Bar tenders in clubs and pubs are routinely flouting UK legislation intended to prevent further sales of alcohol to those who are already drunk, reveals a study of purchase patterns in one UK city, published online in the Journal of Epidemiology and Community Health.

Preventing sales of alcohol to those who are already very drunk would ease the strain on public services and protect long term health, and should be a public health priority, say the authors.

They base their conclusions on the results of a 'sting' in which student actors, aged between 20 and 22, attempted to buy alcohol while pretending to be drunk in 73 randomly selected pubs, clubs and bars in one city in North West England.

All purchase attempts were made between 2100 and 0300 hours, on every day of the week bar Mondays and Tuesdays during May 2013.

One person played the fake drunk, staggering to the bar; the other took the role of the sober friend. The actors took advice from police, who regularly deal with young people who have had too much to drink on the streets of the city, to boost the authenticity of their performance.

The various venues were assessed for 10 well recognised indicators of poorly managed and problematic establishments: few seats; cheap alcohol promotions; young bar tenders; young clientele; crowding; poor lighting; rowdiness; dirtiness; and prevailing levels of drunkenness.

Four out of five attempts were successful (61 out of 73; 84%), and service rates were always high, irrespective of the day or time. Sex of the purchaser made no difference to the response of the bar tender.

Most sales occurred without hesitation, despite the fact that bar tenders often recognised drunkenness by either commenting directly to the actor or to another server, or by rolling their eyes.

And on almost one in five occasions (18%), bar tenders tried to 'upsell' the actor a double rather than a single measure.

The likelihood of being served while drunk rose according to the number of indicators of poor management there were.

All establishments with eight or more of these indicators were prepared to serve compared with two thirds of those with none. Only greater seating provision and older bar staff boosted the odds of service refusal.

The actors were also significantly more likely to be served at venues with door supervisors (where 95% got served). But more than two thirds of venues without door supervisors were still prepared to serve drunk customers.

"Continued provision of alcohol to drunks will increase the risks of acute and long term health and social harms, and consequently, the burdens these place on public services and society," write the authors.

They go on to say that their findings suggest that the law preventing the sale of alcohol to people who are already drunk is routinely being broken, as the figures on prosecutions are low (only three convictions nationally in 2010).

"Although our study focused on one city, a lack of prosecution for sales to drunks throughout England suggests this is typical of nightlife environments nationally," they say.

"With policies to prevent alcohol related harm by increasing alcohol prices failing to be implemented, increased use of legislation preventing sales of alcohol to drunks should be considered a public health priority," they conclude.