Learning Models for Forecasting Hospital Resource Utilization for COVID-19 Patients in Canada

Abstract COVID-19 pandemic has overwhelmed health systems and hospital capacity in Canada. Hospitals are facing a crisis-level shortage of critical supplies (e.g., hospital and ICU beds) and equipment (e.g., ventilators). This motivates our need for models that can accurately forecast regional demand for hospital resources. This work aims to create predictive models that can use Canada COVID-19 data and pandemic-related factors to accurately forecast 5 quantities -- three related to hospital resource utilization (i.e., number of hospital beds, ICU beds, and ventilators that will be needed by COVID-19 patients) and two to the pandemic progress (i.e., number of COVID-19 cases and COVID-19 deaths) -- several weeks in advance. We developed a machine learning method that can use information (i.e., resource utilization, pandemic progress, population mobility, weather condition, and public policy) currently known about a region since March 2020, to learn 116 temporal convolutional network (TCN) models every week; each used for forecasting the weekly average of one of these 5 quantities in Canada (respectively, in six specific provinces) for each of the next 4 weeks -- e.g., on 20 Nov 2020, forecasting the average number of ICU beds that will be needed by COVID-19 patients in the region of Ontario for the week ending on 18 Dec 2020. We compared our method, versus other standard models, on the COVID-19 data and hospital resource data, on the tasks of predicting these 116 values, every week from Oct 2020 to July 2021. Experimental results show that our 4640 TCN models (each forecasting a regional target for a specific future time, on a specific date) can produce accurate forecasts of demand for every hospital resource (i.e., hospital beds, ICU beds, and ventilators) and pandemic progress (i.e., number of COVID-19 cases and COVID-19 deaths) for each week from 2 Oct 2020 to 2 July 2021. Compared to other state-of-the-art predictive models, our TCN models yield the lowest mean absolute percentage error (MAPE). We developed and validated an accurate COVID-19 forecasting method based on the TCN models that can effectively forecast the hospital resource utilization and pandemic progress for Canada and for each of six provinces.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.