Advances in Ontology--Based Mining of Adverse Drug Reactions

TL;DR

This paper introduces an ontology-based statistical model for improved detection of adverse drug reactions in pharmacovigilance, effectively accounting for AE similarities and zero-inflation, and proposes a data splitting method for reliable evaluation.

Contribution

It presents a novel integration of AE ontology into a zero-inflated negative binomial model and introduces a data thinning technique for better model validation in pharmacovigilance.

Findings

Ontology-based model outperforms classical models like GPS.

Data thinning and stratified splitting improve evaluation reliability.

Method effectively accounts for AE similarities and zero counts.

Abstract

Post--marketing pharmacovigilance is essential for identifying adverse drug reactions (ADRs) that elude detection during pre--marketing clinical trials. This study explores a novel approach that integrates an adverse event (AE) ontology into a zero--inflated negative binomial model to improve ADR detection. By accounting for the biological similarities among correlated AEs and addressing the excess of zero counts, this method more effectively disentangles AE associations. Statistical significance is evaluated using a permutation--based maximum statistic that preserves AE correlations within individual reports. Simulations and an application to real data from the Veneto drug safety database demonstrate that the ontology--based model consistently outperforms classical models such as the Gamma--Poisson Shrinker (GPS). For post--selection inference, we furthermore explore a data thinning technique for convolution--closed families, enabling the creation of independent training and validation datasets while retaining all drug--AE pairs. This approach is compared with conventional random train/test splitting, which may leave some drugs or AEs absent from one subset, and stratified splitting, which requires expanding aggregated counts into individual instances. The data--thinning technique and stratified splitting yield very similar results, with stratified splitting showing a slight benefit, and both clearly outperform random splitting in ensuring reliable and consistent model evaluation.