Tracking the national and regional COVID-19 epidemic status in the UK using directed Principal Component Analysis

Swallow, Ben; Xiang, Wen; Panovska-Griffiths, Jasmina

One of the difficulties in monitoring an ongoing pandemic is deciding on the metric that best describes its status when multiple highly inter-correlated measurements are available. Having a single measure, such as whether the effective reproduction number R, has been useful in tracking whether the epidemic is on the incline or the decline and for imposing policy interventions to curb the increase. We propose an additional metric for tracking the UK epidemic across all four nations, that can capture the different spatial scales. This paper illustrates how to derive the principal scores from a weighted Principal Component Analysis using publicly available data. We show the detectable impact of interventions on the state of the epidemic and suggest that there is a single dominant trend observable through the principal score, but this is different across nations and waves. For example, the epidemic status can be tracked by cases in Scotland at a countrywide scale, whereas across waves and disjoint nations, hospitalisations are the dominant contributor to principal scores. Thus, our results suggest that hospitalisations may be an additional useful metric for ongoing tracking of the epidemic status across the UK nations alongside R and growth rate.