Methods Applied to Assess Real-World Effectiveness of Drugs: A Scoping Review

Drug effectiveness evaluations are needed to bridge the gap between premarketing trials and clinical practice, but the specific methods used in such evaluations remain unclear. Therefore, this review aimed to identify pharmacoepidemiological effectiveness studies based on real-world data (RWD) and categorize them based on the methods applied and the drugs and outcomes investigated. We searched PubMed and Embase for relevant records published in the period July-December 2019. Eligible studies: (i) were RWD-based; (ii) evaluated one or multiple drug exposures, and (iii) included at least one effectiveness outcome (as primary outcome). Among 4820 identified records, 1129 passed title-abstract screening, 200 were randomly selected for full-text assessment, and 87 were ultimately included. Of these, 55 included > 1000 patients, 22 included > 10 000 patients, and 22 explicitly restricted to new users. Chemotherapy was the most common exposure (n = 32), all-cause mortality the most prevalent outcome (n = 52), and non-use of the exposure drug (n = 42) the most frequent comparator category. Survival models (n = 55) were the dominating statistical models, and 13 studies used only descriptive statistics. In conclusion, well-known pharmacoepidemiological methods were applied in the included studies, and many had large study populations. Common limitations were use of simple descriptive statistics and absence of active comparators.