Independent Predictors of In-Hospital Mortality in Patients with COVID-19 in London, United Kingdom: An Observational Cohort Study

Background: Coronavirus disease 2019 (COVID-19) has rapidly evolved as a global health crisis, the majority of patients affected with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) will present with mild symptoms, respiratory disease and acute respiratory distress syndrome (ARDS) has been widely reported as one of the most common features of COVID-19. Although many studies have reported the demographics, clinical presentations, and outcomes in COVID-19 all over the world, studies with a large population affected with COVID-19 from the UK are still limited. Therefore, the aim of the study was to describe demographics, baseline clinical characteristics, comorbidities, and outcomes, and furthermore identify risk factors of in-hospital mortality using a large observational study cohort of COVID-19 patients in a busy hospital in London, UK. Methods: We analyze 393 consecutive hospitalized patients with confirmed COVID-19 admitted to Queen Elizabeth Hospital, London from March 12, 2020 to April 18, 2020. Demographics, clinical presentations, laboratory tests, radiological assessments, and in-hospital outcomes were collected and analyzed. Univariate analysis and multivariate Cox regression were performed to determine independent predictors of in-hospital mortality. Findings: In our cohort, 393 patients with COVID-19 were identified (mean age, 67.5 years; 59% males). The majority of patients were Caucasian (66.9%) and most common signs were cough (49.9%), fever (46.8%), and dyspnea (46.1%). A total of 161 (40.9%) of patients died within 30 days of admission. The average age in the non-survival group was older than that in the survival group (75.2 vs. 62.2 years old, q<0.001, family discovery rate corrected). The mortality was higher in Caucasian (48.9%) than African (19.3%) and Asian patients (29.2%) (q<0.001). The median duration from symptom onset to in-hospital outcome was shorter in the non-survival group compared to patients who were discharged home (4 vs. 7 days, q=0.021). Compared to patients who survived, comorbidities that were more commonly found in the non-survival group included cardiovascular diseases such as arrhythmia (66.0%) and pump failure (76.5%), diabetes (50.4%), and dementia (58.3%), (q<0.05). Comparison of biochemical results reveals increased numbers of white blood cells, sodium, chloride, potassium, C-reactive protein, troponin, phosphate, urea, and creatinine in the non-survival group vs. the survival group (q<0.05). Moreover, the cox regression model demonstrated that patients with age over 65 years old (adjusted hazard ratio (HR), 2.36 [95% confidence interval (CI), 1.50-3.70]), ARDS (adjusted HR, 3.49 [95% CI, 2.43-5.00]), AKI stage III (adjusted HR, 2.30 [95% CI, 1.03-5.12]), hypernatremia (adjusted HR, 1.68 [95% CI, 1.00-2.81]), and high level of baseline serum white blood cells (adjusted HR, 1.50 [95% CI, 1.03-2.18]) were independently associated with increased likelihood of in-hospital mortality. Interpretation: This case series provides valuable insights into clinical characteristics and independent risk factors of mortality in hospitalized patients with confirmed COVID-19 within the UK health care system. We are the first to report that hypernatremia is an independent predictor of mortality in COVID-19 patients. Funding: This work was supported by the funding to Yao through the Chinese Postdoctoral Science Foundation (2019M663271) and the Chinese Scholarship Council (201808440461).Declaration of Interests: N/AEthics Approval Statement: The study was approved as a minimal-risk research using data that had already been collected for patients as part of their normal care, was not an interventional study and waived the requirement for informed consent.