The "epidemiar" R Package: Integrating Public Health Surveillance and Environmental Monitoring Data for Early Detection and Early Warning of Infectious Disease Transmission

Public health monitoring of environmentally mediated diseases can benefit from incorporating information on these environmental factors into surveillance activities. This melding of data improves the ability to detect early indication of disease outbreaks, which allows for more efficient and proactive public health interventions. The Epidemic Prognosis Incorporating Disease and Environmental Monitoring for Integrated Assessment (EPIDEMIA) project is using this approach to produce operational malaria forecasts for the Amhara region of Ethiopia. Key environmental predictors include precipitation from NASA GPM along with land surface temperature and vegetation indices derived from MODIS.

The EPIDEMIA modeling of disease transmission for early detection and early warning evaluation is done in R, a free software for statistical computing. We developed the "epidemiar" R package to provide a generalized set of functions for disease forecasting. The package is flexible enough to forecast a wide range of infectious diseases that are affected by climate and other environmental conditions. The main function accepts case count and environmental data, and the users can specify their own time periods, modeling parameters, and early detection/warning settings. An early detection time period is specified to trigger alerts as the algorithm runs over the case data in that period. Modeling is done via general additive model regression of multiple factors, including geographic clusters, lagged environmental drivers, long term trends, and seasonality. Early warning alerts are triggered based on an algorithm that evaluates the model-produced forecasts for a user-specified future time period. The resulting output is fed into formatting documents, such as Sweave files, to create distributable reports with maps and graphs of the results.

Using a GeoHealth interdisciplinary approach, the new epidemiar R package facilitates the combination of earth science data with public health surveillance to support early detection and early warning of disease transmission events for infectious diseases with environmental drivers. The results can be used to help support operational public health monitoring and interventions.