Establishment of a rapid identification of adverse drug reaction program in R language implementation based on monitoring data.
10.3785/j.issn.1008-9292.2020.03.07
- Author:
Dongsheng HONG
1
;
Jian NI
2
;
Wenya SHAN
1
;
Lu LI
1
;
Xi HU
1
;
Hongyu YANG
1
;
Qingwei ZHAO
2
;
Xingguo ZHANG
1
Author Information
1. Department of Pharmacy, the First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou 310003, China.
2. Key Laboratory for Drug Evaluation and Clinical Research of Zhejiang Province, the First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou 310003, China.
- Publication Type:Journal Article
- MeSH:
Databases, Pharmaceutical;
Decision Making, Computer-Assisted;
Drug Monitoring;
Drug-Related Side Effects and Adverse Reactions;
HIV Protease Inhibitors;
adverse effects;
pharmacology;
Humans;
Liver;
drug effects;
Lopinavir;
adverse effects;
toxicity;
Models, Statistical;
Reproducibility of Results;
Software;
standards
- From:
Journal of Zhejiang University. Medical sciences
2020;49(2):253-259
- CountryChina
- Language:Chinese
-
Abstract:
OBJECTIVE:To establish a clinically applicable model of rapid identification of adverse drug reaction program (RiADP) for risk management and decision-making of clinical drug use.
METHODS:Based on the theory of disproportion analysis, frequency method and Bayes method, a clinically applicable RiADP model in R language background was established, and the parameters of the model were interpreted by MedDRA coding. Based on the actual monitoring data of FDA, the model was validated by the assessing hepatotoxicity of lopinavir/ritonavir (LPV/r).
RESULTS:The established RiADP model included four parameters: standard value of adverse drug reaction signal information, empirical Bayesian geometric mean value, ratio of reporting ratio and number of adverse drug reaction cases. Through the application of R language parameter package "phViD", the model parameters could be output quickly. After being encoded by MedDRA, it was converted into clinical terms to form a clinical interpretation report of adverse drug reactions. In addition, the evaluation results of LPV/r hepatotoxicity by the model were matched with the results reported in latest literature, which also proved the reliability of the model results.
CONCLUSIONS:In this study, a rapid identification method of adverse reactions based on post marketing drug monitoring data was established in R language environment, which is capable of sending rapid warning of adverse reactions of target drugs in public health emergencies, and providing intuitive evidence for risk management and decision-making of clinical drugs.
