Background: Repeat testing in emergency clinical chemistry is common, often triggered by critical values, delta checks, or instrument flags. However, its clinical utility is uncertain, and unnecessary repeats may delay reporting and waste resources.
Objective: To assess the necessity of repeat testing by applying three analytical criteria: laboratory-specific observed total allowable error (TEa_Obs), CLIA-based TEa (TEa_CLIA), and reference change values (RCV) to the same cohort.
Methods: In this retrospective study, we analyzed 18,736 repeated result pairs across 23 analytes retrieved from the laboratory information system (LIS) of a tertiary-care emergency laboratory (Jan 2022 to Jun 2025). Each pair was classified as necessary or unnecessary under TEa_Obs, TEa_CLIA, and RCV; agreement between criteria was assessed using Cohen's kappa (κ). Necessary versus unnecessary TAT was compared within each criterion.
Results: TEa_Obs classified 86.2% (16,143/18,736) of repeats as unnecessary, versus 93.7% (17,564/18,736) by TEa_CLIA and 96.7% (18,119/18,736) by RCV. Agreement was generally low to moderate, with κ ranging from <0.2 (e.g. creatinine, CRP, bilirubins) to >0.8 (magnesium). Necessary repeats were associated with longer TAT, although the statistical significance of these differences varied by criterion.
Conclusions: Criterion selection is not neutral; it changes both clinical interpretation and operational outcomes. Rule-driven workflows that combine TEa_Obs, TEa_CLIA, and RCV within LIS-based auto verification may reduce unnecessary repeats, stabilize turnaround times, and optimize resource use without compromising patient safety.
Keywords: Clinical laboratory information systems; biological variation, individual; clinical laboratory techniques; quality assurance, health care; reproducibility of results; workflow.