Good clinical laboratory practices improved proficiency testing performance at clinical trials centers in Ghana and Burkina Faso.

BACKGROUND: The recent drive towards accreditation of clinical laboratories in Africa by the World Health Organization-Regional Office for Africa (WHO-AFRO) and the U.S Government is a historic step to strengthen health systems, provide better results for patients and an improved quality of results for clinical trials. Enrollment in approved proficiency testing (PT) programs and maintenance of satisfactory performance is vital in the process of accreditation. Passing proficiency testing surveys has posed a great challenge to many laboratories across sub-Saharan Africa. Our study was aimed at identifying the causes of unsatisfactory PT results in clinical research laboratories conducting or planning to conduct malaria vaccine trials sponsored by the National Institutes of Health (NIH). METHODOLOGY: PT reports for 2009 and 2010 from the College of American Pathologists (CAP) for the laboratories were reviewed as part of the process. Errors accounting for unsatisfactory results were classified into clerical, methodological, technical, problem with PT materials, and random errors. A training program on good clinical laboratory practices (GCLP) was developed for each center to address areas for improvement. RESULTS: The major cause of PT failure in the four centers was methodological. The application of GCLP improved the success rate in the PT surveys from 58% in 2009 to 88% in 2010. It also decreased the error rate on PT by 35%. CONCLUSION: A previous report from the CAP- PT participating laboratories indicated that the major causes of error were clerical. These types of errors were predominantly made in laboratories in the US, with much more experience in quality control, and varied significantly from what we found. In our centers in sub-Saharan Africa, methodological errors, and not clerical errors, accounted for the vast majority of errors. A process was started for continuous improvement which has decreased methodological errors by 35%, but more improvement is needed.

Abstract licence: CC BY

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.