- 學位論文
男性老年健檢個案以DRE及PSA篩檢攝護腺癌之成果、效益及醫療資源耗用探討
The evaluation of outcome, benefit, medical resources consumption to elder men screened prostate cancer with DRE and PSA
摘要
研究目的 了解接受攝護腺癌篩檢之男性老年健檢個案異常結果與確認為攝護腺癌之比例，與年齡之相關性，篩檢工具彼此之成果差異、效益、醫療資源耗用與篩檢工具效益成本比之情形。探討每年一次之老年健康檢查，是否應針對男性受檢個案，固定增加攝護腺癌之篩檢項目，以做為癌症篩檢的政策參考。 研究方法 本研究屬於縱貫性質研究，以回溯性的方式，並以兩階段比較分析做研究設計。第一階段採用實際受檢個案進入研究醫院，完成個人基本資料記錄及確認個案遵從篩檢流程完成篩檢項目。依篩檢過程的醫療相關直接、間接費用計算，並依受檢個案相關變項與篩檢方式相關變項分析影響篩檢成果的因素。第二階段採取追蹤篩檢異常個案，透過中介變項的介入，以取得最後篩檢成果之敏感性、特異性、準確度，再與國外現有資料做比較。同時，依篩檢過程計算醫療相關直接、間接費用（成本），並將篩檢異常個案的住院花費（效益）做比較，可獲得篩檢異常個案醫療資源耗用情形與不同篩檢方式及篩檢的總效益成本比，藉以了解本研究之成本效益情形。另外，根據篩檢結果可推估每十萬男性人口攝護腺癌發生率並藉由不同篩檢結果用以預測攝護腺癌發生之機率。 本研究係以研究執行醫院所收集、建置的研究對象之第一階段電腦初級資料檔，與存放在研究醫院HIS架構第二階段追蹤篩檢異常個案之健保申報費用檔，以及依個案病理組織切片報告結果所轉置的電腦資料檔，經在EXCEL歸類整合、刪除錯誤及去除遺漏值及極端值後，再轉換為SPSS檔案以利統計分析。 研究結果 第一階段自民國95年4月至95年6月，前來研究執行醫院接受老人健康檢查者，共收案3036位，符合有效收案條件者共有2861人，平均年齡為72.9±6.2歲。篩檢結果為異常的541人（18.9%）中，DRE結果異常者有120人（4.2%），PSA結果異常者有468人（16.3%），兩項篩檢結果同時異常者有47人。第二階段自民國95年4月至97年10月，被通知且回診者有217人（40.1%），回診後經泌尿科專科醫師評估而建議住院執行經直腸超音波攝護腺系統性切片檢查者共有99人（45.6%），執行切片後確認為攝護腺癌罹患者共有23人（23.2%）。 統計分析結果為接受篩檢病患的年齡階層愈高，PSA檢驗結果異常機會也愈高（p＜0.001），肛門指診檢查結果異常機會也愈高（p＝0.002）；被篩檢個案PSA正常，則其DRE正常的機率較高，被篩檢個案PSA異常則其DRE異常的機率也較高（p＜0.001）；接受篩檢者其年齡愈高則其PSA檢驗值也可能愈高（p＜0.001）；四個不同年齡階層其PSA檢驗平均值存在有顯著性的差異（p＜0.001），DRE檢查異常的受檢者其PSA檢驗平均值會顯著高於DRE檢查正常的受檢者（p＜0.001）；PSA檢驗異常的受檢者其平均年齡數顯著高於PSA檢驗正常的受檢者（p＜0.001），DRE檢查異常的受檢者其平均年齡數顯著高於DRE檢查正常的受檢者（p＜0.001）；以PSA篩檢診斷攝護腺癌的敏感性為100%，特異性為9.2%，準確度為30.3%；以DRE篩檢診斷攝護腺癌的敏感性為65.2%，特異性為63.2%，準確度為63.6%；合併PSA和DRE兩者共同來篩檢診斷攝護腺癌的準確度最佳（69.7%），雖然敏感性並非最高（65.2%），卻具有最理想之特異性（71.1%）。四個不同年齡階層其平均住院日數與平均住院費用皆存在有顯著性的差異（p＝0.035；p＝0.042），四個不同主治醫師所照顧的病患其平均住院日數與平均住院費用亦皆存在有顯著性的差異（p＝0.042；p＝0.035）。PSA篩檢的效益成本比＝1.14，DRE篩檢的效益成本比＝0.86，合併兩項篩檢所獲得的總效益成本比＝0.78，以使用PSA篩檢的方式相對較符合成本效益原則。 本研究預估，針對65歲以上的受檢者每十萬男性人口約有804人可被篩檢出有攝護腺癌，已超越亞洲人族群以及我國95年實際發生數和平均值，約莫介於白人族群與西班牙裔族群中間，對照於國內攝護腺癌的低篩檢率，可預測社區中應該尚有為數不少的潛在攝護腺癌患者未被檢出。使用二項式對數回歸分析來預測，研究個案每增加PSA值1 ng/ml，可能得到攝護腺癌之機率相對增加4.7%，而年齡數及PSA檢驗結果、DRE檢查結果與年齡階層等變項皆不能被使用於有意義性的預測攝護腺癌之有或無的機率。 結論與建議 男性老年健檢個案以DRE及PSA篩檢攝護腺癌確實對於民眾面、學術面及醫院面有正面的貢獻，但對於政策面而言、較不符合成本效益原則。整體實施全面篩檢的實際效益尚未明確可知，在仍未有更令人信服的實証醫學證據出現前，是否應於每年一次之老年健康檢查，針對男性受檢個案增加攝護腺癌之篩檢項目，應視政府的財政狀況、攝護腺癌篩檢在公共衛生政策上的角色定位與受重視程度而定。本研究就研究過程中所面臨到篩檢的困境、執行的最終成效、分析之成本效益狀況與國內外文獻的搜尋，綜合評估、判斷，認為目前現階段老年男性攝護腺癌篩檢仍不夠成熟以適合全面列入固定老人健康檢查之癌症篩檢項目。如何突破篩檢後回診率偏低的困境，在兼顧臨床實務、病患隱私、病權自主與減少個案擔憂害怕的綜合情境下，思考更佳的回診對應方式？如何著重醫病溝通，詳盡告知個案切片可能發生的利益或傷害，並建立病人自主與互信溝通討論的基礎？如何看待、處理都市化地區的攝護腺癌發生率確實高於台灣地區攝護腺癌的平均發生率如此之公衛問題？公衛預算的分配使用是否也應有所參酌應對？是往後有志於攝護腺癌篩檢研究之學者該正視、面對之議題。
並列摘要
Objectives To understand and to investigate the outcome, benefit, medical resources consumption of elder men over 65 y/o who received prostate cancer screening with digital rectal examination（DRE）and prostate specific antigen（PSA）combined with their yearly physical examinations as a policy decision making reference. The study was also used to evaluate that if the government should add prostate cancer screening as a new item in annual elder health examination. Methods It was a longitudinal, retrospective study. The study design was divided into 2 stages of comparative analysis. The first stage was used practical cases of their background data records and confirmed by cases to be respected and obeyed all the exams in the research hospital. Counted all direct and indirect medical expenses during the process of exams, and analyzed the influence factors of the result by related variables of the cases and correlated variables of the exam methods. The second stage was used tracker in the abnormal cases by intervening variables, in order to acquire the final sensitivity, specificity and accuracy, and to compare with foreign references. Meanwhile, counting the medical direct and indirect expenses during the process of exams and comparing to the hospitalization expenses of abnormal cases can help us to realize the medical resources consumption, each method and total benefit/cost ratio and cost-benefit condition of this research. In addition, it can be estimated the incidence of prostate cancer in each a hundred thousands male population by the result of the prostate cancer screening, and it can be also used to predict the possibility of prostate cancer by different kinds of screening results. Excel program was used to manage all the subjects’ basic data, health insurance reimbursement information in the healthcare information system and the final computer data transformed by the results of pathological reports. After data classifying, categorizing, deleting errors, missing value and outliers, finally we used SPSS program to perform statistic analyses. Results Total 2861 subjects were enrolled into this study finally. The incidence of abnormal cancer screening result was 18.9% （541/2861）. 40.1%（217/541）subjects were called back and really returned for recheck. 45.6%（99/217）subjects accepted the suggestion of transrectal prostate biopsy from urologists. 23.2% （23/99）patients were approved to have prostate cancer in histology. When the age level of screened group was higher, the abnormal possibilities of PSA and DRE results were getting higher, too. There were significant correlations among age level category, PSA category and DRE category. Significant differences of mean PSA values were noted in the four different age groups. The group had normal PSA result was significantly younger than the group had abnormal PSA result (72.6±6.1 vs. 74.5±6.7 y/o). Also, the group had normal DRE result was significantly younger than the group had abnormal DRE result (72.8±6.2 vs. 75.2±6.2 y/o). Combined with PSA and DRE tests to screen prostate cancer had the best accuracy （69.7%）and specificity（71.1%）. There were significant differences in the length of stay and admission fee among different age groups and among different visiting staffs. The benefit/cost ratio of PSA was 1.14, ratio of DRE was 0.86 and ratio of 2 tests combined was 0.78. Use PSA test to screen prostate cancer was relative cost-benefit. This study estimated that there were 804 prostate cancer patients in each a hundred thousands male population who was over 65 years old. The result was far over the numbers of prostate cancer and the average value in Taiwan and Asian in 2006; the result was between white people and Spanish. Contrast to the lower proportion of prostate cancer screening in Taiwan now, it can be sure that there must be so many potential prostate cancer patients have not been detected in communities. Binary logistic regression analysis was used to test which category can effectively predict the possibility of prostate cancer. Only if PSA value increased 1 ng/ml meant the possibility of prostate cancer also increased 4.7%, the other categories can’t be used in meaningful to predict the final outcome. Conclusions At present, reliable evidence is still not enough to support the policy of routine screened prostate cancer to elder men over 65 y/o, but the government still can take the finance condition and consider the role of prostate cancer in public health policy in order to make the final decision.