Minimum distinguishable difference in concentration: a clinically oriented translation of assay precision summaries

A trend to market-driven health care in many parts of the world is focusing increasing attention on getting maximum value from available resources. Laboratories are not exempted. Well-informed clinical input has a potentially valuable role in any laboratory rationalization process. However, a communication difficulty exists in the sense that, although laboratory workers, commercial developers, regulatory bodies, etc., are thoroughly conditioned to using assay coefficient of variation as a general performance measure (for excellent reasons), this is not necessarily the most intuitive or informative scale from a clinician's perspective. Here we use routine clinical data from an immunoradiometric assay of thyrotropin to illustrate, first, a general approach to estimation and prediction of reproducibility, and second, an alternative summary that expresses the discriminatory power of an assay. This latter measure, our experience suggests, is more suited to the way clinicians perceive assays and assay results. The overall aim is improved clinician/laboratory communication.