A Spreadsheet-Based Short Time Forecasting Method for the COVID-19 Pandemic.

Area of Science:

• Epidemiology
• Computational Biology
• Public Health

Background:

• High COVID-19 active cases strain medical facilities.
• Existing forecasting models often require significant computational resources and expertise.
• Predicting active cases is challenging due to complex dependencies on Non-Pharmaceutical Interventions (NPIs) and social factors.

Purpose of the Study:

• To develop a simple, accessible spreadsheet-based model for forecasting COVID-19 active cases.
• To aid emergency management by predicting active cases in the immediate future (next few days).
• To evaluate the effectiveness of polynomial functions in capturing disease dynamics and peaks.

Main Methods:

• Development of a spreadsheet-based forecasting model for active COVID-19 cases.
• Application of quadratic, cubic, and quartic polynomial functions to model disease peaks.
• Evaluation of prediction accuracy for 1, 3, and 6-day forecasts.
• Implementation of a daily data-updating method, similar to weather forecasting.

Main Results:

• Quadratic polynomial functions demonstrated superior performance in predicting peaks.
• The daily updating prediction method showed good accuracy, even during periods of sharp trend changes.
• The model effectively captured the dynamics of active cases influenced by NPIs.

Conclusions:

• A simple spreadsheet model can effectively forecast COVID-19 active cases.
• The daily updating approach enhances prediction accuracy and adaptability.
• This model offers a valuable tool for public health emergency management.