COVID-19 Nosocomial Transmission Dynamics, a Retrospective Cohort Study of Two Healthcare Associated Clusters in a District Hospital in England during March and April 2020

Summary Objective: To understand the transmission dynamics of SARS-CoV-2 in a hospital outbreak to inform infection control actions. Design: Retrospective cohort study. Setting: General medical and elderly inpatient wards in a hospital in England. Methods: COVID-19 patients were classified as community or healthcare-associated by time from admission to onset/positivity using European Centre for Disease Prevention and Control definitions. COVID-19 symptoms were classified as asymptomatic, non-respiratory or respiratory. Infectiousness was calculated from 2 days prior to 14 days post symptom onset or positive test. Cases were defined as healthcare-associated COVID-19 patients where infection was acquired from the wards under investigation. COVID-19 exposures were calculated based on symptoms and bed proximity to an infectious patient. Risk ratios and adjusted odds ratios (aOR) were calculated from univariable and multivariable logistic regression. Results: Of 153 patients: 65 were COVID-19 patients (45 healthcare-associated). Exposure to a COVID-19 patient with respiratory symptoms was associated with healthcare-associated infection irrespective of proximity (aOR 3.81; 95%CI 1.6.3-8.87), non-respiratory exposure was only significant within 2.5m (aOR 5.21; 95%CI 1.15-23.48). A small increase in risk ratio was observed for exposure to a respiratory patient for >1 day compared to 1 day from 2.04 (95%CI 0.99-4.22) to 2.36 (95%CI 1.44-3.88) Discussion: Respiratory exposure anywhere within a 4-bedded bay was a risk whereas non-respiratory exposure required bed distance ≤2.5m. Standard Infection control measures required beds to be >2m apart; our study suggests this may be insufficient to stop SARS-CoV-2 spread. We recommend improving cohorting and further studies into bed distance and transmission factors.