Wijnberge M, van der Ster BJP, Geerts BF, et al. - Given a machine-learning-derived algorithm created to predict hypotension on the basis of arterial blood pressure (ABP) waveforms significantly decreased intraoperative hypotension, so, researchers undertook this observational cohort study to compare the performance of the Hypotension Prediction Index (HPI) algorithm, using noninvasive vs invasive ABP measurements, at a mathematically optimal HPI alarm threshold (Youden index). In addition, they evaluated the algorithm's performance using a HPI alarm threshold of 85 that is currently applied in clinical trials. Overall 507 adult patients receiving general surgery were included. Similar performance of the algorithm was evident with invasive and noninvasive ABP input. A HPI alarm threshold of 85 demonstrated a median [IQR] time from alarm to hypotension of 2.7 [1.0 to 7.0] min with a sensitivity, specificity, positive and negative predictive values of 92.7, 87.6, 79.9, and 95.8, respectively. Based on findings of this study, the algorithm can be applied using continuous noninvasive ABP waveforms. This unfolds the potential to predict as well as avert hypotension in a larger patient populace.