PRM229 - Do We Need to Strengthen Study Design in Observational Studies?

Abstract

Guidance and standards are made available by the European Network of Centres for Pharmacoepidemiology and Pharmacovigilance (ENCePP) as reference tools for methodology and transparency of post-authorization studies. They are cited by the Guideline on good pharmacovigilance practices (GVP) Module VIII as relevant scientific guidances. Here we report an analysis conducted on Pharmacovigilance Risk Assessment Committee (PRAC) minutes regarding methodological issues. All publicly available minutes of PRAC meetings from July 2012 to March 2015 were considered. The number of submitted post-authorization safety study (PASS) protocols requiring endorsement was recorded (excluding advices), as well as endorsement/ refusal and reason. ENCePP documentation was consulted. From July 2012 to March 2015, 33 PASS protocols submitted to PRAC requiring endorsement were considered, increasing with years: 2 in 2012, 8 (17 considering resubmissions) in 2013, 18 (24 considering resubmissions) in 2014. The total number of evaluations was 54. Thirteen were endorsed at the first step (including endorsement with changes requested), while 20 required at least one amendment reaching approval up to 15 months after submission. The most common reasons for refusal concerned study design (20 cases), mainly reporting designs not allowing to fulfil study objectives (N=14). Furthermore, PRAC review asked for alternatives to reduce bias and confounding (N=2), simplifications aiming at reinforcing the observational nature of the study (N=1), further justification of sample size (N=1) and considerations on feasibility (N=1). From 2013 to 2014 an increasing number of endorsed PASS protocols was observed (7/17 and 12/24 respectively). During the same years ENCePP Checklist for Study Protocols and ENCePP Guide on Methodological Standards in Pharmacoepidemiology were reviewed, and the number of studies included in the ENCePP e-register increased. An increasing use of guidance and standards will allow strengthening robustness of design and results of observational studies.