ANALYSIS OF THE NUMBER OF DAYS REQUIRED FOR FULL-TIME CARE BEFORE DEATH AND THE COST OF CARE IN THE LAST MONTH OF LIFE AMONG THE CHINESE ELDERLY pp. 51-79
Authors: (Danan Gu, Medical Center, Duke Univ., Durham, NC, Donghui Gu, Dept. of Social Work, Fudan Univ., Shanghai, Jie Zhou, Beixinjing Community Center for Health Service, Changning District, Shanghai)
Abstract: A growing number of studies in western nations have found that an important component in improving the quality of end of life is to meet the care needs of individuals approaching death. Information on the cost of these care needs would be useful in determining how to reduce the burden on family and society, but such studies are near-absent in developing countries. This chapter provides an analysis of care needs and expenses before death in China, where the proportion of the elderly population will rapidly increase in the next few decades. Based on the fourth wave of the Chinese Longitudinal Healthy Longevity Survey (CLHLS) in 2005, we estimate the number of days needed for full-time care before death and the care costs in the last month of life among the deceased elderly aged 65 and over by using sequential models in a multi-level context. The CLHLS, a unique specially designed nationwide survey, included data on care needs before death and expenditure in 2005 for those respondents who died between the 2002 and 2005 waves in randomly selected half of counties and cities in 22 of 31 provinces in mainland China. Data on the number of days required for full-time care and care costs in the last month of life were obtained from 5,729 deceased persons in 774 counties by asking their next-of-kin. Both the number of days requiring fulltime care and care expenses are re-classified into eight categories to best describe the data according to their distributions. A two-step estimation approach is applied to estimate the days and costs. Multi-level multi-nominal logit regressions are first applied to estimate the membership probability of each group for each study variable. These probabilities are multiplied by the mean of each corresponding group of days or costs to estimate the weighted average needs or costs. Three sequential models are employed to investigate how individual characteristics and community level factors are associated with these estimates. Model I controls for age, gender, ethnicity, urban/rural residence, education, family economic condition, personal economic independence, accessibility to healthcare plus community socioeconomic development factors measured by education, per capita GDP, and number of hospital beds per thousand population. Model II further controls for proximity to children, marital status, religious participation, health practices (smoking, alcohol use, and exercise), life satisfaction, and psychological disposition. Model III additionally adjusts for overall health condition measured by the frailty index. Our results show that the average number of days required for full-time care before death for a Chinese elder is approximately 80, and this number differs significantly by sex, education, access to healthcare, proximity to children, involvement in religious activities, and baseline overall health condition. However, it is not associated with community factors. Elders who are male, educated, have easy access to healthcare, live farther from children, are religious, and are healthier tend to require fewer days of full-time care before death.