The Physiological Deep Learner: First application of multitask deep learning to predict hypotension in critically ill patients.

Critical care clinicians are trained to analyze simultaneously multiple physiological parameters to predict critical conditions such as hemodynamic instability. We developed the Multi-task Learning Physiological Deep Learner (MTL-PDL), a deep learning algorithm that predicts simultaneously the mean arterial pressure (MAP) and the heart rate (HR). In an external validation dataset, our model exhibited very good calibration: R2 of 0.747 (95% confidence interval, 0.692 to 0.794) and 0.850 (0.815 to 0.879) for respectively, MAP and HR prediction 60-minutes ahead of time. For acute hypotensive episodes defined as a MAP below 65 mmHg for 5 min, our MTL-PDL reached a predictive value of 90% for patients at very high risk (predicted MAP ≤ 60 mmHg) and 2‰ for patients at low risk (predicted MAP >70 mmHg). Based on its excellent prediction performance, the Physiological Deep Learner has the potential to help the clinician proactively adjust the treatment in order to avoid hypotensive episodes and end-organ hypoperfusion.

Prediction of an Acute Hypotensive Episode During an ICU Hospitalization With a Super Learner Machine-Learning Algorithm.

BACKGROUND: Acute hypotensive episodes (AHE), defined as a drop in the mean arterial pressure (MAP) <65 mm Hg lasting at least 5 consecutive minutes, are among the most critical events in the intensive care unit (ICU). They are known to be associated with adverse outcome in critically ill patients. AHE prediction is of prime interest because it could allow for treatment adjustment to predict or shorten AHE. METHODS: The Super Learner (SL) algorithm is an ensemble machine-learning algorithm that we specifically trained to predict an AHE 10 minutes in advance. Potential predictors included age, sex, type of care unit, severity scores, and time-evolving characteristics such as mechanical ventilation, vasopressors, or sedation medication as well as features extracted from physiological signals: heart rate, pulse oximetry, and arterial blood pressure. The algorithm was trained on the Medical Information Mart for Intensive Care dataset (MIMIC II) database. Internal validation was based on the area under the receiver operating characteristic curve (AUROC) and the Brier score (BS). External validation was performed using an external dataset from Lariboisière hospital, Paris, France. RESULTS: Among 1151 patients included, 826 (72%) patients had at least 1 AHE during their ICU stay. Using 1 single random period per patient, the SL algorithm with Haar wavelets transform preprocessing was associated with an AUROC of 0.929 (95% confidence interval [CI], 0.899-0.958) and a BS of 0.08. Using all available periods for each patient, SL with Haar wavelets transform preprocessing was associated with an AUROC of 0.890 (95% CI, 0.886-0.895) and a BS of 0.11. In the external validation cohort, the AUROC reached 0.884 (95% CI, 0.775-0.993) with 1 random period per patient and 0.889 (0.768-1) with all available periods and BSs <0.1. CONCLUSIONS: The SL algorithm exhibits good performance for the prediction of an AHE 10 minutes ahead of time. It allows an efficient, robust, and rapid evaluation of the risk of hypotension that opens the way to routine use.

Bioactive Adrenomedullin, Organ Support Therapies, and Survival in the Critically Ill: Results from the French and European Outcome Registry in ICU Study.

OBJECTIVES: Adrenomedullin has vascular properties and elevated plasma adrenomedullin levels were detected in sepsis. We assessed, in septic and nonseptic ICU patients, the relation between circulating adrenomedullin, the need for organ support and mortality, using an assay of bioactive adrenomedullin. DESIGN: Prospective multicenter observational cohort study. SETTING: Data from the French and euRopean Outcome reGistry in ICUs study. PATIENTS: Consecutive patients admitted to intensive care with a requirement for invasive mechanical ventilation and/or vasoactive drug support for more than 24 hours following ICU admission and discharged from ICU were included. INTERVENTIONS: Clinical and biological parameters were collected at baseline, including bioactive-adrenomedullin. Status of ICU survivors was assess until 1 year after discharge. The main outcome was the need for organ support, including renal replacement therapy and/or for inotrope(s) and/or vasopressor(s). Secondary endpoints were the ICU length of stay and the 28-day all-cause mortality. MEASUREMENTS AND MAIN RESULTS: Median plasma bioactive adrenomedullin (n = 2,003) was 66.6 pg/mL (34.6-136.4 pg/mL) and the median Simplified Acute Physiology Score II score 49 (36-63). Renal replacement therapy was needed in 23% and inotropes(s) and/or vasopressor(s) in 77% of studied patients. ICU length of stay was 13 days (7-21 d) and mortality at 28 days was 22 %. Elevated bioactive adrenomedullin independently predicted 1) the need for organ support (odds ratio, 4.02; 95% CI, 3.08-5.25) in ICU patients whether admitted for septic or nonseptic causes and 2) the need for renal replacement therapy (odds ratio, 4.89; 3.83-6.28), and for inotrope(s) and/or vasopressor(s) (odds ratio, 3.64; 2.84-4.69), even in patients who were not on those supports at baseline. Elevated bioactive adrenomedullin was also associated with a prolonged length of stay (odds ratio, 1.85; 1.49-2.29) and, after adjustment for Simplified Acute Physiology Score II, with mortality (odds ratio, 2.31; 1.83-2.92). CONCLUSIONS: Early measurement of bioactive adrenomedullin is a strong predictor of the need of organ support and of short-term mortality in critically ill patients.