Systematic reviews of therapeutic interventions frequently consider patient-important outcomes

Abstract

ObjectivesTo determine whether recently published and ongoing systematic reviews of therapeutic interventions assess patient-important outcomes. Study Design and SettingFor this methodological review, we searched MEDLINE via PubMed for recently published systematic reviews and online registry of systematic reviews (PROSPERO) for ongoing systematic reviews. We selected systematic reviews with meta-analyses of randomized controlled trials. We extracted all outcomes defined in the methods section and categorized them. Mortality, other clinical events, pain, quality of life, function, and therapeutic decisions were considered patient-important outcomes. ResultsWe included 420 systematic reviews: 90 Cochrane reviews, 200 other published reviews, and 130 registered ongoing reviews. Primary outcomes were defined in 85 Cochrane reviews (95%), 98 (49%) other published reviews and all ongoing reviews. At least one patient-important outcome was defined as a primary outcome in 81/85 Cochrane reviews (95%), 78/98 other published reviews (80%), and 117/130 ongoing reviews (90%). Considering all outcomes assessed, at least one patient-important outcome was evaluated in 90/90 Cochrane reviews (100%), 189/200 other published reviews (95%), and 121/130 ongoing reviews (93%). ConclusionMost recent systematic reviews aim to assess patient-important outcomes, which contrasts with RCTs. These results suggest some important gaps between primary and secondary research.