Reaching Zero-Dose Children in India: Four Family Archetypes

Immunization has quietly saved countless lives across India, shielding children and empowering communities to defeat deadly diseases. Despite these remarkable achievements, a small yet vital group of children known as “zero-dose” children have still not received even a single routine vaccine. These children are often hidden in remote corners of villages and urban pockets, slipping through the cracks of the health care system. Addressing this gap is the core goal of India’s Zero-Dose Implementation Plan, which seeks to better understand and reach those who have been left behind.

Behind every zero-dose child lies a family navigating a complex web of challenges, such as balancing social norms, daily work pressures, migration, and gaps in essential services.

We asked ourselves: Do all these families face identical barriers, or can they be categorized into distinct groups that require tailored support to ensure their children receive vaccines?

To find answers, we collaborated with the William J. Clinton Foundation, engaging with over a thousand caregivers across 42 districts in two states (35 in Uttar Pradesh and 7 in Bihar). These conversations went beyond just recording services received; they aimed to understand how families think about health, who they trust, and what obstacles they encounter on hectic days when vaccination becomes just another task.

Instead of analysing factors in isolation, we adopted a behavioural, data-driven approach. By applying machine learning techniques to more than 200 variables at the household-level – ranging from economic status and social norms to health worker visits — they identified groups of families exhibiting similar patterns. This approach provided a nuanced understanding, moving beyond common explanations like “lack of access” or “low awareness,” to reveal the deeper realities behind missed vaccinations.

Two-Phase Analysis: MCA and K-Means Clustering

The analysis was carried out in two phases. First, a Multiple Correspondence Analysis (MCA) helped simplify the complex data, uncovering hidden relationships among variables such as delivery methods, frequency of visits, and concerns about side effects. Then, K-means clustering (an algorithm that breaks unlabelled data into distinct clusters based on feature similarity) was employed to segment households into four distinct archetypes, each representing a different story of how children fall through the cracks.

What surfaced wasn’t a single homogenous profile, but four distinct categories — each with unique circumstances, challenges, and opportunities for intervention. These were:

1. Systemic Dropouts: “Reached but not retained”:

These families have interacted with health services through facility delivery or early prenatal care, which indicates initial trust. However, support often ceases after childbirth. Home visits become irregular, and counselling fails to address caregivers’ genuine concerns about side effects or the vaccination process.

Without continuous engagement, these families gradually drift away from the health system. Strengthening postnatal follow-up through regular visits, empathetic counselling, and seamless linkages between delivery points and outreach programs can help convert initial contact into complete vaccination coverage.

2. Resource-Limited Refusers: “Willing, but unable”:

On paper, these might appear as vaccine refusers, but their realities are starkly different. They often face extreme financial hardship, dealing with food insecurity, informal employment, and unstable housing. Many births occur at home, and lacking official documents keeps them outside formal systems including immunization programs. Their primary focus is sustenance, making vaccination a lower priority activity.

To reach them, vaccination campaigns must be paired with social services support efforts, bringing vaccines directly to locations where families access food, work, or welfare services, thereby reducing the logistical and economic barriers.

3. Conflicted Caregivers: “Ready in words, hesitant in action”:

These caregivers seek support for vaccinations and often interact with the system for health services such as pregnancy care and childbirth. Yet, vaccination still gets delayed or skipped because of household dynamics. Private doubts or fears, especially from elders or spouses, often influence decisions. Even if a mother is willing, other influential family members might quietly block vaccination afterwards.

 Successful strategies involve empathetic dialogue, building trust, and involving key household decision-makers over multiple conversations to address fears and dispel doubts.

4. Resistant Traditionalists: “Out of reach and out of frame”:

This group remains largely invisible to the health system for years, often across generations. They tend to belong to marginalised communities or tightly knit traditional groups, where health choices are rooted in customs, spiritual beliefs, and community elders rather than official guidance. Distrust of government services, lack of documentation, and beliefs like “the body cannot be injected” or that “child survival is in God’s hands” further reinforces their distance from the health system.

For such families, trust-building requires partnerships with faith leaders, traditional healers, and respected community figures, respecting local customs while gradually introducing vaccination and health services.

Tailored Solutions for Zero-Dose Immunization Barriers

The key takeaway from this exercise is that the zero-dose children are not one uniform group; they face a diverse set of barriers based on their environment, socio-economic status, family dynamics, and cultural beliefs. Effective solutioning strategies must be tailored accordingly.

Simple fixes, like follow-up after hospital births, can be integrated within existing health systems at a low cost. Broader issues, such as economic vulnerability and deep-seated mistrust, require cross-sectoral collaboration and community engagement.

This approach also demonstrates the power of combining advanced data analysis with empathy-based listening. Machine Learning methods can uncover hidden patterns, while conversations with families can ensure these insights are supplemented with the lived experiences. In India, where immunization coverage is generally high, but some children continue to remain unvaccinated, these hyperlocal, behaviorally sensitive strategies can help reach every last child.

What remains clear across all these groups is the overlapping vulnerabilities that push families to the margins of the health system. Closing this gap demands more than just opening clinics and scheduling sessions; it calls for building trust, showing empathy, and crafting solutions that resonate with families’ real lives. When this happens, we can ensure that the child’s first dose is also their last missed opportunity for protection against deadly vaccine-preventable diseases.