Current status and performance goals for serum thyrotropin (TSH) assays

Abstract

Current medical needs dictate that laboratories offer thyrotropin (TSH) assays that can reliably measure low TSH concentrations--a prerequisite for using the more cost-effective TSH-centered strategy currently recommended by the American Thyroid Association. This study reviews the functional performance of the TSH immunometric assay methods currently used in clinical practice. Methodological differences between methods, the rationale for using the 20% interassay CV as the functional sensitivity limit for patient reports, and both TSH-related and non-TSH-related specificity problems are reviewed. We recommend that manufacturers and clinical laboratories use a clinically relevant standard protocol for functional sensitivity assessments. In this protocol, human serum pools are analyzed in random order 10 or more times across a clinically realistic time span (which approximates to 6-8 weeks for TSH measurements used in an outpatient setting). Laboratories should independently establish and periodically check their functional sensitivity by the standard protocol and enlist the manufacturer's help to accomplish this if necessary. Manufacturers should (a) develop promotional material that realistically projects the assay's functional sensitivity expected in a clinical laboratory setting, (b) ensure that new clinical laboratory users achieve the projected functional sensitivity target when using the standard protocol, and (c) focus on the typical functional sensitivity obtained by clinical laboratory users rather than the assay's "generation" achieved under ideal conditions. If manufacturers and laboratories collaborate to solve the sensitivity and specificity problems discussed here, clinical laboratories should be better able to consistently deliver reliable serum TSH measurements across the full range of TSH concentrations encountered in clinical practice.