清华大学、交通大学 统 计 学 研 究 所 专 题 演 讲 题 目: A Semi-Parametric Approach for Estimating Life Expectancy of Patients with A Specific Disease 主讲人: 黄景祥 (中央研究院 统计科学研究所) 时 间: 96年4月27日（星期五）10：40 - 11：30 （上午10：20-10：40茶会於统计所821室举行） 地 点: 清大综合三馆837室 Abstract The life expectancy of patients with a specific disease is not only a major concern of patients, but also essential to public health policies such as estimation of disease burden, resource allocation and national health insurance policy making. Two major determinants of disease specific life expectancy are background life expectancy and disease-associated mortality. The life table method can be used to calculate the life expectancy of a population at a given age, but not for the life expectancy of patients with variable age-of-disease onset. Parametric survival modeling is a popular approach. However, it is very difficult, if not impossible, to take into account mathematically the two major determinants of disease specific life expectancy with a specific functional formula. Long-term estimate is often unreliable if there is a high rate of right-censoring. We propose a relative survival extrapolation approach with information borrowed from national vita statistics through Monte Carlo simulation to estimate life expectancy and years of life loss for patients with a specific disease. The proposed method has been evaluated and successfully applied to real data collected in Taiwan, including patients with major cancers , newly diagnosed HIV infection in the era of highly active antiretroviral therapy and workers with permanent occupational disabilities. The estimated life-long survival can also be applied to integrate with utility functions such as quality of life to obtained expected quality-adjusted survival. --

※ 发信站: 批踢踢实业坊(ptt.cc) ◆ From: 140.113.109.59 ※ 编辑: mangogogo 来自: 140.113.109.59 (04/26 12:34)