Interim Analysis Using Machine Learning

The purpose of this article is to recalculate sample size by adjusting the initial design about trial results and taking uncertainty in power estimations into consideration throughout study preparation. This article examines two primary methods: Pocock and O'Brien/Fleming using R programming. Interim data analysis must be done far in advance to provide enough power and minimize the risk of having Type I errors. These results suggest that robust performance indicators have enabled improved comprehension of the study's development, notably increased control, more efficient use of resources (e.g., monitoring time), quicker enrolment, shorter operational timelines, and less expensive costs.