Infectious Disease · Medical Literature
Bidirectional associations between influenza and COVID-19 vaccination: a systematic review and meta-analysis.
Abstract
This study aimed to investigate the effect of vaccination history on future willingness or behavior to receive vaccinations for influenza and COVID-19. In this systematic review and meta-analysis, the PubMed, Embase, Web of Science, Cochrane Library, APA PsycInfo, CINAHL and Scopus databases were searched up to February 28, 2026, to identify relevant studies in accordance with the inclusion and exclusion criteria. The analysis ultimately included 145 population-based studies conducted in the United States, China, Thailand, the United Arab Emirates, Egypt, and other countries. Prior vaccination histories for influenza and COVID-19 interacted with each other and were associated with an increased willingness and likelihood to receive future vaccines. Specifically, having received the influenza vaccine was associated with an increased willingness to receive the COVID-19 vaccine(odds ratio [OR] = 2.73, 95% confidence interval [CI]: 2.34-3.19, 95% prediction interval:[PI] 0.66-11.07), as well as with an increased likelihood of actually receiving the COVID-19 vaccine (OR = 2.76, 95% CI: 2.32-3.29, 95%PI: 0.90-8.52). Receiving a COVID-19 vaccine was associated with an increased willingness to receive a influenza vaccine (OR = 3.04, 95% CI: 1.25-7.39), and was also associated with an increased likelihood of actually receiving the influenza vaccine (OR = 3.91, 95% CI: 2.45-6.24, 95%PI: 0.81-18.81). In conclusion, prior vaccination history was associated with higher odds of future vaccination, and influenza and COVID-19 vaccination experiences influenced each other. Influenza and COVID-19 vaccine promotion strategies can likely be combined to increase overall vaccination rates. PROSPERO identifier CRD42024594174.
Clinical relevance
Does prior vaccination history for influenza increase the likelihood of receiving or intending to receive a COVID-19 vaccine, and conversely, does prior COVID-19 vaccination history increase the likelihood of receiving or intending to receive an influenza vaccine — and are these associations bidirectional? __SNAPSHOT__ {"studyDesign":"Systematic review and meta-analysis of population-based observational studies","setting":"Multinational; studies from the United States, China, Thailand, the United Arab Emirates, Egypt, and other countries; conducted at Taizhou Hospital of Zhejiang Province Affiliated to Wenzhou Medical University, Linhai, China","sampleSize":"145 included studies. For the influenza history → COVID-19 vaccination willingness arm: 91 studies (sample sizes ranging from 55 to 32,361 per study). For the influenza history → COVID-19 vaccination behavior arm: 44 studies (sample sizes ranging from 196 to 3,474,874 per study). Total participant counts across arms not aggregated in the abstract.","population":"General population, healthcare workers, older adults, pregnant women, patients with chronic diseases, and people living with HIV; studies published between 2020 and 2026","intervention":"Prior vaccination history (influenza vaccine or COVID-19 vaccine) as the exposure of interest","comparator":"Individuals without a prior history of the respective vaccine (influenza-unvaccinated or COVID-19-unvaccinated); comparator definitions varied across included observational studies","primaryOutcome":"Vaccination willingness (intent) or vaccination behavior (actual receipt) for either the influenza vaccine or COVID-19 vaccine, depending on the analysis direction","followUp":"Not reported (cross-sectional and registry-based cohort designs predominate; no standardized prospective follow-up period specified)","fundingSource":"Not reported in the available full-text source"}
Key findings
Direction 1 — Influenza history → COVID-19 vaccination: Among 91 studies, prior influenza vaccination was associated with increased willingness to receive the COVID-19 vaccine (pooled OR = 2.73, 95% CI: 2.34–3.19; 95% prediction interval [PI]: 0.66–11.07). Among 44 studies, prior influenza vaccination was associated with an increased likelihood of actually receiving the COVID-19 vaccine (pooled OR = 2.76, 95% CI: 2.32–3.29; 95% PI: 0.90–8.52). Heterogeneity was extreme for the behavior outcome (I² = 99.3%, p < 0.001). Direction 2 — COVID-19 history → influenza vaccination: The abstract as provided was truncated before these results were fully stated; specific OR, CI, and I² values for the reverse direction were not recoverable from the available text and are therefore not reported here. Subgroup analysis by study population showed substantially reduced heterogeneity in the healthcare workers subgroup (I² = 53.0%) compared with the general and specific population subgroups, which remained highly heterogeneous. Sensitivity analysis using leave-one-out showed pooled OR stability across the behavior arm (range: 2.22–3.44). Meta-regression on study year (p = 0.277) and sample size (p = 0.708) did not explain residual heterogeneity. The Egger test did not detect significant publication bias for the willingness arm (t = 1.91, p = 0.361), with a largely symmetrical funnel plot. The critically wide 95% prediction intervals — spanning below OR = 1.0 for the willingness outcome and approaching OR = 1.0 for the behavior outcome — indicate that in some settings the true association may be negligible or absent. No safety outcomes or adverse events were applicable to this study design. NNT/NNH: not applicable.
Practice implications
This large-scale meta-analysis of 145 observational studies consistently demonstrates that prior influenza vaccination is associated with approximately 2.7-fold higher odds of COVID-19 vaccine willingness and receipt, a finding directionally consistent across diverse populations, geographic settings, and time periods. The bidirectional framing suggests a mutual reinforcement between these two vaccination behaviors, likely reflecting underlying 'vaccine-accepting' attitudes rather than any immunological interaction. However, several critical interpretive constraints apply. First, only the full published article — not the truncated abstract — contains the reverse-direction (COVID-19 history → influenza vaccination) results; those data could not be fully confirmed from the abstract text provided and should be verified directly. Second, extreme heterogeneity (I² ≥ 99%) across the pooled estimates — which subgroup analysis, meta-regression, and sensitivity analysis only partially resolved — means the pooled OR cannot reliably be applied to any specific clinical context. Third, the 95% prediction intervals cross OR = 1.0 for the willingness outcome, confirming that the association is expected to be negligible or absent in some populations or settings. Fourth, because the overwhelming majority of included studies are cross-sectional surveys, no causal inference is warranted; confounding by healthcare engagement, health literacy, socioeconomic status, and systemic vaccine access is likely substantial and largely uncontrolled. Fifth, the included studies span the highly variable COVID-19 pandemic period (2020–2026), during which vaccine availability, mandates, hesitancy climate, and public health messaging changed dramatically, further limiting generalizability of any pooled estimate. The findings do not establish that promoting influenza vaccination will directly cause increased COVID-19 vaccine uptake, nor vice versa. The evidence supports the concept that vaccine-accepting behavior is a consistent predictor across vaccine types and could inform co-promotion strategies, but it does not support deriving specific numeric risk estimates for individual patient counseling. Populations underrepresented or heterogeneously represented include immunocompromised patients, low-income and middle-income country populations, pediatric populations (addressed in a minority of studies), and settings with mandatory vaccination policies. __LIMITATIONS__ Extreme statistical heterogeneity (I² ≥ 99%) across virtually all pooled analyses, which persisted despite subgroup analyses and meta-regression; this severely undermines the interpretability of any single pooled estimate.,Wide 95% prediction intervals that cross OR = 1.0 for the willingness arm, indicating that the true effect may be null in some populations or settings.,Predominance of cross-sectional survey designs with self-reported vaccination history and intent, precluding causal inference and subject to recall bias and social desirability bias.,Uncontrolled confounding by baseline vaccine acceptance attitudes, healthcare access, socioeconomic status, education, and health literacy — all of which are likely to drive both exposures and outcomes simultaneously.,Heterogeneous study periods spanning the full pandemic timeline (2020–2026), during which vaccine policy, availability, and public sentiment varied substantially within and between countries.,Incomplete geographic diversity: overrepresentation of studies from China and the United States, with limited data from sub-Saharan Africa, Latin America, and lower-income settings.,The abstract provided by the user was truncated; the full COVID-19 vaccination history → influenza vaccination direction results (OR, CI, I²) are available only in the full-text publication and could not be verified solely from the submitted abstract.,Funding source not reported, precluding assessment of potential financial conflicts of interest.,Study quality assessment methodology (e.g., NOS for cross-sectional studies) was referenced but the distribution of quality scores across included studies was not available in the abstract.,No adjustment for co-administration of vaccines, vaccine mandates, or temporal proximity between vaccine campaigns, which may confound real-world uptake associations. __BOTTOM_LINE_CATEGORY__ potentially_practice_informing __BOTTOM_LINE__ Prior influenza vaccination is robustly associated with approximately 2.7-fold higher odds of COVID-19 vaccine willingness and receipt across diverse populations, and the reverse association appears directionally consistent, suggesting a shared 'vaccine-accepting' behavioral phenotype rather than a direct causal mechanism. However, extreme heterogeneity (I² ≥ 99%), predominantly cross-sectional evidence, wide prediction intervals, and uncontrolled confounding mean this meta-analysis should inform co-promotion framing in public health communication rather than drive specific clinical protocols. No change in individual clinical practice is warranted on the basis of this evidence alone; future longitudinal and interventional studies are needed to determine whether targeted co-promotion strategies produce durable gains in dual vaccine uptake.