Semiparametric transformation models for panel count data

Abstract

[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI AT REQUEST OF AUTHOR.] Panel count data arise in event history studies. It may not be feasible to monitor subjects continuously and recurrent events can be observed only at discrete time points rather than continuously and thus only the numbers of the events that occur between the observation times, not their occurrence times, are observed. The resulting interval-censored recurrent event data are commonly referred to as panel count data. The first part of this dissertation discusses a class of semiparametric transformation models for regression analysis of panel count data when the observation times or process may differ from subject to subject and more importantly, may contain relevant information about the underlying recurrent event. The second part of this dissertation will consider semiparametric transformation models for regression analysis of multivariate panel count data. The last part of the dissertation considers the same problem studied in Chapter 2 and provides an approach that allows both observation and follow-up times to be correlated with the recurrent event process. In all three parts, extensive simulation studies were conducted and indicate that the proposed approaches work well for practical situations. Illustrative examples are provided.