Interim analysis is typically planned before the trial begins, including specifying the timing and statistical methods to be used. The process involves: - Predefined interim endpoints - Blinded data review to avoid bias - Statistical adjustments to account for multiple looks at the data, often using alpha-spending functions to control the overall type I error rate
