癌症自西元1982年起即被列為國人十大死因之首位。隨著科技的進步環境的惡化和生活方式的改變,癌症發病率逐年增加。然而,生活環境條件中有些致癌危險因子是可以控制的,同時隨著癌症治療技術的進步,癌症病人的存活率也有逐漸提升的趨勢,透過早期診斷及正確治療可以求得較高的治癒率。本研究資料收集來自西元1985~2008年申報醫院申報至行政院衛生署國民健康局,透過癌症登記個案連結重大傷病資料及癌症死亡檔產生之申報醫院癌症死亡檔,經整理比對結果共計1178位癌症個案,探討環境條件可能影響之部分危險因子。藉由存活分析,進一步了解影響存活率的預後因子。研究資料以病歷回溯世代性收集分析,以中部某區域教學醫院之病歷紀錄資料為基礎,探討確診罹癌病患其存活期間之相關影響因素,並在基因與生活條件下建立評估準則,採用多變項因素性別、血型、確診癌症年齡、癌症類別、組織型態、出生地、家屬同住以及治療方式乃至最後真正死亡原因等。藉由資料收集分析探討存活率與變數之相關性,並利用非條件式多變項,以變異數分析-ANOVA分析相關變數之顯著性,再利用T-檢定(T-test)檢定不同癌症部位平均數是否相等,其後用存活分析Cox氏比例危害模式 (Cox proportional hszards model) 針對該項因素的影響探討環境影響因子與病人預後死亡之關聯性分析。
Since 1982, cancer has been in the first place among the top ten leading causes of death in Taiwan. According to previous literatures, the incidence of cancer increases year by year which may be related to environmental deterioration and lifestyle changes as a result of development in technology. However, the survival rate of the cancer patients has risen gradually. It is hypothesized that the rise in cancer survival rates may be because of some controllable environmental and lifestyle risk factors along with early diagnosis and advancement in cancer therapies. In this study, we search the database in the period from 1985 to 2008 of the insurance claims submitted by the provider hospitals to the Bureau of Health Promotion in the Department of Health of the Executive Yuan in Taiwan. There are a total of 1178 registered hospital files of cases of cancer that crosslink to major illness and deaths to look for potential environmental risk factors. Through survival analysis, the prognostic factors that affect survival rates are evaluated further. Then, research materials for a retrospective cohort study is gathered based on medical records in a regional teaching hospital in Taichung area and analyzed for any relevant factors in the survival period of the patients diagnosed with cancer. The criteria for evaluations, under genes and lifestyles, include multi-variable factors of sex, blood type, age of onset, cancer type, histology, birth place, family members living together, treatment, the final causes of death, etc. By analyzing the information collected, we discuss the correlation between survival and variables. We use non-conditional multivariate, ANOVA analysis for the significance of the relevant variables. Moreover, we use T-test to test whether the average number in different cancer is equal. Furthermore, we use survival analysis (Cox proportional hazards model) for the influence of the environmental factors associated with patient outcome of death.