Signal Detection and Causality Evaluation for Pharmacovigilance

약물감시를 위한 실마리정보 파악 및 인과관계 평가

Pharmacovigilance is the science and activities relating to detection, assessment, understanding and prevention of adverse effects or any other possible drug-related problems. Information sources for pharmacovigilance could be premarketing or postmarketing data from pre-clinical experiments, clinical trials, spontaneous reports, large automated databases, pharmacoepidemiologic studies, and meta-analysis studies. Spontaneous reporting data playa key role to identify signals for marketed drugs. Steps in the spontaneous adverse event signaling process consist of signal detection phase and signal evaluation phase, and involve various statistical methods. Through record linkage of large existing automated databases by individual identifiers, it is possible to conduct longitudinal researches on drug safety issues in large population. Carefully designed and conducted pharmacoepidemiologic studies are important tools in pharmacovigilance. Case-control studies and cohort studies are widely used to confirm the causality between drugs and adverse events. In this paper, we discussed possible data sources for pharmacovigilance and introduced some useful methodological approaches. Evidence-based decisions through these methods should be effectively distributed to healthcare professionals, patients and media.