Evaluating laboratory performance. Historical and governmental perspectives.

Historically, the quality of laboratory performance that must be provided to adequately support medical decision-making has been defined in three major ways: by the analytical variance of the state of the practice; by the total variance, including the expected analytical and biological variability; and by medical-usefulness criteria. Quality goals defined by these approaches have influenced instrument manufacturers, professional societies, governmental agencies, and even the precision and accuracy of medical practice. Despite past efforts, a consensus has not emerged on a scientifically and medically supportable set of laboratory performance requirements. Medically relevant goals for analytic performance would help avoid costly and unnecessary regulatory requirements and permit steady improvement in instrumentation and methods.