Abstract: Objective To analyze the epidemiological characteristics and forecast the incidence trends of other infectious diarrhea (OID) in the Shapingba District of Chongqing, in order to provide an empirical basis for evidence-based prevention and control strategies. Methods Utilizing OID surveillance data from 2014 to 2023 in Shapingba District, this study employed descriptive epidemiological and spatial autocorrelation methods to characterize the disease patterns. Joinpoint regression analysis was conducted to assess temporal trends. Furthermore, a Seasonal Autoregressive Integrated Moving Average (SARIMA) model and a Back Propagation Neural Network (BPNN) model were developed and compared for forecasting future incidence. Results A cumulative total of 40 377 OID cases were reported in Shapingba District from 2014 to 2023, corresponding to an average annual incidence rate of 331.78 per 100 000 population. The overall reported incidence exhibited an increasing trend from 2014 to 2018 (APC=10.30%, P=0.14), followed by a significant decreasing trend from 2018 to 2023 (APC=-11.88%, P=0.02). The disease displayed a distinct bimodal seasonal distribution, with a primary peak from November to February and a secondary peak from June to August. The average annual reported incidence was significantly higher in males (361.45/100 000) than in females (312.71/100 000) (χ²=6 740, P<0.05). Children under five years of age constituted the most affected population group, and rotavirus was the predominant etiological agent. In a comparative evaluation of predictive models, the BPNN model demonstrated superior performance, with lower MAE and RMSE values than the SARIMA model. The BPNN model forecasts an epidemic peak between November 2024 and January 2025, with projected monthly incidence rates ranging from 26.54 to 31.73 per 100 000 population. Conclusion The epidemiological burden of OID in Shapingba District remains substantial, warranting sustained implementation of surveillance, early warning systems, risk assessment, and health education initiatives. The BPNN model proves to be a more effective tool for forecasting OID incidence compared to the SARIMA model.

Key words: Diarrhea, Rotavirus, Spatial analysis, Regression analysis, Time series analysis, Neural networks