Adam G Dunn1, Liat Orenstein2, Enrico Coiera3, Kenneth D Mandl4, Florence T Bourgeois5. 1. Centre for Health Informatics, Australian Institute of Health Innovation, Macquarie University, Sydney, NSW, Australia; Computational Health Informatics Program, Boston Children's Hospital, Boston, MA, USA. Electronic address: adam.dunn@mq.edu.au. 2. Computational Health Informatics Program, Boston Children's Hospital, Boston, MA, USA. 3. Centre for Health Informatics, Australian Institute of Health Innovation, Macquarie University, Sydney, NSW, Australia. 4. Computational Health Informatics Program, Boston Children's Hospital, Boston, MA, USA; Department of Biomedical Informatics, Harvard Medical School, Boston, MA, USA. 5. Computational Health Informatics Program, Boston Children's Hospital, Boston, MA, USA; Division of Emergency Medicine, Boston Children's Hospital, Boston, MA, USA; Departments of Pediatrics and Emergency Medicine, Harvard Medical School, Boston, MA, USA.
Abstract
OBJECTIVE: To determine whether certain trial characteristics are associated with faster or more frequent inclusion in systematic reviews for drug interventions in type 2 diabetes. STUDY DESIGN AND SETTING: We examined trials included in systematic reviews published between January 1, 2007 and January 1, 2017. Primary outcomes were time between trial publication and first inclusion in a systematic review and frequency of inclusion in systematic reviews over the study period. Multivariable Cox proportional hazards and regression models quantified associations with funding source, number of participants, trial conclusion, and journal impact factor. RESULTS: Among 668 trials, the median time to inclusion was 76.1 weeks. Time to inclusion was shorter for trials with industry funding (hazard ratio [HR] 1.39; 95% confidence interval [CI] 1.13-1.71), more participants (HR 1.26; 95% CI 1.17-1.36), and published in higher impact factor journals (HR 1.28; 95% CI 1.14-1.45). The median frequency of inclusion was three. Frequency of inclusion was greater for trials with industry funding (relative risk [RR] 2.36; 95% CI 2.11-2.64), more participants (RR 1.51; 95% CI 1.47-1.55), positive conclusions (RR 1.89; 95% CI 1.68-2.13), and published in higher impact factor journals (RR 1.13; 95% CI 1.08-1.18). CONCLUSION: Certain trial characteristics are associated with faster or more frequent trial inclusion in systematic reviews of type 2 diabetes.
OBJECTIVE: To determine whether certain trial characteristics are associated with faster or more frequent inclusion in systematic reviews for drug interventions in type 2 diabetes. STUDY DESIGN AND SETTING: We examined trials included in systematic reviews published between January 1, 2007 and January 1, 2017. Primary outcomes were time between trial publication and first inclusion in a systematic review and frequency of inclusion in systematic reviews over the study period. Multivariable Cox proportional hazards and regression models quantified associations with funding source, number of participants, trial conclusion, and journal impact factor. RESULTS: Among 668 trials, the median time to inclusion was 76.1 weeks. Time to inclusion was shorter for trials with industry funding (hazard ratio [HR] 1.39; 95% confidence interval [CI] 1.13-1.71), more participants (HR 1.26; 95% CI 1.17-1.36), and published in higher impact factor journals (HR 1.28; 95% CI 1.14-1.45). The median frequency of inclusion was three. Frequency of inclusion was greater for trials with industry funding (relative risk [RR] 2.36; 95% CI 2.11-2.64), more participants (RR 1.51; 95% CI 1.47-1.55), positive conclusions (RR 1.89; 95% CI 1.68-2.13), and published in higher impact factor journals (RR 1.13; 95% CI 1.08-1.18). CONCLUSION: Certain trial characteristics are associated with faster or more frequent trial inclusion in systematic reviews of type 2 diabetes.
Keywords:
Clinical trial as a topic; Industry funding; Randomized controlled trials; Systematic review as a topic; Systematic review biases; Type 2 diabetes