Predictive modeling to estimate the demand for intensive care hospital beds nationwide in the context of the COVID-19 pandemic

• Published in: Medwave Vol. 20; no. 9; p. e8039
• Main Authors: Peña, Víctor Hugo, Espinosa, Alejandra
• Format: Journal Article
• Language: Spanish; English
• Published: Chile Medwave Estudios Limitada 05.10.2020
• Subjects: 2019 novel coronavirus disease; Chile - epidemiology; Coronavirus Infections - epidemiology; COVID-19; emergency medicine; epidemiology; Hospital Bed Capacity - statistics & numerical data; hospitalization; Humans; Intensive Care Units - supply & distribution; Models, Statistical; Pandemics; Pneumonia, Viral - epidemiology; public health; viruses; Chile; epidemiology; viruses; 2019 novel coronavirus disease; hospitalization; emergency medicine; public health
• DOI: 10.5867/medwave.2020.09.8039
• ISSN: 0717-6384

SARS CoV-2 pandemic is pressing hard on the responsiveness of health systems worldwide, notably concerning the massive surge in demand for intensive care hospital beds. This study proposes a methodology to estimate the saturation moment of hospital intensive care beds (critical care beds) and determine the number of units required to compensate for this saturation. A total of 22,016 patients with diagnostic confirmation for COVID-19 caused by SARS-CoV-2 were analyzed between March 4 and May 5, 2020, nationwide. Based on information from the Chilean Ministry of Health and ministerial announcements in the media, the overall availability of critical care beds was estimated at 1,900 to 2,000. The Gompertz function was used to estimate the expected number of COVID-19 patients and to assess their exposure to the available supply of intensive care beds in various possible scenarios, taking into account the supply of total critical care beds, the average occupational index, and the demand for COVID-19 patients who would require an intensive care bed. A 100% occupancy of critical care beds could be reached between May 11 and May 27. This condition could be extended for around 48 days, depending on how the expected over-demand is managed. A simple, easily interpretable, and applicable to all levels (nationwide, regionwide, municipalities, and hospitals) model is offered as a contribution to managing the expected demand for the coming weeks and helping reduce the adverse effects of the COVID-19 pandemic.