Abstract: Competing risks data arise when study subjects may experience several different types of failure. It is common that the cause of failure is missing due to various reasons. Analysis of competing risks data with missing cause of failure has received considerable attention recently (Goetghebeur and Ryan (1995), Lu and Tsiatis (2001), Gao and Tsiatis (2005), among others). In this article, we study the semiparametric additive hazards model for analysis of competing risk data with missing cause of failure. Different estimating equation approaches using the inverse probability weighted and double robust techniques are proposed for estimating the regression parameters of interest. The resulting estimators have closed forms and their theoretical properties are established for inference. Simultaneous confidence bands of survival curves are constructed using a resampling technique. Simulations and an example show that the proposed approach is appropriate for practical use.
Key words and phrases: Additive hazards model, Competing risks data, Double robust, Estimating equation, Inverse probability weight, Missing cause of failure.