Rationale and aim: Over the past 12 years, thousands of authors working with the Cochrane Collaboration around the world have produced systematic reviews to reduce uncertainty in health care decision making. We evaluated the conclusions from Cochrane systematic reviews of randomized controlled trials in terms of their recommendations for clinical practice and research.
Methods: In our cross-sectional study of systematic reviews published in the Cochrane Library, we randomly selected and analysed completed systematic reviews published across all 50 Cochrane Collaborative Review Groups.
Results: We analysed 1016 completed systematic reviews. Of these, 44% concluded that the interventions studied were likely to be beneficial, of which 1% recommended no further research and 43% recommended additional research. Also, 7% of the reviews concluded that the interventions were likely to be harmful, of which 2% did not recommend further studies and 5% recommended additional studies. In total, 49% of the reviews reported that the evidence did not support either benefit or harm, of which 1% did not recommend further studies and 48% recommended additional studies. Overall, 96% of the reviews recommended further research.
Conclusions: Cochrane systematic reviews were about evenly split between those in which the authors concluded that at least one of the interventions was beneficial and those in which the evidence neither supported nor refuted the intervention tested. The Cochrane Collaboration needs to include clinical trial protocol summaries with a study design optimized to answer the relevant research questions.