Identification of patients with low risk of severity in confirmed COVID-19 cases. Cuba, 2020- 2021

Abstract

Introduction: The COVID-19 pandemic, a treat to global health, overwhelmed health systems in many countries. Performing and adequate identification of subjects with very low risk to develop a severe form of the disease is of vital importance. Objective: To identify patients with very low risk of transition to severity in confirmed COVID-19 cases in Cuba. Material and Methods: A cohort retrospective study was carried out based on the Cuban National Database of confirmed COVID-19 patients. A total of 10 600 persons were analyzed between March 11th, 2020 and May 31st, 2021. Admission to the Intensive Care Units and mortality related to COVID-19, both measured 30 days after disease confirmation, were the selected response variables. As predictors, age and comorbidities were selected. A classification tree algorithm was applied to identify risk strata. For every stratum, its volume, the risk of transition to severity and the explained risk were calculated. Results: Subjects under 65 years old with no comorbidities (85.8% of the total sample) had a very low risk (0.2%) of admission to ICU and death (1.2%). Adjusted models showed a good calibration for both variables (accuracy= 0.88) and discriminant ability (area under the ROC curve: 0.89 and 0.88). Conclusions: The study allowed the identification of subjects with very low risk of transition to severity who were followed up at primary healthcare services, under home surveillance and symptomatic treatment.

Keywords

• prognosis, triaje, modelos teóricos, covid-19, sars-cov-2.