Lot Quality Assurance Sampling (LQAS)

An Overview

LQAS Conference, Sheraton Kampala Hotel 3rd July 2006

 Purpose of the Presentation

ƒ Explain:
– Basic principles of Lot Quality Assurance Sampling (LQAS)
for carrying out baseline and monitoring surveys of
community programs
– Some statistics behind the LQAS Methodology

ƒ Present some practical examples of LQAS

application in the field
 What is LQAS?

A sampling method that:

ƒ Can be used locally, at the level of a “supervision area,” to
identify priority areas (e.g., county, sub-county) or
indicators that are not reaching average coverage or an
established benchmark

ƒ Can provide an accurate measure of coverage or health

system quality at a more aggregate level (e.g., program
catchment area or district or refugee camp)

ƒ Can be used for quality assurance using a ‘minimal

sample’, ‘maximal security’ principle
 LQAS is a simple, low cost random
sampling methodology
ƒ Originally developed in the 1920s to control the quality of
output in industrial production processes
ƒ Involves taking a small random sample of a manufactured
batch (lot) and test the sampled items for quality
ƒ If the number of defective items in the sample exceeds a
pre-determined criteria (decision rule), then the lot is rejected
ƒ The decision rule is based on the desired production
standards and a statistically determined sample size
ƒ ‘n’ is chosen so that the manager has a high probability of
accepting lots that meet the quality standards and rejecting
lots that fail to meet those standards
 Some Useful LQAS Definitions

Standard LQAS theory UPHOLD ‘LQAS’ Adaptation

Production standard: Coverage:
% of items that must “pass” % of clients who received a
before a lot is accepted service in a defined period of time

Production unit: Supervision Unit:

The machine or team that The district were these services
produced or assembled the lot are delivered

Lot: Supervision area:

A batch of items produced in A county or sub-counties in a
given time by the production unit given district where services are
being delivered
 What are the LQAS Principles?

ƒ Assume a program
covers a whole district
Omoro
ƒ Each county is then
called a ‘supervision
area’ and district a
‘supervision unit’
Gulu MC Kilak
Aswa ƒ LQAS would chose a
minimum of 19 items
(e.g. households,
Nwoya schools, health units)
from each ‘supervision
area’ in order to assess
an indicator
 What are the LQAS Principles?

Good
Omoro

Gulu MC Kilak
Aswa

Below
Nwoya Average
or Desired
Coverage
 What can be done with the
findings?
Maintain the program at the
current level

Good
Identify ‘best practices’ that can
help others programs improve
their performance

Identify the reasons for program

problems
Below Average
or desired coverage
Develop targeted solutions
 A sample of 19 is taken from each
supervision area in each district

S.A. = 19

S.A. = 19
District
S.A. = 19
S.A. = 19

S.A. = 19
 Why use a Sample Size of 19?

ƒ A sample size of 19 provides an acceptable level of error

for making management decisions; at least 92% of the
time, it identifies whether a coverage benchmark has
been reached or whether an SA is substantially below
the average coverage of a program area

ƒ Samples larger than 19 have practically the same

statistical precision as 19. They do not result in better
information, and they cost more
Why use a Sample Size of 19?

ƒ Little is added to the precision of

the measure by using a sample
larger than 19, notwithstanding
the level of coverage to be
assessed

ƒ Sample sizes less than 19

however, see a rapid
deterioration in the precision of
the measure. This is particularly
problematic when coverage
benchmarks vary
 What can a sample of 19 tell us?

ƒ Good for setting priorities within a Supervision Area

ƒ Good for setting priorities among supervision areas
with large differences in coverage
ƒ Good for deciding what are the higher performing
supervision areas to learn from
ƒ Good for deciding what are the lower performing
supervision areas in which to invest resources
ƒ Good for identifying knowledge/practices that have
high coverage from those of low coverage
 What a Sample of 19 Cannot tell us

ƒ Not good for calculating exact coverage in a

supervision area (but can be used to calculate
coverage for an entire program)

ƒ Not good for setting priorities among supervision

areas with little difference in coverage
 Identifying Households for
Interview
ƒ Step 1. List Communities and Total Population

ƒ Step 2. Calculate the Cumulative Population

ƒ Step 3. Calculate the Sampling Interval

ƒ Step 4. Choose a Random Number

ƒ Step 5. Beginning with the random number, use the

sampling interval to identify communities for the 19 sets
of interviews
 Benefits of LQAS as a Sampling Method

ƒ Low sample size needs (n=19 in most cases)

ƒ Simple to apply yet has very specific conclusions

ƒ District level people can be trained to entirely ‘own’ this methodology

ƒ Provides high quality information at low & affordable cost

ƒ Fast – ‘supervision areas’ are able to conduct self-evaluation and obtain results
immediately after the survey

ƒ Results are locally relevant and can be utilized in district level annual planning

and decision-making
 Benefits of LQAS to Districts

ƒ Change of mindsets towards reliance on data for

Evidence-Based planning
ƒ Districts have relied on LQAS data to identify priority
target areas for implementation and to focus work
plans
ƒ Districts can use LQAS to do own evaluations in
different technical areas e.g., education, agriculture
ƒ Strengthened partnership between districts and
NGOs in provision of related services
 Illustrative Costs – Luwero District
2004 Survey
Description Quantity Participants Cost Total cost

LQAS Training 10 Days 6 48,000/- 2,880,000/-

Data collection 5 days 6 48,000/- 1,440,000/-

Allowance
Transport - Car hire 5 days 1 car 180,000/- 900,000/-

Transport - Boda boda 5 days 6 Bodas 15,000/- 450,000/-

Total Implementation 5,670,000/-

costs
Annual Costs 1,890,000/-
(sustainable) + ~ 500,000/-
 NEPAL EXAMPLE

Comparative Cost Analyses for Using

LQAS and Cluster Sampling (USD)
Item Total Essential Minus Opport. Recurrent Cluster Sample
Costs Cost = Marginal Marginal Cost Cost
Cost for next use
Salary 1,328 - - 2,498

Transport 420 420 420 673

Materials 403 403 179 816

Food & Accom. 796 762 146 2,561

Total cost 2,947 1,585 745 6,548

Cost per $7.39 $4.06 $1.87 $21.83

Observation
 Summary of how UPHOLD has
applied the LQAS methodology
ƒ ~ 200 District level Officials (including CDWs, DDHS, DEO
and District Planning personnel) were trained for 2 weeks in
the LQAS methodology in 2004

ƒ 19 villages are sampled from each of the five ‘Supervision

Areas’ in each district

ƒ 5 households are sampled from each village and a different

questionnaire administered to each of the sampled
households (~12,300 households covered in last survey)

ƒ Schools and Health Facilities also surveyed in Baseline (423

Health Units and 1,449 Schools)
 EXAMPLE

After the Baseline achievements are

compared with Annual Targets

BASELINE Yr. 1 Yr. 2 Yr. 3 Yr. 4

10% 30% 50% 70% 80%

Program Costs from Baseline until Year 4 of

the Project
Improvement
 Application of the LQAS
methodology in other countries
ƒ Bangladesh – Immunization Coverage
ƒ Malawi – Malaria Prophylactic Regime
ƒ Kenya, Senegal – Contraception
ƒ Costa Rica – Quality of Healthcare
ƒ Spain – Quality of Health Centres
ƒ Bangladesh, Indonesia, Zimbabwe, Namibia,
Morocco – Neonatal Tetanus Elimination