Clinical and Laboratory Predictors of In-Hospital Mortality in 305 Patients with COVID-19: A Cohort Study in Wuhan, China

• DOI: 10.2139/ssrn.3546115
• Publication Date: 2020-03-24T11:45:37Z
• Authors: Kun Wang; Peiyuan Zuo; Yuwei Liu; Meng Zhang; Xiaofang Zhao; Songpu Xie; Hao Zhang; Xinglin Chen; Chengyun Liu
• Keywords: 2019-20 coronavirus outbreak

Background: COVID-19 has caused a large number of deaths in short period and lacks specific treatment. Early identification poor prognosis patients may facilitate doctors to offer proper supportive treatment advance. This study aimed develop death prediction models for patients. Methods: In this cohort study, participants were who had been hospitalized the First People's Hospi predictive tal Jiangxia District Wuhan from January 7, 2020 February 11, 2020. Clinical characteristics laboratory data collected. We selected baseline variables at admission through ensemble XGBoost model, stepwise Akaike information criterion (AIC) clinical significance. Two mortality built, one with other further included data. Findings: Before 12, 2020, 305 enrolled, whom 22 (7.2%) died during hospitalization 283 (92.8%) cured. The mean age was 47.8 years 53.4% female. Baseline Neutrophil count strongest predictor death, followed by age, plasma D-dimer, lymphocyte count, hsCRP, APTT, WBC, platelet history hypertension fever. model showed good discriminatory power (n=305, AUC 0.85, 95% CI 0.78–0.92, sensitivity 88.89%, specificity 73.98%, NPV 98.91%). addition test significantly (p=0.0493) improved (n=264, 0.92, 0.84–0.96, 94.44%, 76.02%, 99.47%). Interpretation: we developed two in-hospital Wuhan. They can help effectively predict an early stage, provide practical decision-making suggestions on which should be paid close attention given high-level treatments.Funding Statement: supported grant National Natural Science Foundation China (Grants: 81671386 81974222).Declaration Interests: authors declare no competing interests.Ethics Approval protocol approved Medical Ethics Committee Hospital complied Declaration Helsinki. verbally informed that their would used anonymously medical studies permission.