Hu X Joan, Lorenzi Maria, Spinelli John J, Ying S Celes, McBride Mary L
Department of Statistics and Actuarial Science, Simon Fraser University, Burnaby, BC V5A 1S6, Canada.
Lifetime Data Anal. 2011 Apr;17(2):215-33. doi: 10.1007/s10985-010-9183-8. Epub 2010 Aug 22.
In an attempt to provide tools for assessing hospital utilization, this paper extends well-known models for recurrent events to address non-negligible event duration and presents a procedure for estimating the model parameters. The model extension is natural and easy to understand. Asymptotic properties of the associated inferences are derived adapting the well-developed methods based on the counting process formulation. Several specifications of the proposed modeling are illustrated with the hospitalization records of childhood cancer survivors from a health care insurance system that motivated this research. The usefulness and robustness of the proposed approach is demonstrated numerically via simulation.
为了提供评估医院利用率的工具，本文扩展了用于复发事件的著名模型，以处理不可忽略的事件持续时间，并提出了一种估计模型参数的程序。模型扩展自然且易于理解。基于计数过程公式，采用成熟的方法推导了相关推断的渐近性质。通过一个激发本研究的医疗保险系统中儿童癌症幸存者的住院记录，说明了所提出模型的几种规格。通过模拟从数值上证明了所提方法的有用性和稳健性。
