Infection kinetics of Covid-19 and containment strategy

Area of Science:

• Epidemiology
• Computational Biology
• Public Health

Background:

• Covid-19 presents a high mortality-to-infected ratio, necessitating effective containment strategies.
• Limited therapeutics and early vaccination phases underscore the importance of non-pharmaceutical interventions like lockdowns.

Purpose of the Study:

• To develop and validate a mechanistic model, PHIRVD, for predicting Covid-19 mortality dynamics.
• To assess the impact of lockdown strategies on pandemic containment and mortality.
• To identify a more dynamic pandemic descriptor than the reproduction number.

Main Methods:

• Developed PHIRVD, a mechanistic infection propagation model incorporating six infection stages (Healthy, Predisposed, Infected, Recovered, Vaccinated, Deceased).
• Employed Bayesian Markov Chain Monte Carlo (MCMC) for machine learning the model's parameters.
• Trained the model on data from February 10 to June 29, 2020, for mortality prediction up to November 2020.

Main Results:

• PHIRVD accurately predicted Covid-19 mortality profiles for 18 countries, including second-wave kinetics.
• The model demonstrated the effectiveness of early, prolonged, and strategic lockdowns in pandemic control.
• Mortality-to-infection ratio emerged as a more dynamic pandemic descriptor compared to the reproduction number.

Conclusions:

• PHIRVD offers a reliable tool for mortality prediction and understanding pandemic evolution.
• Strategic and sustained lockdowns are crucial for containing Covid-19 relapses.
• The findings support public health policies that integrate non-pharmaceutical interventions with future vaccination efforts.