Hospitalization and Mortality Risk for COVID-19 Cases With SARS-CoV-2 AY.4.2 (VUI-21OCT-01) Compared to Non-AY.4.2 Delta Variant Sublineages

Abstract To investigate if the AY.4.2 sublineage of the SARS-CoV-2 delta variant is associated with hospitalization and mortality risks that differ from non-AY.4.2 delta risks, we performed a retrospective cohort study of sequencing-confirmed COVID-19 cases in England based on linkage of routine health care datasets. Using stratified Cox regression, we estimated adjusted hazard ratios (aHR) of hospital admission (aHR = 0.85; 95% confidence interval [CI], .77–.94), hospital admission or emergency care attendance (aHR = 0.87; 95% CI, .81–.94), and COVID-19 mortality (aHR = 0.85; 95% CI, .71–1.03). The results indicate that the risks of hospitalization and mortality are similar or lower for AY.4.2 compared to cases with other delta sublineages.


METHODS
The study population comprised COVID-19 cases in England with a first positive specimen between 21 June and 7 November 2021 who were infected with AY.4.2 or a non-AY.4.2 delta variant based on whole-genome sequencing. Data on these cases were linked to national hospital care and mortality datasets on 1 December 2021. Before the week commencing 21 June, <0.2% of sequencing-confirmed delta cases had the AY.4.2 sublineage [1]; during the inclusion period the prevalence of AY.4.2 among new sequencing-confirmed cases increased from 0.2% to 15% (Supplementary Figure 1). The data linkage, inclusion criteria, outcome and confounder data sources and definitions, and the analysis strategy have been described in a recent article [2].
Using stratified Cox regression models, we estimated hazard ratios (HRs) of hospital admission and hospital admission or emergency care attendance within 14 days, and of COVID-19 or all-cause mortality within 28 days after a first positive COVID-19 test. These models were stratified for week of specimen and lower tier local authority of residence, to account for reporting delays and unobserved confounders that may differ by calendar time and locality. Regression adjustment was used for age and index of multiple deprivation rank (each modelled using restricted cubic splines with 4 knots), date of specimen (linear term), sex, ethnicity, vaccination status, and recent international travel. We additionally estimated the HRs within subgroups based on symptom or vaccination status. In supplementary analyses, we explored the sensitivity of the HRs to alternative adjustment approaches and to bias due to differences of epidemic phase of the sublineages [3].

Ethics
This surveillance was performed as part of UKHSA's responsibility to monitor COVID-19 during the current pandemic. UKHSA has legal permission, provided by Regulation 3 of The Health Service (Control of Patient Information) Regulations 2002 to process confidential patient information under Sections 3(i) (a) to (c), 3(i)(d) (i) and (ii) and 3(iii) as part of its outbreak response activities. This study falls within the research activities approved by the UKHSA Research Ethics and Governance Group.

Characteristics
A total of 28 736 AY. 4 Table 2). The sensitivity analysis exploring alternative adjustment approaches yielded HRs similar to those from the primary analysis (Supplementary Table 2). The sensitivity analysis adjusting for epidemic phase bias considered multiple scenarios, which suggested that the risks of all considered COVID-19 severity outcomes might be slightly lower for AY.4.2 compared to non-AY.4.2 delta cases (Supplementary Tables 3  and 4), consistent with the primary analysis.

DISCUSSION
Based on record linkage of sequencing-confirmed COVID-19 cases in England, we found that the risks of hospitalization and mortality outcomes were similar or lower for cases infected with the AY.4.2 compared to non-AY.4.2 sublineages of the delta variant of SARS-CoV-2. The results were similar when restricted to symptomatic and likely symptomatic cases, or to vaccinated or unvaccinated subgroups, or after additional adjustment for time since second vaccine dose. Further sensitivity analyses to adjust for the effect of epidemic phase bias [3] consistently suggested that the risks of the hospitalization outcomes are similar or lower for AY.4.2 than non-AY.4.2 delta cases.
Strengths of this analysis include the use of timely population datasets that cover all hospitalization events and deaths for COVID-19 cases in England. Limitations include reporting delays of the outcome events, which may differ over time and by hospital trust. However, after stratification for calendar period and area, the reporting delays should not differ systematically by sublineage. A further limitation is the restriction to cases confirmed through sequencing, due to a lack of other methods capable of distinguishing different delta sublineages. During the study period, the median daily sequencing coverage of new COVID-19 cases was 16.5% (range, 6.5%-27.2%) [1]. More severe cases with higher viral loads may be preferentially selected for sequencing, which may restrict the generalizability of the findings. However, similar cycle threshold counts were reported between individuals infected with AY.4.2 or non-AY.4.2 identified in the REACT-1 random testing survey [4]. Hence, there is no reason to expect that such selection differed systematically by sublineage.
Several variants of the SARS-CoV-2 virus have evolved during the COVID-19 pandemic of 2020-2022. In England, the alpha (Pango lineage B.1.1.7) variant was detected in November 2020 and was found to be associated with higher transmissibility [5], and higher risks of hospital admission [6,7] and mortality [6,8], than previously circulating wild-type SARS-CoV-2. In March 2021, the delta variant was detected in England and soon became the dominant variant in the country. Delta is associated with higher transmissibility [9], partial vaccine escape [10,11], and higher risk of hospitalization [2,10,12] and mortality [12], compared to the alpha variant. Recently, cases with AY.4.2 were reported to be less likely to experience symptomatic disease than cases with other delta sublineages [4]. Although our results indicated similar proportions with symptomatic Omicron has been found to be associated with lower hospitalization and mortality risks than delta [15]. Similar to the emergence of the AY.4.2 delta sublineage, an omicron sublineage (BA.2) with a potential transmission advantage has recently been identified [1]. The findings in our study highlight the importance of assessing severity differences between SARS-CoV-2 variant sublineages, and provide a baseline for future research on the relative severity between delta or delta variant sublineages and other circulating variants such as omicron and its sublineages.  Financial support. This work was supported by the United