Background: To investigate deep vein thrombosis (DVT) in hospitalized patients with coronavirus disease 2019 (COVID-19), we performed a single institutional study to evaluate its prevalence, risk factors, prognosis, and potential thromboprophylaxis strategies in a large referral and treatment center. Methods: We studied a total of 143 patients with COVID-19 from January 29, 2020 to February 29, 2020. Demographic and clinical data, laboratory data, including ultrasound scans of the lower extremities, and outcome variables were obtained, and comparisons were made between groups with and without DVT. Results: Of the 143 patients hospitalized with COVID-19 (age 63±14 years, 74 [51.7%] men), 66 patients developed lower extremity DVT (46.1%: 23 [34.8%] with proximal DVT and 43 [65.2%] with distal DVT). Compared with patients who did not have DVT, patients with DVT were older and had a lower oxygenation index, a higher rate of cardiac injury, and worse prognosis, including an increased proportion of deaths (23 [34.8%] versus 9 [11.7%]; P =0.001) and a decreased proportion of patients discharged (32 [48.5%] versus 60 [77.9%]; P <0.001). Multivariant analysis showed an association only between CURB-65 (confusion status, urea, respiratory rate, and blood pressure) score 3 to 5 (odds ratio, 6.122; P =0.031), Padua prediction score ≥4 (odds ratio, 4.016; P =0.04), D-dimer >1.0 μg/mL (odds ratio, 5.818; P <0.014), and DVT in this cohort, respectively. The combination of a CURB-65 score 3 to 5, a Padua prediction score ≥4, and D-dimer >1.0 μg/mL has a sensitivity of 88.52% and a specificity of 61.43% for screening for DVT. In the subgroup of patients with a Padua prediction score ≥4 and whose ultrasound scans were performed >72 hours after admission, DVT was present in 18 (34.0%) patients in the subgroup receiving venous thromboembolism prophylaxis versus 35 (66.0%) patients in the nonprophylaxis group ( P =0.010). Conclusions: The prevalence of DVT is high and is associated with adverse outcomes in hospitalized patients with COVID-19. Prophylaxis for venous thromboembolism may be protective in patients with a Padua protection score ≥4 after admission. Our data seem to suggest that COVID-19 is probably an additional risk factor for DVT in hospitalized patients.

Full author information is available at the end of the article Xuefeng Zang and Qian Wang contributed equally to this work. Hua Zhou, Sanhong Liu and Xinying Xue contributed equally to this work. The members of the COVID-19 Early Prone Position Study Group are listed in the Acknowledgements.

Study objective The large number of clinical variables associated with coronavirus disease 2019 (COVID‐19) infection makes it challenging for frontline physicians to effectively triage COVID‐19 patients during the pandemic. This study aimed to develop an efficient deep‐learning artificial intelligence algorithm to identify top clinical variable predictors and derive a risk stratification score system to help clinicians triage COVID‐19 patients. Methods This retrospective study consisted of 181 hospitalized patients with confirmed COVID‐19 infection from January 29, 2020 to March 21, 2020 from a major hospital in Wuhan, China. The primary outcome was mortality. Demographics, comorbidities, vital signs, symptoms, and laboratory tests were collected at initial presentation, totaling 78 clinical variables. A deep‐learning algorithm and a risk stratification score system were developed to predict mortality. Data were split into 85% training and 15% testing. Prediction performance were compared with those using COVID‐19 severity score, CURB‐65 score and pneumonia severity index (PSI). Results Of the 181 COVID‐19 patients, 39 expired and 142 survived. Five top predictors of mortality were D‐dimer, O 2 Index, neutrophil:lymphocyte ratio, C‐reactive protein, and lactate dehydrogenase. The top 5 predictors and the resultant risk score yielded, respectively, an area under curve (AUC) of 0.968 ([95% CI:0.87‐1.0]) and 0.954 ([95% CI:0.80‐0.99]) for the testing dataset. Our models outperformed COVID‐19 severity score (AUC = 0.756), CURB‐65 score (AUC = 0.671), and PSI (AUC = 0.838). The mortality rates for our risk stratification scores (0‐5) were 0, 0, 6.7, 18.2, 67.7, and 83.3%, respectively. Conclusions and relevance Deep‐learning prediction model and the resultant risk stratification score may prove useful in clinical decision‐making under time‐sensitive and resource‐constrained environment. This article is protected by copyright. All rights reserved