(239) Performance of Subgrouped Proportional Reporting Ratios in the US Food and Drug Administration (FDA) Adverse Event Reporting System

Publication Number: 1284

Background: Signal detection algorithms are statistical methods used to measure disproportionality and identify signals in adverse drug event (ADE) databases. Many signal detection algorithms give the same weight to information from all products and patients, which may result in signals being masked or false positives being flagged as potential signals. Subgrouped analysis, which is generally used in epidemiology, can be used to help correct for this. This is beneficial in pharmacovigilance to identify potential safety concerns previously missed.

Objectives: The aim of this study was to evaluate PRR analyses subgrouped by either age, sex, ADE report type, seriousness of ADE, or reporter compared to crude PRR analysis.

Methods: The publicly available US Food and Drug Administration Adverse Event Reporting System (FAERS) quarterly data extract files from January 1, 2015 through September 30, 2017 were utilized. The Demographic, Drug, Outcome, and Reaction files were used in this study. The files were linked on the primary ID. A deduplication procedure was performed by selecting the highest primary ID for each report. Only the primary suspect drug from a report was included in the analysis. Reports missing a primary suspect drug or that did not list an ADE were excluded. Generic names were used to identify drugs, and all ADEs and drug names were converted to upper case text for standardization. A PRR analysis subgrouped by either age, sex, ADE report type, seriousness of ADE, or reporter was compared to the crude PRR analysis against a reference set. The reference set included 155 drug-ADE pairs from 60 drugs with 110 positive controls and 45 negative controls. A possible signal was defined by the accepted thresholds of PRR ≥ 2, number of reports ≥ 3, and a χ 2 ≥ 4. All drug-ADE pairs in FAERS during the study period were included in the analysis. The sensitivity, specificity, precision, and c-statistic for the PRR analyses were compared.

Results: Subgrouping by age (n = 78, 34.5% increase) and sex (n=67, 15.5% increase) identified the most signals compared to the crude analysis (n = 58). Subgrouping by report type or seriousness resulted in fewer signals (n = 50 for both). Subgrouping by either age or sex increased both the sensitivity and precision. Subgrouped analyses had higher c-statistic values, and the age subgrouped analysis had the highest (0.468).

Conclusions: Subgrouping by either age or sex increased both the sensitivity and precision compared to the crude PRR analysis, which resulted in identifying a higher percentage of signals that represent true ADEs. Age (children and elderly) and sex (females) increase the risk of a patient experiencing an ADE, and subgrouping by these variables unmask potentially important associations.