惡性腦腫瘤手術後五年資源耗用及存活分析之探討

摘要研究目的台灣地區自1982年惡性腫瘤一直高居國人前十大死因之首，然而隨著醫療科技進步、藥物創新，以及基因療法都開始於治療中，即便如此，相較於其他疾病所產生的負擔，不論是對個人家庭甚至整個國家都是一項沉重的負荷。每年支付多少費用在治療惡性腦腫瘤仍然沒有確切的資料，那未來全民健康保險將預估多少費用在後續追蹤與治療惡性腦腫瘤病患本身，就值得進一步探討；目前台灣並無對此惡性腦腫瘤手術這方面的研究分析，所以惡性腦腫瘤手術之臨床醫療費用的相關研究確實有迫切的需要。綜合上述，提出本研究的目的如下：一、探討惡性腦腫瘤手術發生率及其相關分佈趨勢。二、探討惡性腦腫瘤手術病人住院期間醫療資源耗用及其相關影響因素。三、探討惡性腦腫瘤手術病人術後死亡率及其相關影響因素。研究方法利用國家衛生研究院釋出之「全民健康保險研究資料庫」，進行回溯性(Retrospective)次級資料的縱貫研究分析。資料為承保資料檔(ID)、醫事機構基本資料檔(HOSB)、及住院醫療費用清單明細檔(DD)，進行資料整理與串聯。從「全民健康保險研究資料庫」中，選取國際疾病分類第九版診斷惡性腦腫瘤代碼為(ICD-9-CM為1910-1919、1983、1921) 排除轉移性腦腫瘤（ICD-9-CM 1983），研究對象為惡性腦腫瘤手術病人，研究期間為1998年至2010年共13年，總樣本數7,740人。研究資料以SPSS for Windows 20.0 統計套裝軟體做為分析工具。研究結果惡性腦腫瘤手術病人平均年齡為45歲，惡性腦腫瘤手術病人發生率，從1998年每十萬人口2.7顯著增加至2010年3.2(P＜0.001)。惡性腦腫瘤手術病人之平均住院天數隨著時間有下降之趨勢(29.21天下降至24.90天，改變-14.73%)。研究期間區分三個時期：T1(1998~2000)、T2(2001~2005)、T3(2006~2010)，惡性腦腫瘤手術病人醫療總費用有逐時期增加的趨勢(改變率16.32%)。惡性腦腫瘤手術病人之年齡、合併症嚴重度指標、醫師服務量、呼吸系統疾病與死亡率有顯著相關(P＜0.001)。惡性腦腫瘤手術病人住院期間存活率97%、一年存活率75%、三年存活率61%、五年存活率33%。結論與建議研究發現國內惡性腦腫瘤手術發生率分佈趨勢逐年上升，住院天數隨時間有下降之趨勢，醫療總費用隨時間有顯著增加之趨勢，與國外之研究相似，顯示國內醫療有國際水準。高醫院服務量及高醫師服務量，住院天數、醫療費用顯著愈少，住院中死亡率愈低，依據本研究結果得知醫療服務量愈高，確實能使減少醫療耗用，醫療品質較佳。年齡、嚴重合併症指數(CCI)、醫師服務量、呼吸系統疾病，是惡性腦腫瘤手術病人之死亡率顯著相關影響因子。若疾病嚴重度持續下降帶來更多惡性腦腫瘤存活人數，因此醫療費用也隨之上升。儘管惡性腦腫瘤死亡率下降情形，仍需藉由持續不間斷地化學治療與放射治療來控制病情的惡化。透過本研究瞭解惡性腦腫瘤手術病人住院天數逐年下降，但醫療總費用卻逐漸上升。同時兼顧醫療品質及經營效益，並提供醫師及臨床病患治療模式、衛生主管機關及醫院管理者在規劃醫療資源預算和未來推動衛生政策參考依據。

關鍵字

惡性腦腫瘤手術病人；發生率；醫療資源耗用；死亡率

並列摘要

Abstract Purpose Malignancy has topped the 10 most common causes of death in Taiwan since 1982. With advent of technology progress, drug innovation and gene therapy, economic burden of malignancy therapy grows tremendous, either for individual alone or the society as a whole. Little is known about the health care expense allocating brain tumors patients with surgery by National Health Insurance. This study aims to explore the prevalence, medical utilization and survival analysis of patients with malignant brain tumor undergoing surgery. Research Methodology This longitudinal nationwide population-based study retrospectively evaluated the prevalence, health care utilization and survival between January 1, 1998, and December 31, 2010 in Taiwan. In total, 7,740 subjects extracted from National Health Insurance Research Database (NHIRD) with ICD-9-CM code being 1910-1919 and 1921 were analyzed in 3 time intervals. Data for this observational cohort study consisted of computerized claims from registry for beneficiaries (ID), registry for contracted medical facilities (HOSB), inpatient expenditures by admissions (DD). Survival analysis was examined by Kaplan-Meier method with log-rank test. The multiple linear regression and Cox proportional hazards models were used for multivariate assessment of predictors of outcomes after controlling for possible confounding factors. Results The mean age of this study cohort is 45 years. The annual prevalence per 100,000 subjects is 2.7 in 1998 and 3.2 in 2010, with mean of 2.6 for T1 (1998~2001), mean of 2.5 for T2 (2001~2005) and mean of 2.8 for T3 (2006~2010) (P<0.05). Average lengths of stay decreased over time, with 29.21 days in 1998 and 24.90 days in 2010, with a significant reduction by 14.73%. However, total medical expense increased significantly by 16.32%. The overall in-hospitalization, 1-, 3-, and 5-year survival rates were 97%, 75%, 61% and 33%, respectively. Additionally, multivariate Cox proportional hazard model revealed that age at surgery, Charlson comorbidity index, surgery volume, and respiratory tract diseases are significant predictors of mortality after controlling for all covariables (P<0.05). Conclusions During study period, annual prevalence increased significantly over time. While lengths of stay showed a significant decreasion, total medical expense increased significantly. Five years survival rate was 33%, and age at surgery, Charlson comorbidity index, surgery volume, and respiratory tract diseases are significant predictors of mortality. Two suggestions are as follows: 1. doctor education for early detection of brain tumor to curb expanding expense and pursue integrating outcome. 2. Incentives should be implemented to recruit young doctors to become neurosurgeons.