Severity of Omicron (B.1.1.529) and Delta (B.1.617.2) SARS-CoV-2 infection among hospitalised adults: A prospective cohort study in Bristol, United Kingdom

Abstract

Background: There is an urgent public health need to evaluate disease severity in adults hospitalised with Delta and Omicron SARS-CoV-2 variant infections. However, limited data exist assessing severity of disease in adults hospitalised with Omicron SARS-CoV-2 infections, and to what extent patient-factors, including vaccination, age, frailty and pre-existing disease, affect variant-dependent disease severity.

Methods: A prospective cohort study of adults (≥18 years of age) hospitalised with acute lower respiratory tract disease at acute care hospitals in Bristol, UK conducted over 10-months. Delta or Omicron SARS-CoV-2 infection was defined by positive SARS-CoV-2 PCR and variant identification or inferred by dominant circulating variant. We constructed adjusted regression analyses to assess disease severity using three different measures: FiO₂ >28% (fraction inspired oxygen), World Health Organization (WHO) outcome score >5 (assessing need for ventilatory support), and hospital length of stay (LOS) >3 days following admission for Omicron or Delta infection.

Findings: Independent of other variables, including vaccination, Omicron variant infection in hospitalised adults was associated with lower severity than Delta. Risk reductions were 58%, 67%, and 16% for supplementary oxygen with >28% FiO₂ [Relative Risk (RR) = 0.42 (95%CI: 0.34-0.52), P < 0.001], WHO outcome score >5 [RR = 0.33 (95%CI: 0.21-0.50), P < 0.001], and to have had a LOS > 3 days [RR = 0.84 (95%CI: 0.76-0.92), P < 0.001]. Younger age and vaccination with two or three doses were also independently associated with lower COVID-19 severity.

Interpretation: We provide reassuring evidence that Omicron infection results in less serious adverse outcomes than Delta in hospitalised patients. Despite lower severity relative to Delta, Omicron infection still resulted in substantial patient and public health burden and an increased admission rate of older patients with Omicron which counteracts some of the benefit arising from less severe disease.

Funding: AvonCAP is an investigator-led project funded under a collaborative agreement by Pfizer.

Keywords: COVID-19; Respiratory infection; SARS-CoV-2; Vaccination.

Fig. 1

Flow diagram of adults hospitalised with SARS-CoV-2 infection. Inclusion and exclusion criteria in the cohort stratified by the primary comparison or SARS-CoV-2 variant. Any patient may be excluded for multiple reasons at each stage, hence the counts of reasons for exclusion may not add up to the number of patients excluded. FiO₂, fraction of inspired oxygen; LOS, length of stay; VOC, variant of concern; WHO, World Health Organisation.

Fig. 2

Hospitalisations with SARS-CoV-2 infection throughout the study. (A) Daily rates of hospitalisations with Delta and Omicron SARS-CoV-2 infection, (B) 14-day rolling proportions of hospitalisations by vaccination status (excluding people partially vaccinated on admission) and (C) 14-day rolling proportions of hospitalisations by age groups (in years) throughout the study. Vertical lines represent the earliest time an Omicron case was admitted and the latest time a Delta case was admitted. Before the earliest Omicron detection, cases are assumed to be Delta when sequencing results are not available, and after the last Delta detection, cases are assumed to be Omicron, when sequencing results are not available.

Fig. 3

The detailed breakdown of the three selected indicators of severity of hospital admission stratified by SARS-CoV-2 variant, and the comparison of three binary indicators of hospital burden and their relationship to SARS-CoV-2 variant and patient age, for patients admitted to hospital. The distribution of patients (A) requiring different peak levels of oxygen supplementation, (B) with different ventilation requirements as defined by the WHO outcome or clinical progression score, and (C) with different lengths of stay. The proportion of patients who (D) require high flow oxygen >28% FiO₂, (E) have a WHO outcome score>5 (requiring NIPPV), and (F) have a hospital length of stay greater than three days, as assessed on the seventh day following admission. Error bars show 95% binomial confidence intervals for each outcome, compared to other outcomes. FiO₂, Fraction of inspired oxygen; LOS, length of stay; NIPPV, non-invasive positive pressure ventilation.

Fig. 4

Robust Poisson regression model relative risks for the three indicators of hospital burden. Relative risks describing the effect of different predictors on whether patients require oxygen supplementation with FiO₂ >28% (first column), have a WHO outcome score greater than 5 (second column) or who remain in hospital for more than 3 days (third column) measured on the seventh day after admission. Numerical details of these regression models, including other explanatory variables that are not our primary interest, are given in Supplementary Table S15. A comparison of relative risks computed by different methods and odds ratios by logistic regression is shown in Supplementary Table S18. FiO₂ – fraction inspired oxygen, LOS – length of stay, CCI – Charlson Comorbidity Index.