Empirical Model of Spring 2020 Decrease in Daily Confirmed COVID-19 Cases in King County, Washington.

Area of Science:

• Epidemiology
• Mathematical Modeling
• Public Health

Background:

• Accurate short-term COVID-19 case incidence projections are crucial for public policy and disease outbreak management.
• While complex compartmental models are popular, the fundamental two-parameter exponential model has historical significance in pandemic analysis.
• Short-term predictions are generally unaffected by herd immunity shifts, making simpler models viable.

Approach:

• The study advocates for the routine inclusion of exponential models as a reference or null hypothesis in epidemiological reports.
• Exponential models should be fitted separately for distinct jurisdictions and time intervals with significant changes in factors affecting the reproduction number (R).
• Incidence-count half-life (t1/2) is proposed as a more robust statistic than R for exponential model analysis.

Key Points:

• An exponential model was applied to King County, Washington, during Spring 2020 to demonstrate its utility.
• Model parameters and predictions remained stable over one-to-four-month intervals during the pandemic.
• The exponential model's predictive accuracy surpassed that of more complex models in this application.

Conclusions:

• The COVID-19 pandemic can be effectively modeled as a series of exponential curves, each lasting one to four months.
• The duration of these exponential intervals is difficult to predict precisely but tends to mirror the length of preceding intervals.
• The exponential model provides a stable and accurate forecasting tool for near-future disease incidence.