Intensive care unit admission and mortality risk prediction in covid19 patients with spectrum of comorbidities using clinical laboratory and CT severity score

Abstract

Introduction: Treatment of patients with COVID-19 in intensive care unit (ICU) has been a significant challenge. Early detection of COVID-19 related severity is critical for timely management of such patients. This retrospective observational study aimed to evaluate clinical characteristics such as laboratory parameters, associated comorbidities, computed tomography (CT)-based semi-quantitative score, and cumulative correlative measurement of critically ill COVID-19 patients to assess the requirement for early ICU admission. Materials and methods: Data from the treatment record of COVID-19 patients with severe pneumonia was collected at a tertiary care governament medical institution from 1st May, 2020 to 30th May, 2020. Binary logistic regression analysis, specificities and sensitivities of maximum D-dimer, ferritin, other laboratory measures, and CT-based semi-quantitative score for outcome cutoffs were calculated to predict early indicators of ICU admission and mortality. Results: Of 200 severe COVID-19 patients, 48% had comorbidities while 40% had mortality. The ICU admission rate was 43.5%, in which non-invasive ventillation and mechanical ventilation support were 27.5% and 16% respectively. Hypertension was reported for the most prevalent comorbidities (48%) followed by diabetes mellitus (24%), respiratory diseases (10.5%), renal diseases (5%), and chronic heart diseases (10%). Optimal cutoff for ICU requirement was 17.2 for neutrophil lymphocyte ratio (NLR), 727 ng/mL for ferritin, 2.2 mg/L FEU for D-dimer and 362.5 mg/L for C-reactive protein (CRP). Significantly higher risk of death was reported for patients having CT score of 14.5 (p