Methodological quality of oncology noninferiority clinical trials. (May 2020)

• Record Type: Journal Article
• Authors: Wayant, Cole; Ross, Andrew; Vassar, Matt
• Is Part Of: Critical reviews in oncology/hematology. Volume 149(2020)
• Publication Date: 2020-05
• Subjects: Oncology -- Noninferiority trial -- Clinical trial -- Quality -- Epidemiology; Oncology -- Periodicals; Hematology -- Periodicals; 616.994
• DOI: 10.1016/j.critrevonc.2020.102938
• Languages: English
• ISSNs: 1040-8428

Abstract

Graphical abstract: Highlights: Included noninferiority trials of oncology interventions had key methodological shortcomings. Approximately 15 % of included trials did not contain an alpha value and confidence intervals that matched. Approximately 28 % did not justify the chosen noninferiority margin. Only 13 % of trials provided enough information to calculate the amount of preserved effect size for the chosen margin. Abstract: Introduction: Noninferiority trials can show that new treatments with slightly less efficacy are safer, cheaper, or easier to administer. However, the conclusions of noninferiority trials depend on robust methodology. Methods: We conducted a 6 year cross-sectional investigation of the methodological quality of oncology noninferiority trials published in the top 10 oncology journals. Four key quality criteria were investigated. Results: Nonefficacy benefits of the new treatment were stated in 88/110 (80.0 %) trials. Justification for the noninferiority margin was provided in 79/110 (71.8 %) trials. Authors most often used previous data as justification for the chosen margin (n = 42). In 15 noninferiority trials the percent preserved effect could be calculated and the median effect preserved was 56.8 %. Conclusions: The oncology noninferiority trials included in our study had key methodological shortcomings, counterbalanced by a clear delineation of expected nonefficacy benefits of the new treatment.