An omnibus CUSUM chart for monitoring time to event data.

A non-parametric method is proposed for monitoring time-to-event data. A cumulative sum chart is constructed that is able to detect an unknown out-of-control state. This method exploits the absolute differences between the Kaplan-Meier estimator and the in-control distribution over specific time intervals. The efficiency of the algorithm is studied via a simulation and a real data study. The new method is also tested via the simulation study against existing methods.