Predicting consequences of COVID-19 control measure de-escalation on nosocomial transmission and mortality: a modelling study in a French rehabilitation hospital.

Infection control measures are effective for nosocomial COVID-19 prevention but bear substantial health-economic costs, motivating their "de-escalation" in settings at low risk of SARS-CoV-2 transmission. Yet consequences of de-escalation are difficult to predict, particularly in light of novel variants and heterogeneous population immunity. To estimate how infection control measure de-escalation influences nosocomial COVID-19 risk. An individual-based transmission model was used to simulate SARS-CoV-2 outbreaks and control measure de-escalation in a French long-term care hospital with multi-modal control measures in place (testing and isolation, universal masking, single-occupant rooms). Estimates of COVID-19 case fatality rates (CFRs) from reported outbreaks were used to quantify excess COVID-19 mortality due to de-escalation. In a population fully susceptibility to infection, de-escalating both universal masking and single rooms resulted in hospital-wide outbreaks of 114 (95% CI: 103-125) excess infections, compared to 5 (3-7) excess infections when de-escalating only universal masking or 15 (11-18) when de-escalating only single rooms. When de-escalating both measures and applying CFRs from the first wave of COVID-19, excess patient mortality ranged from 1.57 (1.41-1.71) to 9.66 (8.73-10.57) excess deaths/1,000 patient-days. By contrast, when applying CFRs from subsequent pandemic waves and assuming susceptibility to infection among 40-60% of individuals, excess mortality ranged from 0 (0-0) to 0.92 (0.77-1.07) excess deaths/1,000 patient-days. The de-escalation of bundled COVID-19 control measures may facilitate widespread nosocomial SARS-CoV-2 transmission. However, excess mortality is likely limited in populations at least moderately immune to infection and given CFRs resembling those estimated during the "post-vaccine" era.

Informations de copyright

Copyright © 2024. Published by Elsevier Ltd.

Déclaration de conflit d'intérêts

Declaration of Competing Interest None declared.