Inferring Values of Epidemic Indicators via SEIR Models to Enhance Epidemiological Surveillance in Russia

In this paper, we describe modeling methods which help obtain additional knowledge from the surveillance data related to outbreaks of acute respiratory infections. Based on algorithms of calibration of compartmental SEIR models to incidence data and uncertainty analysis performed with Monte Carlo methods, we develop a computational approach for the interval assessment of epidemic indicators, implement it using Python programming language, and demonstrate the results using data on influenza and COVID outbreaks in Saint Petersburg. The ultimate aim of the presented methods is to contribute to enhancement of epidemiological surveillance in Russia by adding significant model-based insights into weekly epidemiological reports.