A Data Mining Model for Serious Adverse Drug Reaction Signals Based on Real-World Data:Taking Nauclea officinalis Pierre ex Pitard Preparations as an Example for Empirical Analysis
Based on the current signal detection work of Chinese drug regulatory agencies,this paper explores the construction of a new evidence-based model for the timely detection and accurate identification of severe adverse drug reactions(ADRs)utilizing real-world data(RWD)and data mining techniques.This model,based on the spontaneous reporting system database and the electronic medical records system database,divides ADR signal data mining into four stages:initial signal detection,signal screening,signal verification,and signal evaluation.By applying data mining techniques,the model allows for rapid,high-throughput mining of ADR signals,which gradually improves the level of evidence of the signals and improves the efficiency of signal detection.In this paper,an empirical study using this model for the data mining of adverse reactions of Nauclea officinalis Pierre ex Pitard preparations is presented,initially verifying the model's operability and providing a paradigmatic experience for regulatory authorities to carry out data mining on severe ADR signals.
real-world dataserious adverse drug reactiondata miningsignal detectionevidence-based evidence
万方数据
