Modeling the COVID-19 outbreak in Ecuador: Is it the right time to lift social distancing containment measures?

Objective. Model the effect of partial and full reversal of containment measures on COVID-19 morbidity and mortality in Ecuador. Methods. Susceptible, Infected, Recovered (SIR) models were used to simulate the transmission dynamics of COVID-19 before and after the implementation (and reversal) of containment measures. A Healthcare Compartmental Epidemic Model (HeCEM), which accounts for Hospital and Intensive Care Unit admission rates, was developed to also simulate the effect of reversing social distancing containment measures. Reported COVID-19 cases between February 29th and April 23rd, 2020 were obtained from the Servicio Nacional de Gesti&oacuten de Riesgos y Emergencia, the national emergency management office of Ecuador. An ARIMA model was used to forecast reported number of cases based on the reported number of cases. SIR, HeCEM, and ARIMA model prediction errors were estimated. Results. SIR and HeCEM models predict that, at the moment, hospital and ICU bed needs for COVID-19 patients exceed the capacity in Ecuador. Partial or full reversal of containment measures before reaching the point where hospital and ICU beds are enough to meet the expected demand will result in secondary waves that delay reaching this equilibrium, resulting in thousands of excess deaths. Forecasts predict over 50,000 reported COVID-19 cases by July 25th, 2020. Conclusion. Partial reversal of containment measures should occur only after enough hospital and beds are available to meet the demand.