Proficiency testing in laboratory medicine: uses and limitations.

OBJECTIVE: To provide a critical review of recently published literature on the effectiveness, uses, and limitations of proficiency testing (PT) as a mechanism for laboratory improvement, and to explore ways to improve the PT process. DATA SOURCE: All publications identified by a MEDLINE search of the literature dating back to 1987 on the subject of "proficiency testing" in laboratory medicine, as well as selected references cited in recent review articles. STUDY SELECTION: No specific selection criteria were used for inclusion of publications identified by the MEDLINE database as long as they dealt with PT as a mechanism of medical laboratory improvement or a measure of laboratory performance. DATA EXTRACTION: Abstractions of data were made depending on relevance of the data. DATA SYNTHESIS: Proficiency testing data are an indicator, but not a measure, of laboratory performance. Limitations of current PT practices are incomplete assessment of the total testing process, PT materials being treated differently than those from patients, PT performance criteria, and "matrix effect." Proficiency testing performance has been related to length of PT experience, test environment and volume, institutional size, laboratory and analyst workload, difficulty of PT materials, performing quality control, testing methodology, and degree of automation.
CONCLUSIONS: Proficiency testing has a well-established role as both a laboratory improvement and an educational tool. There are, however, several practical and design limitations even for the best-administered PT programs. Suggestions to improve the PT process include increased reliance on PT results in combination with other quality indicators (such as performance in regional surveys), occasional use of "blind" PT, introduction of biological materials to PT participants, electronic grading and reporting of PT results, and introduction of challenging PT materials to fulfill the educational role of PT.