Background In a context of lower than targeted influenza vaccination (IV) coverage, identifying factors associated IV uptake is essential to improve population coverage. Although extensive work has been done in this field, scarcer studies focused on framing these factors within a multifactorial model. Having the social ecological model (SEM) as framework, this study intended to identify and quantify the SEM levels associated to seasonal IV uptake in the Portuguese population aged 65 and more years.

Methods Data from the Portuguese 2014 National Health Survey was restricted to individuals aged 65 and more years (n=5669). The outcome of interest was self-reported IV uptake in the 2 years previous to the interview. Twenty three independent variables were selected from the survey and were allocated in one of the five SEM levels: individual, interpersonal, organizational, community and policy. Sex stratified and age adjusted analysis using Poisson regression were performed for each SEM level. To evaluate statistical significance and marginal contribution of each SEM level for explaining IV uptake, the full model was fitted (all variables from all 5 SEM levels). To determine joint statistical significance of variables of each SEM level, the model without respective SEM level was compared with the full model using likelihood ratio test. Additionally, marginal contribution of each SEM level was measured by relative reduction in magnitude of pseudo R square.

Results For both men and women, older age groups (85 or more vs 65–69 age group; PR=1.59 for men and PR=1.56 for women); having 3 or more chronic conditions (PR=1.39 for men and PR=1.35 for women); number of GP and outpatient visits in the previous 4 weeks were associated to higher IV uptake. For men, only 2 out of the 5 SEM levels were associated to IV uptake (individual and organizational). For women 3 levels were relevant for this preventive measure (individual, organizational and community). Main marginal contribution for explaining the IV uptake, came from individual (17.9% and 16.3%) and organizational (30.7% and 22.7%) levels for both men and women respectively.

Conclusion Besides individual characteristics, like age and health status - known determinants of IV uptake - this study highlights the importance of access and use of health care services for adoption of IV preventive measure. Moreover, it evidences a sex differential behaviour that should be accounted in the definition of the IV campaign strategy.