Time Windows for Predicting Deterioration in COVID-19 Patients

Background: COVID-19 has already become a global pandemic. There are few detailed descriptions of moderate cases and their clinical progression. Our aim was to identify important risk factors and key time windows for predicting disease deterioration. Early intervention is a crucial strategy for reducing mortality.Methods: All moderate cases hospitalized in Guangdong province were included in this retrospective, multi-center, cohort study, and divided into two groups: those showing deterioration or no-deterioration. The clinical course of the disease was divided into stages (incubation, prodromal, exacerbation, fastigium, remission and convalescence) and important time windows for clinical intervention to avoid deterioration were identified. Risk factors associated with disease deterioration were evaluated using multivariate logistic regression.Findings: 1167 of the 1168 patients included in this study were successfully treated and 1 patient died. The median age was 43·5 years (IQR: 32-57), and 472 (40%) were over 50 years old. 148 patients (13%) showed deterioration and developed severe or critical disease. The prodromal stage in patients showing deterioration was longer than those no deterioration (11 days vs 9 days, p <0·0001). 18 patients showing deterioration (12%) developed critical disease, and their symptoms developed more rapidly than patients who developed exacerbation, but not critical disease (3 days vs 8 days, p =0·003). Risk factors for deterioration identified using multivariate regression included age≥50 years, male gender, incubation period<8 days, diabetes, hypertension, fever, cough, and dyspnea.Interpretation: The 9th -11th day in the prodromal stage and the first three days of the exacerbation stage were identified as vital time windows of disease deterioration, which help clinicians to provide intervention beforehand. Risk factors such as older age, male gender, shorter incubation period, comorbidities (diabetes, hypertension), and symptoms (fever, cough, and dyspnea) could also help clinicians to identify patients before hospital admission whose condition is likely to deteriorate.Funding Statement: This study is supported by Special Project on Emergency Response to Control of Novel Coronavirus Infection of Guangdong Province.Declaration of Interests: All authors declare no competing interests.Ethics Approval Statement: The study was approved by the Institutional Review Board of Guangdong Health Commission and Guangdong Provincial People’s Hospital (No. GDREC2020028H). Written informed consent was waived in light of the urgent need to collect data.