Discussion

We describe two linked COVID-19 outbreaks that occurred in a UK county in in July-to-September 2021: the first was associated with a nightclub and the second was associated with five air force bases. Both outbreaks predominantly affected 18–35-year-olds. The outbreaks occurred in the weeks following the final restrictions being lifted after the UK's third national lockdown during the COVID-19 pandemic. The affected population – young adults – represented a group at relatively low risk of developing severe SARS-CoV-2 infection complications [Reference Clift21, Reference Williamson22], and therefore they were also less likely than older adults to have received COVID-19 vaccinations due to older age groups being prioritised for vaccination [23]. It was noted that the nightclub, situated in the county's main town, was particularly busy on the weekend following restrictions being lifted, and many had travelled from outside of the town to attend, which may have increased the risk of high transmission rates. Our analysis of the timing of exposures suggests that the outbreak at the nightclub and those at the bases were likely to have propagated one another to an extent, but that the nightclub outbreak likely led to a greater increase in cases at the bases than vice versa. Base 3 had by far the highest attack rate of the five bases. We cannot be certain of the reasons for this, but of bases with a median case age under 30 years, it was the closest geographically to the nightclub. It might, therefore, have been associated with relatively high levels of nightclub attendance, leading to increased COVID-19 incidence amongst a group who were young and hence relatively unlikely to be fully vaccinated [25].

As of July–August 2021, the B.1.617.2 (‘Delta’) variant was dominant in the UK [Reference Pouwels29]. Although all mandated social restrictions had eased on 19th July, there remained a legal requirement to self-isolate for those who tested positive for COVID-19 for 10 days following onset of symptoms or, if asymptomatic, from the date of the positive test result [30]. Asymptomatic contacts of cases were also legally required to isolate for 10 days from the time of contact with the case [30]. At this time, while most of the general UK population had received two SARS-CoV-2 vaccination doses, over a third of adults under the age of 35 years were yet to receive a single dose [25] (see Fig. 3). Therefore, public health measures that were adopted to control the outbreak focused on health promotion (specifically of vaccination, testing and avoiding close contact and crowds where possible); working with organisations to ensure safe practices; and improving community access to testing and vaccination. Many attendees of the club resided in the county's main town, which was noted to have a relatively low vaccination uptake; this was likely to be largely due to its consisting of high numbers of young adults [31] who were less likely to be in high-priority vaccination groups [23]. Although this population represented a group with few at high risk of severe disease [Reference Clift21, Reference Williamson22], the low vaccination uptake meant that an outbreak was more likely to occur [Reference Chen32].

On 19th July 2021, it became legal to attend nightclubs in the UK [17–19] and our communication strategies needed to consider that the main target population (young adults) were: not acting illegally; not necessarily at high personal risk from serious COVID-19 illness and may have felt they had already made considerable personal sacrifices for others during the previous 16 months of the pandemic [17]. The communication campaign accordingly focused on personal responsibility (the message conveyed stated that ‘COVID-19 is still here’) but rather than dissuade the messages' target audience from going out at all, made the less demanding request to ‘Make space. Avoid crowds. Keep testing’. The target audience may have also been more likely to be motivated to comply with this than a demand to desist socialising altogether [Reference Michie, Atkins and West33, Reference West and Michie34]. Previous evidence has suggested that people may become more motived to perform behaviours that reduce transmission risk when reminded of the responsibility to protect others who are more vulnerable [Reference Lunn35].

The multiagency collaboration within our outbreak response team enabled a coordinated and efficient delivery of multiple interventions. When it became apparent that the two outbreaks were linked and affected similar population demographics, this indicated potential appropriateness of the targeted communications campaign that had been developed by the local authority for use at air force bases. Collaborative surveillance and monitoring of contact tracing data between PHE/UKHSA and the air force was helpful in ensuring validity of incidence and exposure, which was useful for monitoring the ongoing situation. The presence of multiple agencies at OCT meetings facilitated the sharing of intelligence, preventative strategy components and the coordination of actions for respective organisations.

Study limitations include that we do not have complete information on the number of people exposed during the described outbreaks. Furthermore, due to the high levels COVID-19 incidence in the region at the time of data collection, some of the cases presented here as being associated with the air force bases, the nightclub or both may have contracted COVID-19 due to an alternative exposure. We believe, however, that the propagation patterns displayed by the outbreak epi-curves suggest that a considerable proportion of the cases presented resulted from transmission at these settings. On inspection of national contact tracing and cases data, some case variables were missing or contained errors; such errors in the wider dataset may have prevented some relevant cases from coming to our attention. A further limitation is that some of the outbreak interventions may have had beneficial effects that lasted beyond the present outbreaks that are not described here. Also, as our intervention involved the simultaneous implementation of multiple components, we are unable to provide evidence of which of the components was the most effective. It may also be possible that effectiveness was due to there being a particular combination of components, and indeed, ‘packages’ of intervention are often found to be particularly effective forms of COVID-19 prevention [Reference Talic9]. There is some indication, however, to suggest that efforts to promote vaccination uptake amongst young adults may have been particularly effective, as this increased considerably during our investigation, by a greater proportion than the national average, especially for 18–24-year-olds (see Fig. 3).

Previous research has indicated military bases as an example of a congregate setting where there is risk of large COVID-19 outbreaks occurring [Reference Marcus36]. Our findings are consistent with evidence of the potential effectiveness of non-pharmaceutical outbreak–control interventions within [Reference Marcus37] and outside such settings, including social restriction, testing, contact tracing, isolation, vaccination, hygiene and cleaning, PPE [Reference Talic9, Reference Hellewell10, Reference Ingram12] and communicating advice relating to these strategies [Reference Lunn35, Reference Cook and El-Boghdadly38]. Effective communications may involve population-wide health promotion using materials such as posters or online materials [Reference Lunn35], or work-based risk assessments and advice [Reference Cook and El-Boghdadly38]. Strategies such as outreach teams [Reference Denberg39] or patient and public involvement [Reference Brett40] may have improved engagement with local populations and have led to a greater impact of outbreak-related messaging.