Baseline EDSS proportions in MS clinical trials affect the overall outcome and power: A cautionary note. (June 2017)

• Record Type: Journal Article
• Authors: Wang, Guoqiao; Cutter, Gary R; Cofield, Stacey S; Lublin, Fred; Wolinsky, Jerry S; Gustafson, Tarah; Krieger, Stephen; Salter, Amber
• Is Part Of: Multiple sclerosis. Volume 23:Number 7(2017)
• Page Start: 982
• Page End: 987
• Publication Date: 2017-06
• Subjects: Multiple sclerosis -- Expanded Disability Status Scale -- relapsing-remitting multiple sclerosis -- confirmed disability progression rate; Central nervous system -- Diseases -- Periodicals; Myelin sheath -- Diseases -- Periodicals; Inflammation -- Periodicals; Multiple sclerosis -- Periodicals; Central Nervous System Diseases -- Periodicals; Demyelinating Diseases -- Periodicals; Inflammation -- Periodicals; Multiple Sclerosis -- Periodicals; Système nerveux central -- Maladies -- Périodiques; Gaine de myéline -- Maladies -- Périodiques; Inflammation (Pathologie) -- Périodiques; Sclérose en plaques -- Périodiques; Electronic journals; 616.834005
• DOI: 10.1177/1352458516670733
• Languages: English
• ISSNs: 1352-4585

Abstract

Background: In randomized clinical trials, when treatments do not work equally effectively across stratifications of participants, observed event rates may differ from those hypothesized leading to deviations in estimated power. Objectives: To investigate the effect of distributions of baseline Expanded Disability Status Scale (EDSS) proportions in relapsing-remitting multiple sclerosis (RRMS) on the trial outcome, confirmed disability progression rate (CDPR), and power. Methods: We reported CDPRs in the CombiRx trial by baseline EDSS and by groups (1st (0, 1), 2nd (1.5, 2), 3rd (2.5, 3), and 4th (⩾3.5)) and investigated the effect of different combinations of baseline EDSS proportions on the trial outcome and power. Results: There were 244 (25.4%) participants in the 1st group, 368 (38.4%) in the 2nd group, 223 (23.3%) in the 3rd group, and 124 (12.9%) in the 4th group with CDPRs of 40.1%, 13.9%, 11.2%, and 16.9%, respectively. Both CDPR and power increased when the proportion of the 1st group increased in hypothetical trials with equal sample sizes in each arm, and a 10% increase in the 1st group led to a 5% increase in power. Conclusion: Various baseline EDSS proportions yielded different CDPRs and power, suggesting caution in interpretation of treatment effects across trials that enrolled participants with different proportions of baseline EDSS.