Venous thromboembolism in COVID-19 patients and prediction model: a multicenter cohort study

• DOI: 10.1186/s12879-022-07421-3
• Publication Date: 2022-05-13T11:09:48Z
• Authors: Yi Lee; Qasim Jehangir; Pin Li; Deepthi Gudimella; Pooja Mahale; Chun-Hui Lin; Dinesh R. Apala; Geetha Krishnamoo...; Abdul R. Halabi; Kiritkumar Patel; Laila Poisson; Venugopal Balijep...; Anupam A. Sule; Girish B. Nair
• Keywords: Adult; Venous Thrombosis; Pulmonary embolism; Anticoagulants; COVID-19; Infectious and parasitic diseases; RC109-216; Venous Thromboembolism; Risk prediction; 3. Good health; Cohort Studies; 03 medical and health sciences; 0302 clinical medicine; Risk Factors; Deep vein thrombosis; Humans; Pulmonary Embolism; Risk stratification; Venous thromboembolism; Research Article; Retrospective Studies

Patients with COVID-19 infection are commonly reported to have an increased risk of venous thrombosis. The choice anti-thrombotic agents and doses currently being studied in randomized controlled trials retrospective studies. There exists a need for individualized stratification thromboembolism (VTE) assist clinicians decision-making on anticoagulation. We sought identify the factors VTE patients, which could help physicians prevention, early identification, management hospitalized patients improve clinical outcomes these patients.This is multicenter, database four main health systems Southeast Michigan, United States. compiled comprehensive data adult who were admitted between 1st March 2020 31st December 2020. Four models, including random forest, multiple logistic regression, multilinear decision trees, built primary outcome in-hospital acute deep vein thrombosis (DVT) pulmonary embolism (PE) tested performance. study also hospital length stay (LOS) intensive care unit (ICU) LOS non-VTE patients. models assessed using area under receiver operating characteristic curve confusion matrix.The cohort included 3531 admissions, 3526 had discharge diagnoses, 6.68% developed (N = 236). group longer ICU than (hospital 12.2 days vs. 8.8 days, p < 0.001; 3.8 1.9 0.001). 9.8% required more advanced oxygen support, compared 2.7% (p Among all forest model best suggested that blood pressure, electrolytes, renal function, hepatic enzymes, inflammatory markers predictors patients.Patients high VTE, prolonged stay. This prediction identifies aid making judgment empirical dosages