Assessing Impact of Pharmacovigilance Interventions in Europe: Methods and Applications

Le résumé fourni par la source

This thesis examines how the impact of pharmacovigilance interventions is assessed in Europe, with a specific focus on the identification, measurement, and interpretation of unintended effects. It combines a systematic literature review with empirical case studies based on routinely collected healthcare data and concludes with a conceptual framework to strengthen regulatory evaluation. Although pharmacovigilance operates within a harmonised European legal framework and is supported by the Guidelines on Good Pharmacovigilance Practices, adverse drug reactions remain a substantial and partly preventable cause of morbidity, mortality, and healthcare expenditure. Risk Minimisation Measures are mandated across the EU, yet their effectiveness depends on implementation within heterogeneous national healthcare systems. Another challenge is unintended impact, as interventions targeting specific risks may alter prescribing behaviour more broadly. Such effects may reflect harm, benefit, or rational adaptation and require contextual interpretation. The literature review in Chapter 2 demonstrates that unintended effects are infrequently and inconsistently assessed in published impact studies, despite explicit regulatory guidance. Of 96 identified studies published between 2012 and 2022, only 23 evaluated unintended impacts. Methods and outcome measures varied substantially, with most studies relying on drug utilisation metrics as proxies for behavioural change and few examining clinical or system-level outcomes. The empirical component of the thesis comprises three studies evaluating regulatory and policy interventions using real-world data and longitudinal quasi-experimental designs. An evaluation of European and national measures targeting fluoroquinolone prescribing in primary care in the Netherlands and the United Kingdom found an initial reduction in prescribing following implementation in 2018, but no sustained additional decline after further measures in 2020. No consistent substitution with alternative antibiotics was observed, although prescribing persisted among patients with identifiable risk factors. An assessment of risk minimisation measures to address the cardiovascular risks of diclofenac in the United Kingdom showed a substantial decline in initiation following implementation in 2013. However, reductions occurred across risk groups, continued prescribing to patients with contraindications was observed, and treatment initiation was delayed, suggesting broader behavioural effects beyond the intended target population. A national policy introduced in Lithuania in 2020 to promote the rational use of benzodiazepines and related drugs was associated with short-term stabilisation and reduced dispensing volume, yet prevalence and long-term use increased between 2018 and 2024. In addition, a comparative analysis of Estonia, Latvia, and Lithuania demonstrated marked cross-national differences in the incidence and therapeutic intensity of benzodiazepines and related drugs, highlighting the influence of baseline prescribing patterns and national contexts on observable intervention effects. The thesis concludes that impact assessments based primarily on prescribing or dispensing data provide an incomplete basis for evaluating pharmacovigilance interventions. Unintended effects remain underexplored and are inherently difficult to interpret without adequate clinical and system-level context. The thesis proposes a systems thinking approach, conceptualising healthcare as a complex adaptive system in which regulatory actions interact with existing practices, incentives, and constraints. Integrating contextual and system-level perspectives into study design and interpretation is essential for improving the validity, relevance, and policy utility of pharmacovigilance impact assessments in Europe.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.