Re-Analysis of X:BOT using Public Access Data Files

We had been developing new statistical theories and methods to address differential dropouts during trial enrollment, as part of a dissertation project at the University of North Carolina Gillings School of Global Public Health, in the Department of Biostatistics (Elizableth Ajazi, Committee Chair: Todd Schwartz, Advisor: Steve Marshall). The applied example we been working on uses recent publicly released data files from NIDA-CTN-0051 (ClinicalTrials.gov NCT02032433). The primary study report was published in The Lancet on January 27, 2018, entitled "Comparative effectiveness of extended-release naltrexone versus buprenorphine-naloxone for opioid relapse prevention (X:BOT): a multicentre, open-label, randomised controlled trial" by Lee et al. We were able to replicate the findings for the primary endpoints on efficacy. However, our attempt to replicate the primary safety outcome, overdose, failed. Suspecting we were not replicating the study properly, we examined the adverse event data meticulously. To our surprise we found more patients with overdose in the public file than in the published report.