引導式測驗 · 中等 等級
安寧療護與死亡率預測測驗中的人工智慧
測試風險預測與預後、NICE 不確定性、支持性轉診、驗證、患者目標、治療決策的限制。
問題 1 的 8
死亡風險模型的輸出代表什麼?
預覽每個問題8 問題
當您參加測驗時,就會出現答案和解釋。
- 死亡風險模型的輸出代表什麼?
- 需要安寧療護的診斷
- 患者對治療的偏好
- 研究人群中確定結果視窗的機率
- 一個人的明確死亡日期
- 高風險警報該如何使用?
- 作為討論偏好的替代品
- 作為審查需求並考慮姑息性對話的提示
- 作為自動停止治療的理由
- 作為該人已處於生命最後幾天的證明
- NICE對於認識人生的最後幾天強調什麼?
- 模型可以可靠地識別確切的日期
- All patients want to know a predicted timeline
- A score should determine treatment withdrawal
- Prognostic uncertainty and the need for clinical judgment and communication
- Why assess calibration?
- It checks whether predicted risks align with observed outcomes
- It measures only model speed
- It guarantees subgroup fairness
- It establishes patient preferences
- What does a retrospective risk model not prove?
- That it predicts a defined outcome
- That it may be evaluated in a cohort
- That using its alert improves goal-concordant care
- That it was trained on past records
- What should a service monitor after deploying alerts?
- False alerts, missed referrals, alert burden, and care outcomes
- Only the number of predictions made
- Only the model’s training loss
- Only the referral form design
- Which action would be inappropriate based only on a mortality score?
- Assessing whether an alert is accurate
- Restricting treatment or access to care
- Offering a supportive discussion
- Reviewing symptoms and preferences
- Why document a clinician’s reasoning after an alert?
- It makes the prediction certain
- It replaces informed consent
- It guarantees the referral was wanted
- It shows how context, uncertainty, and patient goals informed the response