Pub-goers may be eager for their favourite haunt to reopen amid the UK's coronavirus restrictions, but research suggests getting together for a drink may "significantly" increase the infection's transmission.
The chairman of JD Wetherspoon has called for pubs to reopen alongside non-essential shops, warning that the industry is "on its knees".
Tim Martin has argued pubs and restaurants are "COVID-secure environments", their owners having invested in safety measures like plastic screens at tills and floor markings to guide one-way systems.
Boris Johnson has said he is "optimistic" he will be able to set out plans on 22 February for a "cautious" easing of England's lockdown, starting with schools, then non-essential shops and finally the hospitality sector.
Watch: Pubs boss makes case for safe reopening
To better understand the risk pubs pose amid the pandemic, scientists from the University of Stirling pretended to be customers in 29 licensed premises between July and August 2020, when pubs had reopened following the first lockdown.
Results suggest a "wide range of incidents with potential to increase transmission risk" were observed in all but three of the venues, with most having "multiple incidents", including singing, milling around the bar and embracing others.
Read more: Coronavirus may 'impair' brain function
The scientists also observed customers and staff were failing to maintain social distancing, often due to "alcohol intoxication", which the workers "rarely effectively stopped".
They concluded "despite the efforts of bar operators and guidance from government, potentially significant risks of COVID-19 [the disease caused by the coronavirus] transmission persisted in at least a substantial minority of observed bars, especially when customers were intoxicated".
The study – funded by the Scottish government chief scientist office – is the first in the world to analyse the effectiveness of coronavirus restrictions in licensed premises, with the results said to be informing governments, public health experts and policymakers in the UK and abroad.
"Our study explored and observed business practices and behaviours of customers and staff in licensed premises in summer 2020 with a view to understanding if and how COVID-19 transmission risks could be managed in settings where alcohol is served," said lead author Professor Niamh Fitzgerald.
"Businesses expressed an intention to work within the guidance, but there were commercial and practical challenges to making this a reality.
Read more: Blood-thinning drugs cut COVID death risk
"Upon re-opening, substantial efforts to change the layout of bars were observed and appeared to be working well in many premises, but problems were common including staff not wearing personal protective equipment, or with the management of toilets, queues and other 'pinch points'.
"We also observed several incidents of greater concern – including customers shouting, embracing or repeatedly interacting closely with several households and staff – which were rarely addressed by staff.
"We concluded that, despite the efforts of bar operators and guidance from government, potentially significant risks of COVID-19 transmission persisted in at least a substantial minority of observed bars, especially when customers were intoxicated."
Watch: Can you catch coronavirus twice?
The UK entered its first national lockdown in March 2020, with restrictions being in place to some extent ever since.
Pubs largely reopened their doors that July, with strict measures being introduced to help limit the coronavirus' transmission, including table service and enforced face coverings among staff.
To better understand the effectiveness of these approaches, the Stirling scientists posed as customers for up to two hours in rural and urban pubs in Scotland at various times of the day.
The results – published in the Journal of Studies on Alcohol & Drugs – reveal that while hand-sanitising stations had been set up, these were rarely used, with just two premises routinely administering the gel upon customer entry.
Nine of the businesses did not require customers provide details for contact tracing, with one failing to do so even after this was made mandatory by Scottish officials in August.
Most, but not all, of the venues' staff wore personal protective equipment (PPE), like face coverings.
Some wore coverings "inappropriately", perhaps by not covering their nose, or removed their PPE when talking.
The majority of the venues spaced their tables 1m (3.2ft) apart or had partitions between booths, however, some flouted this guidance.
One-way measures were sometimes ignored by customers, with "pinch points problematic in nearly all venues", including bottlenecks in corridors, entrances and at bar counters. This led to "people congregating, often unchallenged".
Less than half of the venues offered table service only, with one premise having a "continuous queue" between the 1m-distanced tables during the scientists' visit.
A similar number of premises also failed to limit the number of customers entering the toilets at one time, even via a basic measure like a sign on the door.
Most also had no restrictions in place around social distancing at sinks or between cubicles, leading to "overcrowding and poor physical distancing within toilet areas in some premises".
Watch: What is long COVID?
The results reveal that "incidents of greater concern" were observed in 11 of the venues. These included combinations of shouting, groups mixing or customers singing.
In August, scientists from the University of Bristol reported that speaking or singing at a loud volume was linked to a higher risk of coronavirus transmission.
Some customers and staff were also moving around the bar area without observing social distancing.
Others posed for photographs in groups and "shook hands or embraced others who did not appear to be in their household".
Loud singing or shouting was observed in all but one of the venues visited in August, "with just one example of effective staff intervention to suppress customer noise".
"In the majority of premises, no staff intervention in incidents or attempts to enforce restrictions were observed," wrote the scientists.
Some intervened in a "light-hearted way", like "gently or playfully reprimanding customers", but this was "largely ineffective".
Enforcement by external agencies such as environmental health or police officers was not observed in any of the venues.
"Overall, our findings suggest grounds for uncertainty about the extent to which new rules can be consistently and effectively implemented in a sector where interaction between tables, households and strangers is the norm, and alcohol is routinely consumed," said Professor Fitzgerald.
"Despite the efforts of licensed premises, and detailed guidance from government, potentially significant risks of COVID-19 transmission persisted in a substantial minority of observed bars, especially when customers were intoxicated.
"Blanket closures, curfews or alcohol sales bans are more likely to be deemed necessary to control virus spread, if such risks cannot be acceptably, quickly and cost-effectively reduced through support and/or sanctions for premises operators.
"Such blanket actions may also have benefits in terms of protecting staff from occupational exposure and reducing pressure on emergency services from alcohol-related injuries or disorder.
"However, attention also needs to be paid to the impact of closures on businesses, economic activity, employee hardship, and ownership patterns in the sector, as well as any risks posed by diversion of some drinking to the home."