Asian Journal of Probability and Statistics

In most clinical trials, binary outcomes—such as success or failure and presence or absence of side effects—are used to evaluate treatment safety and efficacy. Determining the Maximum Safe Dose (MSD) is essential, as it identifies the highest drug dosage that does not cause harmful effects. Exceeding the MSD can pose serious health risks and undermine the overall benefits of the treatment. This article proposed a confidence interval procedure designed to simplify the complex analysis of binary endpoints as discussed in Thall et al. (2008). To solve this problem, we apply a confidence interval process along with a partitioning approach. therefore, application of reliable statistical approaches in establishing and confirming the range of safe dosages is imperative. It involves thorough examination of data in preclinical as well as clinical trials in the hopes of reducing side effects and optimizing the efficacy of treatment. Thus, our paper introduces a confidence interval approach for estimating MSDs for drugs using binary endpoints. This method was performed using a 100(1 - α)% Wilson (1927) score interval with step-up method for binary endpoints without multiplicity adjustment. We illustrate it through the examples which were published by Neuhauser and Hothorn (1997) in their paper. Additionally, we also observed that this method’s power increases with increasing sample size. Finally, our simulation results shows that Wilson score interval proved to have the shortest length and producing good coverage probability. The results further showed that our newly constructed procedure control the familywise error rate. We advocate that our newly constructed procedure with wilson score interval is suitable for demonstrating MSD when the respose have binary outcomes.