This repository contains code to implement the approximate predictive probability. Primary functions calculate a predictive probability based on interim data from a z-score, p-value, or posterior probability. Helper functions assist in interim information calculation and conversions.
Additional examples and the complete derivation are available on arXiv as Predictive Probabilities Made Simple: A Fast and Accurate Method for Clinical Trial Decision Making.
For some endpoints, information for the predictive probability is best specified as sample size, while others are best defined by exposure time or events. Please see the table below for details.
