癌症醫療費用之推估：馬可夫鏈模型之應用

目前有關癌症發生率的相關文獻中，很少針對罹癌後的存活機率做探討，進而推估罹患癌症後所發生的醫療費用。本研究利用「全民健保資料庫」1996至2000年的癌症醫療統計資料庫，依性別及年齡估計各種不同健康狀態移轉機率模型，建構出不同健康狀態移轉之離散型馬可夫鏈模型，並利用此模型來推估不同性別及年齡之個人癌症醫療所需費用，藉由此醫療費用的推估，進而計算出商業保險中實支實付型癌症保險之合理保費，除此之外，本研究利用未來人口老化推估資料，計算出全民健保在癌症醫療部份的未來成本之估計，本研究之結果可同時做為商業保險及社會保險的參考。
關鍵字：馬可夫鏈模型、醫療成本推估、癌症

Estimation of Cancer Medical Cost：A Markov Model Application

　　　　　　　　　　　　　(Hong-Chih Huang)＊(Chia-Yu Lin)＊(Jack C. Yue)

To date, investigations of survival rates for cancer patients have been limited. In this research we attempt to use Taiwan National Health Insurance Database from 1996 to 2000 to investigate the survival rates and medical expenses for cancer. We then use the information to construct a discrete Markov model to estimate associated medical costs. Three states of health, cancer, and death are constructed in the Markov model. The fair premium for private cancer insurance is calculated. Furthermore, with estimates of the future population structure, we estimate the future expected medical costs of cancer patients under Taiwan National Health Insurance. Our research results are relevant to both private and social insurance.
Keywords: Markov chain model, estimation of medical cost, cancer