| Popis: |
Importance: In randomized controlled trials (RCTs), small sample size and lack of blinding can cause biased and spurious results. Whether and how study characteristics differ based on a trial's funder is an important area to study. Objective: To compare study characteristics of RCTs funded by industry with study characteristics of RCTs not funded by industry. Design, Setting and Participants: We systematically reviewed all RCTs published between 2015 and 2019 in the New England Journal of Medicine (NEJM), Lancet, and Journal of the American Medical Association (JAMA). Our primary data sources were ClinicalTrials.gov and MEDLINE. Data extraction included manual review and use of natural language processing. Main Outcomes and Measures: We compared the rate of blinding, use of placebo, and sample size. We used natural language processing to analyze the sentiment of the study's conclusion as reported in the abstract. As proxies for knowledge dissemination, we calculated the AltMetric scores and number of times the article was cited (citation count). Results: We identified 1533 RCTs published by NEJM, Lancet, and JAMA between 2015 and 2019. Of these RCTs, 697 were funded by industry. Trials funded by industry were more likely to be blinded (n=378, 54% vs n=318, 38%), more likely to include a placebo (n=317, 45% vs n=196, 23%), more likely to post their results on ClinicalTrials.gov (78%, 443 of 570 vs 41%, 207 of 501) compared to trials that were not industry funded. Industry–funded RCTs had a smaller sample size than non–industry–funded RCTs (median=557 [IQR: 230, 1369] vs 648 [IQR: 301, 1916], P |
