A Predictive Model for the Evolution of COVID-19

Area of Science:

• Epidemiology
• Mathematical Modeling

Background:

• The COVID-19 pandemic presented a significant global health challenge requiring predictive modeling for resource allocation and public health interventions.
• Existing logistic models for infectious disease spread have limitations in accurately forecasting pandemic trajectories.

Purpose of the Study:

• To predict the future evolution of the COVID-19 pandemic in multiple countries.
• To estimate the total number of infections and identify peak infection dates using a novel logistic model.

Main Methods:

• A logistic regression model employing least-squares fitting was utilized.
• The model incorporated an exponential decay function for the infection growth rate, differing from previous linear decay models.
• Model validation was performed using data from China and South Korea.

Main Results:

• The model successfully predicted the pandemic's nearing end in China and South Korea.
• Italy, Germany, Spain, and Sweden data indicated that the infection peak had been reached.
• Predictions include approximately 4 million total infections in the USA, 3.2 million in Brazil (peaking July 5, 2020), and 2.4 million in India (peaking August 3, 2020).

Conclusions:

• The developed logistic model provides valuable predictions for COVID-19 pandemic peaks and total infections across diverse countries.
• The model's predictions for India were discussed in relation to government-imposed lockdown measures and population movement restrictions.
• The findings aid in understanding pandemic dynamics and informing public health strategies.