Algorithmic Approach to Determination of Epidemic Thresholds in Infectious Disease Surveillance Systems

Abstract

Introduction: This review is devoted to the algorithmic approach to establishing epidemic thresholds for a wide range of diseases, including influenza and acute respiratory infections. Objective: To compare Russian and foreign approaches to the determination of epidemic thresholds within public health surveillance systems. Materials and methods: To reveal the algorithmic approach to establishing epidemic thresholds in the epidemiological surveillance system, we summarized the results of 14 foreign scientific works and two domestic method guidelines published before December 31, 2023. The literature search was conducted in the eLibrary, CyberLeninka, PubMed, and Google Scholar databases using the keywords “epidemic threshold” and “epidemic”. We compared domestic and foreign algorithms for establishing epidemic thresholds by various characteristics, including the statistical method used, determination of a numerical value of the epidemic threshold, complexity of the algorithm, and the possibility of automating calculations. Results: Here we discuss the classification and comparative characteristics of the basic algorithms for determining epidemic thresholds used in various countries of the world when carrying out epidemiological surveillance (including the syndromic one). We describe the existing methods for establishing and presenting epidemic thresholds, as well as the sequence of steps for performing the Farrington algorithms, the Early Aberration Detection System C1–C3, the Method of Moving Epidemics, the Method of Moving Percentiles, Multi-level identification of increasing activity by indicators taking into account mixed effects, as well as algorithms provided in Russian Method Guidelines MR 3.1.2.0118–17 and MR 3.1.2.0303–22. We also dwell on the problems of development, accuracy assessment and prospects for the implementation of existing and developed algorithms. Conclusions: Current algorithms for establishing epidemic thresholds in epidemiological surveillance systems around the world are diverse; they rely on different statistical methods and vary in complexity. To date, there is no convincing evidence of higher efficiency of any algorithm.