Implementation and validation of an intelligent auto-verification system in a coagulation testing automatic line

Abstract

Objective

To evaluate the implementation and optimization of a fully automated coagulation testing system integrating dual-coagulation-autoline and auto-verification protocols in a high-throughput laboratory, aiming to enhance operational efficiency, standardization, and clinical utility.

Methods

A dual-coagulation-autoline system was deployed, each line equipped with four automated coagulation analyzers. Pre-implementation quality assessments evaluated centrifuge performance, open/closed-cap sampling, robotic arm effects, and sample stability. Auto-verification rules were optimized using several guidelines and validated with 18,373 samples. Key metrics included pass rate, sensitivity, specificity, and turnaround time (TAT).

Results

All centrifuged specimens met platelet-poor plasma criteria (platelet count <10 × 10⁹/L). No significant differences (P > 0.05) were observed between open/closed-cap sampling, robotic arm handling, or room-temperature storage (≤4h). The dual-coagulation-autoline system reduced outpatient and inpatient TAT by 16.78 % (from 49.77 min to 41.42 min) and 33.06 % (from 110.3 min to 73.83 min), respectively, while sample volume increased by 7.9 % (from 74,902 to 80,869). Optimized auto-verification rules achieved a 62.48 % pass rate (11,479/18,373), 100 % sensitivity, 93.50 % specificity, and 0 % false negatives, outperforming initial benchmarks. Auto-verification reduced outpatient and inpatient TAT by 8.29 % (from 41.42 min to 38.25 min) and 16.21 % (from 73.83 min to 63.53 min), respectively, while sample volume increased by 2.8 % (from 80,869 to 83,126).

Discussion

The integration of dual-coagulation-autoline and refined auto-verification significantly improved TAT, reduced manual errors, and standardized workflows in a high-volume setting. Personalized system optimization, including dynamic testing allocation and real-time tracking, enhanced clinical efficiency. The validated auto-verification framework demonstrated robust generalizability for large-scale laboratories. The successful application and personalized setting of our laboratory can provide experience for other laboratories.