Regression analysis of clustered interval-censored failure time data with informative cluster size

Abstract

[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI AT REQUEST OF AUTHOR.] Correlated or clustered failure time data often occur in many research fields including epidemiological, geographical, sociological and medical studies. Sometimes such data arise together with interval censoring and the failure time of interest may be related to the cluster size. Various approaches have been proposed to analyze failure time data with interval censoring. However, these approaches ignore the informativeness of the cluster size. Due to the lack of proper inference procedures for direct analysis, these methods merely simplified or converted interval-censored data into right-censored data, which inevitably resulted in biased parameter estimates. In this dissertation, both parametric and semiparametric approaches are presented for regression analyses of clustered failure time data that allow both interval-censoring and informative cluster size. We further validate these approaches by conducting various simulation studies and apply them to a lymphatic filariasis example.