.e., no air-bone gaps) in the ear with
reach the nontest ear cochlea. If IA is 40 dB, then 50 dB HL normal hearing.
theoretically is reaching the nontest ear: Figure 6.2 illustrates the expected unmasked puretone
air-conduction thresholds in an individual with normal
IA = dB HL Test Ear – dB HL Nontest Ear hearing in the left ear and a profound sensory/neural hearing
= 90 dB HL − 50 dB HL loss in the right ear. Unmasked bone-conduction thresholds,
regardless of bone vibrator placement, are expected at HLs
= 40 dB
consistent with normal hearing in the left ear. If appropriate
If IA is 80 dB, then only 10 dB HL is reaching the nontest contralateral masking is not used during air-conduction test-
ear. It should be apparent that a greater portion of the test ing, then a shadow curve will result in the right ear. Because
signal can reach the nontest ear when IA is small. Depend- cross hearing is primarily a bone-conduction mechanism,
ing on the hearing sensitivity in the nontest ear, cross hear- unmasked air-conduction thresholds in the right ear will
ing can occur. “shadow” the bone-conduction thresholds in the left (i.e.,
IA during earphone testing is dependent on three fac- better) ear by the amount of IA. For example, if IA for air-
tors: Transducer type, frequency spectrum of the test sig- conducted sound is equal to 60 dB at all frequencies, then
nal, and individual subject. There are three major types of unmasked air-conduction thresholds in the right ear theo-
earphones currently used during audiologic testing: Supra- retically will be measured 60 dB above the bone-conduction
aural, circumaural, and insert (American National Stan- thresholds in the better ear. The shadow curve does not
dards Institute/Acoustical Society of America [ANSI/ASA], represent true hearing thresholds in the right ear. Rather, it
2010). Supra-aural earphones use a cushion that makes reflects cross-hearing responses from the better (i.e., left) ear.
contact solely with the pinna. Circumaural earphones use When using supra-aural earphones, IA for puretone
a cushion that encircles or surrounds the pinna, making air-conducted signals varies considerably, particularly across
 CHAPTER 6 • Clinical Masking 79

FIGURE 6.2 Expected unmasked puretone air- and bone-conduction thresholds in an

individual with normal hearing in the left ear and a profound sensory/neural hearing
loss in the right ear.Without the use of appropriate contralateral masking, a shadow
curve will result in the right ear. Unmasked air-conduction thresholds in the right ear
will shadow the bone-conduction thresholds in the better (i.e., left) ear by the amount
of interaural attenuation.(From Yacullo WS. (1996) Clinical Masking Procedures. 1st ed.
Boston, MA: Allyn & Bacon, © 1996, p 7. Adapted by permission of Pearson Education,
Inc., Upper Saddle River, NJ.)

subjects, ranging from about 40 to 85 dB (Berrett, 1973; ified length, a nipple adaptor, and a disposable foam eartip.
Chaiklin, 1967; Coles and Priede, 1970; Killion et al., 1985; A major advantage of the 3A insert earphone is increased
Sklare and Denenberg, 1987; Smith and Markides, 1981; IA for air-conducted sound, particularly in the lower fre-
Snyder, 1973). Your assumption about IA will influence the quencies (Hosford-Dunn et al., 1986; Killion et al., 1985;
decision about the need for contralateral masking. The use Sklare and Denenberg, 1987; Van Campen et al., 1990). This
of a smaller IA value assumes that there is smaller separa- is clearly illustrated in the results of a study by Killion et al.
tion between ears. Consequently, contralateral masking will (1985) (Figure 6.3).
be required more often. When making a decision about the Increased IA with 3A insert earphones is the result of
need for contralateral masking during clinical practice, a two factors: (1) Reduced contact area of the transducer with
single value defining the lower limit of IA is recommended the skull and (2) reduction of the occlusion effect (OE).
(Studebaker, 1967a). Zwislocki (1953) evaluated IA for three types of earphones:
Based on currently available data, a conservative estimate circumaural, supra-aural, and insert. Results suggested that
of IA for supra-aural earphones is 40 dB at all frequencies. IA for air-conducted sound increased as the contact area of
Although this very conservative estimate will take into the earphone with the skull decreased. When an acoustic
account the IA characteristics of all individuals, it will result signal is delivered through an earphone, the resultant sound
in the unnecessary use of masking in some instances. pressure acts over a surface area of the skull determined by
Commonly used insert earphones are the Etymotic the earphone cushion. The surface area associated with a
Research ER-3A (Killion, 1984) and the E-A-RTONE 3A small eartip will result in a smaller applied force to the skull,
(E-A-R Auditory Systems, 1997). The ER-3A and the E-A- resulting in reduced bone-conduction transmission.
RTONE 3A insert earphones are considered functionally Chaiklin (1967) has also suggested that IA may be
equivalent because they are built to identical specifications increased in the low frequencies with a deep insert because
(Frank and Vavrek, 1992). Each earphone consists of a of a reduction of the OE. ANSI/ASA (2010) defines the OE
shoulder-mounted transducer, a plastic sound tube of spec- as an increase in loudness for bone-conducted sound at
80 SECTION I • Basic Tests and Procedures

the “button” transducer (Blackwell et al., 1991; Hosford-

Dunn et al., 1986). Blackwell et al. (1991) compared the IA
obtained with a standard supra-aural earphone (TDH-50P)
and a button transducer fitted with a standard immittance
probe cuff. Although greater IA was observed with the but-
ton transducer, the difference between the insert and supra-
aural earphone did not exceed 10 dB at any frequency.
There are only limited data available regarding IA of
3A insert earphones using deeply or intermediately inserted
foam eartips. IA values vary across subjects and frequency,
ranging from about 75 to 110 dB at frequencies of ≤1,000 Hz
and about 50 to 95 dB at frequencies >1,000 Hz (Killion
et al., 1985; Sklare and Denenberg, 1987; Van Campen et al.,
1990). Based on Studebaker’s (1967a) recommendation, we
will again use the smallest IA values reported when making
a decision about the need for contralateral masking. To take
advantage of the significantly increased IA proved by the 3A
insert in the lower frequencies, a single value of IA will not
be employed across the frequency range.
FIGURE 6.3 Average and range of interaural attenu- Based on currently available data, conservative estimates
ation values obtained on six subjects using two ear- of IA for 3A insert earphones with deeply inserted foam eartips
phones: TDH-39 encased in MX-41/AR supra-aural are 75 dB at ≤1,000 Hz and 50 dB at frequencies >1,000 Hz.
cushion (●) and ER-3A insert earphone with deeply The IA values recommended clinically for 3A earphones
inserted foam eartip (■). (From Killion MC, Wilber LA, assume that deeply inserted foam eartips are used. Maximum
Gudmundsen GI. (1985) Insert earphones for more inte- IA is achieved in the low frequencies when a deep eartip
raural attenuation. Hear Instrum. 36, 34, 36. Reprinted
insertion is used (Killion et al., 1985). The recommended
with permission from Hearing Instruments, 1985, p 34.
deep insertion depth is achieved when the outer edge of the
Hearing Instruments is a copyrighted publication of
Advanstar Communications Inc. All rights reserved.) eartip is 2 to 3 mm inside the entrance of the ear canal. Con-
versely, a shallow insertion is obtained when the outer edge
of the eartip protrudes from the entrance of the ear canal
frequencies below 2,000 Hz when the outer ear is covered (E-A-R Auditory Systems, 1997). An intermediate insertion
or occluded. There is evidence that the OE influences the is achieved when the outer edge of the eartip is flush with the
measured IA for air-conducted sound (e.g., Berrett, 1973; opening of the ear canal (Van Campen et al., 1990). There
Chaiklin, 1967; Feldman, 1963; Killion et al., 1985; Littler are limited data suggesting that IA is similar for either inter-
et al., 1952; Van Campen et al., 1990; Zwislocki, 1953). In mediate or deep insertion of the foam eartip. However, a
fact, there is an inverse relationship between magnitude shallow insertion appears to significantly reduce IA (Killion
of the OE and the measured IA in the lower frequencies. et al., 1985; Sklare and Denenberg, 1987; Van Campen et al.,
Specifically, an earphone that reduces the OE will exhibit 1990). Remember that a major factor contributing to supe-
increased IA for air-conducted sound. Recall that cross rior IA of the 3A insert earphone is a significantly reduced
hearing occurs primarily through the mechanism of bone OE. There is evidence that the OE is negligible when using
conduction. When the nontest ear is covered or occluded either deeply or intermediately inserted insert earphones.
by an air-conduction transducer, the presence of an OE will In fact, the advantage of a greatly reduced OE is lost when
enhance hearing sensitivity for bone-conducted sound in a shallow insertion is used (Berger and Kerivan, 1983). To
that ear. Consequently, the separation between ears (i.e., achieve maximum IA with 3A insert earphones, deeply
IA) is reduced. The increased IA for air-conducted sound inserted eartips are strongly recommended.
observed in the lower frequencies when using 3A insert More recently, E-A-R Auditory Systems (2000a, 2000b)
earphones (with deeply inserted foam eartips) is primarily introduced a next-generation insert earphone, the E-A-
related to the significant reduction or elimination of the OE. RTONE 5A. The lengthy plastic sound tube that conducted
The OE is presented in greater detail later in this chapter sound from the body-level transducer of the 3A has been
in the section on clinical masking procedures during bone- eliminated in the 5A model; rather, the foam eartip is cou-
conduction audiometry. pled directly to an ear-level transducer. Very limited data
If increased IA is a primary goal when selecting an obtained with only two subjects (unpublished research by
insert earphone, then the 3A is the transducer of choice. Evi- Killion, 2000, as cited in E-A-R Auditory Systems, 2000b)
dence suggests that the 3A insert earphone provides signifi- suggest that the average IA for puretone stimuli ranging
cantly greater IA, particularly in the lower frequencies, than from 250 to 4,000 Hz is equivalent (within approximately
 CHAPTER 6 • Clinical Masking 81

5 dB) to the average values reported for the 3A insert ear- ognition thresholds. Rather, a different response task when
phone (Killion et al., 1985). measuring different speech thresholds in each ear (i.e., SDT
IA for speech is typically measured by obtaining speech in one ear and SRT in the other) can affect the measured
recognition thresholds (SRTs) in individuals with unilateral, IA for speech. Comparison of SRTs between ears or SDTs
profound sensory/neural hearing loss. Specifically, the dif- between ears generally should result in the same measured
ference in threshold between the normal ear and impaired IA. Smith and Markides (1981) measured IA for speech in
ear without contralateral masking is calculated: 11 subjects with unilateral, profound hearing loss. IA was
calculated as the difference between the SDT in the better
IA = Unmasked SRTImpaired Ear – SRTNormal Ear ear and the unmasked SDT in the poorer ear. The range
of IA values was 50 to 65 dB. It is interesting to note that
Recall that SRT represents the lowest HL at which speech the lowest IA value reported for speech using a detection
is recognized 50% of the time (ANSI/ASA, 2010; Ameri- task in each ear was 50 dB, a value comparable to the lowest
can Speech-Language-Hearing Association [ASHA], 1988). minimum reported IA value (i.e., 48 dB) for spondaic words
IA for spondaic words presented through supra-aural ear- (e.g., Martin and Blythe, 1977; Snyder, 1973).
phones varies across subjects and ranges from 48 to 76 dB There is also some evidence that it may be appropri-
(Martin and Blythe, 1977; Sklare and Denenberg, 1987; Sny- ate to use a more conservative estimate of IA when making
der, 1973). Again, a single value defining the lower limit of a decision about the need for contralateral masking during
IA is recommended when making a decision about the need assessment of suprathreshold speech recognition. Although
for contralateral masking (Studebaker, 1967a). A conserva- IA for the speech signal remains constant during measure-
tive estimate of IA for spondees, therefore, is 45 dB when ment of threshold or suprathreshold measures of speech
using supra-aural earphones (Konkle and Berry, 1983). The recognition (i.e., the decibel difference between the level of
majority of audiologists measure SRT using a 5-dB step size the speech signal at the test ear and the level at the non-
(Martin et al., 1998). Therefore, the IA value of 48 dB is typ- test ear cochlea), differences in the performance criterion
ically rounded down to 45 dB. for each measure must be taken into account when select-
There is considerable evidence that speech can be ing an appropriate IA value for clinical use. SRT is defined
detected at a lower HL than that required to reach SRT. relative to a 50% response criterion. However, suprathresh-
Speech detection threshold (SDT) is defined as the lowest old speech recognition performance can range from 0% to
HL at which speech can be detected or “discerned” 50% 100%.
of the time (ASHA, 1988). The SRT typically requires an Konkle and Berry (1983) provide an excellent ratio-
average of about 8 to 9 dB greater HL than that required nale for the use of a more conservative estimate of IA when
for the detection threshold (Beattie et al., 1978; Chaiklin, measuring suprathreshold speech recognition. They suggest
1959; Thurlow et al., 1948). Given this relationship between that the fundamental difference in percent correct criterion
the two speech thresholds, Yacullo (1996) has suggested requires the specification of nontest ear cochlear sensitiv-
that a more conservative value of IA may be appropriate ity in a different way than that used for threshold measure-
when considering the need for contralateral masking dur- ment. If suprathreshold speech recognition materials are
ing measurement of SDT. presented at an HL equal to the SRT, then a small percent-
Consider the following hypothetical example. You are age of the test items can be recognized. It should be noted
measuring speech thresholds in a patient with normal hear- that the percentage of test words that can be recognized
ing in the right ear and a profound, sensory/neural hear- at an HL equal to SRT is dependent on the type of speech
ing loss in the left ear. If the patient exhibits the minimum stimuli, as well as on the talker and/or recorded version of
reported IA value for speech of 48 dB, then an SRT of 0 dB a speech recognition test. Regardless of the type of speech
HL would be measured in the right ear and an unmasked stimulus (e.g., meaningful monosyllabic words, nonsense
SRT of 48 dB HL would be measured in the left ear. If an syllables, or sentences) and the specific version (i.e., talker/
unmasked SDT is subsequently measured in the left ear, it is recording) of a speech recognition test, 0% performance
predicted that the threshold would occur at an HL of about may not be established until an HL of about −10 dB rela-
8 to 9 dB lower than the unmasked SRT. An unmasked SDT tive to the SRT. Konkle and Berry (1983) recommend that
would be expected to occur at about 39 to 40 dB HL. Com- the value of IA used for measurement of suprathreshold
parison of the unmasked SDT in the impaired ear with the speech recognition should be estimated as 35 dB. That is,
SRT in the normal ear theoretically would result in mea- the IA value of 45 dB (rounded down from 48 dB) based
sured IA of approximately 39 to 40 dB. When an unmasked on SRT measurement is adjusted by subtracting 10 dB. This
SDT is measured and the response is compared to the SRT adjustment in the estimate of IA reflects differences in per-
in the nontest ear, a more conservative estimate of IA for cent correct criterion used for speech threshold and supra-
speech may be appropriate. threshold measurements.
It should be noted that the actual IA for speech does not The majority of audiologists use an IA value of 40 dB
change during measurement of speech detection and rec- for all air-conduction measurements, both puretone and
82 SECTION I • Basic Tests and Procedures

speech, when making a decision about the need for con- into vibration, both cochleas can be potentially stimulated.
tralateral masking (Martin et al., 1998). The use of a single Consequently, an unmasked bone-conduction threshold
IA value of 40 dB for both threshold and suprathreshold can reflect a response from either cochlea or perhaps both.
speech audiometric measurements can be supported. Given Although a bone vibrator may be placed at the side of the
the smallest reported IA value of 48 dB for spondaic words, test ear, it cannot be assumed that the observed response is in
a value of 40 dB is somewhat too conservative during mea- fact from that ear.
surement of SRT. However, it should prove adequate during Consider the following example. You have placed a bone
measurement of SDT and suprathreshold speech recogni- vibrator at the right mastoid process. A puretone signal of
tion when a more conservative estimate of IA (by approxi- 50 dB HL is presented. If IA is considered to be 0 dB, then it
mately 10 dB) may be appropriate. should be assumed that a signal of 50 dB HL is potentially
Unfortunately, there are only very limited data available reaching both cochleas. It should be apparent that there is
about IA for speech when using insert earphones. Sklare essentially no separation between the two cochleas during
and Denenberg (1987) reported IA for speech (i.e., SRT unmasked bone-conduction audiometry.
using spondaic words) in seven adults with unilateral, pro- Based on currently available data, a conservative estimate
found sensory/neural hearing loss using ER-3A insert ear- of IA for bone-conducted sound is 0 dB at all frequencies.
phones. IA ranged from 68 to 84 dB. It should be noted that
the smallest reported value of IA for spondaic words (i.e.,
68 dB) is 20 dB greater when using 3A insert earphones with WHEN TO MASK
deeply inserted foam eartips (Sklare and Denenberg, 1987) Contralateral masking is required whenever there is the pos-
than when using supra-aural earphones (i.e., 48 dB) (Mar- sibility that the test signal can be perceived in the nontest ear.
tin and Blythe, 1977; Snyder, 1973). Therefore, a value of IA is one of the major factors that will be considered when
60 dB represents a very conservative estimate of IA for evaluating the need for masking. The basic principles under-
speech when using 3A insert earphones. This value is derived lying the decision-making processes of when to mask during
by adding a correction factor of 20 dB to the conservative IA puretone and speech audiometry will now be addressed.
value used with supra-aural earphones (i.e., 40 dB) for all
threshold and suprathreshold measures of speech.
Based on currently available data, conservative estimates Puretone Audiometry: Air
of IA for all threshold and suprathreshold measures of speech Conduction
are 40 dB for supra-aural earphones and 60 dB for 3A insert
When making a decision about the need for masking dur-
earphones with deeply inserted foam eartips.
ing puretone air-conduction testing, three factors need to
be considered: (1) IA, (2) unmasked air-conduction thresh-
Bone-Conduction Testing old in the test ear (i.e., HL at the test ear), and (3) bone-
conduction hearing sensitivity (i.e., threshold) in the non-
There are two possible locations for placement of a bone
test ear. Recall that when cross hearing occurs, the nontest
vibrator (typically, the Radioear B-71) during puretone
ear is stimulated primarily through the bone-conduction
threshold audiometry: The mastoid process of the tempo-
mechanism. When a decision is made about the need
ral bone and the frontal bone (i.e., the forehead). Although
for contralateral masking, the unmasked air-conduction
there is some evidence that a forehead placement produces
threshold in the test ear (ACTest Ear) is compared to the bone-
more reliable and valid thresholds than a mastoid place-
conduction threshold in the nontest ear (BCNontest Ear). If
ment (see Dirks, 1994, for further discussion), the major-
the difference between ears equals or exceeds IA, then air-
ity (92%) of audiologists in the United States continue to
conduction threshold in the test ear must be remeasured
place a bone-conduction transducer on the mastoid process
using contralateral masking. The rule for when to mask dur-
(Martin et al., 1998).
ing puretone air-conduction testing can be stated as follows:
IA is greatly reduced during bone-conduction audiom-
Contralateral masking is required during puretone air-
etry. IA for bone-conducted sound when using a bone vibra-
conduction audiometry when the unmasked air-conduction
tor placed at the forehead is essentially 0 dB at all frequencies;
threshold in the test ear equals or exceeds the apparent bone-
IA when using a mastoid placement is approximately 0 dB at
conduction threshold (i.e., the unmasked bone-conduction
250 Hz and increases to about 15 dB at 4,000 Hz (Studebaker,
threshold) in the nontest ear by a conservative estimate of IA:
1967a). Regardless of the placement of a bone vibrator (i.e.,
mastoid vs. forehead), it is generally agreed that IA for bone- AC Test Ear − BC Nontest Ear ≥ IA
conducted sound at all frequencies is negligible and should
be considered 0 dB (e.g., Dirks, 1994; Hood, 1960; Sanders This rule is consistent with the guidelines for manual
and Rintelmann, 1964; Studebaker, 1967a). When a bone puretone threshold audiometry recommended by ASHA
vibrator, regardless of its location, sets the bones of the skull (2005).
 CHAPTER 6 • Clinical Masking 83

FIGURE 6.4 Audiogram illustrating the need for contralateral masking during
puretone air-conduction audiometry. See text for discussion.

Note that the term “apparent” bone-conduction thresh- First consider the need for contralateral masking
old is considered when making a decision about the need assuming that air-conduction thresholds were measured
for masking. Remember that an unmasked bone-conduc- using supra-aural earphones. A conservative estimate of IA
tion threshold does not convey ear-specific information. It is 40 dB. We will use the following equation when making a
is assumed that the bone-conduction response can originate decision about the need for contralateral masking:
from either or both ears. Therefore, the unmasked bone-
conduction response is considered the apparent or possible AC Test Ear – BC Nontest Ear ≥ IA
threshold for either ear.
Consider the unmasked puretone audiogram* pre- Because it is not possible to measure bone-conduction
sented in Figure 6.4. Because IA for bone-conducted sound threshold at 8,000 Hz, it is necessary to predict an unmasked
is considered 0 dB, unmasked bone-conduction thresholds threshold given the findings at other test frequencies. In this
are traditionally obtained at only one mastoid process. particular example, unmasked bone-conduction threshold
During air-conduction threshold testing, the potential for at 8,000 Hz will probably have a similar relationship with
cross hearing is greatest when there is a substantial differ- the air-conduction thresholds in the better (i.e., left) ear.
ence in hearing sensitivity between the two ears and when a Because there is no evidence of air-bone gaps at the adja-
stimulus is presented at higher HLs to the poorer ear. Con- cent high frequencies, we will assume that a similar relation-
sequently, there is greater potential for cross hearing when ship exists at 8,000 Hz. Therefore, our estimate of unmasked
measuring puretone thresholds in the right ear. bone-conduction threshold is 45 dB HL.
It will be necessary to remeasure puretone thresholds at
*The puretone audiogram and audiometric symbols used throughout all test frequencies in the right ear using contralateral mask-
this chapter are those recommended in ASHA’s (1990) most recent ing because the difference between ears equals or exceeds
guidelines for audiometric symbols (see Chapter 3). our estimate of IA.
84 SECTION I • Basic Tests and Procedures

Right Ear Masking unmasked bone-conduction thresholds before obtaining

(Test Ear) Needed? unmasked air-conduction thresholds. Decisions about the
need for masking during air-conduction testing then can be
250 Hz 55–0 ≥40? Yes
500 Hz 60–5 ≥40? Yes
made using the important bone-conduction responses.
1,000 Hz 80–25 ≥40? Yes 3A insert earphones are often substituted for the supra-
2,000 Hz 90–30 ≥40? Yes aural configuration during audiometric testing. We now
4,000 Hz 95–40 ≥40? Yes will take a second look at the audiogram in Figure 6.4 and
8,000 Hz 100–45 ≥40? Yes assume that air-conduction thresholds were obtained with
3A insert earphones. Recall that conservative estimates of IA
However, contralateral masking is not required when for 3A insert earphones with deeply inserted foam eartips
testing the left ear. The difference between ears does not are 75 dB at ≤1,000 Hz and 50 dB at frequencies >1,000 Hz.
equal or exceed the estimate of IA. Previously, we determined that contralateral masking was
not required when testing the better (i.e., left) ear using
Left Ear Masking supra-aural earphones. Given the greater IA offered by 3A
(Test Ear) Needed? insert earphones, it is easy to understand that contralateral
masking again should not be required when testing the left
250 Hz 20–0 ≥40? No ear. However, a different picture results when considering
500 Hz 25–5 ≥40? No the need for contralateral masking when testing the right ear.
1,000 Hz 30–25 ≥40? No
2,000 Hz 35–30 ≥40? No Right Ear Masking
4,000 Hz 40–40 ≥40? No (Test Ear) Needed?
8,000 Hz 45–45 ≥40? No
250 Hz 55–0 ≥75? No
500 Hz 60–5 ≥75? No
Many audiologists will obtain air-conduction thresh-
1,000 Hz 80–25 ≥75? No
olds prior to measurement of bone-conduction thresholds.
2,000 Hz 90–30 ≥50? Yes
A preliminary decision about the need for contralateral
4,000 Hz 95–40 ≥50? Yes
masking can be made by comparing the air-conduction 8,000 Hz 100–45 ≥50? Yes
thresholds of the two ears.
Contralateral masking is required during puretone air- Because of the greater IA provided by 3A insert ear-
conduction audiometry when the unmasked air-conduction phones in the lower frequencies, the need for contralateral
threshold in the test ear (ACTest Ear) equals or exceeds the air- masking is eliminated at 250, 500, and 1,000 Hz. It should
conduction threshold in the nontest ear (ACNontest Ear) by a con- be apparent that the process of evaluating the need for con-
servative estimate of IA: tralateral masking when using either supra-aural or insert
AC Test Ear − AC Nontest Ear ≥ IA earphones is the same. The only difference is the substitu-
tion of different values of IA in our equations.
It is important to remember, however, that cross hearing for
air-conducted sound occurs primarily through the mecha- Puretone Audiometry:
nism of bone conduction. Consequently, it will be necessary
to re-evaluate the need for contralateral masking during air-
Bone Conduction
conduction testing following the measurement of unmasked Remember that a conservative estimate of IA for bone-
bone-conduction thresholds. conducted sound is 0 dB. Theoretically, masked bone-
Consider again the audiogram presented in Figure 6.4. conduction measurements are always required if ear-specific
Let us assume that we have not yet measured unmasked bone- information is needed. However, given the goal of bone-
conduction thresholds. We can make a preliminary decision conduction audiometry, contralateral masking is not always
about the need for contralateral masking by considering the required. Generally, bone-conduction thresholds are pri-
difference between air-conduction thresholds in the two ears. marily useful for determining gross site of lesion (i.e., con-
Based on the air-conduction responses only, it appears that ductive, sensory/neural, or mixed). The presence of air-bone
contralateral masking is needed only when testing the right gaps suggests a conductive component to a hearing loss.
ear at octave frequencies from 1,000 through 8,000 Hz. Yet, The major factor to consider when making a decision
once unmasked bone-conduction thresholds are measured, about the need for contralateral masking during bone-
it becomes apparent that contralateral masking will also be conduction audiometry is whether the unmasked bone-
required when testing the right ear at 250 and 500 Hz. conduction threshold (Unmasked BC) suggests the presence
It is conventional to obtain air-conduction thresholds of a significant conductive component in the test ear.
prior to bone-conduction thresholds. However, an alter- The use of contralateral masking is indicated whenever
native (and recommended) approach involves obtaining the results of unmasked bone-conduction audiometry suggest
 CHAPTER 6 • Clinical Masking 85

the presence of an air-bone gap in the test ear (Air-Bone mal distribution of the relationship between air- and bone-
GapTest Ear) of 15 dB or greater: conduction thresholds in individuals without significant
air-bone gaps, then an air-bone difference of ±10 dB is not
Air-Bone Gap Test Ear ≥ 15 dB
unexpected.
where If unmasked bone-conduction thresholds suggest air-
bone gaps of 10 dB or less, then contralateral masking is not
Air-Bone Gap = AC Test Ear − Unmasked BC required. Although unmasked bone-conduction thresholds
do not provide ear-specific information, we have accom-
ASHA (2005) recommends that contralateral masking plished our goal for bone-conduction testing. If unmasked
should be used whenever a potential air-bone gap of 10 dB bone-conduction thresholds suggest no evidence of sig-
or greater exists. When taking into account the variability nificant air-bone gaps, then we have ruled out the presence
inherent in bone-conduction measurements (Studebaker, of a significant conductive component. Consequently, our
1967b), however, a criterion of 10 dB may be too stringent. assumption is that the hearing loss is sensory/neural in
There is a certain degree of variability between air- and nature.
bone-conduction threshold, even in individuals without Figure 6.5 provides three examples of the need for
conductive hearing loss. If we assume that there is a nor- contralateral masking during bone-conduction audiometry.

Frequency (Hz) Frequency (Hz)

Hearing level in decibles (dB HL)

Hearing level in decibles (dB HL)

A B
Frequency (Hz)
Hearing level in decibles (dB HL)

FIGURE 6.5 Audiograms illustrating the need for con-

tralateral masking during bone-conduction audiometry.
Example A: Masked bone-conduction thresholds are not
required in either ear. Example B: Masked bone-conduction
thresholds are required only in the right ear. Example C:
Masked bone-conduction thresholds are potentially required
C in both ears. See text for further discussion.
86 SECTION I • Basic Tests and Procedures

Unmasked air- and bone-conduction thresholds are pro- for contralateral masking during speech audiometry: (1) IA,
vided in each case. (2) presentation level of the speech signal (in dB HL) in the
Example A. Contralateral masking is not required dur- test ear, and (3) bone-conduction hearing sensitivity (i.e.,
ing bone-conduction testing in either ear. When we com- threshold) in the nontest ear.
pare the unmasked bone-conduction thresholds to the Contralateral masking is indicated during speech audi-
air-conduction thresholds in each ear, there are no potential ometry whenever the presentation level of the speech signal
air-bone gaps of 15 dB or greater. For example, consider the (in dB HL) in the test ear (Presentation LevelTest Ear) equals or
thresholds at 2,000 Hz. Comparison of the unmasked bone- exceeds the best puretone bone-conduction threshold in the
conduction threshold to the air-conduction thresholds sug- nontest ear (Best BCNontest Ear) by a conservative estimate of IA:
gests a potential air-bone gap of 5 dB in the right ear and
0 dB in the left ear. Because the unmasked bone-conduction Presentation Level Test Ear − Best BC Nontest Ear ≥ IA
threshold does not suggest the presence of significant air-
bone gaps in either ear, our conclusion is that the hearing Because speech is a broadband signal, it is necessary to
loss is sensory/neural bilaterally. Obtaining masked bone- consider bone-conduction hearing sensitivity at more than
conduction thresholds, although they would provide ear- a single puretone frequency. Konkle and Berry (1983) and
specific information, would not provide additional diag- Sanders (1991) recommend the use of the bone-conduction
nostic information. puretone average of 500, 1,000, and 2,000 Hz or some other
Example B. Comparison of unmasked bone-conduc- average that is predictive of the SRT. ASHA (1988) recom-
tion thresholds to the air-conduction thresholds in the left mends that the puretone bone-conduction thresholds at
ear does not suggest the presence of significant air-bone 500, 1,000, 2,000, and 4,000 Hz should be considered. Mar-
gaps. Consequently, masked bone-conduction thresholds tin and Blythe (1977) suggest that 250 Hz can be eliminated
are not required in the left ear. Our conclusion is that the from any formula for determining the need for contralat-
hearing loss is sensory/neural. eral masking when measuring the SRT. Yet, the nontest ear
However, masked bone-conduction thresholds will be bone-conduction threshold at 250 Hz may be an important
required in the right ear. Comparison of unmasked bone- consideration when measuring the SDT. Olsen and Matkin
conduction thresholds to the air-conduction thresholds (1991) state that the SDT may be most closely related to the
in the right ear suggests potential air-bone gaps ranging best threshold in the 250 to 4,000 Hz range when audio-
from 25 to 35 dB. The unmasked bone-conduction thresh- metric configuration steeply rises or slopes. Therefore, fol-
olds may reflect hearing in the better (i.e., left) ear. Bone- lowing the recommendation of Coles and Priede (1975), the
conduction thresholds in the right ear may be as good as the most conservative approach involves considering the best
unmasked responses. They also may be as poor as the air- bone-conduction threshold in the 250- to 4,000-Hz fre-
conduction thresholds in that ear. Because we do not have quency range.
ear-specific information for bone-conduction thresholds, The examples presented in Figures 6.6 and 6.7 illus-
the loss in the right ear can be either mixed or sensory/neu- trate the need for contralateral masking during threshold
ral. To make a definitive statement about the type of hearing and suprathreshold speech audiometry, respectively. First
loss, it will be necessary to obtain masked bone-conduction consider the audiogram presented in Figure 6.6. Audiom-
thresholds in the right ear. etry was performed using supra-aural earphones. Puretone
Example C. There is evidence that contralateral masking testing (using appropriate contralateral masking dur-
will be required when measuring bone-conduction thresh- ing both air- and bone-conduction audiometry) reveals a
olds in both ears. Comparison of unmasked bone-conduc- severe-to-profound, sensory/neural hearing loss of gradu-
tion thresholds to the air-conduction thresholds suggests ally sloping configuration in the right ear. There is a very
potential air-bone gaps ranging from 30 to 35 dB in each mild, sensory/neural hearing loss of relatively flat config-
ear. As in the previous example, bone-conduction thresh- uration in the left ear. Given the difference between ears
olds in each ear may be as good as the unmasked responses. observed during puretone audiometry, it is anticipated that
They may also be as poor as the air-conduction thresholds contralateral masking may be needed during assessment of
in that ear. To make a definitive statement about the type SRT in the poorer ear.
of hearing loss, it will be necessary to obtain masked bone- There are different approaches that can be used when
conduction thresholds in both ears. determining the need for contralateral masking during
measurement of SRT. The most efficient and recommended
approach involves predicting the speech threshold using
Speech Audiometry the puretone threshold data in the poorer ear and, on that
Because speech audiometry is an air-conduction procedure, basis, determining the need for contralateral masking. For
the rules for when to mask will be similar to those used dur- example, SRT is measured at 20 dB HL in the left ear, a find-
ing puretone air-conduction audiometry. There are three ing consistent with the puretone results. Given the relatively
factors to consider when making a decision about the need low HL at which the SRT was established in the better (i.e.,
 CHAPTER 6 • Clinical Masking 87

FIGURE 6.6 Audiogram illustrating the need for contralateral masking during measurement of
speech recognition threshold (SRT). See text for further discussion.
88 SECTION I • Basic Tests and Procedures

FIGURE 6.7 Audiogram illustrating the need for contralateral masking during measurement of supra-
threshold speech recognition. See text for further discussion.
 CHAPTER 6 • Clinical Masking 89

left) ear, it is expected that contralateral masking will not be the right ear (80 dB HL) and the measured SRT in the left
required when measuring the SRT. Specifically, the SRT of ear (20 dB HL) equals or exceeds 40 dB, our estimate of
20 dB HL in the left ear does not equal or exceed the best IA for speech:
bone-conduction threshold of 55 dB HL in the nontest ear
STTest Ear – STNontest Ear ≥ IA
by a conservative estimate of IA (40 dB):
Presentation Level Test Ear – Best BC Nontest Ear ≥ IA 80 dB HL – 20 dB HL ≥ 40 dB ? Yes

20 dBHL – 55 dBHL ≥ 40 dB? No An alternative approach that can be used when mak-
ing a decision about the need for contralateral masking
However, if we predict that an SRT will be measured at about during assessment of SRT involves measuring unmasked
80 dB HL in the right ear (based on the puretone average), speech thresholds in both ears. Consider again the example
then contralateral masking will be required because the esti- presented in Figure 6.6. Assume that unmasked SRTs were
mated speech threshold of 80 dB HL equals or exceeds the measured at 65 and 20 dB HL in the right and left ears,
best bone-conduction threshold of 15 dB HL in the nontest respectively. Again there is an indication that contralat-
ear by 40 dB, our estimate of IA for speech: eral masking will be required when measuring the SRT in
the right ear. The presentation level of 65 dB HL (i.e., the
Presentation Level Test Ear – Best BC Nontest Ear ≥ IA
unmasked SRT) in the test ear equals or exceeds the best
bone-conduction threshold of 15 dB HL in the nontest ear
80 dBHL – 15 dBHL ≥ 40 dB? Yes
by 40 dB, our estimate of IA for speech:
Stated differently, the difference between the predicted pre- Presentation Level Test Ear – Best BC Nontest Ear ≥ IA
sentation level in the test ear and the best bone-conduction
threshold in the nontest ear equals or exceeds our estimate 65 dBHL – 15 dBHL ≥ 40 dB ? Yes
of IA. It is important to note, however, that a decision about
the need for contralateral masking during measurement of Similarly, the difference between the unmasked SRT in the
speech threshold must always take into account not only the right ear (65 dB HL) and the measured SRT in the left ear
presentation level at the measured SRT, but also all supra- (20 dB HL) equals or exceeds our estimate of IA (40 dB).
threshold levels used during threshold measurement. This Although this approach can sometimes provide the audi-
will be discussed further in the section addressing selection ologist with a more accurate estimate of the patient’s IA for
of masking levels during speech audiometry. speech (which may be useful when selecting appropriate
During our earlier discussion of the need for contralat- masking levels), it often just increases the number of steps
eral masking during puretone air-conduction audiometry, needed to establish the true SRT in the test ear.
it was indicated that a correct decision about the need for The audiogram presented in Figure 6.7 illustrates the
contralateral masking can be made sometimes by simply need for contralateral masking during assessment of supra-
comparing the air-conduction thresholds of the two ears. threshold speech recognition. Puretone testing reveals
Similarly, a decision about the need for contralateral mask- normal hearing through 1,000 Hz, steeply sloping to a
ing during measurement of speech thresholds can be often severe-to-profound sensory/neural hearing loss in the high
made by comparing speech thresholds in the two ears. frequencies bilaterally. SRTs were measured at 0 dB HL in
Contralateral masking is required during measurement both ears, a finding consistent with the puretone findings.
of speech threshold when the speech threshold in the test ear Contralateral masking was not required during puretone
(STTest Ear) equals or exceeds the speech threshold in the nontest and speech threshold audiometry. Suprathreshold speech
ear (STNontest Ear) by a conservative estimate of IA: recognition will be assessed using the California Consonant
Test (CCT). This is a closed-set word recognition test that is
STTest Ear − STNontest Ear ≥ IA sensitive to the speech recognition difficulties of individu-
als with high-frequency hearing loss (Owens and Schubert,
Consider again the audiogram presented in Figure 6.6. 1977). If we use the recommended sensation level (SL) of
Recall that we predicted that SRT would be measured at 50 dB (Schwartz and Surr, 1979), then presentation level for
about 80 dB HL in the right ear. In this particular example, both ears will be 50 dB HL (i.e., 50 dB relative to the SRT of
comparison of the two speech thresholds (i.e., the mea- 0 dB HL).
sured SRT of 20 dB HL in the left ear and the predicted Let us now consider the need for contralateral mask-
SRT of 80 dB HL in the right ear) would lead us to a cor- ing during assessment of suprathreshold speech recogni-
rect decision about the need for contralateral masking tion. We will consider the need for masking using two types
when measuring SRT in the poorer ear without the need of air-conduction transducers: Supra-aural and 3A insert
to consider bone-conduction hearing sensitivity in the earphones. The advantage of insert earphones will become
nontest ear. The difference between the predicted SRT in apparent.
90 SECTION I • Basic Tests and Procedures

Let us assume that supra-aural earphones are being Consider the following example. Absolute threshold
used during speech audiometry. Contralateral masking will for a 1,000-Hz puretone stimulus is initially determined to
be required when assessing suprathreshold speech recogni- be 40 dB HL. Another sound, white noise, is now presented
tion in both ears because the difference between the presen- simultaneously to the same ear. Absolute threshold for the
tation level of 50 dB HL in the test ear and the best puretone 1,000-Hz signal is redetermined in the presence of the white
bone-conduction threshold of 0 dB HL in the nontest ear noise and increases to 60 dB HL. Sensitivity to the puretone
equals or exceeds 40 dB, our conservative estimate of IA for signal has been affected by the presence of the white noise.
speech: This increase in threshold of one sound in the presence of
another is defined as masking. Because the puretone thresh-
Presentation Level Test Ear – Best BC Nontest Ear ≥ IA old was raised by 20 dB (i.e., a threshold shift of 20 dB), the
white noise has produced 20 dB of masking.
Right Ear 50 dBHL – 0 dBHL ≥ 40 dB ? Yes There are two basic masking paradigms: ipsilateral and
contralateral. In an ipsilateral masking paradigm, the test
Left Ear 50 dBHL – 0 dBHL ≥ 40 dB ? Yes
signal and the masker are presented to the same ear. In a
contralateral masking paradigm, the test signal and masker
A different outcome results if we substitute 3A insert
are presented to opposite ears. Masking is used clinically
earphones for the supra-aural arrangement. Because of
whenever it is suspected that the nontest ear is participating
the greater IA offered by 3A insert earphones, contralat-
in the evaluation of the test ear. Consequently, masking is
eral masking will not be required when assessing supra-
always applied to the nontest or contralateral ear. Masking
threshold speech recognition in either ear. Specifically,
reduces sensitivity of the nontest ear to the test signal. The
the difference between the presentation level of 50 dB
purpose of contralateral masking, therefore, is to raise the
HL in the test ear and the best puretone bone-conduc-
threshold of the nontest ear sufficiently so that its contribu-
tion threshold of 0 dB HL in the nontest ear does not
tion to a response from the test ear is eliminated.
equal or exceed 60 dB, our conservative estimate of IA
for speech:

Presentation Level Test Ear – Best BC Nontest Ear ≥ IA Masking Noise Selection
Standard diagnostic audiometers provide three types of
Right Ear 50 dBHL – 0 dBHL ≥ 60 dB ? No masking stimuli: narrowband noise, speech spectrum noise,
and white noise. Our clinical goal is to select a masker that is
efficient (Hood, 1960). An efficient masker is one that pro-
Left Ear 50 dBHL – 0 dBHL ≥ 60 dB ? No
duces a given effective level of masking with the least overall
sound pressure level.
The example presented in Figure 6.7 illustrates two
To better understand this concept of masker efficiency,
important concepts related to assessment of suprathreshold
let us review the classic masking experiment conducted by
speech recognition. First, it should not be assumed that con-
Fletcher (1940). White noise is a broadband stimulus that
tralateral masking is never required when assessing individ-
contains equal energy across a broad range of frequencies.
uals with symmetrical sensory/neural hearing loss. Second,
Because of its broadband spectrum, it has the ability to
the need for contralateral masking often can be eliminated
mask puretone stimuli across a broad range of frequencies
by using an air-conduction transducer that provides greater
(Hawkins and Stevens, 1950). Fletcher addressed which fre-
IA (i.e., 3A insert earphone).
quency components of broadband noise contribute to the
masking of a tone.
Fletcher (1940) conducted what is known as a centered
MASKING CONCEPTS masking experiment. Initially, a very narrow band of noise
Before proceeding to a discussion of clinical masking pro- was centered around a puretone signal. The bandwidth of
cedures, a brief review of basic masking concepts, including the noise was progressively widened, and the masking effect
masking noise selection and calibration, will be presented. on the puretone signal was determined. Fletcher observed
Generally, masking relates to how sensitivity for one sound that the masked puretone threshold increased as the band-
is affected by the presence of another sound. ANSI/ASA width of the masking noise was increased. However, once the
(2010) defines masking as follows: noise band reached and then exceeded a “critical bandwidth,”
additional masking of the puretone signal did not occur.
The process by which the threshold of hearing for one
This concept of the critical band as first described by
sound is raised by the presence of another (masking)
Fletcher (1940) consists of two components:
sound. The amount by which the threshold of hearing
for one sound is raised by the presence of another 1. When masking a puretone with broadband noise, the
(masking) sound, expressed in decibels (p 7). only components of the noise that have a masking effect
 CHAPTER 6 • Clinical Masking 91

on the tone are those frequencies included within a nar- Calibration of Effective
row band centered around the frequency of the tone.
2. When a puretone is just audible in the presence of the
Masking Level
noise, the total noise power present in the narrow band When a masking noise is presented to the nontest ear, we are
of frequencies is equal to the power of the tone. interested in how much masking is produced. Consequently,
masking noise is calibrated in EM level (dB EM).
The first component of the critical band concept has clinical ANSI/ASA (2010) defines EM level for puretones as
implications when selecting an appropriate masker during “the sound pressure level of a band of noise whose geomet-
puretone audiometry. The second component has relevance ric center frequency coincides with that of a specific pure
when calibrating the effective masking (EM) level of the tone that masks the pure tone to 50% probability of detec-
masking stimulus. tion” (p 7). (Reference EM levels, calculated by adding an
White noise is adequate as a masker for puretone appropriate correction value to the reference equivalent
stimuli. However, it contains noise components that do threshold sound pressure level [RETSPL] at each frequency,
not contribute to the effectiveness of the masker. The addi- are provided in the current ANSI/ASA specification for
tional noise components outside the tone’s critical band audiometers.) It is also indicated that, in individuals with
simply add to the overall level (and loudness) of the mask- normal hearing, “the amount of effective masking . . . is equal
ing stimulus. Therefore, the most efficient masker for pur- to the number of decibels that a given band of noise shifts a
etone stimuli is a narrow band of noise with a bandwidth pure-tone threshold . . . when the band of noise and the pure
slightly greater than the critical band surrounding the tone. tone are presented simultaneously to the same ear” (ANSI/
It provides the greatest masking effect with the least overall ASA, 2010, p 7).
intensity. Sanders and Rintelmann (1964) confirmed that Stated differently, effective masking (in dB EM) refers to
narrowband noise was a far more efficient masker for pur-
1. The HL (dB HL) to which puretone threshold is shifted
etone stimuli than white noise. For a given sound pressure
by a given level of noise; and
level (50, 70, and 90 dB SPL), narrowband noise centered at
2. The puretone threshold shift (in dB) relative to 0 dB HL
the frequency of the puretone signal (ranging from 250 to
provided by a given level of noise.
4,000 Hz) consistently produced a greater masking effect
(about 10 to 20 dB) than white noise. Although contralateral masking is used clinically dur-
The masking noise typically used during puretone ing hearing assessment, the following examples of ipsilateral
audiometry, therefore, is narrowband noise centered geo- masking will facilitate an understanding of the concept of
metrically around the audiometric test frequency. ANSI/ EM level.
ASA (2010) specifies the band limits (i.e., the upper and Example 1: A puretone air-conduction threshold is
lower cutoff frequencies) of narrowband masking noise. measured at 0 dB HL in the right ear. A narrowband noise
To minimize the perception of tonality that often is associ- geometrically centered at the test frequency is presented to
ated with very narrow bands of noise, the bands specified the same ear at 50 dB EM. This EM level of 50 dB will shift
by ANSI/ASA are somewhat wider than the critical bands puretone threshold to 50 dB HL.
for EM. The goal is to avoid confusion of the masker with Example 2: A puretone air-conduction threshold is
the signal. measured at 30 dB HL in the right ear. A narrowband noise
Speech spectrum noise (i.e., weighted random noise geometrically centered at the test frequency is presented to
for the masking of speech) is typically used as a masker the same ear at 50 dB EM. This EM level of 50 dB will shift
during speech audiometry. Speech is a broadband stimu- puretone threshold to 50 dB HL.
lus that requires a broadband masker. Although white noise These examples illustrate two important points. First, a
is an adequate masker, it is not the most efficient. Speech given level of EM will shift all unmasked puretone thresh-
spectrum noise is white noise that has been filtered to simu- olds to the same dB HL. Of course, if unmasked puretone
late the long-term average spectrum of speech. The average threshold is greater than a particular level of EM, then no
spectrum of speech contains the greatest energy in the low threshold shift will occur. For example, a masker of 50 dB
frequencies with spectrum level decreasing as a function EM will not have a masking effect if the unmasked puretone
of increasing frequency (Dunn and White, 1940). Speech threshold is 70 dB HL. Second, EM refers to the amount of
spectrum noise has a more limited bandwidth than white threshold shift only relative to 0 dB HL.
noise. It is a more efficient masker than white noise, pro- Speech spectrum noise is also calibrated in EM level.
ducing a masking advantage of 8 dB (Konkle and Berry, Just as HL for speech (dB HL) is specified relative to the
1983). ANSI/ASA (2010) specifies that the spectrum of SRT, EM level is also referenced to the SRT. Specifically, EM
weighted random noise for the masking of speech should for speech refers to the dB HL to which the SRT is shifted
have a sound pressure spectrum level that is constant from by a given level of noise. ANSI/ASA (2010) defines EM level
100 to 1,000 Hz, decreasing at a rate of 12 dB per octave for speech as the “sound pressure level of a specified mask-
from 1,000 to 6,000 Hz. ing noise that masks a speech signal to 50% probability of
92 SECTION I • Basic Tests and Procedures

recognition” (p 8). (If the speech spectrum noise has spec- are also referred to as calculation or formula methods.
tral density characteristics as specified by ANSI/ASA and if Psychoacoustic approaches are considered appropriate for
the sound pressure level of the masker is equal to the RET- threshold measurements, whereas acoustic methods are
SPL for speech, then the masker is calibrated in dB EM.) typically most efficient for suprathreshold measurements.
ANSI/ASA (2010) also states that in individuals with nor-
mal hearing, “the amount of effective masking . . . is equal to
the number of decibels that a masking noise shifts a speech
Puretone Audiometry
recognition threshold . . . when the masking noise and Formulas and equations have been presented for the cal-
speech signal are presented simultaneously to the same ear” culation of minimum and maximum masking levels dur-
(p 8). ing puretone audiometry (Lidén et al., 1959; Martin, 1967,
Consider the following example. SRT is measured at 0 1974; Studebaker, 1962, 1964). A brief discussion of these
dB HL. Speech spectrum noise is then presented to the same formulas will facilitate an understanding of the manner in
ear at 50 dB EM. This EM level of 50 dB will shift the SRT which appropriate levels of masking are selected during
to 50 dB HL. puretone threshold testing.
Calibration of masking noise in EM level greatly sim-
plifies clinical masking procedures. When masking noise is
MINIMUM MASKING LEVEL
calibrated in dB EM, then the decibel value indicated on the
masking level control will indicate the masking effect pro- Lidén et al. (1959) and Studebaker (1964) offered formu-
duced in the nontest ear. This clearly facilitates the selection las for calculating minimum masking level during puretone
of appropriate masking levels during clinical testing. air-conduction audiometry that include consideration of
IA, HL of the test signal, and air-bone gaps in the nontest
ear. Although this “formula” approach to calculating mini-
CLINICAL MASKING PROCEDURES mum masking level is necessary during administration of
All approaches to clinical masking address two basic ques- suprathreshold auditory tests (this approach will be dis-
tions. First, what is the minimum level of noise that is cussed later in the section addressing masking in speech
needed in the nontest ear to eliminate its response to the audiometry), it proves somewhat disadvantageous during
test signal? Stated differently, this is the minimum masking threshold audiometry. First, it can be time consuming. Sec-
level that is needed to avoid undermasking (i.e., even with ond, the clinician may not have all required information to
contralateral masking, the test signal continues to be per- accurately calculate minimum masking level at that point in
ceived in the nontest ear). Second, what is the maximum time. (The reader is referred to Yacullo, 1996 for further dis-
level of noise that can be used in the nontest ear that will not cussion of the derivation of these equations and formulas.)
change the true threshold or response in the test ear? Stated The simplified method described by Martin (1967,
differently, this is the maximum masking level that can be 1974) is recommended for clinical use. Martin has suggested
used without overmasking (i.e., with contralateral mask- that formulas are unnecessary during threshold audiome-
ing, the true threshold or response in the test ear has been try and has simplified the calculation of minimum mask-
changed). Because of limited IA for air-conducted sound, ing level. Specifically, the “initial” masking level (in dB EM)
the masking stimulus presented to the nontest ear can also during air-conduction threshold testing is simply equal
cross over to the test ear and produce masking of the test to air-conduction threshold (in dB HL) of the nontest ear
signal (i.e., overmasking). Stated simply, the purpose of (i.e., ACNontest Ear). It should be noted that the initial masking
clinical masking is to make the test signal inaudible in the level is calculated in the same manner regardless of the air-
nontest ear without affecting the true response to the sig- conduction transducer being used (i.e., supra-aural ear-
nal in the test ear. Therefore, the major goal of any clinical phone or 3A insert earphone).
masking procedure is the avoidance of both undermasking The audiometric data presented in Figure 6.8 will be
and overmasking. used to facilitate an understanding of the calculation of
Studebaker (1979) has identified two major approaches masking levels during puretone threshold audiometry.
to clinical masking: psychoacoustic and acoustic. Psycho- Audiometry was performed using supra-aural earphones.
acoustic procedures are “those based upon observed shifts Unmasked air- and bone-conduction thresholds at 500 Hz
in the measured threshold as a function of suprathreshold are provided; masked air- and bone-conduction thresholds
masker effective levels in the nontest ear” (Studebaker, 1979, are also included for later discussion. Unmasked puretone
p 82). These approaches are also identified as threshold shift air-conduction testing suggests that contralateral mask-
or shadowing procedures. Acoustic procedures are “those ing will be required only when measuring air-conduction
based upon calculating the approximate acoustic levels of threshold in the left ear. Specifically, the unmasked air-
the test and masker signals in the two ears under any given conduction threshold of 65 dB HL in the left ear equals or
set of conditions and on this basis deriving the required exceeds the threshold (both air and bone conduction) in the
masking level” (Studebaker, 1979, p 82). These procedures nontest ear by a conservative estimate of IA (i.e., 40 dB).
 CHAPTER 6 • Clinical Masking 93

distribution is about 5 dB, then Studebaker (1979) recom-

mends that a safety factor of not less than 10 dB should be
added to the calculated minimum masking level. Given this
recommendation, Martin’s simplified equation for initial
masking level (in dB EM) during air-conduction threshold
audiometry can be stated as follows:
Initial Masking Level = AC Nontest Ear + 10 dB

Considering again the example presented in Figure 6.8, the

initial masking level is now calculated as 15 dB EM:
Initial Masking Level = AC Nontest Ear + 10 dB
= 5 dB HL + 10 dB
= 15 dB EM

It is important to differentiate the terms minimum

masking level and initial masking level during air-conduc-
tion threshold audiometry. Earlier in this discussion, a gen-
eral definition of minimum masking level was provided.
FIGURE 6.8 An example illustrating the calculation Minimum masking level was defined as the minimum level
of initial and maximum masking levels during puretone of noise needed in the nontest ear to eliminate its response
threshold audiometry. See text for further discussion.
to the test signal. Related to puretone threshold audiom-
etry, a more specific definition of minimum masking level
can be offered: Minimum masking level is the minimum
According to Martin (1967, 1974), the initial masking level level of noise needed to eliminate the contribution of the
(in dB EM) is equal to 5 dB EM (i.e., ACNontest Ear). nontest ear to establish the true or correct threshold in the
Martin (1967, 1974) explains the derivation of this sim- test ear. Initial masking level is simply the first level of noise
plified equation in the following way. A signal detected at introduced to the nontest ear. This initial level of masking
threshold is assumed to have an SL of 0 dB, regardless of is often not sufficient to establish the threshold in the test
whether it is perceived in the test or nontest ear. Therefore, ear; higher levels of masking are often required. This con-
a cross-hearing response during puretone threshold testing cept will be addressed again in our discussion of the rec-
theoretically represents a threshold response in the nontest ommended clinical masking procedure during puretone
ear. Given this assumption, the initial masking level required threshold audiometry.
is one that will just mask a signal of 0 dB SL (i.e., threshold) Lidén et al. (1959) and Studebaker (1964) also have
in the nontest ear. Because of the manner in which mask- offered formulas for minimum masking level during bone-
ing stimuli are calibrated clinically (i.e., EM level, dB EM), a conduction testing that are derived from the same theo-
masker presented at a level (in dB EM) equal to the air-con- retical constructs used during air-conduction testing (see
duction threshold (in dB HL) in the nontest ear should just Yacullo, 1996 for further discussion). Again, the formula
mask the threshold response in the nontest ear. Given the approach during bone-conduction threshold audiometry
example presented in Figure 6.8, a masker level of 5 dB EM is not clinically practical. The use of Martin’s simplified
(which is equal to the air-conduction threshold in the right approach is recommended. Specifically, initial masking level
ear) should be sufficient to just mask a threshold response during bone-conduction audiometry is equal to the air-
to the test signal in the right ear. Martin also indicates that conduction threshold of the nontest ear. However, we will
the simplified approach will lead to the selection of the same need to add the OE to the initial masking level to compen-
masker level as when using the more complex formulas for sate for covering (i.e., occluding) the nontest ear with an
calculating minimum masking level. earphone (Martin, 1967, 1974; Studebaker, 1964). Martin’s
Martin (1974) recommends that approximately 10 dB simplified equation for initial masking level (in dB EM)
should be added to the initial masking level to account for during bone-conduction threshold testing can be stated as
intersubject variability. Remember that dB EM refers to the follows:
HL (dB HL) to which threshold is shifted by a given level of
noise. Calibration of EM is based on the averaged responses Initial Masking Level = AC Nontest Ear + OE + 10 dB
of a group of normal-hearing subjects. Therefore, a given
EM level will not prove equally effective for all subjects. If Bone-conduction thresholds are always obtained with
masked thresholds are normally distributed around the the test ear unoccluded or uncovered. However, when an
average effective level and if the standard deviation of the earphone covers or occludes the nontest ear during masked
94 SECTION I • Basic Tests and Procedures

bone-conduction audiometry, an OE can be created in the Consider again the example presented in Figure 6.8.
nontest ear. The nontest ear consequently can become more Assume that we have subsequently measured a masked air-
sensitive to bone-conducted sound for test frequencies below conduction threshold of 85 dB HL in the left ear. A masked
2,000 Hz, particularly when using supra-aural earphones bone-conduction threshold will also be required in the left
(Berger and Kerivan, 1983; Berrett, 1973; Dean and Martin, ear. Comparison of the unmasked bone-conduction thresh-
2000; Dirks and Swindeman, 1967; Elpern and Naunton, old of 5 dB HL with the masked air-conduction threshold
1963; Goldstein and Hayes, 1965; Hodgson and Tillman, of 85 dB HL in the left ear suggests a potentially significant
1966). During the application of contralateral masking, there air-bone gap (i.e., ≥15 dB). Initial masking level is calcu-
is increased probability that the nontest ear will respond lated in the same manner regardless of the air-conduction
when obtaining a masked bone-conduction threshold in the transducer used for the delivery of the masking stimulus.
test ear. Studebaker (1979) points out that the OE does not The only difference in calculation relates to applying a dif-
actually affect the hearing sensitivity of the occluded ear, but ferent correction factor for the OE when testing in the lower
rather increases the sound pressure level of the signal reach- frequencies. Using the recommended fixed OE values for
ing the cochlea. The reader is referred to Tonndorf (1968, supra-aural earphones, initial masking level during bone-
1972) for further discussion of the contribution of the exter- conduction testing at 500 Hz is calculated as follows:
nal auditory meatus to bone-conduction thresholds.
There is evidence suggesting that the OE is decreased Initial Masking Level = AC Nontest Ear + OE + 10 dB
significantly when using deeply inserted insert earphones = 5 dB HL + 20 dB + 10 dB
(Berger and Kerivan, 1983; Chaiklin, 1967; Dean and = 35 dB EM
Martin, 2000). Berger and Kerivan (1983) and Dean and
Martin (2000) studied the magnitude of the OE in normal- In this particular example, it is appropriate to account for
hearing subjects using E-A-R foam eartips and supra- the OE because there is no evidence of a significant air-bone
aural earphones as occluding devices. Their overall results gap in the nontest (i.e., right) ear. The use of a supra-aural
are remarkably similar. First, the average OEs in the low earphone for delivery of masking in the lower frequencies,
frequencies are greatly reduced when occluding the ear however, will result in greater initial masking levels than
using an E-A-R foam eartip with deep insertion. Second, when using a 3A insert because of a larger OE correction
the advantage of a greatly reduced OE for the E-A-R foam factor.
eartip is lost when a partial or shallow insertion is used.
Third, partial or shallow insertion of an E-A-R foam eartip
yields average OEs that are similar to those measured with MAXIMUM MASKING LEVEL
supra-aural earphones. Different theories have been offered Maximum masking level refers to the maximum level of
to explain the reduced OE for an occluding device that is noise that can be used in the nontest ear that will not shift or
deeply inserted into the ear canal. The reader is referred to change the true threshold in the test ear. Two factors influ-
Berger and Kerivan (1983), Tonndorf (1972), and Yacullo ence maximum masking level during puretone audiometry:
(1996) for further discussion. (1) The bone-conduction threshold of the test ear (BCTest Ear)
The clinician can use either individually determined and (2) IA of the air-conducted masking stimulus (Lidén
(Dean and Martin, 2000; Martin et al., 1974) or fixed OE et al., 1959). Maximum masking level (MMax), based on the
values (i.e., based on average data reported in the literature) original concept described by Lidén et al., can be summa-
when calculating initial masking level. Based on the largest rized using the following equation:
average OEs reported in the literature (Berger and Keri-
van, 1983; Berrett, 1973; Dean and Martin, 2000; Dirks and M Max = BC Test Ear + IA − 5dB
Swindeman, 1967; Elpern and Naunton, 1963; Goldstein
and Hayes, 1965; Hodgson and Tillman, 1966), the follow- If BCTest Ear + IA is just sufficient to produce overmasking,
ing values are recommended for clinical use. then clinically, we want to use a masking level that is some-
When using supra-aural earphones, the following fixed OE what less than the calculated value. Consequently, 5 dB is
values are recommended: 30 dB at 250 Hz, 20 dB at 500 Hz, subtracted from the level that theoretically produces over-
and 10 dB at 1,000 Hz. When using 3A insert earphones with masking. Because we are concerned about an undesired
deeply inserted foam eartips, the following values are recom- masking effect in the test ear, bone-conduction sensitivity
mended: 10 dB at 250 and 500 Hz and 0 dB at frequencies of in that ear must be considered. As a result, overmasking is
1,000 Hz or higher. more of a potential problem when bone-conduction sen-
It should be noted that the OE is decreased or absent sitivity is very good in the test ear. Overmasking, on the
in ears with conductive hearing impairment (Martin et al., other hand, is generally not an issue when bone-conduction
1974; Studebaker, 1979). If the nontest ear exhibits a poten- hearing sensitivity is poor. The poorer the bone-conduction
tial air-bone gap of 20 dB or more, then the OE should not hearing sensitivity is in the test ear, the greater the levels of
be added to the initial masking level at that frequency. masking that can be used without overmasking.
 CHAPTER 6 • Clinical Masking 95

The following two points are important to remember. Clearly in this case, our calculation based on the unmasked
First, the equation for maximum masking level is the same bone-conduction threshold (i.e., 60 dB EM) is an underesti-
for both air- and bone-conduction audiometry. Masking mate of the actual maximum level (i.e., 135 dB EM).
noise is always delivered through an air-conduction trans- Whenever an unmasked bone-conduction thresh-
ducer (e.g., insert or supra-aural earphone) regardless of old is used during determination of maximum masking,
the transducer used for measuring puretone threshold (i.e., the resultant value is typically smaller than the masking
air- or bone-conduction transducer). Second, maximum level that will actually result in overmasking. Although
masking level is generally higher when using 3A insert ear- the actual calculation of maximum masking level during
phones because of increased IA, particularly in the lower puretone threshold audiometry is often of limited use,
frequencies. consideration of the maximum level of noise that can be
Consider again the example presented in Figure 6.8. We used in the nontest ear can alert the audiologist to the pos-
will now calculate the maximum masking level that can be sibility of overmasking, particularly in cases of conductive
used during both masked air- and bone-conduction audi- hearing loss when bone-conduction hearing sensitivity is
ometry: very good.
M Max = BC Test Ear + IA − 5 dB
= 80 dB HL + 60 dB − 5 dB RECOMMENDED CLINICAL PROCEDURE
= 135 dB EM The most popular method for measuring masked puretone
thresholds was first described by Hood in 1957 (Hood,
Rather than using the very conservative IA estimate of 40 dB 1960). The Hood method, also referred to as the plateau,
when using supra-aural earphones, in this case, we will use threshold shift, or shadowing procedure, is a psychoacoustic
the more accurate estimate of 60 dB. If the bone-conduction technique that relies on observations about the relation-
threshold in the right (i.e., nontest) ear is assumed to be ship between masker level in the nontest ear and mea-
5 dB HL (i.e., the unmasked bone-conduction threshold) sured threshold in the test ear. Hood originally described
and the unmasked air-conduction threshold in the left ear a masking procedure that was applicable for measurement
is 65 dB HL, then there is evidence that IA is at least 60 dB. of masked bone-conduction thresholds. However, it proves
If 140 dB EM is just sufficient to produce overmasking (i.e., equally effective for measurement of air-conduction
BCTest Ear + IA), then 135 dB EM is the maximum level of thresholds as well.
noise that can be used in the nontest ear that will not shift or The example presented in Figure 6.9 will help facilitate
change the true threshold in the test ear. It should be noted an understanding of the underlying concept of the threshold
that 135 dB EM is a level that significantly exceeds the out- shift procedure. Unmasked puretone air-conduction thresh-
put limits for standard audiometers. olds, obtained using supra-aural earphones, were measured
Generally, it is neither time efficient nor necessary to at 10 dB HL in the right ear and 60 dB HL in the left ear
calculate maximum masking level during puretone thresh- (Figure 6.9A). Contralateral masking will be required when
old audiometry, particularly when using psychoacoustic or testing the left ear because there is a difference between the
threshold shift masking procedures (which will be described test and nontest ears that equals or exceeds a conservative
shortly). In addition, the estimated maximum masking level estimate of IA (i.e., 40 dB). An initial masking level of 20 dB
is typically very conservative and not an accurate indication EM (i.e., ACNontest Ear + 10 dB) is now presented to the right
of the true maximum. In the above example, we calculated a ear, and puretone threshold is re-established. Recall that the
relatively accurate estimate by using a more accurate value of purpose of contralateral masking is to raise the threshold
IA (rather than the conservative value) and the actual bone- of the nontest ear sufficiently to eliminate its contribution
conduction threshold in the test ear (i.e., 80 dB HL). How- when measuring a response in the test ear. Assuming that
ever, the true bone-conduction threshold (obtained with overmasking is not occurring, then contralateral masking
appropriate contralateral masking) in the test ear is typically should have an effect only on the responsiveness of the non-
not known when maximum masking level is estimated dur- test ear.
ing both air- and bone-conduction threshold audiometry. There are two possible outcomes when puretone thresh-
Because only an unmasked bone-conduction threshold is old is re-established in the presence of contralateral masking:
available at the time that masking levels are determined, we (1) No measured puretone threshold shift (e.g., puretone
are required to use the unmasked threshold as the estimate threshold remains constant at 60 dB HL; Figure 6.9B) or (2) a
of bone-conduction hearing sensitivity in the test ear. Let us measured puretone threshold shift (e.g., puretone threshold
calculate again MMax using the unmasked bone-conduction shifts from 60 to 70 dB HL; Figure 6.9C). If contralateral
response as the estimate of bone-conduction threshold: masking in the nontest ear does not produce a masking
effect, then it is concluded that the unmasked puretone
M Max = 5 dB HL + 60 dB − 5 dB threshold represents a response from the test ear. Conversely,
= 60 dB EM if contralateral masking in the nontest ear does produce a
96 SECTION I • Basic Tests and Procedures

A C

FIGURE 6.9 Example illustrating the underlying concept of the plateau or threshold shift masking
procedure. See text for further discussion. (From Yacullo WS. (1996) Clinical Masking Procedures.
1st ed. Boston, MA: Allyn & Bacon, © 1996, p 69. Adapted by permission of Pearson Education, Inc.,
Upper Saddle River, NJ.)

masking effect, then it is concluded that the unmasked pur- ear. Masking noise is introduced at an initial masking level of
etone threshold represents a response from the nontest ear. 10 dB EM (i.e., ACNontest Ear + 10 dB), and puretone thresh-
The underlying concept of the Hood procedure is that the old is re-established. Threshold shifts to 50 dB HL. When
introduction of masking to the nontest ear will produce a the masker level is raised sequentially to 20 and 30 dB EM,
masking effect (i.e., a threshold shift) only if the nontest ear puretone threshold continues to shift by 10 dB. A shadowing
is contributing to the observed response. Decisions about effect has occurred because the masked puretone threshold
which ear is contributing to the measured threshold are “shadows” the threshold of the nontest or masked ear with
based on whether a threshold shift occurs when masking is each increment in EM level. Because a threshold shift occurs
introduced to the nontest ear. when masking level is raised, it is concluded that the mask-
Hood (1960) outlined two essential steps of the plateau ing noise and the tone are restricted to the nontest ear.
masking procedure: (1) Demonstration of the shadowing When the masker is raised from 30 to 100 dB EM, pur-
effect and (2) identification of the changeover point. The etone threshold no longer shifts and remains stable at 70 dB
hypothetical example presented in Figure 6.10 illustrates HL. A plateau has been reached. Because there is no addi-
basic concepts of the plateau masking procedure. Puretone tional masking effect (i.e., a threshold shift) when masker
testing using supra-aural earphones reveals unmasked level is increased, it is concluded that the nontest ear is no
air-conduction thresholds of 0 dB HL in the right ear and longer contributing to the observed response. Puretone
40 dB HL in the left ear (Figure 6.10A). Unmasked bone- threshold of the test ear (i.e., 70 dB HL) has been reached.
conduction threshold is 0 dB HL. Because there is a 40-dB Hood (1960) refers to the initial point on the masking func-
difference between ears, contralateral masking will be tion where puretone threshold remains stable with increas-
required when measuring air-conduction threshold in the ing masking level as the “changeover point.” In this example,
left ear. (Masked air- and bone-conduction thresholds in the the changeover point of 30 dB EM also corresponds to mini-
left ear are included for later discussion.) mum masking level, the minimum amount of noise required
The masking function presented in Figure 6.10B shows to establish the true threshold in the test ear. Masker levels
the relationship between measured puretone threshold (in that result in no threshold shift (i.e., the plateau) represent
dB HL) in the test ear and EM level (in dB EM) in the nontest adequate masking (i.e., 30 through 100 dB EM). Masker
 CHAPTER 6 • Clinical Masking 97

A B

FIGURE 6.10 Hypothetical example illustrating the concepts of undermasking, ade-

quate masking, and overmasking using the threshold shift or plateau masking proce-
dure. See text for explanation. (From Yacullo WS. (1996) Clinical Masking Procedures.
1st ed. Boston, MA: Allyn & Bacon, © 1996, p 72. Adapted by permission of Pearson
Education, Inc., Upper Saddle River, NJ.)

levels less than 30 dB EM represent undermasking. That is, is increased somewhat by using a masker increment of 5 dB.
there is insufficient masking to establish the true puretone It is somewhat arbitrary whether a 5- or 10-dB step size is
threshold in the test ear. used for increasing masker level. Either step size is accept-
When the masker level exceeds 100 dB EM (i.e., 110 and able. However, the smaller step size of 5 dB is strongly rec-
120 dB EM), however, a puretone threshold shift with each ommended whenever the masking plateau is narrow and
increment in masking level is again observed. Overmasking there is increased risk of overmasking (i.e., cases of bilateral
is now occurring. The masking noise has reached the test ear conductive hearing loss).
through cross hearing, and a masking effect (i.e., a thresh- Hood (1960) did not specify the number of masker
old shift) is observed in the test ear. Assuming that a masked increments needed to establish a masking plateau. Clini-
bone-conduction threshold is measured subsequently in the cally, it is neither time efficient nor necessary to measure the
left ear at 65 dB HL, then an estimate of maximum masking entire masking plateau. It is generally agreed that a mask-
level is 100 dB EM (BCTest Ear + IA − 5 dB: 65 dB HL + 40 dB − ing “plateau” has been established when masker level can be
5 dB). Whereas the plateau and overmasking portions of the increased over a range of at least 15 to 20 dB without shift-
masking function represent responses from the test ear, the ing puretone threshold (Kaplan et al., 1993; Martin, 1980;
undermasking or shadowing portion represents responses Sanders, 1991; Silman and Silverman, 1991).
from the nontest ear. It should be apparent from the masking The recommended clinical procedure (Yacullo, 1996,
function in Figure 6.10 that the width of the masking plateau 2004), based on the major components of Hood’s shadow-
is defined by the minimum and maximum masking levels. ing technique, is summarized as follows:
The clinical goal of the plateau procedure is to estab-
lish the HL at which puretone threshold remains unchanged 1. Masking noise is introduced to the nontest ear at
with increments in masking level. Two important variables the initial masking level. Puretone threshold is then
that relate to the plateau procedure are (1) the magnitude re-established.
of the masker increment and (2) the number of masker 2. Level of the tone or noise is increased subsequently by
increments needed to establish a masking plateau. Although 5 dB. If there is a response to the tone in the presence
Hood (1960) originally recommended that masker level be of the noise, the level of the noise is increased by 5 dB.
changed in increments of 10 dB, others have suggested that If there is no response to the tone in the presence of the
the level should be a 5-dB step size (Martin, 1980; Silman noise, the level of the tone is increased in 5-dB steps until
and Silverman, 1991). Martin (1980) suggests that accuracy a response is obtained.
98 SECTION I • Basic Tests and Procedures

3. A plateau has been reached when the level of the noise occurs, the tone is increased in 5-dB steps until audibility
can be increased over a range of 15 to 20 dB without is achieved. However, the HL of the tone may be increased
shifting the threshold of the tone. This corresponds to a inappropriately because of a decision-making process based
response to the tone at the same HL when the masker is on a single response. This may lead to imprecision when mea-
increased in three to four consecutive levels. suring the masked threshold. Therefore, it is recommended
4. Masked puretone threshold corresponds to the HL of the that masked puretone threshold be re-established using a
tone at which a masking plateau has been established. standardized threshold procedure (e.g., ASHA, 2005) in the
presence of the final level of masking noise that resulted in
If a 10-dB step size is used for increasing masking level, then a plateau. This sometimes leads to a 5-dB improvement in
the plateau corresponds to a range of 20 dB (i.e., a response the masked puretone threshold. However, the decision to re-
to the tone at the same HL when the masker is increased in establish masked puretone threshold at the end of the pla-
two consecutive levels). teau procedure will be influenced by time considerations.
The recommended procedure for establishing a masking Remember that the goal of the plateau procedure is
plateau does not require that puretone threshold be formally to establish the HL at which puretone threshold remains
established each time that the masking level is increased. unchanged with increments in masking level. Given this
This approach would significantly increase the time required goal, there are three major outcomes that can result when
to establish a masking plateau. Rather, the tone is presented measuring puretone threshold. These outcomes are illus-
once at the same HL as the previous response. If no response trated in the three examples presented in Figure 6.11. In

FIGURE 6.11 Examples illustrating the use of the plateau

method for measuring masked puretone air-conduction
thresholds. The unmasked air- and bone-conduction thresholds
are the same in each example. Three different outcomes can
result when using the threshold shift procedure. See text for
discussion. (From Yacullo WS. (1996) Clinical Masking Procedures.
1st ed. Boston, MA: Allyn & Bacon, © 1996, pp 75–77. Adapted by
permission of Pearson Education, Inc., Upper Saddle River, NJ.)
 CHAPTER 6 • Clinical Masking 99

each example, the unmasked puretone thresholds at 2,000 Hz in the nontest ear and a conductive hearing loss in the test
are the same. Unmasked puretone air-conduction thresh- ear. The presence of significant hearing loss in the nontest
olds, obtained using 3A insert earphones, were measured ear requires higher initial masking levels; the presence of a
at 15 dB HL in the right ear and 75 dB HL in the left ear. conductive hearing loss in the test ear (i.e., normal bone-
Contralateral masking will be required when measuring conduction hearing sensitivity) decreases the maximum
air-conduction threshold in the left ear; an initial masking masking level. The consequence of a reduced or nonexistent
level of 25 dB EM is presented to the right ear, and puretone masking plateau is the inability to establish correct masked
threshold is re-established. thresholds in the test ear.
In the first outcome, the unmasked puretone threshold The classic example of a masking dilemma is demon-
of 75 dB HL remains unaffected with increasing masking strated with a bilateral, mild-to-moderate conductive hear-
level. The level of the noise was increased over a range of ing loss. The possibility for overmasking exists when mea-
20 dB without shifting the threshold of the tone. In this exam- suring masked air- and bone-conduction thresholds in both
ple, the initial masking level occurs at the masking plateau. ears. Naunton (1960) states that, in some cases of bilateral
Contralateral masking has confirmed that the unmasked conductive hearing loss, it is not possible to mask the non-
puretone threshold represents a response from the test ear. test ear without simultaneously producing a masking effect
In the second outcome, the initial masking level pro- in the test ear.
duces a puretone threshold shift. A masking plateau is One solution to the masking dilemma is the use of
reached, however, when masking level is increased from insert earphones (Coles and Priede, 1970; Hosford-Dunn
35 to 55 dB EM (i.e., a masking range of 20 dB). Because et al., 1986; Studebaker, 1962, 1964). Recall that the use
masked puretone threshold remains stable at 95 dB HL with of 3A insert earphones significantly increases IA for air-
increasing masking level, puretone threshold is recorded as conducted sound, particularly in the lower frequencies
95 dB HL. Contralateral masking has confirmed that the (Killion et al., 1985; Sklare and Denenberg, 1987). There are
unmasked puretone threshold represents a cross-hearing two advantages of using insert earphones in cases of bilat-
response from the nontest ear. eral conductive hearing loss. First, the need for masking
In the third outcome, the initial masking level again during measurement of air-conduction thresholds is often
produces a puretone threshold shift. However, puretone eliminated because of greater IA for air-conducted sound.
threshold continues to shift to the output limits of the Second, the use of insert earphones reduces the probabil-
audiometer with increasing masking level. A plateau is not ity of overmasking in cases where contralateral masking
obtained. Therefore, it is concluded that there is no measur- is required. The use of an air-conduction transducer with
able hearing in the left ear. This conclusion is correct assum- increased IA increases the range between the minimum and
ing that overmasking has not occurred. maximum masking levels, thereby increasing the width of
Turner (2004a, 2004b) has described a masking method the masking plateau and the range of permissible masking
that can replace the plateau procedure in some masking levels (Studebaker, 1962).
situations. A disadvantage of the plateau method is that it
can be time consuming. The “optimized” masking method CENTRAL MASKING
described by Turner, which is based on the principles of the
masking plateau, can reduce the number of masking levels The introduction of contralateral masking can produce a
required to reach the plateau and establish threshold. The small threshold shift in the test ear even when the mask-
method is optimized because it uses the maximum possible ing level is insufficient to produce overmasking. Wegel
masking levels without overmasking. This is accomplished and Lane (1924) referred to this phenomenon as central
through the use of higher initial masking levels and maxi- masking. It has been hypothesized that threshold shifts
mum masker increments. However, there are some masking in the presence of low levels of masking are mediated
situations where the optimized approach is not appropriate. through central nervous system processes (Lidén et al.,
The reader is referred to the two articles by Turner (2004a, 1959). Central masking has been reported to influence
2004b) for further discussion. thresholds measured during both puretone and speech
audiometry (Dirks and Malmquist, 1964; Lidén et al.,
1959; Martin, 1966; Martin and DiGiovanni, 1979; Martin
THE MASKING DILEMMA
et al., 1965; Studebaker, 1962). Although the threshold
There are clinical situations where minimum masking shift produced by central masking is generally consid-
levels can result in overmasking. Studebaker (1979) states ered to be approximately 5 dB (Konkle and Berry, 1983;
that a “masking dilemma” results when the width of the Martin, 1966), variable results have been reported across
masking plateau is very narrow or nonexistent. Remember subjects and studies. There is also some indication that
that the width of the masking plateau is defined by mini- central masking effects increase with increasing mask-
mum and maximum masking levels. Generally, a masking ing level (Dirks and Malmquist, 1964; Martin et al., 1965;
dilemma results whenever there is a significant hearing loss Studebaker, 1962).
100 SECTION I • Basic Tests and Procedures

There is currently no agreed upon procedure that MASKED AUDIOGRAM INTERPRETATION

accounts for central masking effects during threshold
audiometry. However, it is generally not recommended Unmasked and masked puretone thresholds are typically
that the effect of central masking be subtracted from recorded on the same audiogram. Therefore, audiogram
masked thresholds. First, it is difficult to determine an interpretation will involve consideration of both masked and
appropriate correction factor given the variability of the unmasked responses. ASHA (1990) has published guidelines
central masking effect across subjects. Second, the typical for audiometric symbols and procedures for graphic repre-
central masking effect size of about 5 dB is considered sentation of frequency-specific audiometric findings. These
to be within good test-retest reliability during threshold guidelines have been followed throughout this chapter.
measurements. It is important to remember that the use Figure 6.12 presents an audiogram in which contra-
of contralateral masking can somewhat influence the lateral masking was required when obtaining both air- and
measured masked thresholds and should be taken into bone-conduction thresholds in the left ear. Air-conduction
account when interpreting audiometric test results. For audiometry was performed using supra-aural earphones.
example, a difference of 5 dB between unmasked and Puretone testing reveals a mild conductive hearing loss of
masked thresholds is generally not considered significant. flat configuration in the right ear. Masked air- and bone-
This difference may simply reflect (1) central masking conduction responses indicate a severe-to-profound, sensory/
effects and/or (2) normal variability related to test-retest neural hearing loss of gradually sloping configuration in the
reliability. right ear.

FIGURE 6.12 An example illustrating audiogram interpretation using unmasked and masked
puretone thresholds.
 CHAPTER 6 • Clinical Masking 101

It should be noted that the unmasked air-conduction PSYCHOACOUSTIC MASKING PROCEDURES

thresholds in the left ear are not considered when interpret-
ing hearing status. Because a significant threshold shift (i.e., Recall that the psychoacoustic or threshold shift masking
>5 dB) occurred when contralateral masking was intro- procedures rely on the observation of shifts in the measured
duced to the nontest ear, the unmasked air-conduction threshold in the test ear as a function of masking levels in
responses in the left ear actually represent cross-hearing the nontest ear. The plateau procedure can be applied eas-
responses from the better (i.e., right) ear. In this case, the ily during measurement of speech thresholds (Konkle and
unmasked air-conduction thresholds should not be con- Berry, 1983; Studebaker, 1979). A major advantage of the
nected with lines. In cases where contralateral masking is plateau procedure is that information about bone-conduc-
required, it is acceptable to record only the masked thresh- tion hearing sensitivity in each ear is not required when
olds (ASHA, 1990). selecting appropriate masking levels. Although the plateau
Although the results of unmasked bone-conduction procedure can be applied during measurement of both
audiometry suggested that masked bone-conduction masked recognition and detection thresholds, it proves most
thresholds were required in both ears because of potential efficient during measurement of SDT because of the nature
air-bone gaps, contralateral masking was required only of the response task (i.e., detection rather than recognition).
when testing the left ear. Whenever there is an asymmet- ASHA’s most recent guidelines for determining thresh-
rical hearing loss, it is traditional to first measure masked old level for speech were published in 1988. Recommended
bone-conduction thresholds in the poorer ear. There is the procedures for measuring both detection and recognition
assumption that the unmasked bone-conduction thresh- thresholds are described. Given that determination of SDT
olds may more likely reflect hearing in the better ear. When involves a detection task that is similar to the one used in
masked bone-conduction thresholds were subsequently puretone threshold audiometry, ASHA recommends using a
measured in the left ear, results suggested a sensory/neural test procedure that follows published guidelines for measur-
hearing loss. Consequently, we can correctly assume that ing puretone threshold (e.g., ASHA, 2005). Therefore, the
the unmasked responses are originating from the better plateau masking procedure recommended earlier for use
(i.e., right) ear. Depending on the outcome when measur- during puretone threshold audiometry can be used equally
ing masked bone-conduction thresholds in the poorer ear, effectively when measuring masked SDT.
it is not always necessary to also measure masked thresholds Consider the example presented in Figure 6.13. Audi-
in the opposite ear. As the above example illustrates, ear- ometry was performed using 3A insert earphones. pure-
specific information can be inferred from unmasked bone- tone testing reveals normal hearing in the right ear. There
conduction responses in some cases. is a profound sensory/neural hearing loss of fragmentary
It is traditional to record masking levels when obtain- configuration in the left ear. (Contralateral masking was
ing masked air- and bone-conduction thresholds. Assum- required during measurement of air- and bone-conduction
ing that the clinician has used the recommended threshold thresholds in the left ear.) An SRT of 5 dB HL was measured
shift (i.e., plateau) procedure, then a range of masking lev- in the right ear, a finding that supports the puretone results.
els will be used when establishing threshold. ASHA (1990) When spondaic words were presented at suprathreshold
recommends that the final level of masking used to obtain levels in the left ear, the patient was not able to correctly
masked threshold should be recorded for the nontest recognize any words. Consequently, a decision was made to
ear. A table for recording EM levels to the nontest ear is measure an SDT.
typically provided on an audiogram form. Consider again An unmasked SDT is measured at 75 dB HL in the left
the audiogram presented in Figure 6.12. For example, a ear. Because the difference between the unmasked SDT in
masked puretone air-conduction threshold was measured the test ear (i.e., 75 dB HL) and the SRT in the nontest ear
at 85 dB HL at 2,000 Hz in the left ear; this threshold was (i.e., 5 dB HL) clearly exceeds our conservative estimate of
obtained with a final masking level of 70 dB EM in the IA for speech (i.e., 60 dB) when using 3A insert earphones,
right ear. contralateral masking will be required.
Using the recommended plateau masking procedure,
speech spectrum noise is introduced to the nontest ear at an
Speech Audiometry initial masking level, that is, an EM level (in dB EM) equal
The speech audiometry test battery is traditionally com- to the speech threshold of the nontest ear (SRTNontest Ear) plus
posed of two major components: (1) Measures of hearing a 10-dB safety factor:
sensitivity for speech (i.e., speech threshold) and (2) mea- Initial Masking Level = SRTNontest Ear + 10 dB
sures of suprathreshold speech recognition. Although the
psychoacoustic or threshold shift procedure proves efficient In this example, initial masking level is equal to 15 dB EM.
when measuring SDT, the acoustic masking procedure is SDT is then re-established in the nontest ear in the pres-
the method of choice when assessing threshold and supra- ence of the initial masking level. Depending on the patient’s
threshold speech recognition. response to the speech in the presence of the noise, the level
102 SECTION I • Basic Tests and Procedures

FIGURE 6.13 An example illustrating the use of the threshold shift masking procedure for determin-
ing speech detection threshold.
 CHAPTER 6 • Clinical Masking 103

of the speech or noise is increased by 5 dB until a masking frequency range. Presentation LevelTest Ear – IA, an estimate
plateau has been reached. Remember that it is acceptable to of the HL of the speech signal reaching the test ear, repre-
use a 10-dB masker increment when establishing a masking sents the minimum masking level required. The presence
plateau. In this particular case, the risk of overmasking is of air-bone gaps in the nontest (i.e., masked) ear, however,
essentially nonexistent because of the poor bone-conduction will reduce the effectiveness of the masker. Consequently,
hearing sensitivity in the test ear (and the use of an air- minimum masking level must be increased by the size of the
conduction transducer with increased IA for presenting the air-bone gap.
masking noise). Therefore, the use of a 10-dB masker incre- Lidén et al. (1959) recommended that the average
ment can be easily justified. Masked SDT is subsequently air-bone gap in the nontest ear, calculated using frequen-
measured at 90 dB HL (using 40 dB EM). cies of 500, 1,000, and 2,000 Hz, be accounted for when
Although the plateau masking procedure can be used determining the minimum masking level. Coles and
during assessment of masked SRT, it can prove very time Priede (1975) suggested a more conservative approach
consuming. Recall that only a single detection response to and recommended that the largest air-bone gap at any fre-
speech is required when measuring masked SDT before quency in the range from 250 to 4,000 Hz be considered.
making a decision about increasing the level of the speech Remember that speech is a broadband signal. Therefore,
or masker. The use of the plateau procedure for measuring bone-conduction hearing sensitivity across a range of fre-
masked SRT, however, requires that threshold be re-estab- quencies in the nontest ear must be considered. There is
lished (i.e., 50% correct recognition of spondaic words) at the assumption that the largest air-bone gap will have the
each masking level until a plateau is reached. The acoustic greatest effect on masking level. Following the conservative
method proves to be the method of choice when measuring recommendation of Coles and Priede (1975), it is recom-
masked SRT because of its greater time efficiency (Konkle mended that the largest air-bone gap across the frequency
and Berry, 1983; Studebaker, 1979) and will be discussed in range in the nontest ear be accounted for when determin-
the following section. ing minimum masking level.
Maximum masking level (MMax) for speech, originally
described by Lidén et al. (1959), can be defined using the
ACOUSTIC MASKING PROCEDURES following equation:
Recall that acoustic masking procedures are based on calcu-
lating the estimated acoustic levels of the test and masking M Max = Best BC Test Ear + IA − 5dB
stimuli in the two ears during a test condition and, on this
basis, selecting an appropriate masking level. A major dis- Best BCTest Ear represents the best bone-conduction threshold
advantage of the acoustic or formula approach is that the in the test ear in the frequency range from 250 to 4,000 Hz,
application requires knowledge about air-bone gaps in both and IA is equal to IA for speech. If Best BCTest Ear + IA rep-
test and nontest ears (Konkle and Berry, 1983; Studebaker, resents a level that will just produce overmasking in the test
1979). Knowledge about air-bone gaps in the nontest ear is ear, then a slightly lower masking level should be used clini-
required to calculate minimum masking level. Information cally. Consequently, a value of 5 dB is subtracted from the
about bone-conduction hearing sensitivity in the test ear calculated level.
is required to calculate maximum masking level. Assum- Lidén et al. (1959) originally recommended that the aver-
ing that complete puretone threshold data are available age puretone bone-conduction threshold in the test ear, again
before performing speech audiometry, however, formula calculated using frequencies of 500, 1,000, and 2,000 Hz,
approaches for calculating required masking levels prove should be accounted for when estimating maximum masking
very effective during measurement of both threshold and level. However, a more conservative approach includes con-
suprathreshold speech recognition. sideration of the best bone-conduction threshold in the test
The underlying concepts of minimum and maximum ear over a wider range of frequencies (i.e., 250 to 4,000 Hz).
masking levels for speech are similar to those offered earlier There is the assumption that the best bone-conduction
for puretone stimuli. Minimum masking level for speech threshold in the test ear is the most susceptible to the effects
(MMin), originally described by Lidén et al. (1959), can be of overmasking.
defined using the following equation: The optimal masking level during speech audiometry
is one that occurs above the minimum and below the maxi-
M Min = Presentation Level Test Ear − IA mum masking levels (Konkle and Berry, 1983; Lidén et al.,
+ Largest Air-Bone Gap Nontest Ear 1959; Studebaker, 1979). Minimum and maximum mask-
ing levels represent, respectively, the lower and upper lim-
Presentation LevelTest Ear represents the HL (dB HL) of the its of the masking plateau. Studebaker (1979) states that a
speech signal at the test ear, IA is equal to IA for speech, major goal of the acoustic or formula approach is to apply
and Largest Air-Bone GapNontest Ear represents the largest rules that will place the masking level at approximately the
air-bone gap in the nontest ear in the 250- to 4,000-Hz middle of the range of correct values (i.e., the middle of the
104 SECTION I • Basic Tests and Procedures

masking plateau). This concept was originally discussed by ately for air-bone gaps in the test and nontest ears. In cases
Luscher and König in 1955 (cited by Studebaker, 1979). where there are no air-bone gaps in either ear, the selected
Studebaker (1962) first described an equation for calcu- masking level is simply equal to the HL of the speech signal.
lating midmasking level during puretone bone-conduction To avoid the use of very high levels of contralateral mask-
audiometry. The basic concepts underlying the midplateau ing that can sometimes result, Studebaker indicates that it
procedure, however, can be easily applied during speech is permissible to reduce the masking level by 20 to 25 dB
audiometry. Yacullo (1999) states that a simple approach to below the presentation level of the speech signal. The reader
calculating the midmasking level (MMid) involves determin- is referred to Studebaker (1979) for a more comprehensive
ing the arithmetic mean of the minimum and maximum discussion.
masking levels: According to the results of a survey of audiometric
practices in the United States, many audiologists “base their
M Mid = (M Min + M Max )/2 masking level for word-recognition testing on the stimulus
level presented to the test ear and subtract a set amount,
For example, if MMin is equal to 40 dB EM and MMax is such as 20 dB” (Martin et al., 1998, p 100). Although selec-
equal to 80 dB EM, then MMid, the masking level that falls tion of a masking level that is equal to the presentation level
at midplateau, is 60 dB EM. When a masking level falls at at the test ear minus 20 dB may appear somewhat arbitrary,
the middle of the acceptable masking range (i.e., midmask- it can actually be supported by sound theoretical constructs.
ing level), then the risk of undermasking and overmasking Yacullo (1996, 1999) has described a simplified approach,
is minimized (Studebaker, 1962). It should be noted that based on the underlying concepts of both the midplateau
midplateau actually represents a range of values surround- and Studebaker acoustic procedures, that can be used when
ing the midmasking level. Consequently, there is some flex- selecting contralateral masking levels during speech audi-
ibility in using a somewhat higher or lower masking level. ometry. Although this approach was originally described for
Yacullo (1999) states that there are two major advan- use during assessment of suprathreshold speech recognition
tages of the midplateau masking procedure. First, IA is elim- (Yacullo, 1996), it also proves equally effective during mea-
inated as a source of error when determining an appropri- surement of SRT (Yacullo, 1999). Stated simply, EM level is
ate masking level. Masking levels are often determined using equal to the presentation level of the speech signal in dB HL
very conservative estimates of IA. However, IA has equal yet at the test ear minus 20 dB:
opposite effects on minimum and maximum masking lev-
els. Although the value of IA used for determining mini- dB EM = Presentation Level Test Ear − 20 dB
mum and maximum masking levels will influence the width
of the masking plateau, the midmasking level remains the Given two prerequisite conditions (which will be dis-
same. cussed shortly), the selected masking level will fall approxi-
Second, midmasking level can be determined for both mately at midplateau. Unfortunately, inappropriate use of
threshold and suprathreshold speech measures by using this simplified approach can result in undermasking or
the same formula approach (Konkle and Berry, 1983). The overmasking.
midplateau procedure avoids a potential problem dur- Jerger and associates (Jerger and Jerger, 1971; Jerger et al.,
ing measurement of suprathreshold speech recognition 1966) appear to be the first to report the use of a masking
that is related to calibration of EM level and percent cor- procedure that involved presenting contralateral masking
rect response criterion. Recall that EM level for speech is noise at a level 20 dB less than the presentation level of the
specified relative to the SRT (i.e., 50% correct recognition speech signal at the test ear. Specifically, it was reported that
of spondaic words) (ANSI/ASA, 2010). Suprathreshold “whenever the speech level to the test ear was sufficiently
speech recognition performance, however, can range from intense that the signal might conceivably cross over and be
0% to 100%. Konkle and Berry (1983) indicate that a major heard on the nontest ear, the latter was masked by white
advantage of the midplateau procedure is that the middle of noise at a level 20 dB less than the speech presentation level
the masking plateau (i.e., the optimal masking level) is not on the test ear” (Jerger and Jerger, 1971, p 574). It should be
influenced by different listener response criteria used dur- noted, however, that Jerger and associates used white noise
ing assessment of threshold and suprathreshold speech rec- as a contralateral masker rather than the typically used
ognition. The reader is referred to Konkle and Berry (1983) speech spectrum noise. In addition, the white noise was not
and Studebaker (1979) for more detailed discussion. calibrated in EM level for speech.
Studebaker (1979) has described a somewhat different More recently, Hannley (1986) and Gelfand (2009)
acoustic masking procedure for use during speech audiom- have discussed briefly the simplified approach to masking.
etry that is consistent with the goal of selecting a masking Gelfand indicates, however, that using an EM level equal to
level that occurs at midplateau. Specifically, the recom- the HL of the speech signal at the test ear minus 20 dB gen-
mended masking level is equal to the presentation level of erally proves most effective in cases of sensory/neural hear-
the speech signal in dB HL at the test ear, adjusted appropri- ing loss. In fact, the desired outcome may not occur when
 CHAPTER 6 • Clinical Masking 105

there are significant air-bone gaps in the nontest ear (e.g., M Max = Best BC Test Ear + IA − 5 dB
conductive hearing loss).
= 45 dB HL + 40 dB − 5 dB
Yacullo (1999) states that the simplified masking pro-
cedure when used appropriately can significantly reduce the = 80 dB EM
calculations required for the determination of optimal (i.e.,
midmasking) masking level. Specifically, the method proves M Mid = (M Min + M Max )/2
effective given the following two conditions: (1) There are = (60 + 80)/2
no significant air-bone gaps (i.e., ≥15 dB) in either ear and = 70 dB EM
(2) speech is presented at a moderate SL (i.e., 30 to 40 dB
SL) relative to the measured or estimated SRT. If these two An EM level of 70 dB is appropriate for three reasons.
prerequisites are met, then the selected masking level should First, it occurs at midplateau. Second, it occurs at least
occur approximately at midplateau. 10 dB above the calculated minimum. Remember that a
Acoustic masking procedures are recommended when safety factor of at least 10 dB or greater should be added
assessing threshold and suprathreshold speech recogni- to the calculated minimum value to account for intersub-
tion. The following two examples help illustrate the use of ject variability with respect to masker effectiveness (Martin,
the midplateau masking procedure, as well as the simpli- 1974; Studebaker, 1979). Finally, it does not exceed the cal-
fied approach when applicable, for measurement of supra- culated maximum masking level.
threshold speech recognition and SRT. It should be noted that the width of the masking plateau
The example presented in Figure 6.14 illustrates the use is typically underestimated when a conservative estimate of
of the midplateau masking procedure during assessment of IA is used for determining the minimum and maximum
suprathreshold speech recognition. Puretone testing reveals masking levels. If IA is actually greater than the conservative
a mild, sensory/neural hearing loss of flat configuration in estimate of 40 dB, then the width of the masking plateau
the right ear. There is a moderate-to-severe, sensory/neural will increase. For example, if this patient actually exhibits IA
hearing loss of gradually sloping configuration in the left for speech of 55 dB (rather than the conservative estimate
ear. SRTs were measured at 35 dB HL in the right ear and of 40 dB), then the minimum level will be decreased and
55 dB HL in the left ear, findings that support the puretone the maximum level will be increased by the same amount
results. Suprathreshold speech recognition will be assessed (i.e., 15 dB). Although the width of the masking plateau
at 40 dB SL using Central Institute for the Deaf (CID) W-22 increases, the midmasking level remains the same. As stated
monosyllabic word lists. earlier, a major advantage of the midplateau method is
Let us first consider the situation where supra-aural that IA is eliminated as a source of error when selecting an
earphones are being used during audiometry. Contralateral appropriate masking level.
masking will be required only when measuring suprathreshold We now will take another look at the example presented
speech recognition in the left ear. Specifically, the presentation in Figure 6.14 and substitute 3A insert earphones for the
level of 95 dB HL (i.e., SRT of 55 dB HL + 40 dB SL) exceeds supra-aural arrangement. Contralateral masking will also
the best bone-conduction threshold of 30 dB HL in the nontest be required when assessing suprathreshold speech recog-
ear by a conservative estimate of IA for speech (i.e., 40 dB): nition in the left ear. The presentation level of 95 dB HL
(i.e., SRT of 55 dB HL + 40 dB SL) exceeds the best bone-
Presentation Level Test Ear − Best BC Nontest Ear ≥ IA conduction threshold of 30 dB HL in the nontest ear by a
conservative estimate of IA for speech (i.e., 60 dB). We will
95 dB HL − 30dB HL ≥ 40 dB again use the midplateau masking procedure to select an
appropriate level of contralateral masking. The calculations
65dB HL ≥ 40 dB are the same for both supra-aural and 3A insert earphones
with the exception that an IA value of 60 dB will be substi-
We will use the midplateau masking procedure to select tuted in our equations for minimum and maximum mask-
an appropriate contralateral masking level. Remember that ing levels. Masking levels for use with insert earphones are
the midplateau masking procedure involves a three-step summarized below:
process: Calculation of (1) minimum masking level (MMin),
(2) maximum masking level (MMax), and (3) midmasking M Min = 40 dB EM
level (MMid):
M Max = 100 dB EM
M Min = Presentation Level Test Ear − IA
M Mid = 70 dB EM
+ Largest Air-Bone Gap Nontest Ear
= 95 dB HL − 40 dB + 5 dB It should be apparent that an increase in IA has equal
= 60 dB EM yet opposite effects on the minimum and maximum
106 SECTION I • Basic Tests and Procedures

FIGURE 6.14 An example illustrating the use of the midplateau and simplified masking procedures
during assessment of suprathreshold speech recognition.
 CHAPTER 6 • Clinical Masking 107

masking levels. Because IA is increased by 20 dB when using measured at approximately 0 dB HL in the right ear and
insert earphones, the width of the masking plateau is increased 70 dB HL in the left ear. Prior to measurement of speech
by 40 dB. The midmasking level, however, remains the same. thresholds, we can accurately predict whether contralateral
We now will take one final look at the example pre- masking will be required. Contralateral masking will be
sented in Figure 6.14 and consider the use of the simplified required only when measuring SRT in the left ear because
masking approach. Because the two prerequisite conditions the estimated speech threshold of 70 dB HL exceeds the best
are met, the simplified approach should result in the selec- bone-conduction threshold of 0 dB HL in the nontest ear
tion of an optimal masking level (i.e., midmasking level). by a conservative estimate of IA for speech (i.e., 60 dB). An
Recall that EM level is simply calculated as the presenta- unmasked SRT is subsequently measured in the left ear at
tion level of the speech signal in dB HL at the test ear minus 0 dB HL, a finding consistent with the puretone results.
20 dB. The same equation will be applicable when using any It has already been demonstrated that the simplified
earphone (i.e., 3A and supra-aural): masking procedure proves very effective during assessment
of suprathreshold speech recognition. However, it can also
dB EM = Presentation Level Test Ear − 20 dB be applied effectively during measurement of SRT. When
= 95 dB HL − 20 dB selecting an appropriate contralateral masking level when
= 75 dB measuring SRT, it is important to consider not only the HL
at which the speech threshold is finally established, but also
It should be noted that the masking level of 75 dB EM the highest presentation levels used throughout the thresh-
calculated using the simplified approach is in good agree- old procedure. Regardless of the SRT procedure used, spon-
ment (i.e., ±5 dB) with the value (i.e., 70 dB EM) deter- daic words are presented typically at both suprathreshold
mined using the midplateau procedure. Yacullo (1999) and threshold levels.
indicates that there are two major advantages to using the For example, ASHA (1988) recommends a descend-
simplified masking approach with 3A insert earphones, ing threshold technique for measuring SRT that is based
which are the result of a wider masking plateau. First, there on the earlier work of others (Hirsh et al., 1952; Hudgins
is greater flexibility in deviating somewhat from the cal- et al., 1947; Tillman and Olsen, 1973; Wilson et al., 1973).
culated midmasking level while still remaining within an The initial phase involves familiarizing the patient with
acceptable range of midplateau. It should be noted that the the spondaic words at a comfortable, suprathreshold HL.
midplateau actually represents a small range of values sur- (Familiarization with test words is strongly recommended
rounding the midmasking level. This range of acceptable regardless of the SRT procedure.) The preliminary phase
values essentially increases when using 3A insert earphones. involves setting the HL to 30 to 40 dB above the estimated
The use of the simplified masking approach can sometimes or predicted SRT before descending in 10-dB decrements
result in the selection of high masking levels, even though until two words are missed. In fact, an HL of 30 to 40 dB
overmasking is not occurring. Consequently, the audiolo- above the predicted SRT typically results in a comfortable
gist can justify subtracting a value of greater than 20 dB listening level for most patients during the familiariza-
(e.g., 25 or 30 dB) from the presentation level at the test ear tion phase. The test phase involves initially presenting test
when using insert earphones. In the example presented in words at HLs approximately 10 dB higher than the calcu-
Figure 6.14, an EM level equal to the presentation level in lated SRT. The calculation of threshold, based on a statisti-
the test ear minus 25 or 30 dB (e.g., 65 or 70 dB EM) would cal precedent, takes into account the patient’s responses at
still result in an acceptable masking level that falls within higher HLs.
the vicinity of midplateau, yet clearly exceeds the minimum Consider again the example presented in Figure 6.15. If
level by a sizeable amount. the ASHA-recommended procedure is used to measure SRT,
Second, there is greater flexibility in deviating slightly then the highest HLs employed (during the familiarization
from the recommended prerequisite conditions (i.e., no and preliminary phases) will be about 30 to 40 dB above the
air-bone gaps in either ear, use of moderate SLs) while still estimated SRT. In this example, we will use a moderate SL
remaining within an acceptable range of midplateau and of 30 dB above the estimated SRT (i.e., 70 dB HL + 30 dB
without significantly increasing the potential for overmasking. SL = 100 dB HL) during the familiarization and preliminary
Consequently, there is greater margin for error when select- phases.
ing an appropriate level of masking. The use of the simplified approach to selecting an
The example presented in Figure 6.15 illustrates the appropriate level of contralateral masking should prove
application of the midplateau and simplified masking pro- effective in this case because both prerequisite conditions
cedures during measurement of SRT. Audiometry was per- have been met. First, there are no significant air-bone gaps
formed using 3A insert earphones. Puretone testing reveals in either ear. Second, speech is presented at a moderate SL
normal hearing in the right ear. There is a severe sensory/ (i.e., 30 dB) when the highest HLs are used during the test
neural hearing loss of flat configuration in the left ear. Based procedure (i.e., familiarization and preliminary phases).
on the puretone findings, it is predicted that SRTs will be Assuming that 100 dB HL is the highest presentation level
108 SECTION I • Basic Tests and Procedures

FIGURE 6.15 An example illustrating the use of the midplateau and simplified masking procedures
during assessment of speech recognition threshold.
 CHAPTER 6 • Clinical Masking 109

used during our test procedure, then EM level in the nontest ter suprathreshold word recognition tests at a specified SL
ear is calculated as follows: referenced to the SRT (Martin and Morris, 1989; Martin
et al., 1994, 1998), typically 30 or 40 dB (Martin and Mor-
dB EM = Presentation Level Test Ear − 20 dB ris, 1989; Martin et al., 1994). Finally, it can be applied
= 100 dB HL − 20 dB effectively during both threshold and suprathreshold mea-
= 80 dB sures of speech recognition.
Direct calculation of midmasking level is strongly
We can verify the appropriateness of the selected mask- recommended in cases where there is potential risk of
ing level by using the midplateau method: overmasking. Yacullo (1999) states that any factor that
increases minimum masking level or decreases maximum
M Min = Presentation Level Test Ear − IA masking level will reduce the width of the masking pla-
+ Largest Air-BoneGap Nontest Ear teau and increase the probability of overmasking. For
example, the presence of significant air-bone gaps in the
= 100 dB HL − 60 dB + 0 dB nontest ear and/or the use of higher SLs (i.e., ≥50 dB) will
= 40 dB EM increase minimum masking level. The presence of signifi-
cant air-bone gaps in the test ear will decrease maximum
M Max = Best BC Test Ear + IA − 5 dB masking level. In cases where the masking plateau is either
= 55 dB HL + 60 dB − 5 dB very narrow or nonexistent (e.g., unilateral or bilateral
conductive hearing loss), knowledge about minimum and
= 110 dB EM
maximum masking levels will allow the clinician to make
well-informed decisions about appropriate contralateral
M Mid = (M Min + M Max )/2 masking levels.
= (40 + 110)/2
= 75 dB EM
FOOD FOR THOUGHT
The masking level of 80 dB EM selected using the sim-
1. Discuss how IA influences decisions about the need
plified approach is in good agreement (i.e., ±5 dB) with
for contralateral masking during puretone and speech
the value determined using the midplateau approach (i.e.,
audiometry.
75 dB EM). Although spondees will be presented at lower
2. Discuss how the OE influences measured IA for air-
HLs when measuring the SRT, it is not necessary to decrease
conduction transducers (i.e., supra-aural and 3A insert
the original level of masking. First, the selected masking
earphones with deeply inserted foam eartips). How does
level is appropriate for the highest HLs used during all
the OE influence contralateral masking levels during
phases of threshold determination. Second, the selected
bone-conduction audiometry?
masking level does not exceed the maximum masking level
3. The plateau masking procedure originally was described
(i.e., overmasking will not occur).
by Hood as a method for contralateral masking during
It can be argued that the simplified approach (as well
puretone threshold audiometry. Discuss how the underly-
as the midplateau method) can result in the use of unnec-
ing principles of the masking plateau are applied to proce-
essarily high masking levels during measurement of SRT.
dures for contralateral masking during speech audiometry.
As was discussed earlier, the midplateau represents a range
of masking levels. The audiologist can justify subtracting a
value of greater than 20 dB (e.g., 25 or 30 dB) from the pre-
sentation level at the test ear, particularly when using insert KEY REFERENCES
earphones. In this example, a masking level of 70 or 75 dB A full list of references for this chapter can be
EM (rather than the original level of 80 dB EM) still falls found at http://thePoint.lww.com. Below are the key refer-
within an acceptable range of midplateau, while still occur- ences for this chapter.
ring significantly higher than the minimum.
Yacullo (1999) states that the simplified masking American National Standards Institute/Acoustical Society of
approach during speech audiometry has wide applicability. America (ANSI/ASA). (2010) American National Standard
Specification for Audiometers. ANSI/ASA 3.6–2010. New York:
First, it can be used with a large and diverse patient popu-
ANSI/ASA.
lation, including those with normal hearing and sensory/ American Speech-Language-Hearing Association. (1988) Deter-
neural hearing loss. Second, it can be used equally effec- mining threshold level for speech [Guidelines]. Available
tively when using either supra-aural or insert earphones. online at: www.asha.org/policy.
Third, the procedure can be used in clinical situations American Speech-Language-Hearing Association. (1990) Audio-
where moderate SLs are used. For example, the majority metric symbols [Guidelines]. Available online at: www.asha.
of audiologists in the United States continue to adminis- org/policy.
110 SECTION I • Basic Tests and Procedures

American Speech-Language-Hearing Association. (2005) Guide- Hirsh IJ, Davis H, Silverman SR, Reynolds EG, Eldert E, Benson
lines for manual pure-tone threshold audiometry [Guide- RW. (1952) Development of materials for speech audiometry.
lines]. Available online at: www.asha.org/policy. J Speech Hear Disord. 17, 321–337.
Beattie RC, Svihovec DV, Edgerton BJ. (1978) Comparison of Hodgson W, Tillman T. (1966) Reliability of bone conduction
speech detection and spondee thresholds for half- versus full- occlusion effects in normals. J Aud Res. 6, 141–151.
list intelligibility scores with MLV and taped presentations of Hood JD. (1960) The principles and practice of bone-conduction
NU-6. J Am Audiol Soc. 3, 267–272. audiometry. Laryngoscope. 70, 1211–1228.
Berger EH, Kerivan JE. (1983) Influence of physiological noise and Hosford-Dunn H, Kuklinski AL, Raggio M, Haggerty HS. (1986)
the occlusion effect on the measurement of real-ear attenua- Solving audiometric masking dilemmas with an insert masker.
tion at threshold. J Acoust Soc Am. 74, 81–94. Arch Otolaryngol Head Neck Surg. 112, 92–95.
Berrett MV. (1973) Some relations between interaural attenuation Hudgins CV, Hawkins JE Jr., Karlin JE, Stevens SS. (1947) The
and the occlusion effect. Unpublished doctoral dissertation. development of recorded auditory tests for measuring hearing
Iowa City, IA: University of Iowa. loss for speech. Laryngoscope. 57, 57–89.
Blackwell KL, Oyler RF, Seyfried DN. (1991) A clinical comparison Jerger J, Jerger S. (1971) Diagnostic significance of PB word func-
of Grason Stadler inserts earphones and TDH-50P standard tions. Arch Otolaryngol. 93, 573–580.
earphones. Ear Hear. 12, 361–362. Jerger J, Jerger S, Ainsworth J, Caram P. (1966) Recovery of audi-
Chaiklin JB. (1959) The relation among three selected auditory tory function after surgical removal of cerebellar tumor.
speech thresholds. J Speech Hear Res. 2, 237–243. J Speech Hear Disord. 31, 377–382.
Chaiklin JB. (1967) Interaural attenuation and cross-hearing in Kaplan H, Gladstone VS, Lloyd LL. (1993) Audiometric Inter-
air-conduction audiometry. J Aud Res. 7, 413–424. pretation. 2nd ed. Needham Heights, MA: Allyn & Bacon.
Coles RRA, Priede VM. (1970) On the misdiagnosis resulting from Killion MC. (1984) New insert earphones for audiometry. Hear
incorrect use of masking. J Laryngol Otol. 84, 41–63. Instrum. 35, 28, 46.
Coles RRA, Priede VM. (1975) Masking of the non-test ear in Killion MC, Villchur E. (1989) Comments on “Earphones in audi-
speech audiometry. J Laryngol Otol. 89, 217–226. ometry” [Zwislocki et al., J. Acoust. Soc. Am. 83, 1688–1689].
Dean MS, Martin FN. (2000) Insert earphone depth and the occlu- J Acoust Soc Am. 85, 1775–1778.
sion effect. Am J Audiol. 9, 131–134. Killion MC, Wilber LA, Gudmundsen GI. (1985) Insert earphones
Dirks DD. (1994) Bone-conduction threshold testing. In: Katz J, for more interaural attenuation. Hear Instrum. 36, 34, 36.
ed. Handbook of Clinical Audiology. 4th ed. Baltimore, MD: Konkle DF, Berry GA. (1983) Masking in speech audiometry. In:
Williams & Wilkins; pp 132–146. Konkle DF, Rintelmann WF, eds. Principles of Speech Audiom-
Dirks DD, Malmquist C. (1964) Changes in bone-conduction etry. Baltimore, MD: University Park Press; pp 285–319.
thresholds produced by masking in the non-test ear. J Speech Lidén G, Nilsson G, Anderson H. (1959) Masking in clinical audi-
Hear Res. 7, 271–278. ometry. Acta Otolaryngol. 50, 125–136.
Dirks DD, Swindeman JG. (1967) The variability of occluded and Littler TS, Knight JJ, Strange PH. (1952) Hearing by bone conduc-
unoccluded bone-conduction thresholds. J Speech Hear Res. tion and the use of bone-conduction hearing aids. Proc R Soc
10, 232–249. Med. 45, 783–790.
Dunn HK, White SD. (1940) Statistical measurements on conver- Martin FN. (1966) Speech audiometry and clinical masking. J Aud
sational speech. J Acoust Soc Am. 11, 278–288. Res. 6, 199–203.
E-A-R Auditory Systems. (1997) Instructions for the Use of E-A- Martin FN. (1967) A simplified method for clinical masking. J Aud
RTONE 3 A Insert Earphones. Revised ed. Indianapolis, IN: Res. 7, 59–62.
E-A-R Auditory Systems. Martin FN. (1974) Minimum effective masking levels in threshold
E-A-R Auditory Systems. (2000a) Instructions for the Use of E-A-RTONE audiometry. J Speech Hear Disord. 39, 280–285.
5 A Insert Earphones. Indianapolis, IN: E-A-R Auditory Systems. Martin FN. (1980) The masking plateau revisited. Ear Hear. 1,
E-A-R Auditory Systems. (2000b) Introducing the New E-A-RTONE® 112–116.
5 A Insert Earphone [Brochure]. Indianapolis, IN: E-A-R Auditory Martin FN, Armstrong TW, Champlin CA. (1994) A survey
Systems. of audiological practices in the United States in 1992. Am
Elpern BS, Naunton RF. (1963) The stability of the occlusion effect. J Audiol. 3, 20–26.
Arch Otolaryngol. 77, 376–382. Martin FN, Bailey HAT, Pappas JJ. (1965) The effect of central
Feldman AS. (1963) Maximum air-conduction hearing loss. masking on threshold for speech. J Aud Res. 5, 293–296.
J Speech Hear Disord. 6, 157–163. Martin FN, Blythe ME. (1977) On the cross hearing of spondaic
Fletcher H. (1940) Auditory patterns. Rev Mod Phys. 12, 47–65. words. J Aud Res. 17, 221–224.
Frank T, Vavrek MJ. (1992) Reference threshold levels for an ER-3 Martin FN, Butler EC, Burns P. (1974) Audiometric Bing test for
A insert earphone. J Am Acad Audiol. 3, 51–58. determination of minimum masking levels for bone conduc-
Gelfand SA. (2009) Essentials of Audiology. 3rd ed. New York: tion tests. J Speech Hear Disord. 39, 148–152.
Thieme Medical Publishers, Inc. Martin FN, Champlin CA, Chambers JA. (1998) Seventh survey of
Goldstein DP, Hayes CS. (1965) The occlusion effect in bone-con- audiometric practices in the United States. J Am Acad Audiol.
duction hearing. J Speech Hear Res. 8, 137–148. 9, 95–104.
Hannley M. (1986) Basic Principles of Auditory Assessment. Need- Martin FN, DiGiovanni D. (1979) Central masking effects on
ham Heights, MA: Allyn & Bacon. spondee threshold as a function of masker sensation level
Hawkins JE, Stevens SS. (1950) Masking of pure tones and of and masker sound pressure level. J Am Audiol Soc. 4, 141–
speech by white noise. J Acoust Soc Am. 22, 6–13. 146.
 CHAPTER 6 • Clinical Masking 111

Martin FN, Morris LJ. (1989) Current audiologic practices in the Thurlow WR, Silverman SR, Davis H, Walsh TE. (1948) A statisti-
United States. Hear J. 42, 25–44. cal study of auditory tests in relation to the fenestration opera-
Naunton RF. (1960) A masking dilemma in bilateral conduction tion. Laryngoscope. 58, 43–66.
deafness. Arch Otolaryngol. 72, 753–757. Tillman TW, Olsen WO. (1973) Speech audiometry. In: Jerger J,
Olsen WO, Matkin ND. (1991) Speech audiometry. In: Rintelmann ed. Modern Developments in Audiology. 2nd ed. New York:
WR, ed. Hearing Assessment. Needham Heights, MA: Allyn & Academic Press; pp 37–74.
Bacon; pp 39–140. Tonndorf J. (1968) A new concept of bone conduction. Arch
Owens E, Schubert ED. (1977) Development of the California Otolaryngol. 87, 49–54.
Consonant Test. J Speech Hear Res. 20, 463–474. Tonndorf J. (1972) Bone conduction. In: Tobias JV, ed. Foundations
Sanders JW. (1972) Masking. In: Katz J, ed. Handbook of Clinical of Modern Auditory Theory. Vol II. New York: Academic Press;
Audiology. Baltimore, MD: Williams & Wilkins; pp 111–142. pp 197–237.
Sanders JW. (1991) Clinical masking. In: Rintelmann WF, ed. Hearing Turner RG. (2004a) Masking redux I: An optimized masking
Assessment. Needham Heights, MA: Allyn & Bacon; pp 141–178. method. J Am Acad Audiol. 15, 17–28.
Sanders JW, Rintelmann WF. (1964) Masking in audiometry. Turner RG. (2004b) Masking redux II: A recommended masking
Arch Otolaryngol. 80, 541–556. protocol. J Am Acad Audiol. 15, 29–46.
Schwartz DM, Surr R. (1979) Three experiments on the California Van Campen LE, Sammeth CA, Peek BF. (1990) Interaural attenua-
Consonant Test. J Speech Hear Disord. 64, 61–72. tion using Etymotic ER-3 A insert earphones in auditory brain
Silman S, Silverman CA. (1991) Auditory Diagnosis: Principles and stem response testing. Ear Hear. 11, 66–69.
Applications. San Diego, CA: Academic Press. Wegel RL, Lane GI. (1924) The auditory masking of one pure tone
Sklare DA, Denenberg LJ. (1987) Interaural attenuation for Tube- by another and its probable relation to the dynamics of the
phone insert earphones. Ear Hear. 8, 298–300. inner ear. Phys Rev. 23, 266–285.
Smith BL, Markides A. (1981) Interaural attenuation for pure Wilson R, Morgan D, Dirks D. (1973) A proposed SRT proce-
tones and speech. Br J Audiol. 15, 49–54. dure and its statistical precedent. J Speech Hear Disord. 38,
Snyder JM. (1973) Interaural attenuation characteristics in audi- 184–191.
ometry. Laryngoscope. 73, 1847–1855. Yacullo WS. (1996) Clinical Masking Procedures. Boston, MA: Allyn
Studebaker GA. (1962) On masking in bone-conduction testing. & Bacon.
J Speech Hear Res. 5, 215–227. Yacullo WS. (1999) Clinical masking in speech audiometry: A
Studebaker GA. (1964) Clinical masking of air- and bone- simplified approach. Am J Audiol. 8, 106–116.
conducted stimuli. J Speech Hear Disord. 29, 23–35. Yacullo WS. (2004) Clinical masking. In: Kent RD, ed. The MIT
Studebaker GA. (1967a) Clinical masking of the non-test ear. Encyclopedia of Communication Disorders. Cambridge, MA:
J Speech Hear Disord. 32, 360–371. MIT Press; pp 500–504.
Studebaker GA. (1967b) Intertest variability and the air-bone gap. Zwislocki J. (1953) Acoustic attenuation between the ears. J Acoust
J Speech Hear Disord. 32, 82–86. Soc Am. 25, 752–759.
Studebaker GA. (1979) Clinical masking. In: Rintelmann WF, ed. Zwislocki J, Kruger B, Miller JD, Niemoeller AR, Shaw EA, Stude-
Hearing Assessment. Baltimore, MD: University Park Press; baker G. (1988) Earphones in audiometry. J Acoust Soc Am. 83,
pp 51–100. 1688–1689.
 C H A P T ER 7

Case History

Douglas L. Beck

Researchers have designed decision trees and analysis

INTRODUCTION weightings and other complex models which are powerful
Audiologists must critically and judiciously gather and and accurate and, theoretically, will assist in finding the cor-
examine all information related to sound perception, tin- rect diagnosis. However, when the audiologist is working
nitus, hearing, hearing loss, listening (in quiet and noise), with the patient, assembling the case history is essentially a
dizziness, balance problems, and birth history (for newborns person-to-person event. Frankly, having good people skills
and infants). The audiologist creates and interprets anatomic and adjusting our approach (i.e., course corrections) to the
and physiologic information within the context of a case client we are addressing, matter a great deal.
history, to render an appropriate diagnosis. Audiologists are
charged with the responsibility of diagnosing and “nonmedi-
cally” treating hearing loss. Traditional wisdom reveals two
CASE HISTORY TOOLS
key ingredients to a correct differential diagnosis: An excel- There are three primary tools used to create a case history:
lent case history and a thorough physical examination. Given Interviews, questionnaires, and the subjective, objective,
these two key ingredients, the differential diagnosis “emerges” assessment, and plan (SOAP) format. These three tools are
to the trained professional as the only clear answer (i.e., a often used in tandem, but can certainly be used as preferred
single differential diagnosis) or potentially a series of equally by the professional.
plausible diagnoses emerge, indicating multiple remaining The audiologist, as a licensed or regulated healthcare
questions and avenues to be explored and resolved. Indeed, professional, has a legal obligation to the patient’s health and
case history gathering is an important skill which facilitates well-being. The audiologist must be aware of the warning
the correct differential diagnosis if the clinician obtains rel- signs of dangerous and treatable medical and surgical condi-
evant and focused information. Obtaining and using the case tions and should refer to the appropriate professional when
history requires skill, patience, practice, and knowledge. “red flags” are noticed. Red flags include a sudden hearing
In general, if you’re not looking for something, you won’t loss, ear pain, draining or bleeding ears, unilateral symptoms
find it. However, simply looking for something doesn’t mean of hearing loss or tinnitus, conductive hearing loss, dizziness,
you will find it. For example, if you’re looking for zebras in and other referral criteria. Assembling the case history pro-
a cow pasture, you probably won’t find them. Of course this vides an opportunity to identify red flags while considering a
doesn’t mean zebras don’t exist. However, it may indicate multitude of diagnostic and treatment alternatives.
zebras generally don’t hang out in cow pastures. To find the
correct solution to a given problem, we must pose the right
question, formulate reasonable options and alternatives, and
Interview Techniques
ultimately, choose the most probable alternative. Of course, there is no “one correct way” to interview patients.
When gathering and assembling case histories, health- Flexibility is the key, as professionals, patients, work set-
care professionals must narrow the focus and filter the tings, and the particulars of each situation vary. Nonethe-
information available quickly and efficiently while pulling less, it is always a good idea to proceed in an orderly and
together what is most important. The case history questions professional manner. Interviews should be patient centered,
should be reasonable, and result driven, allowing for an friendly, and private, in accordance with applicable laws,
evidence-based outcome. rules, and regulations.
Across healthcare disciplines (including audiology), While gathering the case history, ascertaining an “index
the method of choice for obtaining the case history is the of suspicion” regarding the CC is an important part of the
“medical model.” The medical model efficiently directs the interview. If the index of suspicion for the item highest on
professional to the “chief complaint” (CC) and helps orga- our list is low, we generally need to look for more probable
nize information into a rational hierarchy with the most alternatives. If the index of suspicion is high, we ask further
important or likely concerns at the forefront. questions to confirm or refute our suspicions.

113
114 SECTION I • Basic Tests and Procedures

For example, a patient presenting with a fluctuating little loud. The same thing happens with the car radio when
low-frequency sensory/neural hearing loss (SNHL) and we’re driving to the store. When she sets the volume, I just
tinnitus in the same ear, with aural fullness and occasional hear noise and can’t tell anything about what they’re saying.
vertigo, has a low index of suspicion for otosclerosis, but has When I was a boy, I could hear a pin drop from 40 paces.”
a reasonably high index of suspicion for Ménière’s disease.
“I understand. How long have you been playing the TV and
The high index of suspicion for Ménière’s disease would
radio louder than your wife likes it?”
lead us to ask probing questions to note whether the pre-
senting symptomatology is in agreement with a Ménière’s “Let’s see, I started working at the steel fabrication factory
diagnosis or would lead us in another direction, such as an 14 years ago, and my son was born 8 years ago . . . . so yeah,
acoustic neuroma. it’s been at least 8 or 10 years. When I let her set the vol-
The competent professional understands the probabili- ume, I can hear the voices, but I really can’t understand what
ties of certain things occurring and the related signs and they’re saying. That drives me nuts. I told her and I’m telling
symptoms of each. Although Ménière’s disease is a relatively you too, I ain’t gonna wear no hearing aids.”
rare disorder, occurring in less than 1% of the general popu-
Given the information presented in this scenario, one
lation, it is a common diagnosis for patients with the symp-
can make several, reasonable, assumptions. We could assume
toms described earlier. Of course, we do not make a medical
that Mr. Smith has a noise-induced SNHL, likely impacting
diagnosis of a disease. However, if the data come together
4,000 Hz, and because he cannot hear the consonant sounds
with what we know about the patient the audiologist might
(high frequencies), he cannot clearly understand the words spo-
include “Consistent with Ménière’s disease,” in the report.
ken to him. We might also assume that Mr. Smith is not going
Three scenarios follow to illustrate the interview
to wear hearing aids and that there is little we can do to assist.
technique.
However, there are other options and protocols to employ:
SCENARIO ONE “Mr. Smith, have you had a hearing test before?”
Review any/all assembled paperwork (chart, lab notes, test “Not since the Army, back some 20 years ago.”
results, history, etc.) before meeting the patient for the ini- “Do both ears seem about the same, or is one ear better than
tial consultation. Shake hands and greet the patient, their the other?”
spouse, significant other, family, and so on, and always
introduce yourself. This is an amazingly simple protocol, “The left ear is terrible—can’t hear thunder with that one.”
but it is often overlooked, and when it is overlooked, it sets “I see. Do you have any ear pain?”
an unprofessional tone for the rest of the encounter. I usu-
ally say, “Good morning. My name is Dr. Beck, I’m an audi- “None at all. My ears feel fine.”
ologist. Please come in Mr. Smith.” “Okay then. May I take a look?”
After exchanging greetings and after sitting down in the
“Sure, help yourself.”
office, inquire as to why the patient scheduled today’s visit.
At this point, the audiologist has a rather low index of
“Thanks for coming in today Mr. Smith. What brings you
suspicion for a tumor, such as an acoustic neuroma, because
to the office?”
they occur in about 0.00001% of the population, but a higher
Mr. Smith: “I would like a comprehensive audiometric eval- index of suspicion for more likely possibilities, including a
uation to confirm my bilateral sensory/neural noise-induced unilateral sudden sensory/neural loss that went undiagnosed
hearing loss that my otolaryngologist diagnosed last week. I (or maybe Mr. Smith works with his left ear toward a loud
am very interested in acquiring two digital hearing aids, and machine while wearing hearing protection only in the right
by the way, I am wealthy and do not have insurance. I pay ear, or perhaps he experienced head trauma on the left or an
cash, and money is no object. I want to hear everything as explosion near his left side during boot camp; there are lots of
best I can.” possibilities). The examination of the pinna, concha, ear canal,
and tympanic membranes is normal. The audiologist says,
Because this patient has already been seen and diag-
“Okay, your ears look fine,” and the interview continues to
nosed by the ear, nose, and throat (ENT) specialist, the index
determine which diagnosis has the highest index of suspicion.
of suspicion for some other disease process or a medical/
surgical issue is extremely low. “Mr. Smith, let me make sure I understand . . . the right ear is
the better ear and the left ear has been bad for a long time.
SCENARIO TWO Have you ever had the left ear checked?”
Mr. Smith: “Well doc, you know how it is. My wife always “Yes. I had the doctor look at it a year or two ago when it
complains I have the TV up too loud and it drives her outta went bad. He put me on antibiotics and that was the end of
the room. I like to be able to hear the darn thing so I keep it a it. It didn’t get better though, so I left it alone.
 CHAPTER 7 • Case History 115

“Okay. What about drainage, anything coming out of so yeah, it’s been at least 8 or 10 years. When my wife sets
your ears?” the TV, it sounds like everyone is mumbling; I can hear the
voices, but I really can’t understand what they’re saying.
“No sir.”
That drives me nuts. I told her and I’m telling you too, I
“Any dizziness or spinning sensations?” ain’t gonna wear no hearing aids.”
Given the information presented above, one can
“Not any more. Well, maybe a little. When my left ear was
make several assumptions. We could assume Mr. Smith
going bad, I had some dizziness, but the doctor looked at it
has a noise-induced SNHL, impacting frequencies around
and put me on antibiotics, and the dizziness got better after
4,000 Hz, and because of the reduced amplitude and dis-
a while.”
tortion affecting mostly the high-frequency consonant
“So the dizziness started and the left ear went bad all sounds, he cannot clearly hear the words spoken to him.
about a year or two ago?” We can also be comfortable in thinking that Mr. Smith is
not going to wear hearing aids, which reduces what we can
“That’s right.”
do to assist him. However, there are other options and pro-
“Okay, very good. Are you on any medications?” tocols to explore.
“Just a cholesterol pill and a baby aspirin, that’s about it.” “Mr. Smith, have you had a hearing test before?”
“Okay, and one last thing I’d like to ask you before we do the “Not since the Army, back some 20 years ago.”
hearing test—do you have any ringing or buzzing noises in
your ears?” “Do both ears seem about the same, or is one ear better than
the other?”
“Yeah, the darn left ear can’t hear anything, but it sure makes
a racket. Kinda like a “shhhhh” noise going on in there. “They’re just about the same”
Keeps me up at night sometimes.” “I see. Any ear pain?”
The audiologist does a comprehensive audiomet- “None at all. My ears feel fine.”
ric evaluation and determines the following audiometric
profile: “That’s good. May I take a look?”
Right ear: Normal peripheral hearing. Tympanogram “Sure doc, knock yourself out.”
normal (type A), ipsilateral reflexes within normal lim-
its (WNL). Word recognition score (WRS) = 96%. Speech The pinna, concha, ear canal, and tympanic membranes
reception threshold (SRT) = 15 dB HL. are normal in appearance. The audiologist says, “Your ears
Left ear: Flat 85 dB sensory/neural (SN) loss. Tympano- look fine,” and the interview continues.
gram normal (type A), ipsilateral reflexes absent @105 dB “Okay, what about drainage? Is there anything coming out
stimulus level. WRS = 8%, SRT = SAT (speech awareness of your ears?”
threshold used because speech understanding was extremely
poor) = 80 dB HL. “No sir.”
The index of suspicion for a left retrocochlear disorder
“Any dizziness or spinning sensations”
is very high at this point. The case history supports this pos-
sibility, and the test results indicate a possible retrocochlear “Nope.”
diagnosis for the left ear.
“Very good. Are you taking any medications?”
The audiologist refers the patient to an otolaryngolo-
gist (preferably an otologist or neurotologist) based on the “Just a cholesterol pill and a baby aspirin, that’s about it.”
high index of suspicion for a retrocochlear hearing loss. The
“The last question I’d like to ask you before we do the
otologist meets with and interviews the patient and refers
hearing test is do you have any ringing or buzzing noises in
the patient for a magnetic resonance imaging (MRI) study
your ears?”
with contrast (gadolinium). A 3-cm vestibular schwannoma
(acoustic neuroma) is diagnosed. Mr. Smith is scheduled “Yeah . . . maybe a little when it’s really quiet, nothing that
for surgery 3 weeks later, and the tumor is removed via the really bothers me though.”
translabyrinthine approach.
The audiologist does a comprehensive audiometric eval-
uation and determines the following audiometric profile:
Right ear: Moderate high-frequency sensory/neural
SCENARIO THREE
hearing loss. Tympanogram normal (type A), ipsilateral
Mr. Smith: “Let’s see, I started working at this really noisy reflexes are within normal limits (WNL). WRS = 96%.
factory 14 years ago, and my son was born 8 years ago . . . SRT = 45 dB HL.
116 SECTION I • Basic Tests and Procedures

Left ear: Moderate high-frequency sensory/neural hear- background page are preferred and appreciated. Black print
ing loss. Tympanogram normal (type A), ipsilateral reflexes on a white background will be the easiest to read. Another
are WNL. WRS = 92%. SRT = 45 dB HL. important consideration is to use and/or design question-
naires that are easily assessed and tabulated, so the profes-
“Mr. Smith, I’d like to review the results of today’s tests with
sional can scan the page to find the “positive” results, which
you. Would you like to have your wife join us while I review
will need to be considered.
the results?”
In 2005, the Centers for Medicare and Medicaid Ser-
“Sure, that would be great. She’s in the waiting room.” vices (CMS) added a new benefit under Medicare Part B
that will likely increase the quantity of pencil and paper–
“Hi Mrs. Smith, please join us while I review the results of today’s
based hearing and balance screenings offered to patients.
examination. This way, the two of you will have the chance to
This benefit is “bundled” within the “Welcome to Medicare”
learn about the results, and I can address your questions.”
examination. The examination has seven screening sections
In this third scenario, the index of suspicion for a noise- for physicians, nurses, or nurse practitioners to employ
induced hearing loss is high, and there are no red flags and when addressing new patients. Importantly, the Medicare
no indications of a medical or surgical problem. In essence, rules state that the screening tests must be in the form of
the same patient, in three different scenarios, has three questions or questionnaires and that the selected screening
separate sets of circumstances, each of which are typically tests must be recognized by a national medical professional
revealed through an interview-based case history, which is organization.
more or less driven by the index of suspicion. In addition to a wealth of other tests, the American
Academy of Audiology (AAA) and the American Speech-
Questionnaires Language-Hearing Association (ASHA) have recommended
that the following questionnaire be used for this purpose:
Another very useful and efficient case history tool is the Hearing Handicap Inventory for the Elderly—Screening
health questionnaire. A well-designed questionnaire is Version (HHIE-S; Ventry and Weinstein, 1982). There is
highly focused, simple, takes just a few minutes to fill out, likely to be greater popularity for screening tests. Therefore,
and quickly directs the professional to the area(s) of great- audiologists should be familiar with the above-noted ques-
est concern. Questionnaires regarding hearing health care tionnaires and their format, scoring, and importance.
can be presented to patients verbally or written. Written
questionnaires are available in electronic and paper-based Subjective, Objective,
formats.
However, it is my personal preference to not have
Assessment, and Plan
patients fill in downloadable, lengthy questionnaires at Another way to gather useful case history information
home. It is terrifically advantageous for the audiologist to quickly is to use the standard subjective, objective, assess-
spend the time working through a well-designed question- ment, and plan (SOAP) format. The SOAP format is
naire with the patient, to establish rapport and trust and essentially a “medical model” case history–gathering tool.
to allow the patient to tell their story. We learn much more There are many variations on the SOAP format used by
about the patient and their situation when we put in the time clinics, medical schools, and, often, military healthcare
to ask questions and listen to the patient and then write chart facilities.
notes reflecting that conversation. Time spent asking ques- Critics believe the SOAP format is impersonal and does
tions and listening to and then questioning and clarifying not recognize the patient as a whole person. In addition,
their response is time well spent. the SOAP format tends to treat the person as if he or she
was the disease/disorder/problem, and it calls for the use of
VERBAL PRESENTATIONS jargon and related abbreviations. Although jargon is com-
monly used in health professions, it can vary from location
Remember, if you are giving a patient a verbal presentation to location, and it may be nearly impossible for many well-
of a hearing questionnaire, there is already a reasonable educated colleagues to interpret. In the following examples,
index of suspicion for hearing loss. Therefore, sit about 3 ft abbreviations will be used along with their explanations,
away from the patient in a well-lit room, face the patient, be which will immediately follow in parenthesis.
sure there is no background noise or visual distractions, and
maintain the patient’s full attention. SUBJECTIVE
The subjective section provides a brief subjective history,
PENCIL AND PAPER PRESENTATIONS often focusing on the CC as well as other clinical observa-
Keep in mind that, because the majority of patients seen by tions. The patient’s medical and audiology history would be
audiologists are over 55 years of age, large font, dark print, placed in this section. Other entries in this section would
and maximal contrast between the printed words and the be notes the patient/relative/friends offer regarding pain or
 CHAPTER 7 • Case History 117

discomfort and related miscellaneous symptoms. An exam- PLAN

ple follows:
The plan is the “plan” as of this moment, moving forward.
Pt (patient) is 56-year-old, Caucasian, divorced female. The physician may write the recommended prescriptions
NKA (no known allergies). or may order blood tests, lab work, or radiology tests, as
Pt has one adult daughter (age 26 years). needed. The audiologist might write
Pt has +BP (high blood pressure) that has been under con-
trol via meds for 3 years. Pt takes daily multivitamin. Refer pt to ENT for AD asymmetric, SNHL to R/O (rule
No other known medical issues. out) retrocochlear origin or other medical/surgical con-
Pt consumes ETOH (alcohol) daily (one glass), stopped cerns. Assuming medical/surgical is R/O, proceed with
smoking 15 years ago. hearing aid evaluation AD.
Previous surgery: C-section 26 years ago. Ingrown toenail Although the SOAP format is a quick and an efficient way
(left big toe) operated on 22 years ago. to gather the history and related information, it may ignore
Today CC: Hearing loss AD (right ear) × 1 mo (1 month more global problems, while attending primarily to the CC.
duration) with tinnitus, no spinning/vertigo, no com-
plaints AS (left ear).
Pt presents for AE (audiometric evaluation). SUMMARY
Gathering an efficient and thorough case history requires
OBJECTIVE understanding, patience, and knowledge of hearing, hear-
ing loss, and related disorders. Although there are options
In medical charts, the objective section often includes mea- regarding the preferred method with which to gather a case
sures of temperature, blood pressure, skin color, swelling, history, there is no alternative to accuracy.
and other “objective” data that can be obtained in the office Whichever protocol(s) is (are) chosen, the clinician has
easily and quickly. This section is where the audiologist the responsibility of assembling the information in a mean-
would write the “objective” test results. An example follows: ingful and relevant way to maximally address the needs,
Puretones: concerns, and well-being of the patient.
65 dB HL SNHL (sensory/neural hearing loss) AD (right)
AS (left) WNL (within normal limits) FOOD FOR THOUGHT
SRT (speech reception threshold):
1. As we move forward, of course the electronic medical
70 dB HL AD, 15 dB HL AS
record (EMR) will play a more prominent role in medi-
SAT (speech awareness threshold):
cine, audiology and will increasingly impact the case
60 dB HL AD
history. Although the EMR will eventually be standard-
15 dB HL AS
ized and comprehensive (at least we can hope!). Do you
WRS (word recognition score):
believe a better “Case History” will come about using a
24% AD at SAT plus 35 dB with contralateral masking
one-on-one dialog, rather than a checklist approach?
100% AS
2. Do you feel that the human side (i.e., information, emo-
OAEs (otoacoustic emissions):
tions and the relationship between the professional and
AD ABS (absent)
the patient) will greatly impact the type of case history
AS WNL
we use?
Tympanograms:
3. Although the standardized EMR will (eventually) enable
WNL AU (within normal limits, both ears)
the gathering and analysis of comprehensive and objec-
tive data, do you feel that this will reduce the value of
ASSESSMENT dialog between the patient and the professional? That is,
how important is the patient’s story more-or-less in their
The assessment section is where the physician or audiologist
own words for the professional?
would make a statement about the probable “working” diag-
nosis, or the final diagnosis, and prognosis. For example,
REFERENCE
Pt presents with probable AD SNHL (right sensory/neural
hearing loss), possibly from untreated sudden SNHL, Ventry IM, Weinstein BE. (1982) The hearing handicap inventory
possibly retrocochlear? for the elderly: A new tool. Ear Hear. 3, 128–134.
 C H A P T ER 8

Diagnostic Audiology

Brian M. Kreisman, Jennifer L. Smart, and Andrew B. John

of hearing loss or the site of lesion. The test battery is use-

INTRODUCTION ful for determining some, but not all, auditory disorders.
Diagnostic audiology is the use of audiologic tests to Following a brief introduction to the cross-check prin-
determine the location of a problem in the auditory sys- ciples employed by audiologists, this chapter will utilize a
tem and, in many cases, further insights about the disor- case study format. Finally, we will address the limitations
der. Diagnostic audiology can be likened to crime shows of the test battery and discuss situations when referrals for
you may watch on television. Each test serves as a clue that other testing are indicated.
points toward a diagnosis of the patient’s hearing disorder.
However, if individual tests (or clues) are examined with-
out taking other evidence into consideration, a wrong con- CROSS-CHECKING TEST RESULTS
clusion might be made. For audiologists to make correct
diagnoses, an audiologic test battery is used. A test battery The major reason that an audiologist uses a diagnostic bat-
is a series or combination of tests used to assess the audi- tery is to be able to check the results of individual tests with
tory system. For most of the examples in this chapter, we each other. The idea that “the results of a single test are
will limit discussion of diagnostic audiology to tests that cross-checked by an independent test measure” is referred
are commonly performed in an audiologic clinic, includ- to as the cross-check principle (Jerger and Hayes, 1976,
ing puretone air-conduction and bone-conduction testing, p. 614). Since the cross-check principle was first proposed,
speech testing, tympanometry, acoustic reflex thresholds many manuscripts have revisited the concept as new diag-
(ARTs, also called middle-ear muscle reflexes [MEMRs]), nostic tests have been developed and different test batteries
and otoacoustic emissions (OAEs). These tests are dis- have been proposed to diagnose specific disorders. The goal
cussed more fully in other chapters in this textbook and we of comparing the results of two or more tests is to increase
refer you to these chapters for specific test procedures and the rate of correct identification of disorders (hit rate) and
norms (see Table 8.1); however, it is important to under- to decrease the rate of diagnosing a disorder when no disor-
stand how to utilize these tests synergistically to arrive at der exists (false alarm rate) (Turner, 2003).
an accurate diagnosis for each patient. Audiometric tests
are used in conjunction with one another to help reinforce Cross-checks for Puretone Air
or, alternatively, rule out the diagnosis of a particular type
Conduction
If you only obtained puretone air-conduction thresholds
then you would not be able to accurately diagnose the type
TA B L E 8 .1 of hearing loss. Air-conduction audiometry is normally
cross checked with bone-conduction audiometry or tympa-
Audiology Procedures Discussed nometry to rule out a conductive component of the hearing
in This Chapter loss. If a difference greater than 10 dB exists between the air-
Test Chapter conduction and bone-conduction thresholds at the same
frequency, a conductive component is indicated. Similarly,
Puretone testing (air and bone 3 air-conduction thresholds for an ear may be within normal
conduction) limits; however, if a tympanogram for that ear falls outside
Speech testing (quiet) 5 of the norms for middle-ear pressure and compliance (e.g.,
Speech-in-noise testing 5 Jerger Type B or Type C), a conductive component may be
Tympanometry 9 present. ARTs can reveal more information about the type of
Acoustic reflex thresholds 10 loss based on the pattern of responses obtained, thus serving
Otoacoustic emissions 19
as an additional cross-check for puretone air conduction.

119
120 SECTION I • Basic Tests and Procedures

Cross-checks for Puretone in noisy environments is to administer a speech-in-noise

test in addition to the word recognition testing in quiet.
Audiometry Although this is technically not a cross-check, the addition
When puretone audiometry (air- and bone-conduction test- of a speech-in-noise test, especially with sentence stimuli,
ing) suggests a significant air–bone gap, tympanometry and will provide a more realistic test environment to evaluate
ARTs can be used to reinforce the diagnosis of the conductive a common patient complaint. The puretone audiogram
element and to contribute to a specific diagnosis. OAEs also does not necessarily correlate with the amount of difficulty
can be used as a cross-check of puretone audiometry. OAEs a listener will have in noise (Killion and Niquette, 2000).
are used to assess the health of the outer hair cells of the In addition, when word recognition testing is performed
cochlea, but their measurement may be affected by disorders in quiet at a single speech presentation level, no guaran-
in the conductive pathway. An audiologist might use OAEs tee exists that the test is measuring the patient’s maximum
as a cross-check to aid in potentially ruling out a nonorganic speech understanding (Wiley et al., 1995).
hearing loss, to verify outer hair cell function and the degree
of cochlear hearing loss, and to further assist with the diag-
nosis of conductive components, auditory neuropathy spec- Cross-check Considerations
trum disorder (ANSD), and other retrocochlear disorders. In for Pediatric Testing
addition, ARTs have been used to cross check puretone audi-
For children, it is imperative that the audiologist utilize the
ometry (Jerger et al., 1974), although other objective tests,
cross-check principle. The behavioral responses obtained
such as tone-burst–stimulated auditory brainstem response
via behavioral observation audiometry (BOA) or visual
(ABR), are considered to be better procedures for estimat-
reinforcement audiometry (VRA) are considered to be accu-
ing hearing thresholds. Acoustic reflexes can be used to help
rate reflections of a child’s true thresholds when these tests
identify the presence of hearing loss in young children as well
are conducted carefully (Madell and Flexer, 2008). How-
as in adults with language and/or cognitive issues that may
ever, because children often do not respond as consistently
reduce the validity and reliability of behavioral measures
or as quickly as adults, it is possible that a child’s behav-
(Hall, 2010). Acoustic reflexes can also be used to determine
ioral responses may still be elevated compared to actual
site of lesion within the auditory pathway, specifically in dif-
thresholds. As a result, the audiologist may judge the child’s
ferentiating between cochlear and retrocochlear pathologies.
responses as unreliable (Baldwin et al., 2010). Regardless of
the judged reliability of such measures, audiologists should
Cross-check for Puretone Average use objective tests such as OAEs and tympanometry as
A puretone average (PTA) is usually calculated as the average cross-checks for pediatric behavioral responses (Baldwin
of the air-conduction thresholds at 500, 1,000, and 2,000 Hz et al., 2010; Littman et al., 1998; Madell and Flexer, 2008).
for each ear. Normally, the PTA should agree with the speech In addition, OAEs and acoustic reflexes have been shown
recognition threshold (SRT), meaning that the PTA and SRT to be good cross-checks for ABR in young children (Berlin
should be within 10 dB of one another in the same ear. One et al., 2010; Stach et al., 1993). The Joint Committee on
instance in which the audiometric thresholds may cause the Infant Hearing Position Statement (JCIH; American Acad-
PTA to be greater than the SRT by 10 dB is when the audio- emy of Pediatrics, 2007) also recommends that electro-
gram configuration is sharply sloping or sharply rising. In physiological measures be employed as a cross-check for
such instances, it is preferable to use a two-frequency PTA behavioral response audiometry for children younger than
by averaging the two lowest (e.g., best) thresholds at 500, 6 months chronological age. The statement further stresses
1,000, and 2,000 Hz. The two-frequency PTA should then the importance of obtaining behavioral thresholds as soon
be in agreement with the SRT. Another instance in which as possible using the most age-appropriate method “to cross
the PTA and SRT may disagree is if a person is malingering check and augment physiologic findings” (American Acad-
or intentionally exaggerating a hearing loss. Outside of these emy of Pediatrics, 2007, p. 910).
special circumstances, we would expect SRTs and PTAs to
be highly correlated (except when language or foreign lan-
guage is a major factor). This allows us to use the SRT to
Electrophysiological Tests
validate the PTA (American Speech-Language-Hearing as Cross-checks
Association, 1988). Although beyond the scope of this chapter, it should be
noted that certain electrophysiological tests can be used to
Considerations for Assessing cross check behavioral measures, as well as to cross check
each other and to help confirm diagnoses of certain disor-
Speech Understanding ders (Bachmann and Hall, 1998; Berlin et al., 2010; Gravel,
One additional step that audiologists may take to address a 2002; Hall and Bondurant, 2009; Stapells, 2011). For exam-
patient’s complaint of not being able to understand speech ple, Berlin et al. (2010) discussed the use of cross-checking
 CHAPTER 8 • Diagnostic Audiology 121

TA B L E 8 . 2

Summary of Cross-checks Used in Diagnostic Audiology

Test Test Cross-check
Air conduction Bone conduction Rule out conductive component (air–bone gap)
Puretone audiometry Tympanometry Helps to verify/rule out middle-ear pathology (air–bone gaps); rule
out nonorganic hearing loss
Puretone audiometry Otoacoustic emis- Helps to verify/rule out middle-ear pathology; helps to confirm
sions outer hair cell function; rule out nonorganic hearing loss
Puretone audiometry Acoustic reflexes Helps to determine site of lesion (e.g., differentiate cochlear from
retrocochlear hearing loss); helps to determine degree of hear-
ing loss and rule out nonorganic hearing loss
Puretone average Speech recognition Verify performance on both measures (SRT should correlate with
threshold PTA)
Speech in quiet (WRS) Speech-in-noise Compare speech perception in quiet (normal audiologic testing) to
tests noise (more realistic test that addresses many patient complaints
of not understanding in noise)
BOA Electrophysiological Better estimate/confirmation of true thresholds
measures
VRA Electrophysiological Better estimate of true thresholds (if VRA responses unreliable)
measures
Note: Tests do not need to be administered in this order.

test results to diagnose ANSD: “. . . the presence of a [cochlear recommend that everyone have tympanometry for the
microphonic] or reversing waves at the beginning of the reasons previously discussed). We think that a nonorganic
trace does NOT make a diagnosis of ANSD . . . without the component is more likely to be present during subjective
cross-check of middle-ear muscle reflexes (MEMR), OAEs, testing and may not be discovered until cross-checked with
and an ABR latency-intensity function” (p. 32). For further objective tests. For these reasons, we recommend objective
information about these tests, the reader is referred to the testing first. A suggested testing order is shown in Figure 8.1.
chapters that discuss electrophysiological tests in the text. Nevertheless, it should be noted that some audiologists
Table 8.2 summarizes many of the cross-check tests that are advocate giving puretone and speech tests first when the
used in audiology. patient may be more alert and can actively respond to these
tests and then relax during the objective tests.
Order of Tests Administered
Beyond the Test Battery
Although we acknowledge that there is considerable vari-
ability in test protocols across clinics, we recommend that Although the use of a test battery is important, it is also
testing begin with the objective tests unless contraindi- vital for the audiologist to remember the case history and
cated. At least two major advantages can be found for test- the patient complaints. In some ways, one may consider this
ing objective measures first. The first advantage is that the patient-reported information to be a cross-check of the test
audiologist will have a good idea of the type and degree of battery itself. The case studies presented below demonstrate
hearing loss before beginning the subjective tests. The sec- examples of diagnostic audiology in action.
ond advantage is the reduced risk for misdiagnosis of dis-
orders such as ANSD and vestibular schwannoma, as well
as failure to detect a patient who is malingering. One caveat
CASE STUDIES
needs to be discussed with conducting objective tests first. The importance of objective testing in conjunction with
With rising costs of health care we need to be cautious that subjective tests can be seen through the use of case exam-
we are doing tests that are necessary. It is possible that, if ples. The following cases are a range of examples that high-
the patient’s hearing is completely normal (with no listen- light the use and benefit of incorporating the cross-check
ing complaints) or the patient is profoundly deaf (with principle into your clinical practice. The cases will be pre-
previous documentation to support the initial diagnosis), sented with a brief history and test results. A discussion
tests such as OAEs and acoustic reflexes will be unlikely to of potential difficulties and challenges in interpreting the
add further information about the patient (but we would audiologic data is incorporated within each case. Although
122 SECTION I • Basic Tests and Procedures

case, we encourage you to make a decision tree based on the

Tympanometry
order of tests presented in the case and then think about
how you might decide to change the test order. It is impor-
Objective Acoustic reflexes tant to review your clinical decision making periodically to
tests ensure that your practice is evidence based.

Otoacoustic emissions
Case 1
CASE HISTORY
Air conduction
Mr. Ang Kim, age 36, is being seen today after he failed the
hearing screening at his company’s health fair. His medical
SRT history is generally unremarkable, though he reports that
he is just getting over a sinus infection and recently under-
Subjective went surgery for a slipped disc in his back. You have back-
tests Bone conduction to-back patients today and because there is nothing remark-
able in his history you decide to do a quick audiogram and
send him on his way. Results from otoscopy, puretone, and
WRS speech audiometry are shown in Table 8.3 and Figure 8.2.
With subjective information alone this audiogram
could indicate many things. For example, you may inac-
Speech in noise
curately diagnose Mr. Kim with a collapsed ear canal, an
FIGURE 8.1 A suggested comprehensive diagnostic
impacted cerumen plug, or a perforated tympanic mem-
test battery. brane without additional tests to cross check your findings.

there is an ideal order to the test sequence as noted above, TA B LE 8 . 3

the order of tests often varies because of tester preference,
patient complaint, age of patient, and so on. Therefore, the Otoscopy Findings for Case 1
test order in the cases below varies to better represent what Right Ear Left Ear
may be actually done in a clinical setting or what a student
Stenotic ear canal, could Stenotic ear canal, could
clinician may see from his/her supervisors. For the sake of
not visualize tympanic not visualize tympanic
simplicity, all of the cases represent testing with standard
membrane membrane
audiologic procedures on adults. As you read through each

Transducer Supra-aural headphones

0 Reliability Good

Results:

Right Left
20 PTA 28 dB HL 5 dB HL

SRT 25 dB HL 5 dB HL

WRS 92% @ 65 dB 100% @ 45 dB

40
dB hearing level

HL HL

60

80

100

120 FIGURE 8.2 Puretone and

250 500 1,000 2,000 4,000 8,000
speech audiometry results
Frequency (Hz) for case 1.
 CHAPTER 8 • Diagnostic Audiology 123

TA B L E 8 . 4 TA B LE 8 . 6

Tympanometry Results (226-Hz Probe Transient-evoked Otoacoustic Emission

Tone) for Case 1 Results for Case 1
Right Left 1,000 1,400 2,000 2,800 4,000
Ear Hz Hz Hz Hz Hz
ECV 0.8 mL 0.7 mL
Compliance NP 0.6 mL Right Absent Absent Absent Absent Absent
Middle-ear pressure NP −50 daPa Left 10.5 dB 10.3 dB 11.4 dB 14.9 dB 13.9 dB

TABLE 8.5

Acoustic Reflexes Results (in dB HL) for Case 1

Ipsilateral Contralateral
Stimulus Ear 500 Hz 1,000 Hz 2,000 Hz 500 Hz 1,000 Hz 2,000 Hz
Right NR NR NR 110 dB 110 dB 105 dB
Left 85 dB 80 dB 85 dB 110 dB 105 dB 105 dB
NR, no response to maximum presentation (110 dB HL).

Despite your busy schedule, you decide you need more and puretone and speech audiometry results are shown in
information to make an accurate diagnosis, so you perform Figure 8.3.
objective testing to cross check your subjective results. The If you decide not to proceed with further tests to cross
results from immittance testing and OAE testing are shown check your results, you might diagnose this patient with
in Tables 8.4–8.6. normal hearing in the right ear and a mild conductive hear-
With this information, you have several different tests ing loss in the left ear. You might then refer Mrs. Jones to
to confirm your finding of a conductive hearing loss. The an Ear Nose and Throat physician who would order more
Type B tympanogram in the right ear reveals normal ear tests.
canal volume but no mobility. The normal ear canal volume Instead, you decide to proceed and include additional
suggests that the TM is not perforated and there is no ceru- tests in your battery that would provide a cross-check. We
men plug. The pattern of the ARTs is consistent with a right will review those results next (see Tables 8.8–8.10).
conductive pathology. TEOAEs in the right ear are absent These results suggest that Mrs. Jones has normal hear-
which is expected with a conductive pathology. ing that contradicts your puretone findings. Normal results
The combination of the subjective and objective test on tympanometry, ARTs, and TEOAEs are not consistent
results correctly leads you to suspect otitis media with effu- with a mild conductive hearing loss. With this information
sion and would require a referral for Mr. Kim to a physician. you review the patient’s case history and puretone findings
In this case, you are able to make an appropriate referral again and realize that the apparent conductive hearing loss
based on the information you obtained from a test battery in the right ear is likely the result of a collapsing ear canal. It
incorporating both objective and subjective measures. is not uncommon for the pressure of the supra-aural head-
phones to cause the canal to collapse, particularly in older
patients for whom the cartilage supporting the ear canal is
Case 2 soft. To confirm this finding you decide to retest Mrs. Jones
with insert earphones. When you repeat your audiogram
CASE HISTORY using the insert earphones, you measure Mrs. Jones’s right-
Mrs. Edith Jones, age 77, is being seen today for a hearing ear air-conduction thresholds at 5 or 10 dB HL for all fre-
test. She does not perceive a listening difficulty but her hus- quencies tested. You are able to report to Mrs. Jones that her
band was recently fit with hearing aids and insisted she have hearing appears to be normal!
her hearing checked too. Her medical history is remarkable Both cases 1 and 2 highlight the importance of using
for high blood pressure and type 2 diabetes which are both objective test results in conjunction with subjective test
controlled by medication. results to avoid misdiagnosis. Both audiograms revealed
You conduct a basic audiometric evaluation on similar test results but very different actual diagnoses, which
Mrs. Jones. Results for otoscopy are displayed in Table 8.7 were only confirmed with the use of objective testing.
124 SECTION I • Basic Tests and Procedures

Transducer Supra-aural headphones

0 Reliability Good

Results:

Right Left
20 PTA 28 dB HL 5 dB HL

SRT 30 dB HL 10 dB HL

WRS 96% @ 70 dB 100% @ 50 dB

40
dB hearing level

HL HL

60

80

100

120 FIGURE 8.3 Puretone and

250 500 1,000 2,000 4,000 8,000 speech audiometry results
Frequency (Hz) for case 2.

TA B L E 8 .7 TA B LE 8 . 8

Otoscopy Findings for Case 2 Tympanometry Results (226-Hz Probe

Right Ear Left Ear
Tone) for Case 2
Stenotic ear canal, could Stenotic ear canal, could Right Left
not tympanic mem- not tympanic mem- ECV 1.3 mL 1.4 mL
brane brane Compliance 0.7 mL 0.8 mL
Middle-ear pressure 0 daPa −5 daPa

TA B L E 8 . 9

Acoustic Reflexes Results (in dB HL) for Case 2

Ipsilateral Contralateral
Stimulus Ear 500 Hz 1,000 Hz 2,000 Hz 500 Hz 1,000 Hz 2,000 Hz
Right 85 dB 85 dB 80 dB 90 dB 95 dB 95 dB
Left 85 dB 80 dB 85 dB 95 dB 90 dB 90 dB

TABL E 8.10

Transient-evoked Otoacoustic Emission Results for Case 2

Ear 1,000 Hz 1,400 Hz 2,000 Hz 2,800 Hz 4,000 Hz
Right 8.9 dB 9.1 dB 12.3 dB 10.4 dB 7.3 dB
Left 9.9 dB 10.4 dB 10.5 dB 9.7 dB 6.1 dB
 CHAPTER 8 • Diagnostic Audiology 125

TA B L E 8 .1 1 TA B LE 8 .1 2

Otoscopy Findings for Case 3 Tympanometry Results (226-Hz Probe

Right Ear Left Ear
Tone) for Case 3
Clear ear canal; intact Clear ear canal; intact Right Left
tympanic membrane tympanic membrane ECV 1.0 mL 1.1 mL
Compliance 0.5 mL 0.6 mL
Middle-ear pressure 5 daPa −20 daPa
Case 3
You receive the following case, accompanied by a
patient-signed medical information release, via fax. A the WRS was only obtained at 20 dB SL in the left ear, yet
recently graduated audiologist at a practice across town Mr. Smith’s WRS is 76%, which is better than expected.
just finished testing Mr. Smith and would like a second According to Dubno et al. (1995), a patient with a PTA
opinion. of 90 dB HL would have an expected WRS of less than
24%. You suggest to the other audiologist that obtaining
TEOAEs would further assist in this diagnosis. The audi-
CASE HISTORY ologist performs TEOAEs (see Table 8.14) to confirm the
Mr. Aaron Smith, 49, reports that he can no longer hear out suspected diagnosis and faxes the results to you.
of his left ear. He works in construction and reports that Based on the pattern of test results, your suspected
a transformer overloaded at his work site yesterday, result- diagnosis is nonorganic hearing loss. Let us review the
ing in a loud explosion which he believes caused his hear- facts. First, the patient’s left ear thresholds are elevated
ing loss. Mr. Smith reported that his hearing was normal above where cross-hearing should have occurred. Sec-
prior to the explosion. He denies any aural fullness, tinnitus, ond, the objective test results (tympanometry, ARTs, and
or dizziness. His medical history is unremarkable and he OAEs) reveal no conductive component and suggest that
denies any other injuries as a result of the explosion. Results outer hair cells are functioning normally. However, the
of the audiologic testing are shown in Tables 8.11–8.13 and puretone and speech audiometry results suggest a severe-
Figure 8.4. to-profound unilateral hearing loss in the left ear, which
You call the audiologist right away and review your is inconsistent with the objective results. Several cross-
concerns with her. Both the air-conduction and bone- checks identified inconsistencies (e.g., ARTs and puretones;
conduction thresholds for the left ear need to be masked. PTA–SRT agreement; puretone thresholds and OAEs). At
Cross-hearing should have occurred before those thresh- this point, you could suggest that the audiologist reinstruct
olds were obtained. Furthermore, you would not expect the patient and then retest the left ear, masking appropri-
to obtain no response for bone-conduction testing with ately. If the thresholds for the left ear are still elevated, a
the bone oscillator on the left side when the hearing in the Stenger test could be performed to confirm the accuracy
right ear is evidently normal. You also note that the PTA of the left puretone thresholds. If the Stenger test result is
and the SRT are not in agreement for the left ear. ARTs are positive (i.e., the patient does not respond to the stimulus),
better than you would expect them to be (given the pur- this would be additional evidence that the apparent hear-
etone thresholds for the left ear). A patient with hearing ing loss is nonorganic. This case highlights the importance
thresholds at 90 dB HL would be expected to have ARTs of a high-quality diagnostic battery (including masking
in the range of 95 to 125 dB HL at 500 Hz and 100 to 125 where appropriate) and use of cross-checks to confirm our
dB HL at 1,000 and 2,000 Hz (Gelfand et al., 1990). Lastly, test results.

TA B L E 8 .1 3

Acoustic Reflexes Results (in dB HL) for Case 3

Ipsilateral Contralateral
Stimulus Ear 500 Hz 1,000 Hz 2,000 Hz 500 Hz 1,000 Hz 2,000 Hz
Right 80 dB 85 dB 85 dB 85 dB 90 dB 90 dB
Left 85 dB 80 dB 85 dB 90 dB 85 dB 90 dB
126 SECTION I • Basic Tests and Procedures

Transducer Insert earphones

0 Reliability Good

Results:

Right Left
20 PTA 10 dB HL 90 dB HL

SRT 10 dB HL 75 dB HL

WRS 100% @ 50 dB 76% @ 95 dB

40
dB hearing level

HL HL

60

80

100

120 FIGURE 8.4 Puretone and

250 500 1,000 2,000 4,000 8,000 speech audiometry results
Frequency (Hz) for case 3.

TABL E 8.14

Transient-evoked Otoacoustic Emission Results for Case 3

Ear 1,000 Hz 1,400 Hz 2,000 Hz 2,800 Hz 4,000 Hz
Right 9.1 dB 12.7 dB 9.2 dB 10.1 dB 12.4 dB
Left 10.5 dB 10.3 dB 11.4 dB 14.9 dB 13.9 dB

Case 4 her puretone thresholds. You would expect a patient with

PTAs in this range to have WRS of 68% or better (Dubno
CASE HISTORY et al., 1995). The next tests that should be performed are
Ms. Ashley Jackson, age 27, has had hearing problems all of ARTs and OAEs. Tables 8.17 and 8.18 show the results of
her life. She has been told by her audiologist that she has those tests.
only a mild sensory/neural hearing loss. Her doctor always Now that you have completed your testing, you cross
tells her that her hearing is really very good. She tried hear- check your test results. First, the patient’s ARTs are not con-
ing aids a few years ago but she says that they did not help sistent with her puretone thresholds. With a mild sensory/
at all. Unfortunately, Ms. Jackson cannot hold a job because neural hearing loss you would expect acoustic reflexes to
of her hearing difficulties. Her bosses always cite miscom- be present (Gelfand et al., 1990). The patient’s TEOAEs are
munication problems as the reason for her dismissal. present and robust which would not be expected based on
Ms. Jackson is here today to see if her hearing has changed. Ms. Jackson’s puretone thresholds. These findings in con-
Tables 8.15 and 8.16 show otoscopy and tympanometry junction with the poor WRS indicate a need for additional
results. Figure 8.5 shows puretone and speech audiometry testing such as an ABR. You suspect that the patient has
results. ANSD. Results of the ABR and a medical evaluation may
Ms. Jackson’s puretone results appear to be consistent help to confirm or rule out your suspected diagnosis. With-
with the previous hearing tests in her medical record. There out the addition of both ARTs and OAEs to the test battery,
are some red flags that warrant additional testing, though. Ms. Jackson’s disorder might have been missed again. The
First, her reports of listening difficulties and communica- recommendations for patients with ANSD or other retroco-
tion problems in her case history suggest that she may have chlear pathologies are often very different from the recom-
more than a mild sensory/neural hearing loss. Additionally, mendations for those who have a peripheral hearing loss.
her word recognition scores are poorer than expected given Misidentification of the site of lesion for a hearing loss like
 CHAPTER 8 • Diagnostic Audiology 127

TA B L E 8 .1 5 TA B LE 8 .1 6

Otoscopy Findings for Case 4 Tympanometry Results (226-Hz Probe

Right Ear Left Ear
Tone) for Case 4
Clear ear canal; intact Clear ear canal; intact Right Left
tympanic membrane tympanic membrane ECV 1.3 mL 1.4 mL
Compliance 0.7 mL 0.5 mL
Middle-ear pressure 20 daPa −10 daPa

Transducer Insert earphones

0 Reliability Good

Results:

Right Left
20 PTA 33 dB HL 35 dB HL

SRT 30 dB HL 35 dB HL

WRS 48% @ dB HL 52% @ 75 dB HL

40
dB hearing level

60

80

100

120 FIGURE 8.5 Puretone and

250 500 1,000 2,000 4,000 8,000 speech audiometry results for
Frequency (Hz) case 4.

TA B L E 8 .1 7

Acoustic Reflexes Results (in dB HL) for Case 4

Ipsilateral Contralateral
Stimulus Ear 500 Hz 1,000 Hz 2,000 Hz 500 Hz 1,000 Hz 2,000 Hz
Right NR NR NR NR NR NR
Left NR NR NR NR NR NR
NR, no response to maximum presentation (110 dB HL).

TABL E 8.18

Transient-evoked Otoacoustic Emission Results for Case 4

Ear 1,000 Hz 1,400 Hz 2,000 Hz 2,800 Hz 4,000 Hz
Right 20.1 dB 22.9 dB 19.5 dB 18.4 dB 19.3 dB
Left 22.5 dB 20.6 dB 20.1 dB 22.9 dB 20.3 dB
128 SECTION I • Basic Tests and Procedures

Transducer Insert earphones

0 Reliability Good

Results:

Right Left
20 PTA 28 dB HL 18 dB HL

SRT 30 dB HL 20 dB HL

WRS 40% @ 70 dB 85% @ 60 dB

40
dB hearing level

HL HL

10% @ 85 dB

60 HL

80

100

120 FIGURE 8.6 Puretone and

250 500 1,000 2,000 4,000 8,000 speech audiometry results
Frequency (Hz) for case 5.

Ms. Jackson’s might result in delayed or inappropriate reha- ing. The patient’s asymmetry in high-frequency audiometric
bilitation recommendations. Unfortunately, ANSD can be thresholds and poor WRS in the right ear combined with
missed easily in clinical practice if the audiologist does not the objective results suggest a retrocochlear pathology and
perform a thorough test battery (Smart and Kelly, 2008). warrant an ENT referral for additional testing. The patient’s
report of unilateral tinnitus, the abnormal ART pattern, the
asymmetry in puretone thresholds, and the apparent rollover
Case 5 in the patient’s right-ear word recognition are all suggestive
CASE HISTORY of retrocochlear dysfunction. Taken in isolation, each might
be sufficient for you to make a referral for a medical evalu-
Mr. Don Warner, age 58, is being seen today with his pri- ation. However, having consistent results from several tests
mary complaint being a constant ringing in his right ear. He allows you to be more confident in your recommendation
notes that the ringing has been present off and on for over and provide the ENT with as much information as possible.
3 years but it has become more bothersome recently. When
asked about his hearing he admits that he has difficulty
understanding what others are saying in noisy places. He Case 6
denies aural fullness and dizziness. He plays tennis regularly
CASE HISTORY
and generally is in great health. Results from your testing are
shown in Tables 8.19–8.22 and Figure 8.6. Mr. José Gonzalez, age 24, was seen today for an audiologic
The order of test administration is important. Because evaluation. He was just hired as a clerk for a federal judge and
you performed immittance testing and TEOAEs first, you therefore has to undergo a rigorous physical examination,
knew that the patient’s tympanic membrane is mobile, that including a hearing test. Mr. Gonzalez denies any hearing dif-
the ART pattern is abnormal in the right ear, and that the ficulties, tinnitus, dizziness, or aural fullness. He reports that
outer hair cells of the right ear do not appear to be function- he is in great health and is currently training for a marathon.
ing normally. You were able to obtain this information before
the patient provided any information through subjective test-
TA B LE 8 . 20

TA B L E 8 .1 9
Tympanometry Results (226-Hz Probe
Tone) for Case 5
Otoscopy Findings for Case 5 Right Left
Right Ear Left Ear ECV 1.6 mL 1.8 mL
Clear ear canal; intact Clear ear canal; intact Compliance 0.7 mL 0.9 mL
tympanic membrane tympanic membrane Middle-ear pressure 0 daPa −10 daPa
 CHAPTER 8 • Diagnostic Audiology 129

TA B L E 8 . 2 1

Acoustic Reflexes Results (in dB HL) for Case 5

Ipsilateral Contralateral
Stimulus Ear 500 Hz 1,000 Hz 2,000 Hz 500 Hz 1,000 Hz 2,000 Hz
Right 105 dB 110 dB 110 dB 110 dB NR NR
Left 85 dB 90 dB 85 dB 95 dB 95 dB 95 dB
NR, no response to maximum presentation (110 dB HL).

TABL E 8.2 2

Transient-evoked Otoacoustic Emission results for Case 5

Ear 1,000 Hz 1,400 Hz 2,000 Hz 2,800 Hz 4,000 Hz
Right Absent Absent Absent Absent Absent
Left 8.0 dB 9.3 dB 9.1 dB 6.2 dB 6.1 dB

Transducer Supra-aural headphones

0 Reliability ?

Results:

Right Left
20 PTA 10 dB HL NR

SRT 5 dB HL NR

WRS 100% @ 45 dB CNT

40
dB hearing level

HL

60

80

100

120 FIGURE 8.7 Puretone and

250 500 1,000 2,000 4,000 8,000 speech audiometry results
Frequency (Hz) for case 6.

TABLE 8.24

Tympanometry Results (226-Hz Probe

TA B L E 8 . 2 3
Tone) for Case 6
Otoscopy Findings for Case 6 Right Left
Right Ear Left Ear ECV 1.5 mL 1.7 mL
Clear ear canal; intact Clear ear canal; intact Compliance 0.9 mL 0.7 mL
tympanic membrane tympanic membrane Middle-ear pressure −10 daPa −15 daPa
130 SECTION I • Basic Tests and Procedures

TA B L E 8 . 2 5

Acoustic Reflexes Results (in dB HL) for Case 6

Ipsilateral Contralateral
Stimulus Ear 500 Hz 1,000 Hz 2,000 Hz 500 Hz 1,000 Hz 2,000 Hz
Right 85 dB 85 dB 90 dB 95 dB 90 dB 95 dB
Left 85 dB 90 dB 85 dB 90 dB 95 dB 95 dB

TABL E 8.2 6

Transient-evoked Otoacoustic Emission Results for Case 6

Ear 1,000 Hz 1,400 Hz 2,000 Hz 2,800 Hz 4,000 Hz
Right 12.3 dB 14.6 dB 10.2 dB 11.1 dB 8.9 dB
Left 13.5 dB 12.8 dB 12.4 dB 10.1 dB 9.9 dB

Based on the testing you have completed thus far you

would expect this patient has normal hearing. His case his-
LIMITATIONS OF THE
tory and all objective tests suggest hearing within normal AUDIOLOGIC TEST BATTERY
limits. You proceed with your puretone and speech testing. The combination of well-validated test measures, precise
Results from your testing are shown in Tables 8.23–8.26 patient instruction, careful scoring, and application of the
and Figure 8.7. cross-check principle should result in accurate diagnostic
The puretone and speech audiometry results are sur- and rehabilitative decisions for most patients. It is impor-
prising because they conflict with the results from the objec- tant to remember, however, that real-world patients usually
tive tests. Specifically, ARTs and TEOAEs within normal lim- do not present as textbook cases. The case studies contained
its are not consistent with a measured profound hearing loss in this chapter and the diagnostic criteria published in the
in the left ear. Your first thought is nonorganic hearing loss. audiologic test literature should be treated as guidelines
You decide to reinstruct Mr. Gonzalez and retest the left ear rather than absolute rules. High-quality diagnosis depends
air-conduction thresholds. This time you tell Mr. Gonzalez on both the construction of a high-quality test battery and
that you are going to play some more beeps in his left ear and skill in interpreting ambiguous or seemingly contradictory
all he has to do is press the button when he hears the tone. He test results. A good rule for daily practice is this: When test
nods and appears to understand the instructions. You begin results seem in disagreement, first check the tester (rule out
retesting at 1,000 Hz and Mr. Gonzalez does not respond the clinician’s own mistakes); then, check the equipment
at the maximum limits of the audiometer. As you enter the (rule out malfunction or equipment performing out of
booth to reinstruct again, Mr. Gonzalez informs you that he calibration); and finally, check the patient (rule out patient
never heard a beep and has been waiting for you to present error or pseudohypacusis).
the tone. In fact, he notes that he has not heard anything
from the left earphone. You check the headphone jack con-
nections and find that the left headphone jack is unplugged.
MAKING REFERRALS
After you plug in the jack and retest Mr. Gonzalez’s left ear, No audiologist is an island. A team approach to the treat-
you obtain thresholds within normal limits. It is important ment of hearing and balance disorders, particularly in pedi-
to note that the patient history and the objective test results atric patients, is often indicated. Appropriate treatment
were not consistent with the subjective test results. Although of a patient seen for audiologic evaluation may require
having a well-constructed test battery is important, you also consultation with specialists including (but not limited
want to be flexible with your test order and be vigilant to to) allergists, endocrinologists, neurologists, occupational
notice inconsistencies between test results as you go. This therapists, ophthalmologists, psychiatrists, rheumatologists,
flexibility would allow you to notice the unplugged head- and speech-language pathologists. Referral of pediatric
phone jack sooner and save time and frustration for you and patients with hearing loss to an ophthalmologist is particu-
Mr. Gonzalez. larly important; approximately 50% of children born with
 CHAPTER 8 • Diagnostic Audiology 131

severe-to-profound hearing loss also have abnormalities of With respect to the audiologic test battery, disagree-
vision (American Academy of Pediatrics, 2007). ment among objective and subjective test measures which
cannot be resolved as tester, equipment, or patient error is
Referral for Medical Otolaryngologic indicative of need for medical referral. Abnormally poor
Evaluation speech scores relative to the audiogram, asymmetric hearing
loss, and reports of aural fullness and/or tinnitus are other
The most common referral made by audiologists is to a signs of possible serious ear disease which should be evalu-
medical doctor. Sending a patient to an otolaryngologist, ated by a physician.
primary care physician, or pediatrician is indicated if the
audiologic evaluation reveals evidence of an underlying Referral for Auditory Processing
medical condition. Symptoms may include ear pain, bleed-
ing or drainage from the ear (otorrhea), tympanometric
Evaluation
abnormality without known etiology, or physical abnormal- Disagreement between objective and subjective hearing tests
ity observed during otoscopy. Patients who report frequent may be reason to refer a patient for an evaluation of audi-
ear infections, fluctuating or sudden hearing loss, or bal- tory processing. Patients with apparently normal peripheral
ance disturbance should also be seen by a medical profes- auditory function may still have difficulty processing com-
sional (see Table 8.27). Newly identified hearing loss is also plex signals such as speech. These individuals often report
reason for referral. Although some audiologists undertake that they can hear well, but have difficulty understanding
cerumen management in their own practice, many others what others are saying, particularly in the presence of noise.
prefer to refer to an otolaryngologist or the patient’s pri- Tests of speech perception in noise such as the Bamford–
mary care physician for removal of impacted wax. Children Kowal–Bench Speech-in-Noise Test (BKB-SIN; Etymo-tic
who exhibit a previously undiagnosed hearing loss or who Research, 2005), Quick Speech-in-Noise Test (QuickSIN;
exhibit delays in speech or language development should be Etymo-tic Research, 2001), and Hearing in Noise Test (HINT;
seen by a pediatric otolaryngologist or developmental pedi- Nilsson et al., 1994) may help to confirm this difficulty. If
atrician prior to any audiologic m