Risk prediction system for dengue transmission based on high resolution weather data.

Area of Science:

• Environmental Science
• Epidemiology
• Data Science

Background:

• Dengue is a rapidly spreading vector-borne disease with significant global health impact.
• Predicting dengue transmission dynamics, including risk and outbreak timing, remains challenging.
• Accurate prediction is crucial for effective public health interventions.

Purpose of the Study:

• To develop a robust model for predicting dengue transmission risk.
• To utilize high-resolution weather data and entomological surveillance for risk assessment.
• To improve the accuracy of predicting vector (Aedes aegypti) density and dengue risk.

Main Methods:

• A two-stage risk prediction system was implemented.
• Stage one used logistic regression with weather data to predict Aedes aegypti larvae presence.
• Stage two employed a zero-inflated negative binomial model to estimate larvae counts in positive areas.

Main Results:

• The two-stage model achieved 71% accuracy in identifying larvae-positive locations.
• It predicted larvae numbers with 98% coverage probability over 95% prediction intervals.
• This approach improved accuracy by 29% and reduced mean squared error by 9.6% compared to single-stage models.

Conclusions:

• High-resolution weather data can effectively inform dengue risk prediction.
• A two-stage modeling approach enhances risk assessment in geographically diverse regions.
• The system provides valuable insights into localized dengue transmission risk distribution.