A data-driven dynamic evaluation model for the quality management level of clinical trial institutions

DONG Wen-bin; JIANG Min; ZHAO Shu-hua; ZHANG Xue-mei; CHEN Yi-fei

doi:10.20251/j.cnki.1003-3734.2025.09.008

Abstract：Objective:To improve evaluation effectiveness,an evaluation model was established from the perspectives of quality management and risk identification,which comprises three modules including quality risk monitoring,quality management issue warning and quality management level evaluation.Methods:According to the relevant regulations and guidelines for drug clinical trials,an evaluation system was established through 5M1E analysis and fishbone diagrams.The Delphi method,failure mode and effect analysis,and analytic hierarchy process were employed to quantify the evaluation system.This model was tested using analogue data as well as real data.Additionally,the test results of both single module and multi-module were interpreted.Results:After two rounds of consultation and semi-quantitative analysis,we developed a quantitative and visualization model for assessing the quality management level of drug clinical trial institutions.This model enables linking of quality risk,quality management issue and quality management level.The quality risk monitoring module comprises 6 first-level indicators and 18 second-level indicators,and the quality management issue warning module consists of 5 indicators.Additionally,the quality management level evaluation module includes 3 first-level indicators and 11 second-level indicators.Conclusion:Compared to the traditional method of counting defect items in inspection of drug clinical trials,the evaluation model for the quality management level evaluation of drug clinical trial institutions established in this study is more intuitive.Furthermore,the multi-module linkage analysis method enables a comprehensive representation of information,thereby effectively enhancing decision-making efficiency.