C Wayant1, D Margalski2, K Vaughn3, M Vassar4. 1. Oklahoma State University Center for Health Sciences, Department of Biomedical Sciences, 1111 West 17th Street, Tulsa, OK, 74107, USA. Electronic address: cole.wayant@okstate.edu. 2. Wake Forest Baptist Medical Center, Department of Internal Medicine, 1 Medical Center Blvd, Winston-Salem, NC, 27157, USA. Electronic address: daniel.margalski@okstate.edu. 3. University of Oklahoma School of Community Medicine, Department of Medicine/Pediatrics, 4502 E. 41st Street, Tulsa, OK, 74135, USA. Electronic address: kaleb.vaughn@okstate.edu. 4. Oklahoma State University Center for Health Sciences, Department of Psychiatry, 1111 West 17th Street, Tulsa, OK, 74107, USA. Electronic address: matt.vassar@okstate.edu.
Abstract
PURPOSE: Spin, the misrepresentation of research findings, in clinical trial abstract has been shown to influence how oncologist rate a drug's efficacy. MATERIALS AND METHODS: We searched PubMed for clinical trials published in ten key journals in 2017. Our primary objectives were to assess the frequency and manifestations of spin in the abstracts of those clinical trials that measured both overall survival and at least one surrogate efficacy endpoints. RESULTS: 124 trials were included for analysis. We found evidence of spin in 46 of 124 (37.1%, 95% CI 29.1%-45.9%) trial abstracts. Spin in the abstract results was most often due to authors emphasizing a statistically significant subgroup analysis (n = 6). Spin in the abstract conclusions was most often due to authors relying on a statistically significant surrogate endpoint to highlight the bioefficacy of the intervention (n = 17). CONCLUSION: Spin is prevalent in the abstracts of oncology clinical trials that measure OS and a surrogate endpoint. The conclusion sections of abstracts were most prone to contain spin. When OS was the primary endpoint, spin was primarily used to distract from the nonsignificant OS data. To mitigate unintentional hype for cancer therapies, we recommend authors structure their conclusions around patient-important outcomes.
PURPOSE:Spin, the misrepresentation of research findings, in clinical trial abstract has been shown to influence how oncologist rate a drug's efficacy. MATERIALS AND METHODS: We searched PubMed for clinical trials published in ten key journals in 2017. Our primary objectives were to assess the frequency and manifestations of spin in the abstracts of those clinical trials that measured both overall survival and at least one surrogate efficacy endpoints. RESULTS: 124 trials were included for analysis. We found evidence of spin in 46 of 124 (37.1%, 95% CI 29.1%-45.9%) trial abstracts. Spin in the abstract results was most often due to authors emphasizing a statistically significant subgroup analysis (n = 6). Spin in the abstract conclusions was most often due to authors relying on a statistically significant surrogate endpoint to highlight the bioefficacy of the intervention (n = 17). CONCLUSION:Spin is prevalent in the abstracts of oncology clinical trials that measure OS and a surrogate endpoint. The conclusion sections of abstracts were most prone to contain spin. When OS was the primary endpoint, spin was primarily used to distract from the nonsignificant OS data. To mitigate unintentional hype for cancer therapies, we recommend authors structure their conclusions around patient-important outcomes.