Predicting health literacy in Brazil: a machine learning approach.

Area of Science:

• Public Health
• Health Communication
• Health Services Research

Background:

• Health literacy is crucial for informed health decisions and overall well-being.
• Understanding health literacy levels is vital, especially in low- and middle-income countries.
• Brazil's context offers insights into challenges faced by populations with similar socioeconomic backgrounds.

Purpose of the Study:

• To investigate the prevalence of health literacy in Brazil.
• To identify significant factors predicting health literacy levels within the Brazilian population.
• To contribute data relevant to the Global South and Latin America's health literacy landscape.

Main Methods:

• Descriptive analysis of health literacy levels.
• Application of the Random Forest machine learning algorithm with bootstrap aggregation.
• Utilized Shapley's Additive exPlanation values for result interpretability.

Main Results:

• A predominant level of problematic health literacy was observed in the Brazilian population.
• Medication use, dependence on the Unified Health System (Sistema Único de Saúde), and educational attainment were identified as significant predictors.
• The study provides a detailed analysis of health literacy determinants in a major Latin American country.

Conclusions:

• The findings underscore the urgent need for targeted public policies to address health literacy.
• Interventions should focus on reducing socioeconomic disparities and enhancing the public health system.
• Improving access to and comprehension of health information is critical for population well-being.