Developing a forecasting model for cholera incidence in Dhaka megacity through time series climate data.

Area of Science:

• Epidemiology
• Environmental Health
• Time-Series Analysis

Background:

• Cholera is a significant public health threat, exacerbated by environmental factors and contaminated water.
• Understanding the link between climate variables and cholera incidence is crucial for developing predictive models.

Purpose of the Study:

• To establish correlations between climatic variables and cholera incidence.
• To develop and compare single-variable and multi-variable Seasonal Auto-Regressive Integrated Moving Average (SARIMA) models for cholera forecasting.

Main Methods:

• Time-series analysis of cholera incidence data from 2000-2013.
• Development and comparison of single-variable SARIMA models (SVMs) and multi-variable SARIMA models (MVMs).
• Evaluation of model performance using metrics like AIC, BIC, RMSE, and MAE, with a focus on lead time.

Main Results:

• Maximum temperature (r=0.56) and rainfall (r=0.43) showed the strongest correlation with cholera incidence.
• A 1°C increase in maximum temperature predicted a 7% rise in cholera incidence with a one-month lead time (p < 0.001).
• Multi-variable SARIMA models (MVMs) outperformed single-variable models (SVMs), with an MVM using rainfall and temperature providing the best fit for forecasting.

Conclusions:

• Climate variables, particularly rainfall and maximum temperature, are strong predictors of cholera incidence.
• Multi-variable SARIMA models with a one-month lead time offer a reliable approach for cholera forecasting.
• These findings can enhance public health preparedness and cholera risk management strategies.