查詢結果分析
來源資料
相關文獻
- 人工智慧介入改善COVID-19插管病人脫離呼吸器的結果--某醫學中心專責病房的經驗
- 加護病房護理人員更換呼吸器管路過程標準化
- 外科加護病房呼吸器管路更換頻率與其肺炎相關感染率之比較
- 縮短兒科加護病房呼吸器管路裝置時間
- 不同鼻胃管灌食方法對重症呼吸器使用病人照護成效之探討
- Ventilator Related Nosocomial Infection in MICU Patients
- 某內科加護病房搬遷前後呼吸器相關肺炎之比較分析
- 呼吸治療技術與醫院空氣品質
- 組合式感染控制措施(Bundle Intervention)降低呼吸器相關肺炎感染發生:某區域醫院內科加護病房推行經驗
- 降低某兒科加護病房開心術後呼吸器相關性肺炎感染密度專案
頁籤選單縮合
| 題 名 | 人工智慧介入改善COVID-19插管病人脫離呼吸器的結果--某醫學中心專責病房的經驗=Artificial Intelligence Intervention Improving Ventilator Weaning Outcomes of COVID-19 Intubated Patients: A Case Study from a Dedicated Ward in a Medical Center |
|---|---|
| 作 者 | 王文靖; 林泱瀚; 張庭嘉; 劉忠峰; 陳欽明; 周偉倪; | 書刊名 | 台灣急重症醫學雜誌 |
| 卷 期 | 8:4 2023.12[民112.12] |
| 頁 次 | 頁96-103 |
| 分類號 | 415.415 |
| 關鍵詞 | 新冠肺炎; 人工智慧; 加護病房; 機械通氣; 呼吸器; 脫離時機; COVID-19; Artificial intelligence; Intensive care unit; Mechanical ventilation; Ventilator; Weaning timing; |
| 語 文 | 中文(Chinese) |
| 中文摘要 | 目的:加護病房(intensive care unit, ICU)接受機械通氣(mechanical ventilation, MV)的病人可能會出現與呼吸器相關的不良事件和延長的 ICU 住院時間。然而在新冠肺炎(COVID-19)疫情時間,臨床上醫護人員避免與病人接觸感染,亦可能降低訓練病人脫離呼吸器之機會。我們進行了一項人工智慧(artificial intelligence, AI)介入的研究,以期改善 COVID-19 疫情期間,使用呼吸器患者脫離呼吸器之結果。 材料與方法:我們使用大數據和人工智慧(AI)/ 機器學習(machine learning, ML)技術,建立了嶄新的 AI 預測模型,用於提示成功脫離呼吸器的最佳時機(48 小時內不再插管)。跨領域團隊將 AI 整合到呼吸器脫離的標準流程(protocol)中。我們根據臨床需要建立兩階段的脫離預測程序(try weaning and weaning MV, 嘗試脫離與拔管),每階段根據不同時間點之成功率進行預測,提供預測結果給臨床團隊。我們在 COVID-19 疫情期間,使用 AI 來協助某醫學中心 COVID-19 專責加護病房插管病人脫離呼吸器之情形。我們調查使用呼吸器的所有成年病人的臨床結果,並比較了 AI 介入前(2022 年 9 月 1 日至 10 月 31 日)和介入後(2022 年 11 月 1 日至 12 月 31 日)之效益。 結果:研究結果顯示,介入後組(n=22)的平均呼吸器使用時間(185.5 小時 vs. 344.6 小時,p 值=0.017)、加護病房住院天數(9.8 天 vs. 15.8 天,p 值 =0.006)和醫院住院天數(20.3 天 vs. 39.3 天,p 值 =0.004)均低於介入前組(n=11),然而呼吸器脫離率則未能達到顯著的差異(86.4% vs. 81.8%,p 值 =1.000)。 結論:我們初步發現將 AI 與呼吸器脫離方案整合可提高病人的醫療品質。我們認為 AI 預測模型在呼吸器脫離之應用上可成為一個實用且有用之工具,尤其是在 COVID-19 疫情期間。 |
| 英文摘要 | Objective: Patients in intensive care units (ICUs) who require mechanical ventilation (MV) may experience ventilator-associated adverse events and prolonged ICU stays. However, during the COVID-19 pandemic, minimizing direct patient contact could also reduce opportunities for training patients to wean from ventilators. In this study, we conducted an artificial intelligence (AI) intervention to improve the outcomes of patients using ventilators during the COVID-19 pandemic. Materials and Methods: We investigated the impact of AI intervention on all adult patients using ventilators in a specialized COVID-19 ward at a medical center. We employed the big data and AI/machine learning (ML) technologies to build the novel predictive models for determining the optimal timing to successfully weaning from ventilators (extubation within 48 hours). An interdisciplinary team integrated AI intervention into the standard weaning protocol. We established a two-step weaning prediction process (try weaning and weaning), with multiple timeperiod predictive models, providing decision assistance to the respiratory therapy team. During the COVID-19 pandemic, we used AI to assist with ventilator weaning decision in the specialized COVID-19 ward and compared the outcomes before AI intervention (September 1 to October 31, 2022) with those after AI intervention (November 1 to December 31, 2022). Results: The study demonstrated that the intervention group (n=22) exhibited lower mean MV duration (185.5 hours vs. 344.6 hours, p=0.017), ICU length of stay (9.8 days vs. 15.8 days, p=0.006) and hospital length of stay (20.3 days vs. 39.3 days, p=0.004) compared to the preintervention group (n=11). However, the extubation weaning rate did not show a significant difference (86.4% vs. 81.8%, p = 1.000). Conclusion: Preliminary findings suggest that integrating AI into the MV weaning protocol can enhance the quality of patient care. The AI prediction model is recognized as a practical and valuable tool for ventilator weaning, especially in the context of the COVID-19 pandemic. |
本系統中英文摘要資訊取自各篇刊載內容。