背景：醫療不良事件(AMEs)的管理是促進病人安全的基石。根本原因分析(RCA)是探討重大病人安全不良事件問題，但是由於無法解釋更廣泛的因素而受到限制。本研究使用人為因素分類與分析工具(HFACS)回溯審視醫療不良事件，以了解影響醫療不良事件的人為因素及各層面中錯誤原因之分佈或系統性問題。方法：本研究回溯2012-2016年間有做過根本原因分析之醫療不良事件，以連續性取樣方式(continuous sampling)收集40例醫療不良事件，以人為因素分類與分析工具(HFACS)去系統性的挖掘每件案例發生錯誤之原因。結果:在行為層面中，分析結果主要為基礎技能錯誤(95°%)、判斷和決策錯誤(87.5°%)，其中在基礎技能錯誤當中以未依程序作業執行(81.6°%)及未如預期設備操作(50°%)佔多數；判斷和決策錯誤中以現場風險評估不確實(68.6%)及採取錯的步驟/行為(68.6%)佔多數。而在先決條件層面中，以團隊合作(27.5%)為主要因素包含溝通不良(81.8%)及重要訊息未以正確的方式即時傳達(72.7%)佔多數。在監督層面當中，以監督不周(92.9%)或監督不妥當(92.9%)為主要因素。在組織政策和流程方面(50%)，以組織系統的政策/風險未充分評估(35%)為主要因素。結論：本研究進一步了解醫療不良事件在過去使用根本原因分析未探討到的人為因素，反映出潛在的監督、組織行為或系統性的問題，促進病人安全預防錯誤再發生。

Background: Appropriate management of adverse medical events (AMEs) is the cornerstone of patient safety. Root cause analysis (RCA) is used to investigate serious or high-frequency errors, but is limited by its inability to explain a broader scope of factors. We re-reviewed RCA cases to understand the underlying human, organizational, or systemic factors affecting AMEs. Methods: A total of 40 consecutive RCA cases (2012-2016) were retrieved from the AMEs database. A panel was organized to retrospectively re-review these cases using the Human Factor Analysis and Classification System HFACS. Results: For active failures, errors stemmed largely from performance-based (95%), judgment and decision- making errors (87.5%). Incorrectly followed procedures (81.6%) and accidental equipment operation (50%) were the most common types of performance-based errors. Inadequate real-time assessment (68.6%) and inappropriate operative actions (68.6%) were the most common decision-making errors. For sources of latent failure, teamwork problems (27.5%), including failure to effectively communicate (81.8%), and communicate critical information (72.7%) were common. Inadequate supervision (92.9%) or command oversights (92.9%) were the most common problems related to inadequate supervision (35%). Organizational program/policy risks not adequately assessed (50%) were the most common problems related to policy and process problem (25%). Conclusions: The HFACS review enhances our understanding of human factors underlying AMEs. The HFACS reveals latent supervisory, organizational, or systematic problems that cannot be addressed through traditional RCA.