Governmental perspectives on evaluating laboratory performance.

The quality of laboratory analytical performance required to support medical decision-making has been defined in four major ways: (a) by the analytical variance of the state of the practice; (b) by the total variance, including analytical and biological variability; (c) by the loss of diagnostic efficiency attributable to analytical error; and (d) by medical-usefulness criteria. From the federal government's perspective, the answer to the question "How good must a laboratory test result be to be medically relevant?" must take into account the clinical context of the test, with accompanying concerns about access, timeliness, and cost, as well as limits for precision and accuracy in the analytical process and the frequency and potential patient-care impact of error in the pre- and postanalytical steps of the total testing process. Therefore, medically relevant goals should encompass not only analytical precision and accuracy but also goals to provide access to clinically effective tests and to reduce errors in the total testing process that can lead to medically misleading information. Development of more appropriate regulatory requirements for laboratories, as well as any needed improvements in instrumentation and methodology, should focus on ensuring that goals for medically relevant results are met by appropriate design and management of the entire process of laboratory testing.