| Review: |
Time-to-event data are often analysed for sample units that experience only one event: death; complete break down of a machine, etc. Recurrence events data analysis looks at repeated events, where a sample unit may accumulate over time, such as the number and repeated cost of machine repairs, or the repeated cost of recurrent disease episodes. This book introduces non-parametric methods for such data, in particular, the plot of the non-parametric estimate of the population mean cumulative function (MCF), which yields most of the information that is sought. Topics covered include: recurrence data and information sought; the basic non-parametric population model; estimating, plotting and interpreting the sample MCF; confidence limits for the MCF; analysis of a mix of events; comparison of samples; and a survey of related topics. |