Cost and Survival Prognosis Model for Lung Cancer Patients: A Continuous Gaussian Bayesian Network Approach

• Main Authors: Jyun-Lin Chen, 陳俊霖
• Other Authors: Kung-Jeng Wang
• Format: Others
• Language: en_US
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/16087186742646333410

碩士 === 國立臺灣科技大學 === 工業管理系 === 103 === In Taiwan, cancer has always become one of the leading cause of death since 1982. Ministry of Health and Welfare mortality statistics showed that 44,791 people died of cancer in 2013, accounting for 29 percent of all deaths. Furthermore, lung cancer is the leading cause of mortalities no matter in men or women in 2013, which accounted 19.77% of all cancer deaths. The resources for lung cancer patients’ medical care should be considered much deep. Risk adjustment deals with the issues of equity and efficiency separately by establishing a risk equalization, which is seen as an effective way to evaluate individual medical requirement. This study presented a continuous Gaussian Bayesian network model to evaluate lung cancer patients’ survival time and expenditure from Taiwan’s National Health Insurance databank. Based on previous literatures, we summarized related risk adjustment outcomes, and also provided an overview of factors selection of lung cancer. In addition, this study presented the severity stages of risk adjustment model. For survival time estimation, the adjusted R2 performed 93.574% of stage I, 86.827% of stage II, 67.222% of stage III, and 52.940% of stage IV. For expenditure estimation, the adjusted R2 performed 32.63% of stage I, 50.301% of stage II, 50.363% of stage III, and 66.578% of stage IV. Compared with previous literatures, this study successfully increased the predictive power of risk adjustment model by using a continuous Gaussian Bayesian network. This study also performed the probability density function for all factors, as well as healthcare expenditure and overall survivability prediction. Public decision maker can utilize the proposed model to measure the lung cancer patients. According to this study, requirement planning of lung cancer patients can be evaluate properly.