Evidence of a Cohort Effect on Survival Function for Patient with Catastrophic Illnesses Based on NHIRD Uremia Patients

• Main Authors: Yin-Ju Lin, 林音汝
• Other Authors: Shing-Her Juang
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/36673338906516262914

碩士 === 東吳大學 === 商用數學系 === 95 === 　　Possibly due to a birth cohort effect, the mortality rate of catastrophic illnesses in Taiwan seems to be improving. This paper offers evidence of this effect through examining survival time and incidence among end stage renal disease (ESRD) patients gathered from Taiwan’s National Health Insurance Data Base (NHIRD). Data from nearly 9 years – ranging from December 2003 to March 1995 – was extracted from the NHIRD. As uremia typically affects elderly patients, the age-at-onset for this study has been observed to be 60 to 74 years old, with patients born from 1926 to 1937. 　　For survival time of ESRD patients, we observe the followings: (1) patients in the later cohorts display higher survival rates; (2) as age of disease onset increases, so does survival function; (3) female patients display better survival rates than male patients; and, (4) a three-factor – birth cohort, age-at-onset, and sex – and no interaction accelerated failure-time model is effective for building models for survival time. For the incidence of ESRD, a probit model with interaction terms indicates that the birth cohort, age, and sex are important factors for the incident rates.