The impact of local and national restrictions in response to COVID-19 on social contacts in England: a longitudinal natural experiment

Abstract

Background: England's COVID-19 response transitioned from a national lockdown to localised interventions. In response to rising cases, these were supplemented by national restrictions on contacts (the Rule of Six), then 10 pm closing for bars and restaurants, and encouragement to work from home. These were quickly followed by a 3-tier system applying different restrictions in different localities. As cases continued to rise, a second national lockdown was declared. We used a national survey to quantify the impact of these restrictions on epidemiologically relevant contacts.

Methods: We compared paired measures on setting-specific contacts before and after each restriction started and tested for differences using paired permutation tests on the mean change in contacts and the proportion of individuals decreasing their contacts.

Results: Following the imposition of each measure, individuals tended to report fewer contacts than they had before. However, the magnitude of the changes was relatively small and variable. For instance, although early closure of bars and restaurants appeared to have no measurable effect on contacts, the work from home directive reduced mean daily work contacts by 0.99 (95% confidence interval CI] 0.03-1.94), and the Rule of Six reduced non-work and school contacts by a mean of 0.25 (0.01-0.5) per day. Whilst Tier 3 appeared to also reduce non-work and school contacts, the evidence for an effect of the lesser restrictions (Tiers 1 and 2) was much weaker. There may also have been some evidence of saturation of effects, with those who were in Tier 1 (least restrictive) reducing their contacts markedly when they entered lockdown, which was not reflected in similar changes in those who were already under tighter restrictions (Tiers 2 and 3).

Conclusions: The imposition of various local and national measures in England during the summer and autumn of 2020 has gradually reduced contacts. However, these changes are smaller than the initial lockdown in March. This may partly be because many individuals were already starting from a lower number of contacts.

Keywords: COVID-19; Contact survey; Disease outbreak; England; Lockdowns; Non-pharmaceutical interventions; Pandemic; United Kingdom.

Fig. 1

a Number of restrictions by English Lower Tier Local Authority (LTLA) over time. b Mean contacts for adults from CoMix over time. c Heatmap of number of LTLA areas affected by each restriction over time. Graph A shows the geographical distribution of the restriction by LTLA region in England overtime. The number of restrictions represents the maximum number of restrictions applied at any time during a single month. Graph B shows the mean number of contacts in all settings with a cap of 200 contacts per person, 2000 bootstrap samples were used to generate 95% uncertainty intervals with sampling done at the participant level to reflect clustering of repeated measures. Graph C shows a temporal heatmap of the number of LTLA areas that were affected by each restriction. *Rule of Six and 10 pm rules are included in the first national lockdown as more stringent rules were applied with all restaurants and bars being closed and no meetings allowed with any other households

Fig. 2

a The distribution of the number of setting-specific contacts before and after each restriction came into place. b Change in contacts for each restriction. c The proportion of changes comparing before and after the restrictions started. Graph A shows the distribution of setting-specific contacts before and after each restriction came into place. Colours represent whether the change in contacts increased (brown), decreased (yellow), or stayed the same (orange). Values greater than 10 are not shown on the graph. Graph B shows the distributions of the difference in contacts, with zero values removed to ease comparison between increases and decreases. Differences of magnitude greater than 10 are not shown. Graph C shows the proportions of changes in contacts due to each restriction. For most participants, the number of contacts remained unchanged (yellow) after restrictions. Additional file 1 Figure S1A and S1B show graphs 2A and 2B without the restrictions on the axes and with zero values included

Fig. 3

a Change in contacts after entry into each tier. b Change in contacts after entry into national lockdown from each tier. c Proportion of changes in contacts after entry into each tier. d Proportion of changes in contacts after entry into lockdown from each tier. Graph A and B show the distributions of the difference in contacts. Zero values are not shown to ease comparison between increases and decreases. Differences of magnitude greater than 10 are not shown. Graph C and D show the proportion of change in contacts due to each restriction. For the majority of people, the number of contacts remained unchanged after entry into each tier or exit from each tier to lockdown. Additional file 2 Figure S2A and S2B show graphs 3B and 3B without the restrictions on the axes and with zero values included