Interim 2023/2024 Season Influenza Vaccine Effectiveness in Primary and Secondary Care in the United Kingdom

ABSTRACT Background We report 2023/2024 season interim influenza vaccine effectiveness for three studies, namely, primary care in Great Britain, hospital settings in Scotland and hospital settings in England. Methods A test negative design was used to estimate vaccine effectiveness. Results Estimated vaccine effectiveness against all influenzas ranged from 63% (95% confidence interval 46 to 75%) to 65% (41 to 79%) among children aged 2–17, from 36% (20 to 49%) to 55% (43 to 65%) among adults 18–64 and from 40% (29 to 50%) to 55% (32 to 70%) among adults aged 65 and over. Conclusions During a period of co‐circulation of influenza A(H1N1)pdm09 and A(H3N2) in the United Kingdom, evidence for effectiveness of the influenza vaccine in both children and adults was found.


| Introduction
In the United Kingdom, seasonal influenza vaccination is offered freely to those aged ≥ 65, aged 16-64 within clinical groups and at increased risk of severe influenza-related outcomes and children aged 2-15 [1].In Scotland, this offer was extended to healthy adults aged 50-64 [2].The 2023/2024 season northern hemisphere influenza vaccine included an updated A(H1N1) pdm09 strain: a cell culture-propagated A/Wisconsin/67/2022 or egg-propagated A/Victoria/4897/2022 (H1N1)pdm09-like virus, while the A(H3N2) and B components remained the same as for 2022/2023 [3].
Throughout the United Kingdom, indicators of influenza activity suggested low levels of circulation prior to December 2023 [4][5][6].Influenza activity rose throughout December, reaching a modest peak by late December 2023, before dropping slightly early-to mid-January 2024.However, by late January 2024, influenza activity was increasing again.To January 2024, the United Kingdom has primarily seen cocirculation of influenza A(H1N1)pdm09 and A(H3N2).
We report interim vaccine effectiveness (VE) in primary and secondary care settings within the United Kingdom.

| Methods
We brought together primary care sentinel swabbing data from England, Scotland and Wales (GB-PC) and conducted two separate data-linkage studies on hospitalised patients in England (EN-H) and Scotland (SC-H).Study characteristics are summarised in Table 1.Methods for all three studies have been described elsewhere [8][9][10][11]; a protocol for GB-PC is provided in the supporting information.
All studies used the test negative design, applying logistic regression with adjustments as outlined in Table 1 to estimate the odds ratio (OR) of vaccination in influenza positive cases and influenza negative controls.VE is reported as a percentage: 100 × (1 − OR).For the GB-PC study, patients presented to primary care with acute respiratory infection (ARI), swabs known to be taken beyond 7 days of onset were not used to estimate VE (see supporting information for full detail).Symptom onset date was not available within hospital records.In EN-H, influenza A subtype VE analyses were restricted to data from labs that carried out subtyping.In SC-H, potential vaccine contaminants and influenza codetections were removed from all analyses.SARS-CoV-2-positive controls were excluded in the GB-PC and EN-H studies [12].
Note that for Scotland and England, the primary source of vaccination histories is immunisation databases; these capture vaccinations offered freely on the National Health Service well but may miss vaccines administered privately, for example, via workplaces, likely leading to a small amount of exposure misclassification among working age adults.In Wales the primary source of vaccination histories was through questionnaire administered by the GP at the point of swabbing.
A sample of influenza virus positive primary care specimens was further characterised, as described in the supporting information.

| Results
The GB-PC study included 1193 cases and 12,098 controls, including 6343 samples from England, 6108 from Scotland and 840 from Wales; the number of specimens positive for influenza A(H1N1)pdm09 was 461, 475 A(H3N2), 215 influenza A (untyped) and 46 influenza B, with 4 dual infections.The EN-H study included 1359 cases and 22,539 controls; 770 influenza A samples were untyped, 161 were influenza A(H1N1)pdm09, 395 were influenza A(H3N2) (note that more laboratories subtyped influenza A/H3 than A/H1) and 32 were influenza B, including 5 dual infections.The SC-H study included 1977 cases and 34,476 controls; 1567 influenza A samples were untyped, 172 were influenza A(H1N1)pdm09, 188 were influenza A(H3N2) and 50 were influenza B.

| Discussion
During a period of co-circulation of influenza A(H1N1)pdm09 and A(H3N2) in the United Kingdom, we found evidence of moderate VE in both children and adults.Our results concur with those reported for Canada during the early 2023/2024 season, who reported moderate VE during an influenza A(H1N1) pdm09-dominated period [13].
During the 2022/2023 season, estimates of interim VE against influenza A(H1N1)pdm09 in Europe, including some corresponding results for EN-H and SC-H, were typically lower than those we report for the 2023/2024 season, especially among those aged 65 and above [10].Our results suggest that VE may have been improved by the update to the A(H1N1)pdm09 component of the northern hemisphere vaccine for 2023/2024.The 2023/2024 A(H3N2) northern hemisphere vaccine strain belongs to subgroup 2a of genetic subclade 3C.2a1b.2a.2.
Experiments to evaluate the vaccine's ability to recognise subgroup 2a.3a.1 viruses similar to those circulating have reported mixed findings [14].Our 2023/2024 VE estimates against A(H3N2) are higher than those reported for 2022/2023 in SC-H as well as among children and adults aged 65 and above in EN-H [8], suggesting that the differences in subgroup of the vaccine and circulating viruses have not resulted in reduced VE.However, no effectiveness was demonstrated in the 2023/2024 EN-H for those aged 18-64 years.Healthy adults aged 50-64 were influenza vaccine eligible in 2022/2023, but in 2023/2024, this was withdrawn in England and Wales while continuing in Scotland; evidence for residual protection

TABLE 1 |
[7]mary of methods and characteristics of three 2023/2024 season influenza vaccine effectiveness studies in Great Britain.: aQIV, adjuvanted QIV; ARI, acute respiratory infection; GB, Great Britain; GP, general practitioner; LAIV, quadrivalent live attenuated influenza vaccine; QIVc, cell-grown quadrivalent influenza vaccine; QIVe, egg-grown quadrivalent influenza vaccine; QIVr, recombinant quadrivalent influenza vaccine; SIMD, Scottish Index of Multiple Deprivation; TND, test-negative design.aTheEU-ARIdefinition is sudden onset of symptoms AND ≥ 1 of cough, sore throat, shortness of breath or coryza AND a clinician's judgement that the illness is due to an infection.bVariesaccordingtoSARS-CoV-2/influenza testing practices by Health Board.cTheQCOVIDriskgroups are defined as the number of generic comorbidity conditions of a patient, and are used as a measure of comorbidity.The list of conditions is found in the study of Clift et al.[7].
AbbreviationsEN-H; hospitalisation definitions differ to SC-H and further exploration of this is planned.In a hospital setting, testing is among patients presenting with ARI symptoms, but a limitation is that this may not be the main reason for hospitalisation.Unlike previous seasons, Scotland successfully collected individual-level vaccination data for children from all health boards, resulting for the first time in a national hospital study population.