Risk factors for clinical deterioration in patients admitted for COVID-19: A case-control study

Uranga, A.; Villanueva, A.; Lafuente, I.; González, N.; Legarreta, M. J.; Aguirre, U.; España, P. P.; Quintana, J. M.; García-Gutiérrez, S.

Introduction There is controversy regarding the best predictors of clinical deterioration in COVID-19. Objective This work aims to identify predictors of risk factors for deterioration in patients hospitalized due to COVID-19. Methods design Nested case-control study within a cohort. Setting: 13 acute care centers of the Osakidetza-Basque Health Service. Participants: Patients hospitalized for COVID-19 with clinical deterioration—defined as onset of severe ARDS, ICU admission, or death—were considered cases. Two controls were matched to each case based on age. Sociodemographic data; comorbidities; baseline treatment; symptoms; date of onset; previous consultations; and clinical, analytical, and radiological variables were collected. An explanatory model of clinical deterioration was created by means of conditional logistic regression. Results A total of 99 cases and 198 controls were included. According to the logistic regression analysis, the independent variables associated with clinical deterioration were: emergency department O2 saturation ≤90% (OR 16.6; 95%CI 4–68), pathological chest X-ray (OR 5.6; 95%CI 1.7–18.4), CRP > 100 mg/dL (OR 3.62; 95%CI 1.62–8), thrombocytopenia with <150,000 platelets (OR 4; 95%CI 1.84–8.6); and a medical history of acute myocardial infarction (OR 15.7; 95%CI, 3.29–75.09), COPD (OR 3.05; 95%CI 1.43-6.5), or HT (OR 2.21; 95%CI 1.11–4.4). The model’s AUC was 0.86. On the univariate analysis, female sex and presence of dry cough and sore throat were associated with better clinical progress, but were not found to be significant on the multivariate analysis. Conclusion The variables identified could be useful in clinical practice for the detection of patients at high risk of poor outcomes.