Influenza Vaccine Effectiveness Against Mild and Severe Outcomes: A Comparison Study in Portugal Authors: Francisco Mourao, Verónica Gómez, Patrícia Soares, Ana Paula Rodrigues, Carlos Matias Dias, Ausenda Machado Published: 2026 ABSTRACT ABSTRACTBackgroundInfluenza vaccination is an effective public health intervention to lessen influenza effect on healthcare and has been recommended to prevent severe outcomes. Previous studies present higher influenza vaccine effectiveness (IVE) estimates against primary care medically attended (outpatients) compared to hospitalised cases (inpatients). Inconsistencies in study designs may hinder comparisons of IVE in preventing mild (outpatient consultations) versus severe influenza (inpatient admissions). We aimed to compare IVE against mild and severe influenza within primary care and hospital networks in Portugal from 2015/16 to 2023/24.MethodsUsing the test‐negative design, we analysed 3393 RT‐PCR‐tested patients with influenza‐like illness in outpatients or severe acute respiratory infection in inpatients. IVE was estimated using a generalised linear mixed model with fixed effects for sex, age category, comorbidities, month of symptom onset and random effects for season. For difference estimation, 95% confidence intervals (CI) were obtained through bootstrapping.ResultsIVE was 40.4% (22.7; 54.0) for outpatients and 38.7% (17.6; 54.5) for inpatients. The 1.7% difference was not statistically significant (−22.0; 26.8). Subgroup analyses for individuals aged ≥ 65 years, patients with comorbidities and by influenza A subtype showed similar non‐significant differences, ranging from −1% to 17%.ConclusionWe found no statistically significant difference in IVE against mild or severe influenza infection. However, comparison across care settings should be interpreted cautiously, as patients differ in sociodemographic characteristics and comorbidity burden. Although limited sample size may reduce statistical power, the small magnitude of observed differences suggests that patients benefit from current vaccination strategies across care settings.Keywords: hospital, influenza, influenza like illness, influenza vaccine, primary care, severe acute respiratory infection, vaccine effectiveness Introduction Seasonal influenza has a major impact on health systems, causing substantial morbidity, hospitalisations and mortality worldwide [1]. Vaccination is the most effective preventive public health intervention to reduce influenza burden, offering a favourable benefit–risk profile [2]. In Portugal, the General‐Directorate of Health (Direção‐Geral da Saúde) aligns vaccine composition with World Health Organization (WHO) recommendations and provides free‐of‐charge seasonal vaccination for priority groups—namely pregnant women, adults over 64 years, immunocompromised individuals and healthcare workers—through vaccination centres in primary care facilities and community pharmacies (since 2020/21) [3]. The primary goal of targeting high‐risk groups is to prevent severe outcomes, such as hospitalisation and death, by inducing immune responses that attenuate disease severity, though neither infection nor vaccination confers sterilising immunity [4]. Vulnerable populations often exhibit reduced vaccine‐induced immunity due to immunosenescence, immunosuppression or waning protection from prior vaccinations [5, 6]. This raises a particular interest in influenza vaccine effectiveness (IVE) across disease severity gradients, from mild (outpatient) to severe (inpatient) cases [7]. Previous studies have shown higher estimates for outpatients than for inpatients [8, 9, 10]. For example, during the 2022/23 season in Europe, IVE was 27% (95% confidence interval (CI): 13; 39) in inpatients whereas IVE in outpatients was 53% (95% CI: 48; 58) [9, 10]. A similar pattern was also observed in Great Britain (2023/24), although with 95% CI overlap [8]. However, methodological heterogeneity and limited multiseasonal data hinder consensus in IVE comparisons across settings [7, 11, 12, 13, 14]. To address these gaps, we used data from Portugal's participation in European influenza vaccine effectiveness monitoring networks, collaborative initiatives designed to better inform public health decision making. I‐MOVE (Influenza—Monitoring Vaccine Effectiveness in Europe), VEBIS (Vaccine Effectiveness, Burden and Impact Studies), I‐MOVE + (Integrated Monitoring of Vaccines in Europe) and EuroEVA (Influenza Vaccines Effectiveness in Europe) have implemented standardised test‐negative design (TND) protocols to estimate IVE since 2008 for mild disease and since 2015 for severe disease [15, 16, 17]. This study compares IVE against mild and severe influenza, using data collected in primary care and in hospital settings, respectively, across seven influenza seasons (2015/16 to 2023/24), providing within‐country evidence over time, including subgroup analyses by age, comorbidity status and influenza subtype to control for these factors. Methods We estimated IVE using a test‐negative case–control design, the gold standard for assessing real‐world vaccine performance [18]. This approach consists of comparing the odds of vaccination among influenza laboratory‐confirmed cases and test‐negative controls [19]. The study uses data from Portugal's participation in the I‐MOVE and VEBIS networks, adhering to a common protocol developed by the network consortium and the European Centre for Disease Prevention and Control (ECDC) across seasons [20]. In the primary care setting, general practitioners from the national sentinel network (EuroEVA) recruited outpatients who presented with influenza‐like illness (ILI), defined as a sudden onset of symptoms including at least one systemic symptom (fever, malaise, headache, or myalgia) and at least one respiratory symptom (cough, sore throat, or dyspnoea) [20]. In the hospital setting, inpatients were included if admitted for severe acute respiratory infection (SARI), defined as hospitalisation for at least 24 h with at least one systemic symptom (fever, myalgia, malaise, headache or general deterioration) and at least one respiratory symptom (cough, sore throat or dyspnoea) [17, 20, 21]. During the 2015/16 to 2019/20 seasons, inpatient enrolment was restricted to individuals aged 65 years and older and was expanded to include all ages since the 2020/21 season. No restrictions were applied for outpatient enrolment. All participants underwent reverse transcription (RT)‐PCR testing for influenza virus detection and subtyping. Samples from outpatients were analysed at the Portuguese Reference Laboratory for influenza and other respiratory viruses, whereas inpatient samples were processed at the respective hospital laboratories. Influenza cases were defined as individuals meeting the ILI or SARI criteria with a positive RT‐PCR result for influenza, whereas controls met ILI or SARI criteria but tested negative for influenza. Vaccination status was ascertained through consultation of the national vaccination registry by participating medical practitioners. This registry is centralised and accessible to all healthcare providers; all patients recruited were covered by this system, ensuring near‐complete ascertainment of vaccination status. Vaccines provided across seasons followed WHO composition recommendations, with more than one brand available each season (Table S1 in the Supporting Information). All vaccines were non‐adjuvanted, egg‐based and standard dose, except for a high‐dose vaccine available during 2022/23 and 2023/24 seasons only available for institutionalised individuals. Information on comorbidities was obtained directly from participating physicians. Inclusion and Exclusion Criteria This study included community‐dwelling adult patients, aged 18 years or older, who gave their explicit consent to participate. Among vaccinated individuals, only those who had received the vaccine at least 14 days before symptom onset were included in the analysis. Additionally, we retained only months in which at least one case and one control were identified, following the studies' protocols. Patients were excluded if they met any of the following criteria: microbiological sampling performed eight or more days after symptom onset; symptom onset more than 7 days before outpatient consultation or inpatient admission; onset of SARI symptoms more than 48 h after inpatient admission; pregnancy; controls recruited before the first seasonal influenza case reported within the Portuguese IVE monitoring network; missing data on age, sex, ILI/SARI criteria, symptom onset date, swab date, vaccination status or comorbidities. The end of the study for each season was defined if during two consecutive weeks no cases were detected. Seasons 2020/21 and 2021/22 were intentionally excluded from this study due to the low influenza circulation among community‐dwelling individuals, resulting from COVID‐19 pandemic restrictions [22]. Statistical Analysis We described the study population using frequencies and percentages for categorical variables (sex, age group, comorbidity status, vaccination status) and means for continuous variables (age, days from vaccination to symptom onset, days from symptom onset to swab). Stratified summaries by cases and controls were generated for outpatients and inpatients to highlight differences in demographic and clinical characteristics. To estimate IVE, we fitted a generalized linear mixed model with a logit link function, modelling the binary outcome of case/control status as a function of vaccination status. The model included fixed effects for sex (male or female), age group (18–44, 45–64 and ≥ 65 years), presence of comorbidities (yes or no) and month of symptom onset (categorical), along with a random effect for study season to account for inter‐seasonal variability. The adjustment set was selected according to a directed acyclic graph, previously developed by Machado et al. [23]. IVE was calculated as (1—adjusted odds ratio) × 100 for each setting separately. We conducted stratified analyses to assess IVE by age group (≥ 65 years), comorbidity status and influenza A subtype (H1N1pdm09 and H3N2). Standard dose vaccines from different manufacturers were assumed to have equivalent effectiveness. To compare IVE estimates, we calculated the absolute difference between IVE against mild and severe influenza and quantified uncertainty using a parametric bootstrapping approach. Specifically, we assumed that the log of the odds ratios for vaccination status followed a normal distribution, with means and standard deviations derived from the coefficient and standard error estimates of the adjusted models for each setting. Then, we generated 10,000 simulated samples from these distributions and computed the difference in IVE. This process was repeated 100 times to account for simulation variability and the 2.5th and 97.5th percentiles of the resulting differences were averaged to construct the final 95% CI. A difference was considered statistically significant if the 95% CI excluded the null value. All statistical analyses were performed using R version 4.4.2 (2024‐10‐31) [24]. Ethical Considerations and Data Protection This study used anonymised data originally collected by the Portuguese National Health Institute as part of its participation in the I‐MOVE+ and VEBIS projects. For those studies, ethical approval was obtained for data collection, which included informed consent from all participants. The current analysis was conducted in accordance with good research practices. Results From the 5804 patients initially screened across all study seasons, 2411 (41.5%) were excluded according to the study criteria (Figure 1). The final sample size of outpatients is nearly twice that of inpatients, with 916 cases for 1250 controls and 359 cases for 868 controls, respectively. Flowchart of study patient inclusion for main influenza vaccine effectiveness analyses, Portugal, influenza seasons 2015/16 to 2023/24. Patients presented different sociodemographic characteristics across settings (Table 1). Inpatients were older, with 84.8% aged 65 years or older, with a mean age of 76 years. Whereas outpatients were younger, with only 18.8% aged 65 years or older, with a mean age of 48 years. Additionally, more inpatient males (53.5%) were observed compared to outpatients (35.0%). The presence of comorbidities was also higher for inpatients (95.9%) than for outpatients (41.0%). TABLE 1Demographic, clinical and vaccination characteristics of influenza cases and test‐negative controls from outpatients (primary care) and inpatients (hospital) in the main analysis, Portugal, influenza seasons 2015/16 to 2023/24.CharacteristicsOutpatientsInpatientsCases (n = 916)Controls (n = 1250)Cases (n = 359)Controls (n = 868)Age group (years), n (%)18–44416(45.4%)522(41.8%)7(1.9%)25(2.9%)45–64358(39.1%)462(37.0%)39(10.9%)115(13.2%)≥ 65142(15.5%)266(21.3%)313(87.2%)728(83.9%)Age mean (years)47.0149.3876.1476.27Sex, n (%)Female556(60.7%)852(68.2%)166(46.2%)404(46.5%)Male360(39.3%)398(31.8%)193(53.8%)464(53.5%)Any chronic condition a , n (%)No585(63.9%)693(55.4%)22(6.1%)28(3.2%)Yes331(36.1%)557(44.6%)337(93.9%)840(96.8%)Vaccination status, (%)Unvaccinated778(84.9%)960(76.8%)212(59.1%)357(41.1%)Vaccinated138(15.1%)290(23.2%)147(40.9%)511(58.9%)Days to swab after symptom onset (mean)2.212.483.243.05Days from vaccination until symptom onset (mean)72.1277.4074.12110.64Month of symptoms onsetSeptember< 5(< 1%)15(1.2%)————October22(2.4%)99(7.9%)< 5(< 1%)33(3.8%)November126(13.8%)178(14.2%)18(5%)102(11.8%)December280(30.6%)286(22.9%)146(40.7%)177(20.4%)January278(30.3%)410(32.8%)129(35.9%)203(23.4%)February151(16.5%)177(14.2%)50(13.9%)147(16.9%)March46(5.0%)63(5.0%)6(1.7%)39(4.5%)April10(1.1%)16(1.3%)< 5(< 1%)82(9.4%)May————< 5(< 1%)18(2.1%)June< 5(< 1%)6(0.5%)< 5(< 1%)67(7.7%) a Reporting of chronic conditions varied slightly across seasons and settings, check Table S2 in the Supporting Information for more details. Demographic, clinical and vaccination characteristics of influenza cases and test‐negative controls from outpatients (primary care) and inpatients (hospital) in the main analysis, Portugal, influenza seasons 2015/16 to 2023/24. Reporting of chronic conditions varied slightly across seasons and settings, check Table S2 in the Supporting Information for more details. Vaccination rates were higher in the inpatient setting with 53.6% of inpatients being vaccinated, against only 19.8% vaccinated outpatients. There was a longer mean delay until swab analysis after symptom onset in inpatients (3.1 days versus 2.4 days for outpatients) and a longer mean interval from vaccination to symptom onset in inpatients (100.0 days versus 75.1 days for outpatients). The period of symptom onset was concentrated in the autumn‐winter months, following the known influenza circulation pattern. Influenza strain circulation varied greatly across seasons (Table S2 in the Supporting Information). In the 2015/16 season, A(H1N1)pdm09 was the predominant strain. The 2016/17 season was characterised by the dominance of A(H3N2). During the 2017/18 season, there was a mix of A(H1N1)pdm09, A(H3N2) and B/Yamagata, with few hospitalisations due to B/Victoria infections in our sample. The national surveillance system identified B/Yamagata in 54.4% of inpatients for the 2017/18 season [25]. The 2018/19 season saw the presence of both influenza A subtypes in our analysis, with a higher proportion of A(H3N2) identification; this subtype had an antigenic drift from the seasonal vaccine strain [26]. In the 2019/20 season, A(H1N1)pdm09 and B/Victoria were prevalent, with no influenza B infections identified in inpatients of our sample despite B/Victoria circulation and identification in hospitalised patients through the national surveillance system [27]. The 2022/23 season was dominated by A(H3N2), although A(H1N1)pdm09 and B/Victoria were also present. Finally, the 2023/24 season saw a resurgence of A(H1N1)pdm09 as the dominant strain. These findings are consistent with the Portuguese National Health Institute annual epidemiologic descriptions, except for the absence of influenza B/Victoria in inpatients for seasons 2017/18 and 2019/20 [25, 26, 27, 28, 29, 30, 31]. Results of vaccine effectiveness are summarised in Figure 2. In the main analyses, IVE was 40.4% (95% CI: 22.7; 54.0) against mild influenza (measured in the outpatient setting) and 38.7% (95% CI: 17.6; 54.5) against severe influenza (measured in the inpatient setting). These results indicate overall significant protection against influenza among vaccinated individuals. The absolute difference of 1.7% (95% CI: −22.0; 26.8) between the two settings was not statistically significant. Comparison of influenza vaccine effectiveness between outpatients and inpatients, main analyses and stratified by patient characteristics and virus subtype, Portugal, influenza seasons 2015/16 to 2023/24. (A) vaccine effectiveness estimates; (B) difference in vaccine effectiveness (outpatients versus inpatients). Restricting the analysis to individuals aged 65 years or older improved comparability between settings, given the higher proportion of this age group among inpatients. Under this restriction, the estimated difference in IVE increased to 16.9% (95% CI: −15.73; 45.63), but remained statistically non‐significant. Stratification by the presence of comorbidities did not differ from the main analyses IVE estimates in both settings. Despite a reduction of more than half of cases and controls among outpatients when restricting to patients with comorbidities, the IVE estimates remained unchanged. When examining IVE by influenza A subtypes, significant protection was observed only against mild A(H1N1)pdm09 influenza, with an IVE of 42.7% (95% CI: 13.4; 62.0). For the A(H3N2) subtype, IVE estimates were not statistically significant for both outcomes, with 28.3% (95% CI: −3.8; 50.4) in outpatients and 29.8% (95% CI: −18.7; 58.4) in inpatients. Although the confidence intervals for these estimates included the null value, the point estimates were consistently positive. Due to the low incidence of influenza B in the hospital setting, IVE was not stratified by this influenza type. Figure 3 shows IVE estimates and IVE differences within season without stratification. Season 2017/18 is the only one presenting a significant difference between the two settings of 75.0% (95% CI: 9.7; 175.4). Remaining seasons show no significant differences in IVE between outpatients and inpatients. Comparison of influenza vaccine effectiveness between outpatients and inpatients without stratification, by season, Portugal, influenza seasons 2015/16 to 2023/24. (A) vaccine effectiveness estimates; (B) difference in vaccine effectiveness (outpatients versus inpatients). During seasons in which influenza A(H3N2) dominated (2016/17, 2018/19 and 2022/23), IVE estimates against severe influenza were higher than those against mild influenza, although the differences were not statistically significant (Figure 3; Table S2 in the Supporting Information). A similar pattern was observed in the 2023/24 season, although influenza A(H1N1)pdm09 was the dominant subtype (Figure 3; Table S3 in the Supporting Information). Discussion The objective of this study was to compare the IVE against different levels of disease severity. We found no statistically significant difference in IVE against mild (outpatient) or severe disease (inpatient) in the main and restricted analyses. Stratification by comorbidity status and age ≥ 65 years did not alter IVE estimates in the inpatient setting, likely due to the high prevalence of comorbidities (95.9%) and older adults (84.8%) in this population. The exceptional 2017/18 season, where IVE was significantly different between inpatients and outpatients (75.0%, 95% CI: 9.7; 175.4), likely stems from strain‐specific disparities; outpatient cases were predominantly infected with influenza B/Yamagata, whereas inpatients had more A(H3N2) infections. This aligns with previous studies showing higher IVE for influenza B/Yamagata than for influenza A(H3N2) despite the vaccine influenza B mismatch that season [32, 33]. The lower IVE for mild influenza in season 2018/19 might be explained by the antigenic drift for A(H3N2) observed during this season. The lower IVE against influenza A(H3N2) reported in the literature may partly explain the observed pattern during seasons in which this subtype predominated, with higher—though not statistically significant—protection against severe influenza compared with mild infection (2016/17, 2018/19 and 2022/23) [34]. However, a similar trend was observed during the 2023/24 season, despite the predominance of influenza A(H1N1)pdm09 in both settings. This suggests that, although influenza type and subtype may contribute to differences in IVE between severe and mild clinical presentations, other factors, such as viral cocirculation, may also influence IVE estimates. Stratified seasonal IVE estimates by influenza subtype were not feasible due to limited sample sizes, but such analyses could be informative for exploring within‐season subtype‐specific differences in IVE. Our aggregated findings are consistent with previous studies comparing IVE across outpatient and inpatient settings. Kelly et al. [11] reported non‐significant trends suggesting higher IVE in outpatients over three seasons (2011/13) in Victoria, Australia [11]. Similarly, Castilla et al. (2018) found no significant differences in IVE by disease severity in a comparative analysis conducted in Navarre, Spain, between 2010 and 2016 [12]. The latter argue that their results might contradict the hypothesis that influenza vaccination mitigates illness severity and reduces the risk of hospital admission in individuals for whom it did not prevent influenza infection [12]. This has been stated elsewhere [11, 13]. However, this comparison between settings should be considered carefully, given that: 1) primary care and hospital patients are inherently different in terms of sociodemographic characteristics and comorbidity burden, requiring caution when comparing cases in the two settings, 2) seasons with viral cocirculation may be characterized by milder infections caused by one virus type or subtype/lineage and more severe infections caused by another, reflecting differences in pathogenicity patterns, 3) in our study, IVE estimates are consistently positive for inpatients although not always statistically significant, 4) lower IVE estimates in the hospital setting do not imply lack of protection against severe outcomes such as the need for ventilatory support or intensive care unit admission [35, 36]. The highest IVE estimate was observed among outpatients aged 65 years and older with at least one comorbidity, with an IVE of 59.8% (95% CI: 25.6; 78.3). Notably, no differentiated vaccines were provided to this age category during the seasons and population included in our study, meaning this effect cannot be attributed to preferential use of high‐dose or adjuvanted formulations. This finding contrasts with the immunosenescence hypothesis, which suggests that older and comorbid patients are expected to have lower vaccine effectiveness—a conclusion also reached by Machado et al. [23] and Tenford et al. [7]. However, this result may partly reflect differences in the composition of vaccinated populations across age groups. Influenza vaccination among younger individuals is primarily recommended for healthcare workers and for those with underlying medical conditions. As a result, the vaccinated population under 65 years constitutes a highly selected subgroup with existing comorbidities, who may have a reduced immune response and a higher baseline risk of infection or greater exposure to infected individuals. Furthermore, strong adherence to vaccination programmes in Portugal has resulted in high vaccination coverage among older and comorbid individuals [37]. In addition, specific health‐seeking behaviours within the Portuguese population, including low out‐of‐pocket spending for outpatient consultations (free of charge since 2021), may contribute to timely access to vaccination and to healthcare services [38]. Regarding potential information bias, misclassification of vaccination status is considered negligible because vaccination records are routinely filled in health centres and community pharmacies, where the majority of the population typically receives influenza vaccination. Misclassification of the outcome is also expected to be residual in both settings because of the high sensitivity and high specificity of RT‐PCR testing used to classify the outcome. This study has limitations: the fact that we pooled data from different seasons may introduce bias and neglect variations across seasons. We tried to counter this problem by fitting a generalised linear mixed model with random effects for study season and nuance IVE estimations according to season characteristics. However, the influence of previous seasons in immunoattenuation (through vaccination or infection) was not considered. The differences in sociodemographic characteristics between inpatients and outpatients should be kept in mind when interpreting this study's results. IVE estimations were adjusted for these potential confounders; however, some residual confounding may persist, as the study collects information on a limited number of confounders. Small and sparse data bias is a recognised concern when using maximum likelihood‐based estimation models [39]. In our main analysis, the number of vaccinated cases and controls was sufficient to support the covariates included in the models. The wider confidence intervals observed in some restricted subgroup analyses may reflect the smaller sample sizes in those strata and those specific estimates should be interpreted with caution. Also, other outcomes to study IVE according to illness severity could have been used, such as intensive care unit (ICU) admission, ventilatory support or death. Although information on ICU admission and respiratory support was available for some seasons, it was systematically collected only regarding patients' status before and at recruitment; no follow‐up was ensured by protocol thereafter, whether due to transfer to another hospital unit or facility or simply changes in clinical status. Consequently, post‐recruitment events of this kind would not have been captured, precluding their use as outcomes in our study. Arriola et al. [40] found that, in intensive care units, vaccinated patients aged between 50 and 65 years were more likely to be discharged earlier than unvaccinated patients [40]. To assess if this protection is more important than for general wards hospital admissions or primary care attended patients remains a topic of interest for future research. The geographic distribution between the two settings is also uneven. The outpatient setting is based on a sentinel network nationally distributed, whereas few selected hospitals participate in the inpatient setting, resulting in different geographical coverage. For most of the studied period, the two participating hospitals are located in Lisbon and Setúbal, within the Lisbon metropolitan area. The introduction of a third hospital in Porto metropolitan area in the study only occurs from 2022 onwards. Although the national sentinel network covers most of the mainland districts and the two autonomous—insular—regions. Vaccine effectiveness is known to wane over time, both within a single season and between seasons. This study does not account for within‐season waning of vaccine effectiveness, nor for potential differential waning by disease severity. Whether waning occurs at different rates in outpatient and inpatient settings remains unknown, but if protection against mild disease wanes faster than against severe disease, this could partially attenuate the difference in IVE estimates across care settings. The widespread use of barrier methods, vulgarized after the COVID‐19 pandemic, may have altered influenza transmission patterns post‐COVID‐19. Additionally, healthcare‐seeking behaviour may have been impacted, with increased use of antigenic testing for influenza and SARS‐CoV‐2 virus when ILI symptoms occur. The strengths of this study include the use of data from the same country, ensuring a similar source population; the systematic study design applied consistently across seasons and healthcare settings and the use of RT‐PCR to classify cases and controls. Our results did not observe a difference in IVE estimates against mild and severe influenza in Portugal from 2015/16 to 2023/24 seasons. Although a difference was not established, future studies should continue to estimate IVE separately in each setting, avoiding joint analysis of inpatient and outpatient IVE. Public health policies must prioritise annual influenza vaccination due to its effectiveness in preventing both mild and severe disease. Therefore, efforts to achieve high coverage rates should persist. Author Contributions Francisco Mourao: conceptualization, formal analysis, methodology, writing – original draft. Verónica Gómez: data curation, writing – review and editing. Patrícia Soares: validation, writing – review and editing, data curation. Ana Paula Rodrigues: writing – review and editing. Carlos Matias Dias: writing – review and editing. Ausenda Machado: conceptualization, methodology, supervision, writing – review and editing. Funding PS work was partially supported by the Foundation for Science and Technology (reference: CEECINST/00049/2021/CP2817/CT0001 and 10.54499/CEECINST/00049/2021/CP2817/CT0001). Ethics Statement This study utilised anonymized data originally collected by the Portuguese National Health Institute as part of its participation in the I‐MOVE+ (2009–2020) and VEBIS (2021 to present) projects. For those studies, ethical approval was obtained for data collection, which included informed consent from all participants. The current analysis was conducted in accordance with good research practices. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Table S1: Influenza vaccines by season, Portugal, influenza seasons 2015/16 to 2023/24. Table S2: Reported chronic conditions by setting and season, Portugal, influenza seasons 2015/16 to 2023/24. Table S3: Seasonal characteristics of influenza cases and test‐negative controls, virus type/subtype distribution and vaccine effectiveness estimates, by healthcare setting, Portugal, influenza seasons 2015/16 to 2023/24. Data Availability Statement Research data are not shared. References Associated Data Supplementary Materials Table S1: Influenza vaccines by season, Portugal, influenza seasons 2015/16 to 2023/24. Table S2: Reported chronic conditions by setting and season, Portugal, influenza seasons 2015/16 to 2023/24. Table S3: Seasonal characteristics of influenza cases and test‐negative controls, virus type/subtype distribution and vaccine effectiveness estimates, by healthcare setting, Portugal, influenza seasons 2015/16 to 2023/24. Data Availability Statement Research data are not shared.