Abstract / Summary
Objective: The pre-analytical phase is the stage at which laboratory errors most frequently occur and is therefore of critical importance for patient safety. During the COVID-19 pandemic, the marked increase in workload, organizational restructuring, staff and resource redeployment, changes in patient volume and clinical workflows, and disease-related pathophysiological alterations affecting specimen characteristics created unprecedented conditions that could influence laboratory quality indicators. This study aimed to characterize changes in specimen rejection rates across the pre-pandemic, pandemic, and normalization periods between 2018 and 2023, and to determine whether pandemic-related disruptions in health care delivery were associated with differences in the number, rate, and distribution of specimen rejections according to clinical unit, rejection reason, and time interval, with emphasis on the magnitude of these differences rather than statistical significance alone. The study also sought to generate evidence that could inform the development of preventive systems and strengthen laboratory preparedness for future large-scale disruptions to health care services. Materials and Methods: This retrospective observational study used routinely collected, aggregated laboratory information system (LIS) data from the Central Laboratory of Istanbul Atlas University Hospital for 1 January 2018 through 31 December 2023. The LIS extract was restricted to eligible specimen types, including spot and 24 h urine, serum, and whole-blood or plasma specimens collected in EDTA-, citrate-, or heparin-containing tubes; all other specimen types were excluded. Specimen rejection rates were evaluated by study period, clinical unit, time interval, and rejection reason. For cross-period comparisons, the total number of specimens within each respective study period was used as the denominator for calculating rejection rates, whereas for rejection-reason-specific analyses, the number of specimen types relevant to each rejection category was used as the denominator. Negative-binomial regression with denominator-based offsets estimated adjusted rejection-rate ratios (RRs) with 95% confidence intervals (CIs), with Bonferroni adjustment for selected pairwise comparisons. No missing or unclassifiable values were identified; no imputation was required. Reporting follows STROBE and the RECORD extension for studies using routinely collected health data. Results: The overall rejection rate increased progressively from 0.23% in the pre-pandemic period to 0.49% during the pandemic and 0.65% in the normalization period, and monthly observed rates supported this progressive increase. In adjusted terms, the rejection rate was 2.20-fold higher during the pandemic than in the pre-pandemic period (95% CI: 1.96–2.46) and increased a further 1.42-fold in the normalization period compared with the pandemic period (95% CI: 1.27–1.59). Across clinical units, the emergency department had the highest observed rejection rate, whereas intensive care units showed the highest adjusted rejection rates during the pandemic (RR = 1.34 versus the emergency department, 95% CI: 1.00–1.80) and normalization periods. Hemolysis and clotting were the dominant rejection reasons; although clotting had the highest observed rates in the pre-pandemic and pandemic periods, hemolysis ranked highest after adjustment in all periods and separated significantly from clotting only in the normalization period (RR = 1.97, 95% CI: 1.65–2.35). Time-interval effects were context-dependent rather than uniform and became informative mainly in combination with clinical unit or rejection reason: intensive care unit rejections peaked in the daytime interval during the pandemic (RR = 1.84 versus the evening interval, 95% CI: 1.06–3.19), and in the normalization period emergency department–hemolysis emerged as the leading adjusted clinical unit–rejection reason combination (RR = 1.89 versus intensive care unit–hemolysis, 95% CI: 1.19–3.00), while clotting remained centered on intensive care units and became more concentrated in the overnight and early-morning interval. Conclusions: Pre-analytical specimen rejection increased progressively across the pre-pandemic, pandemic, and normalization periods, with adjusted effect sizes indicating a substantial and sustained deterioration rather than a transient pandemic-related disturbance, and with persistent differences in rejection patterns according to clinical unit, time interval, and rejection reason. The further increase observed in the normalization period suggests that pandemic-related disruptions were associated with sustained changes in pre-analytical performance but were unlikely to be the sole explanation. Emergency department–hemolysis during the normalization period and intensive care unit–related clotting emerged as the most concrete quality-monitoring targets. Integrating rejection rates and counts with stratified monitoring by clinical unit, time interval, and rejection reason in combination may support earlier identification of emerging risks, more targeted preventive interventions, and stronger laboratory preparedness for future large-scale disruptions to health care services.