What works in immunization programming?

Background: Implementation of stakeholder-informed interventions is critical to address childhood immunization barriers that may affect male and female caregivers differently. Objective: Adapt, implement, and evaluate a family-focused intervention to promote fathers' engagement and improve childhood immunization in Ghana. Methods: Mixed-methods evaluation: 1) formative assessment to adapt intervention, 2) implementation and monitoring, 3) summative evaluation to assess intervention effect and implementation feasibility and acceptability. Phase 1: baseline household survey (n=481 caregivers), eight caregiver focus groups, and 20 interviews (n=10 health workers, n=10 caregivers). Stakeholders reviewed baseline findings to develop a tailored intervention, including interpersonal communication training of health workers, community meetings, and community-based quality improvement teams. During Phase 2, we implemented and monitored the intervention, using rapid surveys (n~250 caregivers) of community meeting participants' relevant vaccination perceptions and experiences. In Phase 3 we conducted a summative evaluation, including caregiver focus groups and interviews (n=14 health workers, n=9 caregivers). Results: baseline assessment findings suggested fathers needed vaccine knowledge and mothers needed social support to seek vaccination. Thus, our intervention aimed to increase fathers' involvement in childhood immunization by building their knowledge, awareness, and support for vaccination. During the workshop, stakeholders reported acceptability of proposed intervention components. Evaluation suggests improved joint decision-making between caregivers and social support among fathers for attending the vaccination visit, sharing reminders, and providing money for transport to vaccination sites. Discussion: Findings demonstrate how community engagement in vaccine program design and implementation can address persistent sex-related barriers to vaccine demand, potentially improving uptake.

The "Keluarga Berimun" digital campaign was designed to increase parents' knowledge, confidence and participation in childhood immunization (0–18 months) across urban Depok City and suburban Bekasi Regency. To reach parents who actively consume online news and health content, the campaign used programmatic advertising across 15 networks of national publishers known for high credibility and engagement among family audiences.
Using A/B testing across six iterative cycles (March–Semptember 2025), our campaign compared variations in visuals, tone, and call-to-action. Metrics measured include impressions, clicks, and click-through rate (CTR). After six months of implementation, our campaign reached over 30 million impressions and generated more than 200,000 clicks.
We identified that ads mentioning "immunization," using question-style headlines, and featuring human imagery consistently performed best across both areas. Message tone also influenced performance: friendly messages resonated more in Depok City (0.78% CTR), while empathic tones were more effective in Bekasi Regency (0.76% CTR). Call-to-action phrasing shaped engagement as well, with "Free Consultation Here" performing better in Depok City (0.69% CTR) and "Ask for Free Here" in Bekasi Regency (0.70% CTR). Non-descriptive banners slightly outperformed in Depok City (0.67% CTR), whereas descriptive banners achieved higher engagement in Bekasi Regency (0.75% CTR).
Findings indicate that local audience context shapes how digital messages are received and acted upon. Differences in tone and message detail between urban and suburban settings highlight the value of continuous testing and locally attuned, human-centered communication to enhance digital immunization impact.

SBCC campaigns are widely used to motivate vaccination, yet often rely on broad messaging or assumptions about what drives behavior. This limits effectiveness and leaves programs guessing which barriers to address. This study applies causal modeling and segmentation to provide partners with precise, actionable insights.
In priority areas of Karachi, Pakistan, where one in three children remain zero-dose or under-immunized, we used machine learning-based causal models to examine two key transitions: moving from zero-dose to under-immunized, and from under-immunized to fully immunized. We then segmented caregivers of non-fully immunized children into five distinct groups, ranging from those with entrenched mistrust to those with high desire but poor follow-through. 
At the first stage of the vaccination journey, vaccination status was driven by gaps in motivation, household context, and lack of trust in vaccines and vaccination program. Drivers in the later stage were more numerous and diverse, including experiences and interactions with the vaccination system. At both stages, desire to receive recommended vaccines and spousal support were crucial. 
Findings highlight the value of tailoring SBCC strategies. While causal models identified priority targets for intervention, segmentation underscored that segment-specific strategies are crucial, as not all drivers were important for all subpopulations. For example, engaging fathers and religious leaders is critical for segments where opposition is strongest, whereas strategies emphasizing follow-though are more effective for those already supportive. Our approach provides partners with actionable evidence to design more precise, high-impact SBCC strategies.

Para combatir la disminución de la cobertura de inmunización en América Latina y el Caribe (LAC), UNICEF LACRO implementó un protocolo de investigación formativa basado en el Diseño Colaborativo. Este marco integra la Ciencia del Comportamiento (BeSci) y el Diseño Centrado en el Ser Humano (DCH) con adaptaciones para la región en base a una reflexión crítica sobre los modelos. El análisis se llevó a cabo con datos recogidos en cinco países prioritarios de América Latina, para mapear barreras y facilitadores en el viaje de vacunación. La presentación se centrará en el proceso de transformar estos hallazgos cualitativos en tres narrativas de viajes de usuario (Acto de balance, Lo moderno se encuentra con lo ancestral, y La última milla) y perfiles de persona (47 en total). Estos modelos, arraigados en la experiencia vivida, actúan como escenarios concretos y aplicables que empoderan a los Programas Ampliados de Inmunización (PAI) para mejorar la prestación de servicios.