Early Epidemiological and Clinical Analysis of the First 200 Patients with COVID-19 Admitted via the Emergency Department in Kings College Hospital, London: A Retrospective Cohort Study

Abstract

Background: In December 2019, SARS-CoV-2 was identified as the cause of a cluster of cases of pneumonia subsequently referred to as coronavirus disease 2019 (COVID-19) described in Wuhan, Hubei Province, in China which rapidly spread leading to a worldwide pandemic. We describe the epidemiological, clinical, laboratory, and current clinical outcomes of patients admitted via the Emergency Department at Kings College Hospital, London.

Methods: A single centre retrospective observational study of the first 200 patients admitted via the Emergency Department at Kings College Hospital, London, that subsequently tested positive for SARS Co-V in any nasopharyngeal swabs. The first patient was admitted on March 3 and the last patient on March 26. Data were obtained from the admission records and electronic patient records. Chest X-rays were scored using the adapted RALE-score for COVID-19 as introduced by Wang 1 . The destination from the Emergency Department was recorded, in relation to the ceiling of care. Outcomes were followed up until March 28 and compared between admission destination and mortality.

Findings: A total of 200 patients admitted to King’s College Hospital via the Emergency Department with laboratory confirmed SARS-CoV-2 were included. Data collection ended at March 28. Median age of the patients admitted was 63 years, 115 (57.5%) were male. 103 (51.5%) patients were of black ethnicity (51.5%), compared to 47 (23.5%) of white and 29 (14.5%) of any other ethnic background. Most common comorbidities were hypertension (n=98, 49%) and diabetes (n=77, 38.5%). The most common clinical symptoms at presentation were fever (n=147, 73.5%), cough (n=139, 69.5%), shortness of breath (n=110, 55%) and lethargy (108, 54%). Median NEWS2 was 5, median Radiological Severity Score was 3 (range 0-8). Of the 200 patients, 108 (54%) went to the ward with a ceiling of care, 70 (35%) to the ward with a ceiling of care and 22 (11%) to Critical Care. Mortality rate is estimated to be 16% at 10 days.

There was a significant difference found in laboratory markers such as neutrophil count, creatinine, urea and CRP between different admission destinations. The Radiological Severity Score differed significantly with people admitted on Critical Care having a score of 5.5 compared to admission on the ward (without ceiling of care) of 2.9. When comparing ethnicities, the gender distribution in these groups differed significantly (p=0.049) with a more equal distribution between sexes in people from black ethnicity. Furthermore, patients of black and other ethnicities were significantly younger (60.6 and 59.8 years respectively) in contrast to patients of white ethnicity (74.2 years).

Analyses demonstrated that ethnic minorities (OR 3.17), having 1 or more comorbidity (OR 2), higher neutrophil count (OR 3.92), higher creatinine (OR 2.50, SD 1.14) and higher CRP (OR 2.82) appear to be predictors of Critical Care admission. After multivariable regression analysis of this small sample, only Radiological Severity Score is a significant predictor with an adjusted OR of 1.64 respectively 1.28.

Interpretation: Admission for COVID-19 was correlated to race and the prevalence of hypertension, diabetes and underlying lung disease. The likelihood of the need for Critical Care was associated with both NEWS2 and Radiological Severity Score. Further follow up of all admissions as well as new research is needed to address possible other attributing factors such as ethnicity and laboratory markers.

Funding Statement: None.

Declaration of Interests: None.

Ethics Approval Statement: The study was approved as a service evaluation, and formal ethics was not obtained. In line with the Health Research Authority Design tool, the study uses only data collected during routine care, the analysis was undertaken by the clinical team who already have access to all data, and the primary reason for conducting the study was to inform local service delivery. Data was stored anonymously. In addition, reporting used aggregate findings only ensuring patient anonymity.