Changes in Patient Attitudes and Vaccination Intention After Pneumococcal Vaccination Counselling by Trained Family Physicians: A Multicentre Implementation Study in Romania

Introduction: Improving adult pneumococcal vaccination is a public-health priority in Romania, and evidence regarding patient-level outcomes following different approaches to physician training remains limited. Materials and methods: This non-randomised, multicentre, quasi-experimental implementation study examined patient-level changes across three geographically distinct Romanian settings between August and November 2025. Physicians in Iași and Dolj counties received in-person training combining medical content with communication techniques based on motivational interviewing, whereas physicians in the webinar group received online medical training without a dedicated communication module. Eligible adult patients completed questionnaires before and after counselling and reported whether they had been vaccinated or had scheduled a vaccination appointment. Results: The analysis included 295 questionnaire pairs from patients nested within 29 physicians. Baseline attitudes differed markedly across settings (composite score of 5.64 in the webinar group, 5.29 in Dolj County, and 4.09 in Iași County; p < 0.001), as did strong baseline vaccination intention (75.6%, 48.1%, and 27.2%, respectively). Using a composite score comprising all seven attitude items, with item 7 reverse-coded and vaccination intention analysed separately, the cluster-robust group-by-time model indicated different mean changes between settings (global p = 0.001): +0.28 in the webinar group, +1.02 in Iași, and +0.60 in Dolj. Strong vaccination intention increased within each setting, but the magnitude of change did not differ between settings (group-by-time p = 0.929). The combined self-reported vaccination-or-scheduling outcome was observed in 81.5% of patients in the webinar group, 49.4% in Iași, and 73.3% in Dolj; after adjustment for baseline intention, there was no longer evidence of an overall group difference (p = 0.651). Discussion and conclusions: Because training approach, delivery modality, geographical setting, physician cohort, and patient population were not independently allocated, and because baseline levels differed markedly between settings, the findings should be interpreted as exploratory implementation patterns rather than causal comparative effects.