How good are clinical laboratories? An assessment of current performance

The Clinical Laboratory Improvement Act of 1967 and Amendments of 1988 (CLIA '67 and CLIA '88) were enacted to ensure that clinical laboratories within the U.S. provide a quality of service that meets clinical needs for good patient care. Approved proficiency-testing programs are to judge the quality of laboratory testing by promulgated performance criteria. We examine the quality of analytical results reported in 1991 to the New York State Department of Health Proficiency Testing program in light of these criteria and analytical goals, based on medical usefulness. Analytical performance is examined for cholesterol, potassium, sodium, calcium, glucose, aspartate aminotransferase, digoxin, and theophylline. In general, proposed CLIA '88 performance standards are compatible with the current state of practice for the population of laboratories examined. Exceptions appear to be digoxin and sodium (failure rate exceeding average) and most therapeutic substances (low failure rate). Sources of analytical bias relative to an accuracy-based target value must be characterized as method-, laboratory-, or matrix-dependent if regulatory programs are to achieve the objective of improving analytical accuracy across all testing sites.