Taking AI to the Last Mile: Transforming Maternal and Child Health through AI-Enabled, Community-Led Anaemia Screening

Anaemia remains a major public health challenge, particularly among adolescent girls, pregnant women and young children in resource-constrained settings. Its effects extend across generations: anaemic adolescent girls may enter pregnancy with depleted nutritional reserves, increasing the risk of adverse pregnancy outcomes and low birth weight, while poor maternal and early-childhood nutrition perpetuate cycles of illness and vulnerability. Conventional haemoglobin screening often requires blood collection, laboratory infrastructure and trained personnel, creating barriers to timely detection in remote communities. Every Infant Matters developed a community-led model integrating AI-enabled, non-invasive haemoglobin screening with existing public health systems. The objective is to decentralize early anaemia screening, improve access among underserved populations, and establish a continuum from community-level detection to treatment, nutritional support and follow-up. Methodology The programme recruits and trains women from local communities to deliver screening using portable, smartphone-enabled EzeRx AI technology. The non-invasive platform enables haemoglobin screening without blood collection or laboratory infrastructure. Screening is conducted through door-to-door household outreach, community health camps and Government Anganwadi centres, targeting adolescent girls, pregnant and lactating women, and children aged 1–5 years. Individuals identified through screening are referred to appropriate government or local health facilities for clinical evaluation, treatment and follow-up. The model has been implemented across diverse Indian settings, including Kargil (Ladakh), Kota (Rajasthan), Silvassa (Dadra & Nagar Haveli), and Vidisha (Madhya Pradesh), in partnership with government health systems, frontline health workers, community leaders and technology and private-sector partners. Results and Discussion To date, the programme has screened more than 23,000 adolescent girls and women, screened more than 19,000 children aged 1–5 years, supported more than 52,000 pregnancies, and trained more than 3,600 frontline health workers. Implementation across geographically and socioeconomically diverse settings demonstrates the feasibility of decentralizing haemoglobin screening beyond conventional laboratory-based pathways. The non-invasive approach facilitates community acceptance, particularly among children and pregnant women, while delivery through trusted local women strengthens engagement and continuity of care. Importantly, the intervention integrates digital diagnostics into existing public health structures rather than creating parallel systems. This combination of technology, community ownership and government integration creates a potentially scalable pathway for addressing anaemia and its downstream consequences, including poor maternal nutrition, low birth weight and childhood malnutrition. Conclusions This experience demonstrates that AI-enabled diagnostics can extend beyond technological innovation to become a practical tool for health equity when combined with community-led delivery and public health system integration. Taking screening to households and community settings can help overcome geographic, infrastructural and behavioural barriers that limit conventional testing. The model offers a scalable, people-centred approach for strengthening maternal and child health services in low-resource settings. Its broader lesson is that the impact of frontier health technologies depends not only on their technical capabilities, but on whether they are designed around the communities they are intended to serve.