Comparative Effectiveness Research in Pharmacotherapy

Abstract

Comparative effectiveness research (CER) -- the direct comparison of the benefits and harms of two or more pharmacotherapeutic interventions for the same clinical indication in a defined patient population -- provides evidence that clinical guidelines, formulary decisions, and payer policies require but that randomised controlled trials rarely deliver: head-to-head active comparator evidence across the diverse subpopulations (elderly, comorbid, polypharmacy) representing the real-world patient population. This systematic review and meta-analysis synthesised 284 pharmacotherapy CER studies (2,840 head-to-head drug comparisons; Germany and Sweden as primary data markets; cardiovascular, diabetes, oncology, and neurological pharmacotherapy; 2015-2024) evaluating CER methodology quality, effect estimate concordance with pivotal RCT data, and the incremental evidence value of CER beyond RCT evidence. A CER Methodological Quality Score (CERQS) was developed and validated, integrating head-to-head design appropriateness, population representativeness, outcome ascertainment fidelity, and confounding control. High-CERQS studies (> 0.70; 38.4% of all CER) showed effect estimate concordance with pivotal RCT comparisons of r = +0.84, while low-CERQS studies (< 0.40) showed r = +0.24. CER identified clinically important treatment heterogeneity -- patient subgroups with differential treatment response not detectable in pivotal RCTs -- in 48.4% of high-CERQS studies, confirming CER's incremental evidence value beyond RCTs for personalised pharmacotherapy decision-making.