Pre-analytical phase errors constitute the vast majority of errors in clinical laboratory testing

Résumé fourni par la source

OBJECTIVES: Clinical laboratory errors pose a threat to patient safety and previous studies have demonstrated that pre-analytical error is the most common error type. Our study aimed to determine the types and frequency of errors occurring in clinical laboratory testing in contemporary practice. METHODS: Errors occurring in a core laboratory between 01/2022 and 05/2023 were recorded retrospectively. Errors were quantified using multiple data-streams including real-time manual technologist intervention, incidence reports filed by hospital staff/physicians, and retrospective assessment using automated reports from the lab information system (LIS). Errors were adjudicated and binned into pre-analytical, analytical, and post-analytical phases. Total test volumes were assessed in the LIS and electronic medical record. RESULTS: There were 37,680,242 billable results reported from approximately 11,000,000 specimens during the study period. In total, 87,317 errors occurred impacting 0.23 % (2,300 ppm) of billable results and approximately 0.79 % (7,900 ppm) of specimens. Among these errors, 85,894 (98.4 %, 984,000 ppm) were in the pre-analytical, 451 (0.5 %, 5,000 ppm) were in the analytical, and 972 (1.1 %, 11,000 ppm) occurred in the post-analytical phase. Hemolysis impacting specimen integrity (60,748/87,317, 69.6 %, 696,000 ppm) was the most common error. When excluding hemolysis, there were 26,569 errors documented (0.06 %, 600 ppm of billable results), among which 94.6 %, 1.7 % (17,000 ppm) and 3.7 % (37,000 ppm) were in the pre-analytical, analytical and post-analytical phase respectively. CONCLUSIONS: Observed error rates were consistent with previous studies with pre-analytical errors comprising most errors. High prevalence of pre-analytical errors implies a need for enhanced tools for error detection and mitigation in the pre-analytical phase of testing.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.