Abstract

The dynamics of directly transmitted infectious diseases may be chaotic or have high degrees of unpredictability, making it difficult to assess the long-term evolution. Chaotic dynamics has been observed in several infectious diseases. The COVID-19 pandemic might follow nonlinear and chaotic dynamics, explaining its low predictability. We study the unpredictability of the dynamics generated by COVID-19 2020-22 pandemic in a sample of eleven countries that have generated very good quality data according to WHO reports. We compute the Maximum Lyapunov Exponent (MLE) profile in moving windows of 250 days along the epidemic for the series of the incidence of reported cases per 100,000 inhabitants. The MLE profile for incidence of reported cases reveals that, regardless of local measures to mitigate the virus, this exponent becomes extremely high (above 0,4 Bits/month on average), close to one order of magnitude above the values ??previously reported for other infectious diseases. When evaluating the evolution of the MLEs profiles, after a transient that is country dependent, this index becomes very stable and similar in clusters of countries somehow independently of their own sanitary measures. In periods of strong pharmacological and non-pharmacological interventions, the MLE values were reduced probably due to the restriction of the degrees of freedom of the trajectories, recovering in a very short period the pandemic values when the interventions decreased. This behavior suggests a highly chaotic type of dynamics that requires a fine synchronization of classical epidemiological measures with the processes of vaccination and immunity gain. © Prof. Dr. Jan van der Hoeven stichting voor theoretische biologie 2025.