Quantitative Methods in Pharmacovigilance

Abstract

Pharmacovigilance serves to detect previously unrecognised adverse events associated with the use of medicines. The simplest method for detecting signals of such events is crude inspection of lists of spontaneously reported drug-event combinations. Quantitative and automated numerator-based methods such as Bayesian data mining can supplement or supplant these methods. The theoretical basis and limitations of these methods should be understood by drug safety professionals, and automated methods should not be automatically accepted. Published evaluations of these techniques are mainly limited to large regulatory databases, and performance characteristics may differ in smaller safety databases of drug developers. Head-to-head comparisons of the major techniques have not been published. Regardless of previous statistical training, pharmacovigilance practitioners should understand how these methods work. The mathematical basis of these techniques should not obscure the numerous confounders and biases inherent in the data. This article seeks to make automated signal detection methods transparent to drug safety professionals of various backgrounds. This is accomplished by first providing a brief overview of the evolution of signal detection followed by a series of sections devoted to the methods with the greatest utilisation and evidentiary support: proportional reporting rations, the Bayesian Confidence Propagation Neural Network and empirical Bayes screening. Sophisticated yet intuitive explanations are provided for each method, supported by figures in which the underlying statistical concepts are explored. Finally the strengths, limitations, pitfalls and outstanding unresolved issues are discussed. Pharmacovigilance specialists should not be intimidated by the mathematics. Understanding the theoretical basis of these methods should enhance the effective assessment and possible implementation of these techniques by drug safety professionals.

Drug dependence associated with triptans and ergot derivatives: a case/non-case study

“…Several studies using the case/non-case method have been published on this database applying to different fields of drug safety [24][25][26]. One limitation of the pharmacovigilance database and consequently of the case/non-case method [20,26] lies in the selection and notoriety bias due to the spontaneous notification system, which depends on the cooperation and goodwill of healthcare professionals [25][26][27][28]. The association between a given drug and an SED may be artificially decreased if another specific SED is more often reported (and inversely, the association may be increased if there are only a few other SEDs associated with the study drug).…”

“…Case/non-case studies [20,21] based on database data rely on principles similar to case/control studies. This method may be used to generate signals from a pharmacovigilance database [21,22].…”

“…The association between drug exposure and SED was assessed by calculating the relative risk (RR), also called proportional reporting ratio [20,21]. As the number of drug dependence cases is much lower than the total number of SED reports recorded in the database, the relative risk may be evaluated by calculating the odds ratio (OR).…”

Data Mining Techniques in Pharmacovigilance: Analysis of the Publicly Accessible FDA Adverse Event Reporting System (AERS)

“…A signal usually requires that the lower 95% CI of the IC exceed zero [40]. A comprehensive review of the principles subtending calculation of Bayesian methods is beyond the aim of this chapter and the reader is referred to Hauben & Zhou [41] for sophisticated, yet intuitive discussion of this issue.…”

Informatic Tools and Approaches in Postmarketing Pharmacovigilance Used by FDA

“…The data must be evaluated to determine whether or not the disproportionately high adverse eventdrug product pair represents a true safety signal for the drug product (5). The biases present in the AERS database are still present when the database is "mined" for disproportionately high adverse event-drug product pairs, and these biases must be considered when evaluating the data.…”