應用實證於臨床醫療 臺北巿立中興醫院副院長 沈希哲

現象：某大城巿，機動車輛肇事　　 中有30%，是因為駕駛喝酒　　 過量。 結論：某大城巿機動車輛肇事，　　 70%是因為駕駛喝酒未過　　　 量！？

實證醫學 Evidence Patient Doctor

醫療行為的決策(Decision-Making) 永遠的真理 ……. Everlasting truth 教科書理論…….. Proved theory 專業訓練……….. Practice training 臨床經驗 ………. Personal experience 文獻探索 ………. Critical appraisal Research Clinical Evidence Practice GAPS EBM

實證醫學 各醫學相關領域專家以其良知，明確且 明智地應用世上現有衛生醫療領域中可 取得之最佳文獻（證據）來處理個別患 者的醫療問題。 EBM is a supplement to, not a substitute for, professional skills and experience.

實證醫學的效益 　　　　　　Medical Care Quality Cost Containment

實證醫學照顧 Evidence-based Care Better medical care Cost containment Reduce uncertainty of prognosis Encourage clinical research

彰基經驗 1996.07 成立EBM推動小組 2002.06 協助中區健保局完成-證據醫學專案，共享醫學知識資料 推動步驟 → 舉辦研討會 未來方向 將圖書館改為EBM-based的資料中心，重整或增加EBM雜誌圖書，結合現階段人力資源與本院臨床醫師共同學習

萬芳經驗 2001.02 成立實證醫學推動小組及實證醫學中心 2002.07 參加國家品質獎評鑑 2002.01 由臺灣證據醫學中心網站票選為優質及熱門實証醫學網站 2003.01 建立病患E-mail疾病群組收集流程，以期提高醫療服務品質 未來方向 實證醫學結合問題導向教學為主軸，助全院醫療單位學習實證醫學，進而應用於臨床工作

實證醫學－五大步驟 整理出一個可以回答的問題 尋找文獻證據 嚴格評讀文獻 應用於病人身上 對過程進行稽核及評估 　（Asking an answerable question） 尋找文獻證據 　（Tracking down the best evidence） 嚴格評讀文獻 　（Critical appraisal） 應用於病人身上 　（Integrating the appraisal with clinical expertise 　& patients’ preference） 對過程進行稽核及評估 　（Auditing performance and evaluation）

實證知識的難題 一、臨床醫療的新“證據”不斷地被發表 二、沒有時間去讀所有的論文 三、許多“證據”的解析非常困難 四、無法順利應用這些證據於臨床醫療

EBM的推動為何爭論這麼多？ EBM有落入新教條主義的危險，為了 流行的議題，增加醫療人員之工作負 擔

面對EBM的幾個常見問題 EBM是否與經驗傳承，無法相容？ EBM的推動是在協助醫療的進步還是 推翻現今醫療行為的正當性？ EBM是否改變了“ART of Medicine” 的境界？

How to ask a answerable question?

問題那裡來？ Clinical finding Etiology Differential diagnosis Diagnostic test Prognosis Therapy Prevention Self-improvement

如何檢視問題的完整性？ 是否完全表述病患或其疾病的問題？ 是否考慮所有的醫療流程？ 是否與醫療病程的結果有關？

How to search for the best evidence? 有多少證據，做多少決定！ 但文獻實在太多了！

證據那裡來？ 建立檢索策略：KEYWORD個別檢索再 組合或KEYWORDS合併檢索 Pub Med, HINT, The Cochrane Library, Medline, World-Wide Web(WWW)

文獻資料有那些？ Meta-analysis：集合許多文獻加以統計分析 Systemic review：包含 a thorough literature and critical appraisal of individual studies Overview：代表作者的expert opinion Evaluating clinical practice guideline

實證醫學的基礎－多重分析的文獻整理 （Meta-analysis） A statistical analysis of the results from independent studies, which generally aims to produce a single estimate of a treatment effect. 　為了確立單一療效，而引用許多獨立且 有統計分析之資料所做之文獻。 Is quantitative

實證醫學的基礎－系統化的文獻回顧（Systemic Reviews） 特色： 1.文內有清楚的主題與完整地內含或除外之定義。 2.引用多源之文獻資料。 3.引用之各文獻中皆有分析評估其個別之研究品質。 4.有將各文獻結果依其品質程度或統計分析水準做『統合』 。

實證醫學的基礎－ 綜論 (Overview) Summary of the medical literature on a focused clinical problem Represent the author’s expert opinion on the basis of clinical experience and nonsystematic review of the literature Is qualitative

實證醫學的基礎－臨床醫療指引 (Practice Guidelines) Systematically developed statements to assist practitioner and patient decisions about appropriate health care for specific clinical circumstances Attempt to identify diagnostic and treatment strategies for common clinical problems Are recommendations

How to appraising the evidence?

如何評估證據？ Assessing study design: Concealed, Blinding, Equal use , Complete follow-up , intention-to-treat analysis Interpreting the results:Determining treatment effect, Precision of the results, Enrollment of adequate sample size Applying the results to a specific patient

Applying the results to a specific patient Are the results applicable to the patient? Are the potential treatment benefits worth the potential side effects, costs, and inconvenience? (NNT is a useful tool)

Critical Appraisal Topic Background: Objectives: Search strategy: Selection criteria: Data collection and analysis: Main result: Reviewer’s conclusion:

如何運用實證於醫療行為？ 各單位發展實證之取得方式 靈活運用電腦網路資訊（eEverything!） 運用小群組討論的方式，進行群體學習 製作醫療的標準流程（Practice guideline）

醫護人員該如何來看待及運用EBM 1.首先必須先了解什麼是EBM? Medical decision-making based on evidence 2.不要懷疑EBM的必要性，所需要的是該如應 用在臨床實務上！ 3.要懂得如何搜尋資料 4.瞭解實證強度（Level of Evidence）

Level of Evidence Level I:從多個設計優良含對照組研究之統計學合成（meta-analysis）或從high-power隨機有比較標準的臨床試驗得到的證據。 Level II:從至少一個設計優良的驗證(experimental)研究或從low-power隨機有比較標準的臨床試驗得到的證據。

Level of Evidence Level III:從設計優良的半驗證（quasiexperimental）研究如非隨機，controlled single-group, pre-post, cohort, time或matched case-control series得到的證據。

Level of Evidence Level IV:從設計優良的非實驗證明之研 究，如comparative和correlational描述 性和個案研究得到的證據。 Level V:從個案報告和臨床範例得到的證 據。

評讀文獻 (Evaluating integrative literature) Is the problem framed in a clinically relevant manner? Does the integrative framework incorporate all valid information? Is the process of integrating information rigorous? What is the finding or recommendation and is it presented appropriately? Is the finding or recommendation applicable to the care of my patient?

Is the problem framed in a clinically relevant manner? Are the population, predictor, and outcome clearly individual variable and relevant?

Does the integrative framework incorporate all valid information? Are prospective inclusion and exclusion criteria identified? Is the search strategy comprehensive and explicitly described? Is there an assessment of validity of the individual studies? Is the process of study selection, searching, assessing validity and data abstraction reliable?

Is the process of integrating information rigorous? Are the individual studies sufficiently similar that they can be combined? Is the summary finding representative of the largest and most rigorously performed studies?

What is the finding or recommendation and is it presented appropriately? Are the key elements of each individual study clearly displayed? What is the magnitude of the finding and is it statistically significant? Is the finding homogeneous or heterogeneous? Is a sensitivity analysis performed?

Is the finding or recommendation applicable to the care of my patient? Are the subjects adequately described and similar to my patient? Is the predictor variable adequately described and applicable to my patient? Will the finding result in an overall net benefit for my patient?

評讀文獻常犯的錯誤 隨便找一篇文章。 花很時間，但找到品質不是很好的文章。 先看結果，再看討論，對於研究背景和方法卻不詳讀。 將上述原則當作“食譜”一樣，一項一項對著打勾，而不瞭解其為什麼重要。 看完文章對照顧病人並沒有幫助。

All men by nature desire to learn. －Aristotle 人類本質上都有學習的慾望。 －亞里士多德

謝謝您的聆聽！