Interventions Aimed at Controlling Antisocial Behaviour Associated with On-Sale Alcohol

Interventions Aimed at Controlling Antisocial Behaviour Associated with On-Sale Alcohol Establishments

Prepared by Niamh Fingleton and Catriona Matheson, CAPC, University of Aberdeen for CERGA, December 2011

Key findings

·  Three systematic reviews and two intervention studies were identified. Only three studies were based in the UK, all of which came from one review; therefore the generalisability of the findings are limited.

·  The evidence base consisted of numerous interventions and outcomes measures. Results varied greatly between studies.

·  Studies were generally of inadequate methodological quality.

·  There is some evidence to support the use of multi-component interventions.

·  Whilst there is strong evidence supporting the effectiveness of dram shop liability at reducing alcohol-related crash fatalities in the US, this finding may not be transferrable to the UK due to cultural differences.

·  Overall, the evidence base is weak and further UK-based research of high methodological quality is required.

1.  Background

Alcohol plays a major role in antisocial behaviour in the UK. Alcohol-related disorder not only affects the victims, but also places a major burden on health services and police resources, and is estimated to cost England £7.3 billion annually (Strategy Unit 2004). In England and Wales, over 20% of all violence occurs in or around bars and nightclubs (Nicholas et al. 2007). Data from Cardiff revealed that 61% of alcohol-related incidents involved physical violence, with the remainder involving disorder. Over half of these incidents occurred in, or adjacent to, on-sale establishments (Maguire & Nettleton 2003). A Scottish audit of Accident and Emergency admissions over a six-week period found that alcohol was thought to be a contributory factor in over 70% of assault cases (Scottish Emergency Department Alcohol Audit Group 2006).

This rapid review sought to identify and evaluate the literature on interventions aimed at controlling antisocial behaviour associated with on-sale establishments. On-sale establishments are those in which customers must consume their alcohol at the point of sale e.g. bars and nightclubs.

2.  Methods

An electronic database search was carried out using MEDLINE and SCOPUS. The search was conducted using the concepts of alcohol, intoxication, antisocial behaviour, and intervention studies/program evaluations. The SCOPUS search also included the concept of on-sale establishments due to the high volume results initially retrieved. The final search retrieved a total of 352 results, 302 of which were unique.

Due to the presence of two previously identified relevant systematic reviews published in 2011 (Jones et al. and Brennan et al.), the decision was made to exclude retrieved studies prior to 2009, as the two reviews included studies published until 2008 and 2009 (Jones et al. 2011 and Brennan et al. 2011, respectively). This left a total of 93 studies to be screened for relevance.

Studies were included if they:

a) aimed to reduce antisocial behaviour or intoxication, and

b) contained interventions which occurred in or around on-sale establishments.

3.  Results

Three systematic reviews and two intervention studies were deemed to be relevant. The two studies came from the USA and Finland. The evidence is discussed below, according to the type of study.

3.1. Systematic reviews

Brennan et al. 2011

Brennan et al. (2011) evaluated 14 studies aimed at reducing disorder or intoxication in or around licensed premises. Interventions included responsible beverage service training, server violence prevention training, enhanced enforcement of licensing regulation, multi-level interventions, licensee accords, and a risk-focused consultation. The review found that the effectiveness of the interventions varied, even when similar interventions were compared across studies. They concluded that whilst there is a lack of evidence to support the use of server training courses to reduce intoxication, there is some evidence to support its use for reducing disorder. However, the supporting evidence is weak and limited and therefore a robust conclusion cannot be made.

Jones et al. 2011

Jones et al. (2011) systematically reviewed 39 studies assessing interventions implemented in drinking environments which aimed to reduce the use of alcohol and its associated harms. The authors concluded that there was some evidence to support the use of multi-component programmes which combined community immobilisation, responsible beverage service training, house policies and stricter enforcement of licensing laws. However, the effectiveness of other interventions, i.e. server training, patron targeted interventions, police campaigns and other approaches regarding the enforcement of alcohol sales, was limited. Interventions were found to vary greatly, and were generally of inadequate methodological quality.

Rammohan et al. 2011

Rommohan et al. (2011) identified eleven studies assessing the effectiveness of dram shop liability with outcomes pertaining to various alcohol-related harms. Dram shop liability holds the owner or server(s) of an on-sale establishment (e.g. pubs and nightclubs), where a patron consumed their last alcoholic beverage, responsible for the harms subsequently inflicted by the patron on others. There was strong evidence to support that dram shop liability is effective at reducing alcohol-related crash fatalities, but insufficient evidence regarding its effect on excessive alcohol consumption.

Over-service laws prohibit the sale of alcoholic beverages to intoxicated patrons. Only two studies evaluated the effectiveness of enhanced enforcement of over-service laws. Both assessed its effects on service to pseudo-intoxicated pseudo-patrons and driving under the influence (DUI) arrestees, and found conflicting results.

3.2. Intervention studies

Warpenius et al. 2010

Title: Effects of a community intervention to reduce the serving of alcohol to intoxicated patrons.

Study design: Controlled before and after study design.

Intervention: Community-based programme combining law-enforcement, responsible beverage service training, information campaigns and policy initiatives, in one Finnish town.

Outcomes measure: Frequency of serving alcohol to male actor pretending to be intoxicated.

Findings: There was a statistically significant increase in the number of refusals to serve alcohol to the actor in the intervention area compared to the control area.

Kazbour & Bailey 2010

Title: An analysis of a contingency program on designated drivers at a college bar.

Study design: ABCA experimental design (explained below).

Intervention: Stage A was baseline condition consisting of free soft-drinks for designated drivers, as was current policy at the bar. Stage B consisted of in-bar prompts advertising free pizza to groups with a designated driver. Stage C consisted of advertising free pizza and $5 dollar gas cards with advertisement posters in the community, a newspaper story and a radio interview. Following stage C, the stage A condition was restored.

Outcomes measure: Percentage of patrons functioning as, or travelling with, a designated driver.

Findings: The number of patrons functioning as, or travelling with, a designated driver whose blood alcohol concentration was zero, increased at each subsequent stage, and decreased upon return to the baseline condition.

Additional notes: No significance testing was reported and there are concerns regarding the validity of the study.

3.3. Overview of evidence from the UK

Only three UK-based studies were identified, all of which were contained in the systematic review by Jones et al. (2011). These are particularly highlighted here due to their increased relevance and not necessarily their methodological quality.

Warburton & Shepherd 2000

Title: Effectiveness of toughened glassware in terms of reducing injury in bars: a randomised controlled trial.

Study design: Randomised controlled trial.

Intervention: Replacement of pint glasses with annealed (control) or toughened (intervention) glassware in 57 on-sale establishments in England and Wales.

Outcome measure: Bar staff injuries recorded monthly: number, site and severity of injuries.

Findings: The rate of injury was significantly higher for the intervention group when adjusted for hours worked, with no significant difference in severity. Injuries in the intervention group tended to occur simultaneously in more than one body part due to spontaneous disintegration of the glassware, thereby leading to wider dispersion of the fragments.

Additional notes: Jones et al. (2011) rated the quality of the trial as ‘weak’ as the data collection methods were not reported as valid or reliable, and due to the high rate of withdrawals and drop-outs. Furthermore, the intervention was aimed at reducing the injuries caused by both antisocial behaviour and accidents, as opposed to actually reducing antisocial behaviour.

The Tackling Alcohol-related Street Crime (TASC) project

The TASC project was reported by both Maguire & Nettleton (2003) and Warburton & Shepherd (2006) which are described in more detail below.

Maguire & Nettleton 2003

Title: Reducing alcohol-related violence and disorder: an evaluation of the 'TASC' project.

Study design: Uncontrolled before and after study design.

Intervention: Police-led multi-agency scheme aimed at reducing alcohol-related incidents and disorder in Cardiff and Cardiff Bay. It involved a range of interventions including focused dialogue between the police and the licensed trade, measures aimed at improving the quality and behaviour of door staff, publicising the problem of alcohol-related violent crime, targeted policing operations at crime and disorder ‘hot spots’, a cognitive behavioural programme for repeat offenders, a training programme for bar staff, a programme of education regarding alcohol for school age children, and support for victims of alcohol-related assaults attending hospital.

Outcome measures: Alcohol-related violence and disorder in the 12 months prior to the project, to the 12 months after the project launch.

Findings: The majority of assaults occurred in establishments, whereas disorder occurred more frequently on the street. Operations targeting individual establishments resulted in significant reductions in assaults and disorder occurring in or adjacent to these establishments, whereas operations targeting whole streets were less successful. There was a 4% reduction in violence compared to the previous 12 months, and a 49% increase in disorder, which was accounted for by one street which had the densest concentration of on-sale establishments and several newly opened establishments.

Additional notes: The quality of this study was classed as being weak by Jones et al. (2011), as few details were reported regarding the study methodology, and whilst data were compared with trends elsewhere in South Wales, the authors were unable to make direct comparisons due to differences in data collection.

Warburton & Shepherd 2006

Title: Tackling alcohol related violence in city centres: effect of emergency medicine and police intervention.

Study design: Longitudinal controlled intervention delivered in three stages.

Intervention: As described above in the above study by Maguire & Nettleton 2003. Additional emergency department intervention involving two high risk on-sale establishments, whereby the managers of these establishments were visited by two consultants who presented in graphic detail the injuries sustained, treatment, and numbers of assaults there. The managers were also informed that they were auditing violence on the premises, and that the results would be made available to local media.

Outcome measures: Rates of emergency department treatment due to assault, inside on-sale establishments and on the street, over three nine-month periods: pre-intervention, and intervention periods one and two.

Findings: The overall number of assaults in both on-sale establishments and on the street area increased by 24% in the intervention area, compared to a 6% increase in the control area. Relative to total capacity, there was no change in the number of assaults occurring in on-sales establishments in the intervention area, but a slight reduction in the control area. In the intervention area, assaults in the street increased by 34%, whereas in the control area there was an increase of only 8%. One street accounted for a disproportionate increase when compared to other streets, which was thought to be due to the increased number and capacity of establishments on the street.

Data from these two high risk premises which underwent the additional intervention delivered by consultants indicated that the combination of this intervention with the police led intervention was associated with a significantly greater reduction in assaults in these premises compared to those not in receipt of the additional emergency department intervention.

Additional notes: Jones et al. (2011) regarded the overall quality of the study as weak, and considered the reporting of evaluation methods to be unclear.

4.  Considerations

One important factor to consider is that a minority of evidence came from the UK. Whilst all the evidence came from developed countries, i.e. the USA, Australia, Sweden, Canada, Finland and the UK, only three studies, all of which were contained in the review by Jones et al. (2011), were UK-based. Therefore, interventions found to be effective elsewhere are likely to require adaptation before implementation in the UK due to behavioural, cultural and environmental differences (Jones et al. 2011). For example, the strongest evidence base was for dram shop liability, which has been found to be largely effective in the US at reducing alcohol-related crash fatalities (Rammohan et al. 2011). However, this may not be transferrable to the UK due to cultural differences regarding legal action. Furthermore, litigation is likely to be a costly method of increasing bar-server responsibility.

Many of the studies measured processes or surrogate outcomes, rather than the actual outcome of interest to the current review. For example, refusals to serve pseudo-intoxicated patrons, and alcohol use and intoxication were used regularly. Whether these have a significant impact on the levels of antisocial behaviour is unclear, therefore more appropriate outcome measures such as police reports, hospital admissions or number of incidents witnessed by door staff should also be considered, though these are not without problems themselves.

5.  Conclusion

There is a great deal of literature regarding interventions aimed at reducing alcohol consumption, and its related harms, as identified by the three systematic reviews. However, the interventions, their outcome measures, and their findings varied considerably between studies which made comparison between studies difficult. Furthermore, studies were generally of poor methodological quality, thereby raising concerns regarding their validity.

The strongest evidence base was for dram shop liability, which may not be transferrable to the UK. A small evidence base supporting the effectiveness of multi-component programs was found by Jones et al. (2011). This finding was also demonstrated by Warpenius et al. (2010) who found an increase in the number of refusals to serve alcohol by a pseudo-drunk actor.

The limited evidence base highlights the need for UK-based studies of high methodological quality in order to establish the effectiveness of interventions aimed at reducing antisocial behaviour associates with on-sale establishments.

References

Brennan, I., Moore, S.C., Byrne, E. & Murphy, S. 2011, "Interventions for disorder and severe intoxication in and around licensed premises, 1989-2009", Addiction, vol. 106, no. 4, pp. 706-713.