Early prediction model of hemodynamic instability has the potential to improve the critical care, whereas limited external validation on the generalizability. We aimed to independently validate the Hemodynamic Stability Index (HSI), a multi-parameter machine learning model, in predicting hemodynamic instability in Asian patients. Hemodynamic instability was marked by using inotropic, vasopressor, significant fluid therapy, and/or blood transfusions. This retrospective study included among 15,967 ICU patients who aged 20 years or older (not included 20 years) and stayed in ICU for more than 6 h admitted to Taipei Veteran General Hospital (TPEVGH) between January 1, 2010, and March 31, 2020, of whom hemodynamic instability occurred in 3053 patients (prevalence = 19%). These patients in unstable group received at least one intervention during their ICU stays, and the HSI score of both stable and unstable group was calculated in every hour before intervention. The model performance was assessed using the area under the receiver operating characteristic curve (AUROC) and was compared to single indicators like systolic blood pressure (SBP) and shock index. The hemodynamic instability alarm was set by selecting optimal threshold with high sensitivity, acceptable specificity, and lead time before intervention was calculated to indicate when patients were firstly identified as high risk of hemodynamic instability. The AUROC of HSI was 0.76 (95% CI, 0.75–0.77), which performed significantly better than shock Index (0.7; 95% CI, 0.69–0.71) and SBP (0.69; 95% CI, 0.68–0.70). By selecting 0.7 as a threshold, HSI predicted 72% of all 3053 patients who received hemodynamic interventions with 67% in specificity. Time-varying results also showed that HSI score significantly outperformed single indicators even up to 24 h before intervention. And 95% unstable patients can be identified more than 5 h in advance. The HSI has acceptable discrimination but underestimates the risk of stable patients in predicting the onset of hemodynamic instability in an external cohort.

中文翻译：

---

机器学习模型的外部验证以预测重症监护病房的血流动力学不稳定性

血流动力学不稳定的早期预测模型有可能改善重症监护，但对普遍性的外部验证有限。我们旨在独立验证血流动力学稳定性指数 (HSI)，一种多参数机器学习模型，用于预测亚洲患者的血流动力学不稳定。通过使用正性肌力药物、血管加压药、显着的液体治疗和/或输血来标记血流动力学不稳定。这项回顾性研究纳入了 2010 年 1 月 1 日至 2020 年 3 月 31 日期间在台北荣总医院 (TPEVGH) 收治的 15,967 名 20 岁或以上（不包括 20 岁）并在 ICU 停留超过 6 小时的 ICU 患者，其中 3053 名患者出现血流动力学不稳定（患病率 = 19%）。不稳定组患者在入住ICU期间至少接受过一次干预，干预前每小时计算稳定组和不稳定组的HSI评分。使用受试者工作特征曲线下面积 (AUROC) 评估模型性能，并与收缩压 (SBP) 和休克指数等单一指标进行比较。通过选择具有高灵敏度、可接受的特异性的最佳阈值来设置血流动力学不稳定警报，并计算干预前的提前时间，以指示何时首次将患者识别为血流动力学不稳定的高风险。恒指的 AUROC 为 0.76（95% CI，0.75-0.77），明显优于休克指数（0.7；95% CI，0.69-0.71）和 SBP（0.69；95% CI，0.68-0.70）。通过选择 0.7 作为阈值，HSI 预测了所有 3053 名接受血流动力学干预的患者中的 72%，特异性为 67%。时变结果还表明，即使在干预前 24 小时，HSI 评分也显着优于单一指标。95% 的不稳定患者可提前 5 小时以上识别。HSI 具有可接受的辨别力，但低估了稳定患者在预测外部队列中血流动力学不稳定发作的风险。