Nonparametric Prediction of Event Times for Analysis of Failure-Time Data

In trials with failure-time outcomes, statistical information is determined by accumulated events. Interim and final analyses are performed after a prespecified number of events have been observed. It is of interest to predict when a prespecified number of events will be observed based on accumulating data. We propose a fully Bayesian nonparametric approach in modeling the survival probabilities. We compare the accuracy and precision of this approach to proposed parametric and semi-parametric methods. In summary, the proposed method offers greater flexibility and based on our studies has the ability to match or outperform existing methods.