行業指南

為醫生提供人工智慧臨床決策支持

Clinical decision support (CDS) is software, usually built into or connected to the electronic health record, that gives clinicians patient-specific alerts, reminders, order sets and risk predictions while they care for a patient.

聆聽此指南AI 語音敘述 · ~3 min

本頁4 分鐘閱讀
  1. 概述
  2. 深入探討
  3. 現實世界的實施
  4. The Future of AI Clinical Decision Support for Physicians
  5. 付諸實踐
  6. 不斷探索
  7. 常見問題

概述

It matters because well-designed CDS catches drug interactions, dosing errors and missed diagnoses. Poorly designed CDS causes alert fatigue, where clinicians override most warnings, including the important ones.

深入探討

CDS falls into two broad families. Knowledge-based CDS runs explicit rules written by clinicians and pharmacists, such as drug-drug interaction checks, allergy checks, dose ranges, preventive care reminders and order sets. Predictive CDS uses statistical or machine learning models trained on past patient data to estimate risk of events such as sepsis, deterioration, readmission or no-shows. Delivery can be interruptive (a pop-up that stops the workflow), passive (a flag, a banner or a default choice in an order), or routed to someone else, such as a pharmacist or a rapid response team. Informaticists often use the "five rights" framework: the right information, to the right person, in the right format, through the right channel, at the right time. Alerts that break these rules produce alert fatigue. Many studies of drug interaction alerts have found that clinicians override the large majority of them, often well above 90 percent in some settings, largely because so many alerts are clinically irrelevant. Once clinicians learn that most alerts are noise, they also dismiss the rare one that matters. Predictive models add another risk: they can perform worse in practice than their developers reported. In 2021 a University of Michigan team published an external validation of the Epic Sepsis Model in JAMA Internal Medicine. It found much weaker discrimination than advertised (an AUC of about 0.63). The model missed about two thirds of sepsis cases while generating alerts on many patients who did not have sepsis. On regulation, the 21st Century Cures Act of 2016 excluded some CDS from FDA device regulation, mainly tools whose basis a clinician can independently review, and the FDA issued final CDS guidance in 2022. ONC's HTI-1 rule, finalized in late 2023, requires certified EHRs to expose information about decision support interventions, including predictive ones. The common misconception is that more alerts mean safer care. Past a point, more alerts make care less safe.

現實世界的實施

When a physician orders an antibiotic for a patient on warfarin, the EHR shows an interruptive interaction alert suggesting closer INR monitoring. The physician must acknowledge it or give an override reason.

A renal dosing rule reads the patient's latest eGFR and suggests a lower dose of a renally cleared drug inside the order itself, without a pop-up.

A hospital's sepsis prediction model scores inpatients every few hours and sends a best practice advisory to the rapid response nurse when the score passes a threshold.

An emergency physician opens a community-acquired pneumonia order set that preselects guideline-recommended antibiotics and a severity score calculator.

The Future of AI Clinical Decision Support for Physicians

Language models are starting to appear in CDS as chart summarizers and question-answering tools, not only as rule engines. That raises new questions about how to validate a tool whose output is free text. Transparency rules such as HTI-1 push vendors to document how predictive tools were trained and tested, which should make local evaluation easier. The lesson from sepsis models and interaction alerts is likely to hold for newer tools as well: local validation, measured alert burden and routing to the right person decide whether CDS helps. Health systems that retire low-value alerts may gain as much as those that add new ones.

付諸實踐

為什麼這很重要

  • 背景與規則. 產業背景決定了人工智慧創意能否與現實接觸。
  • 品質管控. 領域約束會影響可接受的錯誤率和監督模型。
  • 配裝選擇. 成功的部署使技術能力與第一線工作流程保持一致。

留意

  • 監理要求可能會使原本強大的原型失效。
  • 歷史資料可能會編碼損害特定社區的偏見。
  • 遺留系統可能會造成整合瓶頸和隱性成本。

後續步驟

  1. 讓領域專家參與從問題框架到評估的整個過程。
  2. 在啟動前設計審計追蹤和文件。
  3. 儘早驗證合規性和安全義務。
  4. 分階段推出,並有明確的停止和回滾標準。

不斷探索

免費電子報

獲取每日 AI 簡報

每個工作日上午提供三則經過驗證的 AI 新聞,以淺顯易懂的英文撰寫。永久免費,無廣告。

每個工作日一封電子郵件。一鍵取消訂閱。我們絕不出售或分享您的地址。

測試一下自己

Take the AI Clinical Decision Support for Physicians quiz

每個答案皆有即時回饋,通過課程後可獲得具備可驗證 ID 的可分享證書。

開始測驗

支持免費的 AI 教育。AI Understanding 是一個 501(c)(3) 非營利組織——絕無廣告,絕無付費牆。 捐款

常見問題

What is AI Clinical Decision Support for Physicians?

Clinical decision support (CDS) is software, usually built into or connected to the electronic health record, that gives clinicians patient-specific alerts, reminders, order sets and risk predictions while they care for a patient. It matters because well-designed CDS catches drug interactions, dosing errors and missed diagnoses. Poorly designed CDS causes alert fatigue, where clinicians override most warnings, including the important ones.

藥物交互作用警報導致警報疲勞的主要原因是什麼?

高覆蓋率主要來自低價值警報。一旦大多數警報都是噪音,臨床醫生也會忽略那些重要的警報。

2021 年密西根大學 Epic 敗血症模型的外部驗證發現了什麼?

研究發現 AUC 約為 0.63。該模型漏掉了大多數膿毒症病例,同時對許多沒有膿毒症的患者發出了警報。

腎臟給藥規則建議在訂單內使用較低劑量,而不會彈出視窗。那是什麼樣的交付?

在不停止臨床醫生的情況下在工作流程中顯示建議是被動的、不間斷的交付。

在 CDS Hooks 中,外部 CDS 服務將什麼發送回 EHR?

EHR 在諸如醫囑簽名之類的掛鉤點呼叫服務,服務透過臨床醫生可以讀取或操作的卡片進行應答。

為什麼在上線前以靜默模式運行預測模型?

靜音模式對患者進行評分而不提醒任何人,因此團隊可以先檢查本地準確性和警報量。