Early warning signals predict emergence of COVID-19 waves

Abstract

Early warning signals (EWSs) aim to predict changes in complex systems from phenomenological signals in time series data. These signals have recently been shown to precede the initial emergence of disease outbreaks, offering hope that policy makers can make predictive rather than reactive management decisions. Here, using daily COVID-19 case data in combination with a novel, sequential analysis, we show that composite EWSs consisting of variance, autocorrelation, and return rate not only pre-empt the initial emergence of COVID-19 in the UK by 14 to 29 days, but also the following wave six months later. We also predict there is a high likelihood of a third wave as of the data available on 9th June 2021. Our work suggests that in highly monitored disease time series such as COVID-19, EWSs offer the opportunity for policy makers to improve the accuracy of time critical decisions based solely upon surveillance data.

Overlapping Time Scales Obscure Early Warning Signals of the Second COVID-19 Wave

“…Early warning indicators based on critical slowing down have been studied extensively and led to a promising research line that aims to utilize them as a tool to forecast the (re)emergence as well as the elimination of infectious diseases (e.g., Brett et al, 2017;Brett et al, 2018;Dessavre et al, 2019;Dibble et al, 2016;Drake et al, 2019;Harris et al, 2020;Miller et al, 2017;O'Dea et al, 2018;O'Regan & Burton, 2018;O'Regan & Drake, 2013;O'Regan et al, 2020;Southall et al, 2020). Accordingly, some researchers have suggested that they could have been used to anticipate the second wave of COVID-19 (Liu et al, 2021;O'Brien & Clements, 2021;Proverbio et al, 2021).…”

Early warning signals of infectious disease transitions: a review

“…This general dynamic threshold is a key advantage of this method when compared with the constant threshold used in the logistic composite measure. However, although the 2σ method has been demonstrated to detect the onset of environmental deterioration [13], population collapse [54], recovery of an ecosystem [55] and the emergence of COVID-19 [43], it has not been formally validated with simulation studies of infectious data, which remains an open topic.…”

“…Emerging diseases can surge rapidly on a fast time scale, for instance the emergence of the SARS-COV-2 pathogen which causes the disease COVID-19. Although COVID-19 has been the focus of recent studies on EWSs [40][41][42][43], further work is required to understand the performance of EWSs subject to factors affecting the speed of a disease, sampling frequency and time scale of data. Indeed, the speed of disease has been explored in simulation studies, finding that the rising autocorrelation trend typically reaches 1 at the transition unless the system has a fast speed of emergence [28].…”

An experimental COVID‐19 messaging study in a representative sample of the Scottish population: Increasing physical distancing intentions through self‐efficacy

“…However, it is possible that over the course of the pandemic physical distancing became more difficult as feelings of loneliness and social isolation increase (Williams et al, 2021 ), and as the proportion of the people who are vaccinated increases, the perceived necessity of physical distancing may decline. That said, modelling studies (Iacobucci, 2021 ; O'Brien & Clements, 2021 ) accurately predicted multiple waves of the pandemic and the consequent increases in hospitalizations and deaths, which, in late 2020, required the UK governments to reimpose physical distancing restrictions that had previously been relaxed. Subsequently, restrictions on physical distancing were again relaxed and the requirement to wear face coverings were removed in England during the summer of 2021.…”