Clin Chem Lab Med. 2026 May 11. doi: 10.1515/cclm-2026-0153. Online ahead of print.
ABSTRACT
OBJECTIVES: Few studies have systematically evaluated the factors contributing to laboratory errors in the analytical phase. This study analyzed five years of retest data to determine the frequency and characteristics of analytical errors across routine blood chemistry parameters, identify analytes most prone to error, and provide insights for improving test result accuracy.
METHODS: We retrospectively reviewed retest records for 23 blood chemistry parameters performed between January 2018 and June 2022. Of 18,015,323 total tests, 58,637 (0.31 %) underwent repeat analysis following delta or panic check alerts. Acceptance limits for repeated measurements were established based on CLIA proficiency testing (PT) regulations and the precision of each analyte. Differences in error rates among analytes were assessed using chi-square tests.
RESULTS: The overall analytical error rate among retested samples was 1.66 %. Carbon dioxide (19.67 %) and ammonia (16.53 %) demonstrated the highest error rates, whereas alanine aminotransferase (0.04 %), albumin (0.09 %), aspartate aminotransferase (0.11 %), sodium (0.18 %), and chloride (0.24 %) showed very low rates. Electrolytes measured using ion-selective electrodes (sodium, potassium, chloride) exhibited lower error rates than those quantified colorimetrically (calcium, phosphate, magnesium). Analytes prone to evaporation or instability, such as carbon dioxide and ammonia, displayed substantial discrepancies upon retesting.
CONCLUSIONS: Analytical error rates are influenced by assay complexity, physiological analyte concentration, or specimen stability. Routine retesting may offer limited benefit for high-concentration analytes or methodologically simple assays. In contrast, assays for low-concentration or methodologically complex assays should require more vigilant monitoring. Implementing analyte-specific retesting strategies may enhance laboratory efficiency and improve result accuracy monitoring.
PMID:42105300 | DOI:10.1515/cclm-2026-0153