Influenza vaccine effectiveness against influenza A subtypes in Europe: Results from the 2021–2022 I‐MOVE primary care multicentre study

Abstract Background In 2021–2022, influenza A viruses dominated in Europe. The I‐MOVE primary care network conducted a multicentre test‐negative study to measure influenza vaccine effectiveness (VE). Methods Primary care practitioners collected information on patients presenting with acute respiratory infection. Cases were influenza A(H3N2) or A(H1N1)pdm09 RT‐PCR positive, and controls were influenza virus negative. We calculated VE using logistic regression, adjusting for study site, age, sex, onset date, and presence of chronic conditions. Results Between week 40 2021 and week 20 2022, we included over 11 000 patients of whom 253 and 1595 were positive for influenza A(H1N1)pdm09 and A(H3N2), respectively. Overall VE against influenza A(H1N1)pdm09 was 75% (95% CI: 43–89) and 81% (95% CI: 45–93) among those aged 15–64 years. Overall VE against influenza A(H3N2) was 29% (95% CI: 12–42) and 25% (95% CI: −41 to 61), 33% (95% CI: 14–49), and 26% (95% CI: −22 to 55) among those aged 0–14, 15–64, and over 65 years, respectively. The A(H3N2) VE among the influenza vaccination target group was 20% (95% CI: −6 to 39). All 53 sequenced A(H1N1)pdm09 viruses belonged to clade 6B.1A.5a.1. Among 410 sequenced influenza A(H3N2) viruses, all but eight belonged to clade 3C.2a1b.2a.2. Discussion Despite antigenic mismatch between vaccine and circulating strains for influenza A(H3N2) and A(H1N1)pdm09, 2021–2022 VE estimates against circulating influenza A(H1N1)pdm09 were the highest within the I‐MOVE network since the 2009 influenza pandemic. VE against A(H3N2) was lower than A(H1N1)pdm09, but at least one in five individuals vaccinated against influenza were protected against presentation to primary care with laboratory‐confirmed influenza.

Discussion: Despite antigenic mismatch between vaccine and circulating strains for influenza A(H3N2) and A(H1N1)pdm09, 2021-2022 VE estimates against circulating influenza A(H1N1)pdm09 were the highest within the I-MOVE network since the 2009 influenza pandemic. VE against A(H3N2) was lower than A(H1N1)pdm09, but at least one in five individuals vaccinated against influenza were protected against presentation to primary care with laboratory-confirmed influenza.  6 Influenza virus A(H3N2) dominated in all European countries. 6 In countries participating in I-MOVE, influenza vaccination is recommended for older adults (those aged ≥50, ≥55, ≥60, or ≥65 years, depending on the country), among those with medical risk conditions and among children in In this article, we present the I-MOVE end-of-season estimates of 2021-2022 influenza VE among patients presenting with a respiratory infection at primary care level.

| METHODS
The methods of the multicentre case-control study have been described previously and are based on the ECDC generic case-control study protocol and the I-MOVE+ generic study protocol. 1,5,[8][9][10] Briefly, study sites in nine European countries took part in the primary care-based I-MOVE multicentre study in the 2021-2022 influenza season: Croatia, France, Germany, Republic of Ireland, the Netherlands, Portugal, Romania, Spain (contributing with two distinct study sites), and Sweden. General practitioners (GPs), or paediatricians, systematically selected patients, or selected all patients presenting with influenza-like illness (ILI) or acute respiratory infection (ARI) to include in the study. Physicians either sampled patients or referred them to a medical laboratory or a COVID-19 testing centre. Physicians obtained information on patients using interviews and linkage to electronic health records. All study sites collected the symptoms, date of onset and swabbing, 2021-2022 influenza vaccination status and date of vaccination, sex, age, presence of chronic conditions, and influenza and SARS-CoV-2 test results.
In the pooled analysis, we included patients with a specimen taken less than 8 days after symptom onset. Using the test-negative design, a case of confirmed influenza was an ARI or ILI patient who was sampled and tested positive for influenza A(H1N1)pdm09 or A(H3N2) virus (for an influenza A subtype-specific analysis) using realtime reverse-transcription polymerase chain reaction (RT-PCR). 11 Controls were ARI/ILI patients who tested negative for any influenza virus.
For each study site, we included patients presenting symptoms 14 or more days after the start of national influenza vaccination campaigns. In each study site, controls were excluded if presenting in weeks of onset prior to the first influenza (sub)type positive case for each (sub)type-specific analysis.
For one study site not collecting date of symptom onset, we imputed it as 2 days before the sample date, as 2 days was the median delay between onset and sample in the pooled data.
We defined a person as vaccinated if he or she had received 2021-2022 influenza vaccine 14 or more days before symptom onset.
Patients vaccinated fewer than 14 days before symptom onset were excluded.
One of the Spanish study sites (Navarra) is a comprehensive surveillance system, where patients from all GPs in the region are included, compared with the sentinel system of other study sites. We included all cases, but only a random sample of 20% of controls by onset week for Navarra, to account for the differences in absolute numbers between the systems.
We excluded from the pooled analysis any study site that had less than 10 influenza (sub)type-specific cases for each (sub)typespecific analysis. We combined individual patient data and used a one-stage model, with study site as a fixed effect. We carried out a complete case analysis and used a logistic regression model to calculate VE including potential confounding factors: Age (modelled as a restricted cubic spline with four or five knots, age groups or age as a linear term depending on the analysis, as determined by the Akaike Information criterion [AIC]), sex, presence of at least one chronic condition (including pregnancy and obesity where available and applicable), and date of onset (modelled as a restricted cubic spline with four or five knots, depending on the AIC).
We stratified the data into patients aged 0-14 years, 15-64 years, and those aged 65 years and older to obtain age groupspecific VE. We estimated VE among the target group for influenza vaccination. We also estimated VE by time since vaccination, comparing unvaccinated to those vaccinated <90, 90-119, 120-149, and ≥150 days before symptom onset.
In a sensitivity analysis, we dropped COVID-19 positive controls, 12 as a correlation between influenza and COVID-19 vaccination could potentially bias VE estimates, by violating a fundamental criterion of the test-negative design: Controls are not affected by the vaccination under study. 11 In nine study sites, all or a random sample of influenza viruses were selected for sequencing (haemagglutinin genome segment and/or whole genome). In one study site (France), only samples going to one of three labs used within the study carried out sequencing.

| Participant profiles
We excluded 961 patients (7.6%) from the complete case analysis, due to missing data for age, sex, chronic condition or influenza vaccination status, including one study site with too few influenza cases for analysis (<10 influenza positive cases). We included 253 influenza  (Table 1). Among controls 19% were aged 0-4 years old compared with 22% and 8% among influenza A(H1N1)pdm09 and A(H3N2) cases, respectively.

| Genetic characterisation
Among study sites included in the pooled analysis, eight study sites systematically sequenced influenza positive specimens or sequenced all specimens technically possible. Among the 153 influenza A(H1N1) pdm09 cases included in the analysis from those countries/labs sequencing viruses, and part of the A(H1N1)pdm09 analysis (DE, FR, NL), 53 (35%) were sequenced (
The overall adjusted VE against influenza A(H3N2) when excluding SARS-CoV-2 controls was 36% (95% CI: 21-49) (   studies also suggest a protection against current season A(H3N2) viruses from previous season vaccines. 26 F I G U R E 2 All influenza (upper figure, in blue) and influenza A(H3N2) (lower figure, in red) vaccine effectiveness by days since vaccination. I-MOVE primary care multicentre case control study, influenza season 2021-2022 VE against A(H3N2) was lower among the target group for vaccination (20%) and also lower among those with presence of at least one chronic condition (21%) compared with all ages (29%). It may be that changing guidelines around personal protective behaviour related to the COVID-19 pandemic meant that persons at risk were less likely to be exposed to the influenza virus early in the season.

PEER REVIEW
The peer review history for this article is available at https://publons. com/publon/10.1111/irv.13069.

DATA AVAILABILITY STATEMENT
Data available only on request due to privacy/ethical restrictions: The data that support the findings of this study may be available on request from the corresponding author, depending on the request.
The data are not publicly available due to privacy or ethical restrictions. ORCID