532135

research-article2014
PENXXX10.1177/0148607114532135Journal of Parenteral and Enteral NutritionColtman et al

2014 Premier Research Paper

Journal of Parenteral and Enteral
Nutrition
Use of 3 Tools to Assess Nutrition Risk in the Volume 39 Number 1
January 2015 28­–33
Intensive Care Unit © 2014 American Society
for Parenteral and Enteral Nutrition
DOI: 10.1177/0148607114532135
jpen.sagepub.com
hosted at
online.sagepub.com
Anne Coltman, MS, RD, CNSC1; Sarah Peterson, MS, RD, CNSC1;
Kelly Roehl, MS, RD, CNSC1; Hannah Roosevelt, MS, RD, CNSC1;
and Diane Sowa, MBA, RD1

Abstract
Background: Identifying patients at nutrition risk proves difficult in the intensive care unit (ICU) due to the nature of critical illness. No
consensus exists on the most appropriate method to identify these patients. Traditional screens and assessments are often limited due to
their subjective nature. The purpose of the quality improvement project was to compare proportions of ICU patients deemed at nutrition
risk using 3 different tools. Material and Methods: A convenience sample of 294 patients admitted to the ICU was used. Patients were
assessed using the institution’s routine nutrition screening method, the Subjective Global Assessment (SGA), and the NUTrition Risk in
Critically ill (NUTRIC) score. Information was collected on demographics, severity of illness, hospital and ICU length of stay (LOS),
and disposition. Descriptive statistics were used to examine counts/proportions of risk categories; means ± SD were used to summarize
demographic and clinical variables. Results: A total of 139 patients (47%) were deemed at nutrition risk or malnourished by at least 1 tool.
Patients identified were older and had a lower body mass index, more weight loss, more fat and muscle wasting, more fluid accumulation,
and lower average handgrips than those not at nutrition risk; they also had longer hospital and ICU LOS, higher rates of requiring further
rehabilitation upon discharge, and higher mortality during hospitalization. Conclusion: Traditional screening and assessment tools did not
uniformly identify patients as malnourished or at nutrition risk in the ICU and therefore may be inappropriate for use in this population.
Inclusion of physical assessment, functional status, and severity of illness may be useful in predicting nutrition risk in the ICU. (JPEN J
Parenter Enteral Nutr. 2015;39:28-33)

Keywords
adult; life cycle; outcomes research/quality; nutrition support practice; nutrition assessment; nutrition

Malnourished patients have worse clinical outcomes than their Many of these criteria may be difficult to obtain in critically ill
well-nourished counterparts; unfortunately, this relationship is patients. For example, because ICU patients may require
often exacerbated in the intensive care unit (ICU) due to the mechanical ventilation or present with altered mental status,
hypermetabolic nature of critical illness.1 Accurately identify- food intake histories are often challenging to obtain. Similarly,
ing patients at risk for malnutrition is essential to decrease obtaining information on functional status and gastrointestinal
negative outcomes during hospitalization.2 Recently, a consen- (GI) symptoms prior to admission may also be difficult. If
sus statement by Academy of Nutrition and Dietetics (AND) weight histories are available, changes in weight may actually
and American Society for Parenteral and Enteral Nutrition be more reflective of fluid status than actual changes in weight,
(A.S.P.E.N.) recognized the importance of inflammation in the since many ICU patients are given large volumes of fluid to
characterization of malnutrition and recommended an etiol- maintain hemodynamic stability.2 Physical assessment and
ogy-based definition categorizing patients with the presence of assessment of muscle tone can be used as a more objective
2 or more of the following characteristics: insufficient energy tool, since it does not require patient interview or previous
intake, weight loss, loss of muscle mass, loss of subcutaneous
fat, localized or generalized fluid accumulation, or decreased
functional status in the context of acute illness or injury, From 1Rush University Medical Center, Chicago, Illinois.
chronic diseases or conditions, and starvation-related malnutri-
Financial disclosure: None declared.
tion.3 However, there is no consensus on the best tool to iden-
tify these patients, particularly in the ICU. Received for publication December 17, 2013; accepted for publication
A number of tools employing a variety of criteria are used to March 26, 2014.
identify nutrition risk, including clinical diagnosis, laboratory This article originally appeared online on April 18, 2014.
data, physical examination, anthropometric data, food/nutrient
Corresponding Author:
intake, and functional assessment.4-8 These indicators were Anne Coltman, MS, RD, CNSC, Rush University Medical Center,
primarily validated in outpatients or a general hospitalized pop- Chicago, 1700 W. Van Buren St, Suite 425, Chicago, IL 60612, USA.
ulation; they were not specifically designed for use in the ICU.9 Email: [email protected]

Downloaded from pen.sagepub.com at WASHBURN UNIV on January 13, 2015

Coltman et al 29

knowledge of body habitus. Fluid status can also be assessed severe), and functional assessment (defined as no change,
through examination of areas where edema or ascites often decreased ability to complete activities of daily living, or bed-
occur.7 In addition, muscle and fat wasting may be masked by ridden) and low handgrip strength. Anthropometric data,
severe edema or a patient’s overall body habitus, as seen in including, height, weight, and BMI, were collected from the
cases of overweight and obesity. Many traditional tools do not electronic medical record. Three consecutive hand dynamom-
provide information regarding inflammatory status. The eter measurements were obtained from patients able and will-
NUTrition Risk in Critically ill (NUTRIC) score, a tool intro- ing to participate using the patient’s self-identified dominant
duced by Heyland et al,10 uses a unique scoring method for hand while the patient was sitting up in bed or in a chair with
identifying patients who would most benefit from aggressive the elbow bent at a 90° angle. Patients squeezed with maxi-
nutrition support in the ICU by linking starvation, inflamma- mum strength for 5 seconds; the highest force exerted was
tion, and outcomes. However, this tool includes no traditional recorded for each of the 3 measurements and the average was
markers of nutrition risk, such as body mass index (BMI), calculated. Patients were classified as follows: normal (A),
weight status, oral intake, or physical assessment, and may have mild to moderate malnutrition (B), or severe malnutrition (C)7;
limited clinical application due to its exclusion of nutrition his- for the purpose of analysis, patients with scores of B or C were
tory variables. classified as malnourished.
The use of traditional screening and assessment tools in Patients were also deemed at nutrition risk via a nutrition
the ICU may not provide accurate results in determining screening tool used on non-ICU floors at the current institu-
patients at risk for malnutrition. Clinical observations of the tion. The institution’s routine nutrition screen included the fol-
limitations of current tools in the ICU prompted a quality lowing criteria: recent unintentional weight loss (5% in
improvement project in an effort to identify the most appro- 1 month, 10% in 6 months), BMI <18.5 or >40, presence of
priate tool for identifying nutrition risk or malnutrition in the dysphagia/inadequate food intake prior to admittance, or use of
ICU. While different tools measure different constructs enteral nutrition (EN)/parenteral nutrition (PN). Patients meet-
(nutrition risk vs malnutrition), the purpose of the project was ing at least 1 criterion were deemed at risk for malnutrition.
exploratory in nature; therefore, tools identifying both nutri- A NUTRIC score was calculated for each patient using
tion risk and malnutrition were used. The purpose of this age, number of comorbidities, days from hospital to ICU
quality improvement project was to describe the proportion admission, and Acute Physiology and Chronic Health
of ICU patients deemed at increased nutrition risk or Evaluation II (APACHE II) and Sequential Organ Failure
malnourished via the institution’s routine screening method, Assessment (SOFA) scores from admission. NUTRIC scores
the NUTRIC score, and the Subjective Global Assessment were calculated without using interleukin (IL)–6 values; the
(SGA). creators of the tool allow for exclusion of this variable when
not clinically available. Therefore, patients were classified as
having a high score if the sum was 5 or greater; these patients
Methods were classified as having a higher risk of malnutrition. Higher
A convenience sample of 294 patients admitted to the medical, scores (≥5) have been associated with worse clinical out-
surgical, and neuroscience ICUs over a 3-month period in a comes; in addition, patients with a score ≥5 have been pro-
large, urban academic medical center was used as part of a posed to be most likely to benefit from aggressive nutrition
quality improvement project. Patients were not included if they therapy.
were younger than 18 years or unable to communicate in Additional information collected included: demographics
English. The hospital’s institutional review board approved the (age, sex, and race), hospital and ICU length of stay (LOS),
evaluation and use of outcomes from this quality improvement and disposition (defined as discharged to home, rehabilitation,
project; no informed consent was required due to the nature of or patient death).
the project.
At our institution, completing SGA on all ICU admittances
is standard practice to identify malnutrition. The SGA was
Statistical Analysis
completed within 24 hours of admission. The 4 ICU dietitians Descriptive statistics were used to examine counts/proportions
have been previously trained according to a protocol described of nutrition risk categories using the different screening and
by Sheean et al.11 No formal interrater reliability studies were assessment tools; means ± standard deviations were used to
performed. The SGA is composed of the following criteria: summarize demographic and clinical variables. Patients may
intake prior to admission (categorized as no change, decreased, have been identified at nutrition risk by the institution’s screen-
or unable to eat), presence of GI symptoms (including anorexia, ing tool, malnourished by the SGA, and/or more likely to ben-
nausea, vomiting, and diarrhea), weight loss, physical assess- efit from nutrition support by the NUTRIC score; because
ment (including change in fat wasting, muscle wasting, and patients may belong to more than 1 group, statistical compari-
fluid accumulation defined as normal, mild to moderate, and sons were not made between nutrition risk groups, since

Downloaded from pen.sagepub.com at WASHBURN UNIV on January 13, 2015

30 Journal of Parenteral and Enteral Nutrition 39(1)

membership groups would violate the independence rule. project, the standard of care was changed in the ICUs. All
Analysis was performed using SPSS version 18 (SPSS, Inc, an patients admitted to an ICU are assessed by a dietitian using
IBM Company, Chicago, IL). the SGA. While there are some limitations to the use of the
SGA in the ICU, it was deemed the most valuable tool to
assess nutrition status in this population.
Results Patients identified at nutrition risk using the NUTRIC score
A total of 139 patients (47%) were deemed at nutrition risk or had the longest hospital and ICU LOS. This is likely because
malnourished by at least 1 tool. Of these patients, a total of the NUTRIC score encompasses severity of illness, since these
63% were deemed at nutrition risk using the institution’s rou- patients also had higher APACHE II and SOFA scores than did
tine screening method (87/139), 80% were malnourished patients in other groups. Severity of illness has been shown to
according to the SGA (111/139), and 26% were deemed candi- be a major contributing factor to LOS.12 Accurate identifica-
dates for nutrition support with the NUTRIC score (36/139). tion of these patients may allow for more appropriate and
Many patients met criteria for more than 1 tool and therefore timely administration of nutrition support, thus decreasing
may have been at nutrition risk using the institution’s screening LOS. In addition, all components of this tool can be objectively
tool, malnourished using the SGA, and more likely to benefit obtained through chart review without the necessity for a
from nutrition support using the NUTRIC score. Nutrition risk patient or family interview. Inclusion of severity of illness may
rates and associated demographic information are described in be a necessary component in accurately identifying patients at
Table 1. Regardless of the tool used, patients who were at nutrition risk in the ICU.
nutrition risk or malnourished were older and had a lower More patients classified as malnourished using the SGA
BMI, more weight loss, more fat and muscle wasting, more alone or in combination with any other screening tool were
fluid accumulation, and lower average handgrip than those not discharged to a rehabilitation facility. This is likely due to the
at nutrition risk; of note, handgrips strength was collected from tool’s inclusion of functional status in assessing overall nutri-
only 83% of patients (242/294) due to limitations in mental tion status. Similarly, patients deemed malnourished using the
status. Patients at nutrition risk also had longer hospital and SGA had the lowest average handgrip strength among those
ICU LOS, higher rates of requiring further rehabilitation after investigated. Decreased handgrip values have been associated
discharge, and higher rates of mortality during hospitalization with decreased functional status.13 These patients also had the
than those not at risk. highest proportion of muscle and fat wasting of all the risk
Because many patients met criteria for more than 1 tool, groups. Loss of muscle and fat mass often occurs as the result
further investigation into risk classification was needed to of a prolonged disease course, contributing to decreased grip
accurately identify trends. Only 9 (6%) patients met criteria for strength and overall functional status. These patients may not
all 3 tools (Figure 1). Rates of death, rehabilitation, and hospi- have been deemed at nutrition risk using the NUTRIC score
tal and ICU LOS were evaluated among the further delineated alone, since they may have lacked signs of acute severity of
risk groups. illness. Nonetheless, this population may benefit from aggres-
Patients determined at nutrition risk using the NUTRIC sive nutrition intervention in an effort to attenuate the loss of
score alone or in combination with any other tool had the high- lean body mass during hospitalization and prevent further loss
est rates of death. A larger proportion of patients requiring of functional status. Although completing a patient’s interview
additional rehabilitation after discharge were seen with both or obtaining a handgrip measurement may be difficult in the
NUTRIC and SGA scores classifying patients at risk. Patients ICU, assessing functional status upon admission may better
identified as at nutrition risk or malnourished using both identify patients in need of aggressive nutrition support during
NUTRIC and SGA had the longest hospital LOS and ICU LOS hospitalization.
(Figure 2). Patients at nutrition risk using only the institution’s Conversely, patients identified at risk using the institution’s
screening tool and NUTRIC had the shortest ICU LOS. routine institution screening had the shortest hospital and ICU
LOS. This may have occurred because most of the information
required for this tool is subjectively obtained; therefore,
Discussion patients meeting the tool’s criteria may have been missed due
There was a great deal of variability among the institution’s to the clinician’s inability to obtain the information as a result
routine screening tool, the SGA, and the NUTRIC score in of a patient’s clinical status. Patients unable to complete nutri-
identifying different groups of patients at risk for malnutri- tion interviews are often the most critically ill; exclusion of
tion; however, there was a great deal of overlap between such patients may have potentially contributed to a higher pro-
groups. A much larger number were identified using the insti- portion of patients with a shorter LOS.
tution’s routine screening tool (63%) and SGA (80%), com- Limitations of the study must be addressed. Data were
pared with those identified using the NUTRIC score (26%). obtained as part of a quality improvement project intended to
Interestingly, only 9 patients were deemed at nutrition risk and investigate nutrition screening and assessment practices at 1
malnourished using all 3 screening tools. As a result of this institution. Therefore, results may not apply to all institutions,

Downloaded from pen.sagepub.com at WASHBURN UNIV on January 13, 2015

 Table 1.  Demographic Data Classified Using Routine Screening, the Subjective Global Assessment (SGA), and the NUTRIC Score.

Any Nutrition Risk (n = 139)

Risk With Routine Malnourished With Risk With NUTRIC

Total Population No Risk Screeninga SGAb Scorec
Characteristic (n = 294) (n = 155) (n = 87/139, 63%) (n = 111/139, 80%) (n = 36/139, 26%)
Age, mean ± SD, y 59.0 ± 16.4 56.5 ± 16.4 61.0 ± 15.6 61.8 ± 15.0 69.5 ± 12.4
Sex, No. (%)  
 Male 146 (50) 78 (50) 44 (51) 50 (45) 20 (56)
 Female 148 (50) 77 (50) 43 (49) 61 (55) 16 (44)
Race, No. (%)  
 White 142 (48) 88 (57) 38 (44) 41 (37) 11 (31)
 Black 113 (39) 48 (31) 41 (47) 57 (51) 17 (47)
 Hispanic 32 (11) 15 (10) 6 (7) 11 (10) 7 (19)
 Other 7 (2) 4 (3) 2 (2) 2 (2) 1 (3)
BMI, mean ± SD, kg/m2 28.6 ± 7.6 29.8 ± 7.4 26.3 ± 7.4 27.2 ± 8.1 27.1 ± 8.2
Weight loss, mean ± SD, % 3.1 ± 6.9 0.0 ± 0.2 10.4 ± 9.2 8.1 ± 9.3 3.5 ± 8.6
Fat wasting, No. (%) 49 (17) 2 (1) 33 (38) 46 (41) 10 (28)
Muscle wasting, No. (%) 62 (21) 2 (1) 38 (44) 59 (53) 15 (42)
Fluid accumulation, No. (%) 61 (21) 16 (10) 31 (36) 42 (38) 10 (28)
Average handgrip, mean ± SD 21.1 ± 12.0 (n = 242) 24.7 ± 11.1 (n = 133) 17.1 ± 12.3 (n = 62) 14.5 ± 10.0 (n = 89) 15.8 ± 12.1 (n = 30)
Hospital LOS, mean ± SD, d 8.5 ± 8.1 6.9 ± 6.7 10.7 ± 9.0 9.9 ± 8.6 12.1 ± 10.7
ICU LOS, mean ± SD, d 4.3 ± 4.3 3.7 ± 3.5 4.5 ± 4.2 5.4 ± 5.3 6.6 ± 7.2

Downloaded from pen.sagepub.com at WASHBURN UNIV on January 13, 2015

APACHE score, mean ± SD 13.0 ± 6.2 10.9 ± 5.0 13.5 ± 6.1 15.0 ± 6.8 22.9 ± 4.3
SOFA score, mean ± SD 2.7 ± 2.6 2.2 ± 1.9 2.7 ± 2.7 3.0 ± 2.6 5.8 ± 3.8
NUTRIC score, mean ± SD 2.5 ± 1.5 2.0 ± 1.1 2.7 ± 1.6 2.8 ± 1.6 5.4 ± 0.6
Died, No. (%) 21 (7) 5 (3) 10 (11) 13 (12) 5 (14)
Discharged to rehabilitation, No. (%) 38 (13) 16 (10) 14 (16) 19 (17) 6 (17)

APACHE II, Acute Physiology and Chronic Health Evaluation II; BMI, body mass index; ICU, intensive care unit; LOS, length of stay; NUTRIC, NUTrition Risk in Critically ill; SGA, Subjective
Global Assessment; SOFA, Sequential Organ Failure Assessment.
a
Routine nutrition screening included the following criteria: significant weight loss (5% in 1 month, 10% in 6 months), BMI <18.5 or >40, presence of dysphagia, or use of enteral/parenteral nutrition
prior to admission. Patients meeting at least 1 criterion were deemed at risk for malnutrition
b
Patients were classified as follows: normal (A), mild-moderate malnutrition (B), or severe malnutrition (C); for the purpose of analysis, patients with scores of B or C were classified as malnourished.
c
NUTRIC scores were calculated without using interleukin-6 values. Therefore, patients were classified as having a high score if the sum was 5 or greater; these patients were classified as having a
higher risk of malnutrition.

31
32 Journal of Parenteral and Enteral Nutrition 39(1)

Figure 1.  Venn diagram of patients determined at nutrition risk or malnourished using the Subjective Global Assessment (SGA),
routine screening, and/or NUTrition Risk in Critically ill (NUTRIC) score.

Figure 2.  Hospital and intensive care unit (ICU) length of stay (LOS) categorized by nutrition risk tool. NUTRIC, NUTrition Risk in
Critically ill; SGA, Subjective Global Assessment.

including those with trauma or burn units, since none of these limited by both body habitus and presence of edema.
patients were included. Results of the SGA between the 4 clini- Examination of body composition through more objective
cians completing assessments may vary due to the subjective measures, such as ultrasound or computed tomography, would
nature of the tool; assessment of fat and muscle wasting was be ideal but is limited by training and expertise needed. It

Downloaded from pen.sagepub.com at WASHBURN UNIV on January 13, 2015

Coltman et al 33

should be noted, however, that all the clinicians were trained in patient: Society of Critical Care Medicine (SCCM) and American Society
a uniform manner and all assessments were completed accord- for Parenteral and Enteral Nutrition (A.S.P.E.N.). JPEN J Parenter
Enteral Nutr. 2009;33(3):277-316.
ing to institution protocol, thus limiting variability between 3. White JV, Guenter P, Jensen G, Malone A, Schofield M. Consensus
assessments. Use of handgrip strength in this population was statement of the Academy of Nutrition and Dietetics/American Society
also limited due to patients’ mental status and mechanical ven- for Parenteral and Enteral Nutrition: characteristics recommended for the
tilation; this criterion requires additional examination prior to identification and documentation of adult malnutrition (undernutrition). J
use in the ICU. In addition, due to the nature of statistical anal- Acad Nutr Diet. 2012;112(5):730-738.
4. Kruizenga H, Seidell J, De Vet H, Wierdsma N, van Bokhorst-de van
ysis, comparisons made should be discussed in terms of clini- der Schueren MAE. Development and validation of a hospital screening
cal differences, rather than statistical. Further research is tool for malnutrition: the Short Nutritional Assessment Questionnaire
warranted. (SNAQ©). Clin Nutr. 2005;24(1):75-82.
5. Guigoz Y, Vellas B, Garry PJ. Assessing the nutritional status of the
elderly: the Mini Nutritional Assessment as part of the geriatric evalua-
Conclusion tion. Nutr Rev. 1996;54(1):S59-S65.
6. Ferguson M, Capra S, Bauer J, Banks M. Development of a valid and
Traditional screening and assessment tools identified different reliable malnutrition screening tool for adult acute hospital patients.
patients as malnourished or at nutrition risk in the ICU. Nutrition. 1999;15(6):458-464.
Differences may have resulted from inability to obtain subjec- 7. Detsky AS, McLaughlin JR, Baker JP, et al. What is subjective global
tive data due to the nature of critical illness; therefore, these assessment of nutritional status? JPEN J Parenter Enteral Nutr.
tools may be inappropriate for use in the ICU. Currently, no 1987;11(1):8-13.
8. Buzby GP, Mullen JL, Matthews DC, Hobbs CL, Rosato EF. Prognostic
one tool encompasses severity of illness along with traditional nutritional index in gastrointestinal surgery. Am J Surg. 1980;139(1):
markers of nutrition status. Tools using primarily objective 160-167.
measures may be the most useful in this population due to a 9. van Bokhorst-de van der Schueren MAE, Guaitoli PR, Jansma EP, de Vet
frequent inability to obtain subjective information, such as HC. Nutrition screening tools: does one size fit all? A systematic review
weight history and changes in oral intake. Inclusion of physical of screening tools for the hospital setting. Clin Nutr. 2014;33(1):39-58.
10. Heyland DK, Dhaliwal R, Jiang X, Day AG. Identifying critically ill
assessment, functional status, and severity of illness may be patients who benefit the most from nutrition therapy: the develop-
useful in screening and assessing nutrition risk in the ICU due ment and initial validation of a novel risk assessment tool. Crit Care.
to their objective nature. Additional research is needed to 2011;15(6):R268.
determine if changes in the nutrition screening process affect 11. Sheean P, Peterson S, Gurka D, Braunschweig C. Nutrition assessment:
outcomes in the critically ill. the reproducibility of subjective global assessment in patients requiring
mechanical ventilation. Eur J Clin Nutr. 2010;64(11):1358-1364.
12. Higgins TL, McGee WT, Steingrub JS, Rapoport J, Lemeshow S, Teres
References D. Early indicators of prolonged intensive care unit stay: impact of illness
1. Jensen GL, Bistrian B, Roubenoff R, Heimburger DC. Malnutrition severity, physician staffing, and pre-intensive care unit length of stay. Crit
syndromes: a conundrum vs continuum. JPEN J Parenter Enteral Nutr. Care Med. 2003;31(1):45-51.
2009;33(6):710-716. 13. Norman K, Stobäus N, Gonzalez MC, Schulzke J, Pirlich M. Hand grip
2. McClave SA, Martindale RG, Vanek VW, et al. Guidelines for the provi- strength: outcome predictor and marker of nutritional status. Clin Nutr.
sion and assessment of nutrition support therapy in the adult critically ill 2011;30(2):135-142.

Downloaded from pen.sagepub.com at WASHBURN UNIV on January 13, 2015