Data-driven modeling reveals a universal dynamic underlying the COVID-19 pandemic under social distancing

Area of Science:

• Epidemiology
• Statistical modeling
• Network theory

Background:

• The COVID-19 pandemic presented unprecedented challenges in understanding disease dynamics.
• Social distancing measures were implemented globally, necessitating the study of their impact on transmission.
• Accurate forecasting models are crucial for public health response and resource allocation.

Approach:

• Analyzed cumulative COVID-19 infection and death data from various countries and US states.
• Utilized the error function (erf) to model the observed growth and slowing patterns.
• Performed data collapse to demonstrate universal dynamics across different populations and policies.
• Simulated disease transmission on scale-free and random networks with non-Markovian dynamics.

Key Points:

• COVID-19 pandemic dynamics, including infections and deaths, exhibit universal behavior under social distancing.
• The error function accurately describes the transition from exponential to power-law growth and eventual slowing.
• Data collapse across diverse geographical and policy contexts confirms this universal outbreak pattern.
• Predictive power of models is limited until near the curve flattening stage.

Conclusions:

• The universal dynamics observed in the COVID-19 pandemic can be modeled using the erf function.
• This universality suggests common underlying mechanisms in disease spread, irrespective of population or policy specifics.
• The findings provide a framework for improved forecasting and understanding of epidemic trajectories.
• Simulations support the consistency of this universal dynamics with transmission on complex networks.