Forecasting Covid-19 in the United Kingdom: A dynamic SIRD model

Area of Science:

• Epidemiology and Public Health
• Mathematical Modeling
• Biostatistics

Background:

• Traditional SIRD models often assume constant rates, limiting their ability to capture real-world pandemic dynamics.
• Time-varying parameters and reinfection are crucial factors in understanding disease transmission and population impact.
• Accurate forecasting requires models that can adapt to evolving epidemiological conditions.

Purpose of the Study:

• To develop a stochastic generalization of the SIRD model with time-varying rates for infection, mortality, and underreporting.
• To incorporate a mechanism for reinfection within the compartmental framework.
• To estimate and forecast key epidemiological variables using real-world data from the UK.

Main Methods:

• Utilized a state-space framework to create a stochastic SIRD model.
• Incorporated time-varying parameters for infection, mortality, and underreporting rates.
• Applied the model to UK daily data from April 2020 to September 2021, using reported cases and deaths as inputs.

Main Results:

• Estimated infection rates showed wave-like patterns correlating with new variants and social measures.
• Mortality rates declined significantly in 2021, likely due to vaccination programs.
• Underreporting rates were volatile but showed a downward trend, indicating increased testing over time. Significant reinfection was observed in 2021.

Conclusions:

• The stochastic SIRD model provides a robust framework for analyzing and forecasting infectious disease dynamics with time-varying parameters.
• The model highlights the impact of vaccination on mortality and the increasing significance of reinfection.
• Estimated effective reproduction rates align well with observed case and death data, validating the model's predictive power.