;The purpose of thesis is to explore the relationship between salary income and the choice of hospice care for terminal cancer patients. We take cancer patients in Taiwan as a research subject to examine the salary income and other factors of demand and supply side whether influence the termimal cancer patients to receive hospice services.The source of thesis is the 2000 and 2005 National Health Insurance Million Personnel Samples, and the choice of terminal cancer patients’s medical decision in the end-of-life is divided into two stages for analysis. The first stage is take all terminal cancer patients as research subject to examine whether patients will choose to use hospice care in the end-of-life. In the second stage, we take hospice care patients as research subject. We want to examine terminal cancer patients has chosen to use hospice care, what kind of hospice care type will be chosen. We also use the number of days of hospice care in the period before the end-of-life to examine the short-term to long-term impact of terminal cancer patients hospice care utilization. In addition, this thesis used logit regression model, multinomial logit regression model and zero-inflated negative binomial regression model for empirical analysis. Most of the empirical results show that the cancer patients with high socioeconomic status and low-income households are most inclined to receive hospice care. The former hopes to improve the quality of end-of-life, which means that terminal cancer patients with high socioeconomic status pay more attention to the quality of end-of-life is better than the extension of life; the latter is due to subsidies from social assistance such as the government and non-governmental organization, which increases the financial incentives for cancer patients with low-income households to use hospice care to alleviate the financial burden of poor families. In terms of other demand side factors, males, patients under the age of 30 and over the age of 75 are the least inclined to use hospice care. Terminal cancer patients with high clinical comorbidity index scores, cancer with high mortality, long cancer with total inpatient days and high cancer with total outpatient service visits who is inclined to use hospice care. In terms of supply side factors, patients who are diagnosed in medical centers, religious hospitals, and those hospitals with higher degrees of urbanization where they are visiting hospitals tend to use hospice care.
