Multistate models provide a natural and powerful framework for characterizing life history processes. They can play a central role in medical decision making in complex diseases where the prediction of outcomes is necessary under different treatment strategies, and in health economics where the specification of health policies may be based on projections of long-term costs. Conceptualizing processes in terms of multistate models is also useful when considering the impact of sample selection schemes and complex observation processes, both of which must be addressed in observational studies to avoid biased inferences.

Recurrent event data arise in fields such as medicine and public health, business and industry, reliability, the social sciences, and insurance. The literature on the statistical analysis of recurrent events has grown rapidly over the past twenty years and a variety of models and methods has been developed. This book provides a comprehensive treatment of the area. We describe important models, explain their underlying assumptions and properties, consider settings where they are appropriate, and discuss in detail how to fit and base inferences on these models. Parametric, nonparametric, and semiparametric methods are covered. Many illustrative examples are given, most of which are taken from health or industrial settings.