A data-driven test approach to identify COVID-19 surge phases: an alert-warning tool

Abstract

The effective monitoring of the pandemic emergency and, specifically, the early detection of surge phases are crucial to define proper health policies. We propose a statistical testing approach to identify the acceleration in contagion growth that potentially marks the start of new waves, based on the study of the reproduction rate dynamics. The proposed method can be considered as a supplementary early warning system that can assist policymakers in the attempt to anticipate and tailor countermeasures. It can also be used as an ex-post tool to date-stamp surge phases and evaluate the impact of the implemented strategies on their timing. The effectiveness of our approach is exemplified on ten countries' contagion data, reaching robust and insightful results in assessing the timing and severity of COVID-19 surge phases.

Keywords:

• COVID-19 pandemics
• reproduction rate
• explosivity tests
• epidemic surge detection

2020 Mathematics Subject Classifications:

• 62P10
• 62M10

Acknowledgments

A.A. and P.C. developed the methodology, A.A. conducted the empirical analysis, A.A. and P.C. analyzed the results. All authors reviewed the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 We verified that our results are stable to small changes in the $r_0$ parameter.

2 https://github.com/owid/covid-19-data/tree/master/public/data

3 Due to the technical characteristics of the test, the first results can be calculated starting from 29 July 2020, as the previous days are used to populate the first expanding window in order to have statistically significant results. Nevertheless, starting from 1 June 2020 avoids the inclusion of the very first wave of the pandemic, which does not offer a previous time series useful for fitting the model as no pandemic was of course in place before January 2020.)

Additional information

Funding

The authors acknowledge the European HORIZON 2020 PERISCOPE Project (Contract Number 101016233), funded by European Union – Next Generation EU.