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 emergence of disease outbreaks, offering hope that policy makers can make predictive rather than reactive management decisions. Here, using a novel, sequential analysis in combination with daily COVID-19 case data across 24 countries, we suggest that composite EWSs consisting of variance, autocorrelation, and skewness can predict non-linear case increases, but that the predictive ability of these tools varies between waves based upon the degree of critical slowing down present. 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 urgent intervention decisions but best characterise hypothesised critical transitions. Dataset The deposited dataset contains scripts used in the early warning signal and generalised additive model analysis, the generation of figures, and the custom R functions underpinning the work. Raw COVID-19 case data is also provided if users prefer to access files directly rather than sourcing from the host repositories (all credit is provided to the original publishers).

All scripts are written in the R language (R Core Team. 2021. R: A language and environment for statistical computing).