AUTHOR=Xu Rong , Hou Keke , Zhang Kun , Xu Huayan , Zhang Na , Fu Hang , Xie Linjun , Sun Ran , Wen Lingyi , Liu Hui , Yang Zhigang , Yang Ming , Guo Yingkun 

TITLE=Performance of Two Risk-Stratification Models in Hospitalized Patients With Coronavirus Disease

JOURNAL=Frontiers in Medicine

VOLUME=Volume 7 - 2020

YEAR=2020

URL=https://www.frontiersin.org/journals/medicine/articles/10.3389/fmed.2020.00518

DOI=10.3389/fmed.2020.00518

ISSN=2296-858X

ABSTRACT=Background 
Despite the increasing familiarity of the medical community with the epidemiological and clinical characteristics of coronavirus disease 2019 (COVID-19), there is currently a lack of rapid and effective risk stratification indicators in clinical practice to predict the poor clinical outcomes of COVID-19 especially in severe patients.
Methods
In this retrospective single-center study, we included 117 cases confirmed with COVID-19. The clinical, laboratory, and imaging features were collected and analyzed during admission. The MuLBSTA scale was used to assess the death and intensive care unit (ICU) risks in all patients. 
Results 
Of all 117 hospitalized patients, 21 (17.9%) patients were admitted to the ICU care, and 5 (4.3%) patients died. The median hospital stay was 12 (10–15) days. There were 18 patients had a MuLBSTA score greater than 12 points and were all of severe type. The ROC curve showed good efficiency of diagnosis death (area under the curve [AUC], 0.956; cutoff value, 12; specificity, 89.5%; sensitivity, 100%) and ICU (AUC, 0.875; cutoff value, 11; specificity, 91.7%; sensitivity, 71.4%). The K–M survival analysis showed that patients with a MuLBSTA score ≥ 12 had higher ICU (log-rank, P = 0.001) and high death (log-rank, P = 0.000) risks. 
Conclusions 
The MulBSTA score can effectively screen high-risk patients at admission. The higher score at admission have greater need for ICU care and higher risk of death in patients with COVID.