RAISE评分预测蛛网膜下腔出血患者延长机械通气的外部有效性分析:美国队列
该单中心回顾性队列研究纳入2017年7月至2025年7月神经重症监护病房收治的202例成人非创伤性蛛网膜下腔出血患者,评估RAISE评分预测机械通气>7天的外部有效性。49例(24.3%)机械通气>7天;RAISE评分的AUC为0.85,校准斜率为0.60,提示过拟合。5个RAISE评分组分中仅Hunt-Hess分级(p<0.001)与延长机械通气显著相关。作者认为该评分虽能区分不同风险类别,但可能过拟合于原始研究人群,或存在插管、通气及拔管实践差异等混杂因素。
要点
- 在202例非创伤性蛛网膜下腔出血患者中,49例(24.3%)机械通气>7天。
- RAISE评分预测机械通气>7天的AUC为0.85,校准斜率为0.60,提示过拟合。
- 5个RAISE评分组分中仅Hunt-Hess分级(p<0.001)与延长机械通气显著相关。
- 作者认为该评分可能过拟合于原始研究人群,或存在插管、通气及拔管实践差异等混杂因素。
- 作者建议未来研究可考虑利用另一队列评估外部有效性,或设计新的评分系统预测延长机械通气。
适用人群与场景
- 成人非创伤性蛛网膜下腔出血患者,收治于神经重症监护病房。
研究设计与关键结果
- 单中心回顾性队列研究,纳入2017年7月至2025年7月神经重症监护病房收治的202例成人非创伤性蛛网膜下腔出血患者。
- 49例(24.3%)机械通气>7天。
- RAISE评分预测机械通气>7天的AUC为0.85,校准斜率为0.60,提示过拟合。
- 5个RAISE评分组分中仅Hunt-Hess分级(p<0.001)与延长机械通气显著相关。
证据类型、等级与局限
- 证据类型:单中心回顾性队列研究。
- 局限:原文未报告局限性。
对床旁实践的改变
- 作者认为RAISE评分虽能区分不同风险类别,但可能过拟合于原始研究人群,或存在插管、通气及拔管实践差异等混杂因素。
- 作者建议未来研究可考虑利用另一队列评估外部有效性,或设计新的评分系统预测延长机械通气。
提炼说明
- 本条依据论文摘要整理;摘要未涉及「诊断标准与关键阈值」「处置流程与药物剂量」「红旗征与不能漏的情况」「指南或共识出处」,故未列出这些栏目;全文请点「阅读原文」。
英文原文(摘要)· 点击展开
BACKGROUND/OBJECTIVE: Outcomes for patients with subarachnoid hemorrhage (SAH) have improved substantially through the years, but there remains room for improvement. The RAISE (pRolonged ventilAtion In Subarachnoid hEmorrhage) score was designed to predict which SAH patients will require prolonged mechanical ventilation. We sought to assess the external validity of the RAISE score in our patient population.
METHODS: This was a single center retrospective cohort study of adult patients admitted to the neurointensive care unit with non-traumatic SAH from July 2017-July 2025. The ability of the RAISE score to differentiate between patients with varying risk profiles was assessed. The predicted probabilities of mechanical ventilation duration > 7 days were compared to observed frequencies, and a calibration plot was generated. Multivariable logistic regression was also employed to examine if the components of the RAISE score were significantly associated with prolonged mechanical ventilation in our cohort.
RESULTS: A total of 202 patients met study criteria. Of these, 49 (24.3%) were mechanically ventilated > 7 days. The area under the receiver operating characteristic curve was 0.85, and the calibration slope was 0.60, suggestive of overfitting. Of the 5 RAISE score components, only Hunt-Hess grade (p < 0.001) was significantly associated with prolonged mechanical ventilation in logistic regression.
CONCLUSION: While the RAISE score performed well in its original cohort and can discriminate between various risk categories for prolonged mechanical ventilation, results indicate that the score is overfit to the original study population or that there are other confounders present such as differences in practices surrounding intubation, ventilation, and extubation. Future studies could consider utilizing another cohort for assessing external validity or designing a novel scoring system for prediction of prolonged mechanical ventilation.
参考来源
- Nguyen AV, Edwards AS, Reed LK, et al. Analysis of the External Validity of the RAISE Score for Prediction of Prolonged Mechanical Ventilation in a United States-Based Cohort of Subarachnoid Hemorrhage Patients. Neurocrit Care. 2026. PMID: 41845147. doi:10.1007/s12028-026-02465-8
供执业医师抢救与诊疗参考,不替代临床判断;剂量请结合体重、肝肾功能和本院药品规格复核,以指南原文为准。