Receive a weekly summary and discussion of the top papers of the week by leading researchers in the field.

In Journal of the American Medical Informatics Association : JAMIA

OBJECTIVE : Extracorporeal membrane oxygenation (ECMO) resource allocation tools are currently lacking. We developed machine learning (ML) models for predicting COVID-19 patients at risk of receiving ECMO to guide patient triage and resource allocation.

MATERIAL AND METHODS : We included COVID-19 patients admitted to intensive care units for >24 hours from March 2020 to October 2021, divided into training and testing development and testing only holdout cohorts. We developed ECMO-deployment timely prediction model ForecastECMO using Gradient Boosting Tree (GBT), with pre-ECMO prediction horizons from 0-48 hours, compared to PaO2/FiO2 (PF) ratio, Sequential Organ Failure Assessment (SOFA) score, PREdiction of Survival on ECMO Therapy-Score (PRESET) score, logistic regression (LR), and 30 pre-selected clinical variables GBT Clinical GBT models, with area under the receiver operator curve (AUROC) and precision recall curve (AUPRC) metrics.

RESULTS : ECMO prevalence was 2.89% and 1.73% in development and holdout cohorts. ForecastECMO had the best performance in both cohorts. At the 18-hour prediction horizon, a potentially clinically actionable pre-ECMO window, ForecastECMO had the highest AUROC (0.94 & 0.95) and AUPRC (0.54 & 0.37) in development and holdout cohorts in identifying ECMO patients without data 18-hours prior to ECMO.

DISCUSSION AND CONCLUSION : We developed a multi-horizon model, ForecastECMO, with high performance in identifying patients receiving ECMO at various prediction horizons. This model has potential to be used as early alert tool to guide ECMO resource allocation for COVID-19 patients. Future prospective multi-center validation would provide evidence for generalizability and real-world application of such models to improve patient outcomes.

Xue Bing, Shah Neel, Yang Hanqing, Kannampallil Thomas, Payne Philip Richard Orrin, Lu Chenyang, Said Ahmed Sameh


COVID-19, ECMO, early alert, machine learning, prediction, resource allocation