estimateR

estimateR estimates the time-varying effective reproductive number (R_t) from delayed observations of infection events (e.g., confirmed cases, hospitalizations, deaths) to support epidemiological surveillance and retrospective analysis.

Key Features:

• Modular design: A modular computational framework that allows implementation and selection of alternative estimation methods.
• Estimation from delayed observations: Produces R_t estimates over time from delayed infection indicators such as case-confirmation incidence, hospitalizations, deaths, and wastewater measurements.
• Speed and efficiency: Validated with simulation studies and reported to deliver estimates comparable to other methods while being approximately two orders of magnitude faster.
• Versatility in data types: Accepts multiple observational data types including clinical incidence data and environmental measurements like wastewater SARS-CoV-2 concentrations.
• Integration with epidemic models: Generates R_t estimates that can be directly interpreted or incorporated into more complex models for purposes such as forecasting.

Scientific Applications:

• Epidemic surveillance and retrospective analysis: Monitoring transmission dynamics and investigating past outbreak behavior through time-varying R_t estimation.
• COVID-19 analysis: Applied to empirical COVID-19 datasets across nine countries to estimate transmission dynamics.
• Dengue epidemiology: Applied to dengue fever data in Brazil for transmission assessment.
• SARS-CoV-2 wastewater surveillance: Used to analyze wastewater measurements of SARS-CoV-2 for community-level transmission inference.
• Influenza transmission studies: Applied to investigate influenza transmission using clinical and environmental data.
• Model-informed forecasting: Provides inputs for more complex epidemic models to support forecasting efforts.

Methodology:

Estimates R_t over time from delayed observational data using a modular computational framework and was validated via simulation studies reporting comparable estimates and ~100× speed improvements; supports input data types including case-confirmation incidence and wastewater measurements.