檢查後疾病可能性之估計：臨床實證醫學的即時應用與教學

The Estimation of Post-Test Probability of Disease: Real-Time Application and Teaching of Clinical Evidence-Based Medicine

臨床上，應用流行病學的觀念，來評估得知檢查結果後，疾病的可能機率，以提供後續如觀察、治療或進一步檢查等決策參考，是實證醫學運作中，重要的一環。其過程，需要利用實證文獻資料中之敏感度、特異度及疾病盛行率等測量值依公式加以計算。如果能引用勝算及概似比的概念，更能簡化公式。實務上，不只是實驗室檢驗，其他包括病史中的症狀、及理學檢查中的病徵，都可加以運用。然而，醫療專業人員多半對此數學預測公式，尤其是勝算及概似比的概念，覺得難以理解；至於客觀計算的預測數字，如何妥適地搭配實務經驗累積之直覺，也不易釐清其觀點；在忙碌的工作中，更常因沒有攜帶專用的電腦軟體或計算圖表，以致於無法即時加以運用。本文作者嘗試以概念圖示，提供容易理解的證明，說明經驗與直覺在此計算公式中的角色，並在網站提出一個實用的EXCEL檔案，供讀者免費下載，在臨床即時運用時參考；隨後，在兩個說明實例中，作者提議，運用分數型式或四捨五入簡化之勝算值及概似比，可以很容易使用紙筆，計算出誤差不大的機率估計。

Estimating the post-test probability of diseases with relevant epidemiological concepts such as sensitivity, specificity and prevalence are important clinical practices of evidence-based medicine. These practices are not only applied in laboratory tests, but also in clinical symptoms or physical signs. Transformed estimating formulas with terms of odds or likelihood ratio are simplified but tough to understand for some health care professionals. Besides, how to integrating clinical experience and its instinct in differentiating clinical nuance in these formulas is still a perplexing issue. Without real-time solutions is also important obstacle for physicians in hurry clinical practice. In trying to solve the issues, we provide, in this article, an easy-to-understand proof, clearly explain the roles of clinical experiences and, in our website, provide a real-time EXCEL solution for estimating post-test probability of diseases. After two explaining clinical examples in this article, we suggest our readers trying to round odds and likelihood off and using paper-and-pencil method to reach answers without significant bias.