Laboratory standardization from the patient's point of view.

Optimal patient care is best defined in terms of outcome. Reliable laboratory test results are important to patient safety. The laboratory must use the tools available to it to minimize the uncertainty of measurement. Three sources of error contribute to uncertainty: Intermethod bias, which is minimized and trueness of measure maximized when laboratories use calibrators and methods traceable to higher order, reference standards; imprecision inherent in the analysis, which is seen as small differences between replicate tests; interference from factors external to the test itself, which are seen as erroneous values markedly deviant from trueness. Although improbable, such contributions to total analytical error may be the most misleading. Risk is best managed by identifying the sources of error and controlling for those sources most likely to contribute to total analytical error. Comprehensive control of error requires the laboratory scientist and physicians caring for patients to work together to ensure interpretability of results. Practice guidelines are available from the Clinical and Laboratory Standards Institute to address each of these factors.