Epidemiological models for predicting Ross River virus in Australia: A systematic review

Area of Science:

• Epidemiology
• Public Health
• Environmental Science

Background:

• Ross River virus (RRV) is Australia's most prevalent arbovirus.
• Epidemiological models aid in understanding RRV transmission and predicting outbreaks.
• A systematic review and analysis of existing RRV predictive models are lacking.

Purpose of the Study:

• To systematically review and analyze epidemiological models used for predicting Ross River virus disease.
• To identify drivers of RRV incidence and transmission patterns through model analysis.
• To assess the performance and identify research gaps in current RRV predictive modeling.

Main Methods:

• Systematic literature search across major databases (PubMed, EMBASE, Web of Science, Cochrane Library, Scopus).
• Inclusion of studies using population-based data, epidemiological models, and exposure-disease association analysis for RRV.
• Analysis of model types, covariates used (climate, weather, mosquito data), and reported performance metrics.

Main Results:

• Forty-three high/medium quality articles were included, analyzing 140 models.
• Generalized linear models (51.2%) and time-series models (25.6%) were most common.
• Rainfall, temperature, and tide height were the most frequent environmental covariates; only 23.3% of studies reported model performance.

Conclusions:

• Current knowledge of RRV modeling is summarized, highlighting a significant gap in comparative analysis of predictive methods.
• Environmental factors like rainfall and temperature are key drivers in current models.
• Further investigation into advanced forecasting methods, including non-linear mixed models and machine learning, is recommended to enhance predictive accuracy for RRV outbreaks.