Prediction of pulmonary tuberculosis case trends among older adults in Chongqing based on time series models

Tuberculosis is a major global public health issue. Older adult individuals, due to factors like immunosenescence and comorbidities, are at high risk for TB. Chongqing's significant aging population poses severe challenges for TB control in this group.

This study is based on the monthly case counts of pulmonary tuberculosis among older adults aged 65 and above in Chongqing from January 2020 to June 2024. It constructs and compares the Seasonal Autoregressive Integrated Moving Average (SARIMA) model, the Nonlinear Autoregressive Neural Network (NNAR) model, and the hybrid SARIMA-NNAR model to predict the monthly number of PTB cases in 2025.

The study data were extracted from the National Tuberculosis Surveillance System (TBIMS). Data collection and organization for pulmonary tuberculosis cases among individuals aged 65 and above were performed using Microsoft Excel 2019 (Microsoft Corp). Statistical analysis and predictive modeling were conducted using R software, version 4.5.2 (Network Theory Ltd., Bristol, United Kingdom). Data from January 2020 to June 2024 formed the training set, while July to December 2024 data served as the testing set. Model performance was evaluated using Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE).

The number of pulmonary tuberculosis cases among older adults in Chongqing exhibited distinct seasonal fluctuations, with peaks consistently occurring in March and May each year. On the testing set, the hybrid model achieved the lowest MAE (38.39%) and MAPE (11.77%), whereas the NNAR model produced the lowest RMSE (45.52%). Overall, the hybrid model demonstrated a more balanced performance across evaluation metrics. The forecasted case counts for 2025 maintained a similar seasonal pattern, with projected peaks in March and May.

The SARIMA-NNAR hybrid model improves the prediction accuracy of pulmonary tuberculosis case counts in older adults by integrating linear and nonlinear components, providing a scientific basis for optimizing resource allocation and seasonal interventions in Chongqing.