Laboratory quality control (QC) strategies in resource-constrained settings

in MSF–OCA programs

Daniel Orozco, Pamela Hepple, Derryck KlarkowskiLaboratory Specialists - MSF OCA

Basic health care laboratories Malaria/TB microscopy

Secondary level labs for TB/ HIV

programs

High technology labs MDR - TB program

Uzbekistan

Background

•Pre-2005 quality of lab performance within MSF OCA labs was unknown

•We identified the need to set up a standardized QC system

• Goals of the programme– Monitor performance

– Identify poor performers

– Provide clinical staff with information on the accuracy of laboratory results

• Program design - Field/HQ input for practicality

• QC protocol implemented in 2005

• Key element is the central reporting system, which enables effective management

• Immediate finding:

– Wide diversity of performance

– Unacceptable number of centers were performing poorly

• 2005 – 2007: Significant improvement across the lab network

Background

• Malaria, TB, Kala Azar, Cutaneous Leishmaniasis and HIV applying similar principles

• We also have a separate QC methodology for Chemistry, Hematology and CD4 testing

Scope

QC Methodology

QC Methodology

• Monthly random selection

• Based on crosschecking

• Results are sent to Amsterdam

• Analyzed using % agreement, false positivity and false negativity

QC Methodology

• Based on a small sample size –5 weak positives + 5 negatives

•To limit workload for laboratory staff

•Limited reference laboratory capacity

– The small sample size is compensated for by• Targeting weak positives, where errors are most likely to occur

• 4-month cohort analysis (40 slides)

– % agreement

– % false positives/false negatives

– Quartile analysis

– Compliance

The statistical analysis has been validated by WHO (LSHTM).

Analysis:

Results

• Between 2005 – 2007:

- 748 QC monthly reports for malaria from 70 centres

- 33,346 slides

• Currently, 40 centers centrally reporting (20 countries)

False Positive Rates

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

8M 2005

12M

2005 4M 2006 8M 2006

12M

2006 4M 2007 8M 2007

%FP

of w

eak

posi

tives median

1st Quartile

3rd Quartile

Results: Improvement

• QC alone does not improve performance, but is part of a package – auditing system: Standard Indicators, Equipment, reagents, Infrastructure,Training, in-situ support

• Progressively strengthening Quality Assurance (QA). E.g. malaria manual, Standard Indicators Reports (SIR), upgrading QC database software

Improvement•Identification of weak performers and corrective action taken

•Our QC programme created a healthy “competition” by comparing centres against each other, which are working under similar circumstances

OVERALL QC COMPLIANCE PER TEST

0%

20%

40%

60%

80%

100%

8M 2006 12M 2006 4M 2007 8M 2007 12M 2007

COHORT PERIOD

CO

MP

LIA

NC

E

Malaria

Tuberculosis

HIV

Kala Azar

CutaneousLeishmaniasis

Conclusions

•Sustainable method for resource-constrained settings. Practical and with minimal investment

•Statistically valid

•Successful - e.g. for Malaria

– Median increased from 90% to 100%

–1st quartile increased from 86% to 97%

•Supports accuracy of diagnosis and operational research

Demands commitment:

• from field staff for continuous operation

• lab specialists for central monitoring by issuing compliance and benchmarking reports

• Depends upon support from Medical Director, Head PHD and CMTs for its implementation and follow up.

Conclusions