Pharmaceutical innovation is often portrayed as a learning process characterized by uncertainty reduction and risk quantification over the product life-cycle. Knowledge on adverse drug reactions (ADRs) is often limited upon market entry and continues to accumulate once pharmaceuticals are used in clinical practice based on feedback from users and coordinated transfer of knolwedge between system actors. However, while knowledge accumulation on ADRs is actively pursued by institutional actors through pharmacovigilance activities, there is limited understanding of the institutional factors that facilitate ADR discovery. This study therefore aimed to provide insight in the factors associated with the discovery of new ADRs once innovative pharmaceuticals enter the market.
We conducted a longitudinal cohort study of all innovative drugs authorized by the European Commission in the period 1995-2016 (n=465). A web-crawler was built to extract information on all post-marketing procedures that resulted in a license change as indicated by a product label update. A fuzzy text matching algorithm was used to extract all listed ADRs and indications from each label version and to determine at what moment during the life-cycle new ADRs and indications were added to the label. Pharmaceuticals were followed from approval up to 01-01-2018 or market withdrawal, with median follow-up time being > 8 year. ADRs and indications were mapped onto the standardized language of the Medical Dictionary for Regulatory Activities to facilitate comparisons between pharmaceuticals and over time.
Preliminary results show that during follow-up, there were 2,160 additions of one or more ADRs to drug labels, corresponding to a median [IQR] of 3 [1-6] additions over the life-cycle. ADRs were most likely to be added in the second to sixth year after drug approval, while ADR addition rapidly declined after generic entry of competitors. We estimated a panel logistic model predicting the moment of ADR addition. Results demonstrate that once controlled for drug and year-specific fixed effects, ADR addition was significantly more likely following approval of new indications and significantly less likely following generic entry of competitors. These findings suggest that the exploration of new markets by pharmaceutical companies provides a major incentive to discover and report new ADRs.