How to sufficiently use the limited resources becomes controversial owing to aging. It is then important to know how to estimate medical cost accurately and efficiently. The naïve estimators ignoring the unobservable data may be biased owing to dropouts. Lin (1997) suggested partitioning the study duration and then constructing the estimate by summing up the cost from each interval. Furthermore, to take into account of the unobservable data, Lin (1997) and Band and Tsiatis (2000) proposed weighted estimators that used the survival probability and uncensored probability as the weight, respectively.

Furthermore, the medical cost may be related to many covariates. Baser (2006) suggested using the general linear model for the longitudinal data to model the partitioned cost, where a random intercept is included. This paper extends the model to a more general parametric model. Furthermore, we suggest using the survival probability as the weight adjustment which is estimated by the Cox proportional hazards model. This paper compares the performance of the unadjusted and adjusted cost estimators under various scenarios using simulations. Finally, the proposed model is implemented on the data extracted from Health Insurance database for patients with the colorectal cancer.

Keywords: Inverse probability; Medical cost; Proportional hazards model; Random effect model

Biography: I am a professor in Depeartment of Statisitcs in National Taipei University. My research interest includes survival data analysis, longitudinal data analysis and computational statistics.