The project created an index called the Early Childhood Development Friendly Index (ECD-FI), which uses evidence-based service indicators to assess factors that contribute to creating environments conducive to promoting early childhood development at the municipal level, through monitoring and identifying opportunities to expand ECD programs. As a product of machine learning, this index runs analytical models that consider demographic data and risk factors. Disaggregated data are not available in Brazil.
The Food Insecurity Index in Early Childhood Development (ECD-FI) is established through a comprehensive set of indicators categorized into five domains: Health, Nutrition, Early Learning Opportunities, Responsive Parental Relationships, and Safety and Protection. Additionally, municipal-level metrics such as population size, annual births, infant mortality rates, child poverty rates, and domestic violence incidences were included. Data sources include multiple databases and national surveys. To evaluate the ECD-FI, outcomes such as stunting and malnutrition data were used. The analysis combined supervised and unsupervised machine learning techniques to classify municipalities and identify patterns, resulting in a more accurate ECD-FI than traditional methods.
The Municipal Index Friendly to Early Childhood (IMAPI) was created to assess and monitor the environment favorable to child development across Brazil’s 5,570 municipalities. Based on the Care Framework recommended by the World Health Organization and UNICEF, IMAPI highlights the essential role of early childhood development systems in mitigating adverse experiences during early childhood.
To help children reach their full potential, it is crucial to provide the five interrelated and indivisible components of attentive care: good health, adequate nutrition, safety and protection, responsive caregiving, and learning opportunities. An eight-step methodology was developed to systematically and multisectorally identify valid and reliable indicators. This process used data science engineering techniques to create municipal-level indices of Attentive Care, using routine information systems.
Additionally, a user-friendly platform was developed, allowing stakeholders to easily access IMAPI indices. This platform includes materials to support effective use and dissemination of the indices, facilitating informed policy decisions that promote child development and address inequalities from the earliest stages of life. Thus, IMAPI has the potential to facilitate decision-making among Brazilian stakeholders and generate significant global impacts (https://imapi.org/).
Implementing the ECD-FI at the municipal level represents an innovative approach both in Brazil and globally. Around the world, effective indices have been developed to monitor progress on key indicators that assess the environment conducive to early childhood development (ECD), particularly in areas such as breastfeeding and newborn care.
A distinctive feature of this proposal is the use of machine learning methods to create this index. This approach allows for the development of analytical models that take into account demographic information and specific risk factors for each municipality, especially in contexts with challenging ECD environments. As a result, a personalized index is generated, tailored to the unique circumstances of each municipality. This innovation facilitates comparisons among peers at a level of data disaggregation that was previously unavailable in Brazil.
The first 1,000 days, starting from conception, represent a particularly sensitive period for child development. Investing in attentive care for Early Childhood Development (ECD) during this critical timeframe is essential to enhance health, productivity, and social cohesion throughout life, providing benefits that can be passed down through generations. In Brazil, the diversity and disparities in environmental, political, and socioeconomic conditions among the 5,570 municipalities are likely contributing to inequalities in ECD. This means that the poorest children are at risk of not reaching their full physical, psychosocial, emotional, and cognitive potential, which perpetuates the cycle of intergenerational poverty.
IMAPI represents an initial effort to quantify nutrition care indicators at the municipal level, addressing the global demand for a decentralized, cost-effective monitoring system. This initiative has the potential to enhance decision-making processes among stakeholders in Brazil.
The next steps involve testing the internal validity and sensitivity of IMAPI in Brazil by utilizing existing population-level measures of child development. The team will also assess IMAPI’s sensitivity and specificity through the development and pre-testing of a municipal decision-making process based on IMAPI results. This will help to investigate how the use of IMAPI can lead to improvements in investment decisions regarding nutrition care at the municipal level.
Additionally, external validity will be evaluated by applying this methodology to examine child development in three low- to middle-income countries. The data collected will not only inform programmatic actions but also facilitate meaningful comparisons over time and across countries. This initiative aims to contribute to the formulation of guidelines for a standardized global monitoring system that supports evidence-based decision-making in child development.