From metrological comparability to clinical continuity: authorising laboratory result reporting when changing measurement procedures

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Abstract Objectives To develop and apply an operational framework integrating metrological, analytical, clinical and risk evidence to authorise an output-specific reporting rule after a change in measurement procedure. Methods Four measurement procedure changes were evaluated: alanine aminotransferase (ALT) procedures with and without pyridoxal-5′-phosphate activation; follicle-stimulating hormone (FSH) reagent replacement; routine vs. short-turnaround-time (STAT) parathyroid hormone (PTH) procedures; and tacrolimus procedure replacement with retention of the previous procedure for contingency use. The pathway integrated metrological comparability and compatibility, operational equivalence, controlled correction to a target reporting scale, independent clinical verification, uncertainty-aware classification, risk assessment and surveillance. Results The framework produced four qualitatively distinct outcomes: no cross-procedure continuity for ALT; direct continuity for raw FSH results; PTH continuity only through a corrected STAT result; and routine use of the new tacrolimus procedure, with the previous one restricted to contingency testing. FSH showed 100 % pair-level compatibility (95 % CI 91.2–100 %) and nominal percentage of clinically acceptable concordance (PCAC) of 96.6 % (82.2–99.9 %). PTH raw compatibility was assumption-sensitive at 90.0 % (76.3–97.2 %) vs. 87.5 % (73.2–95.8 %); the corrected output increased compatibility to 95.0 % (83.1–99.4 %), with nominal PCAC of 100 % (83.2–100 %). Tacrolimus met the mean-difference criterion but showed excess residual scatter, raw compatibility of 70.0 % (53.5–83.4 %) and nominal PCAC of 65.0 % (40.8–84.6 %); exploratory correction was not authorised. Conclusions Changes in measurement procedure should culminate in an authorised reporting rule integrating metrological, analytical, clinical and risk evidence. Regression alone cannot establish continuity or interchangeability.