Multidimensional algorithm for anthropometric assessment of children, on a population scale, through the offline app "App Pastoral da Criança + Gestante"

Abstract

Child nutritional assessment is essential for monitoring growth, especially in contexts of nutritional transition, such as that of Brazil. The use of anthropometric indices based on the World Health Organization's growth curves is the conventional procedure employed in Primary Health Care within the SUS. Limitations related to the complexity of interpretation or improper use of the Child Health Record have been observed. To overcome these limitations, the “Pastoral da Criança+Gestante” app was developed, incorporating an advanced multidimensional algorithm. This algorithm prioritizes over 1,090 filter combinations, with standardization in 700 integrated responses, taking into account anthropometric measurements (weight, height), demographic variables (age and gender), food recalls, and growth trends such as changes in Z-scores. This approach allows for more precise identification of nutritional deviations and prioritization of critical situations, such as severe obesity or malnutrition, through personalized electronic guidance cards. Additionally, the app operates offline, facilitating its use in areas with low connectivity. Data from 2024 indicate that, of the more than 106,000 children evaluated, 7.7% presented conditions requiring urgent care. The tool not only reduces errors in manual assessments but also enhances the integration of technology and public health, demonstrating its value as a support in primary care and the improvement of public health policies.