Machine learning algorithms can outperform older methods in predicting clinical deterioration, but rigorous prospective data on their real-world efficacy are limited. We hypothesized that real-time machine learning generated alerts sent directly to front-line providers would reduce escalations.

中文翻译：

---

实时机器学习警报以防止护理升级：一项非随机集群实用临床试验。

机器学习算法在预测临床恶化方面可以优于旧方法，但有关其现实世界疗效的严格前瞻性数据有限。我们假设实时机器学习生成的警报直接发送给一线提供商将减少升级。