A comparative study of selected nonparametric multi-state approaches for estimating the survival function

In this paper, a comparative study is made on some selected nonparametric multistate approaches along with a parametric approach to analyze competing risks and multiple failure data. The objective of this paper is to find a suitable nonparametric approach for analyzing follow-up data. In the study, the asymptotic efficiency of the Kaplan-Meier estimator relative to a parametric estimator of survival function and nonparametric survival functions are examined in the presence of competing risks. Finally, an application of these approaches was shown using diabetes mellitus data in which different states of complications made available that occurred once after the diabetes were detected were made available.