Geo-helmintíases em Alagoas : insights de séries temporais para o planejamento em saúde

Abstract

Geo-helminthiasis represent a persistent challenge to public health, especially in regions with limited sanitation infrastructure. This study analyzed the temporal evolution of these infections in the state of Alagoas between 2010 and 2021, applying predictive models based on time series. Prevalence data were extracted from the Schistosomiasis Control Program, focusing on individuals aged 7 to 17. Three analytical approaches were used: Autoregressive Integrated Moving Average (ARIMA) models, Error, Trend, and Seasonality (ETS) model, and Holt's Exponential Smoothing (AEHolt). The performance of the methods was compared using accuracy metrics such as MAE, RMSE, and MAPE. The results indicated a significant decline in prevalence, from values over 30% in 2010 to less than 10% in 2021. The AEHolt model excelled in short-term forecasts (1 and 3 steps), while ARIMA showed greater robustness in long-term horizons (6 and 12 steps). ETS presented intermediate and consistent performance. It is concluded that the use of predictive techniques can enhance strategic planning in public health, guiding intervention policies to mitigate regional inequalities. Strengthening basic sanitation and health education remains essential for the effective control of geo-helminthiasis.