The review assessed the effectiveness of brief interventions to reduce alcohol-related outcomes for injury patients in emergency settings, but specific conclusions could not be drawn due to substantial clinical and methodological heterogeneity. The conclusions should be interpreted with caution, given the potential for bias as a result of the poor quality of reporting in the review.

To assess the effectiveness of brief interventions (BIs) for injury patients in emergency care settings.

MEDLINE, PsycLIT, CINAHL and The Cochrane Library were searched up to January 2007; search terms were reported. The starting dates for the searches were not reported. Studies were restricted to the English language. Handsearches of specialist journals were also performed.

Eligible studies had to be randomised and include only injured patients in emergency care settings receiving BIs (defined according to protocols in individual studies) or in follow-up outpatient care following emergency care. Eligible outcomes included: reduced alcohol intake; risky drinking practices; alcohol-related negative consequences; and injury frequency. Alcohol-related variables were required to be measured at baseline and changes in these measures assessed at later follow up.

In the included studies, injuries to participants included motor vehicle injury, facial injury and other types of injury. Most studies included both male and female patients and most patients were 18 years old or more. Participants were mostly required to be "at risk" or "hazardous" drinkers, but definitions varied. The majority of interventions involved one BI session. BIs were compared with either control groups or other BI groups of varying intensity with the primary aim of achieving reductions in alcohol intake. Interventions were delivered by various health personnel or by computer in different settings and for different durations. Outcomes were assessed using various different measurement tools. The included studies were conducted in the US and Europe.

Abstracts of potentially relevant studies were retrieved and checked against the inclusion criteria. The authors did not stated neither how many reviewers performed the selection nor how disagreement was resolved.

The authors did not state whether they performed a validity assessment.

Data were extracted by the first author and discussed with all the authors to reach agreement.

Studies were combined in a narrative synthesis and in tables.

Fourteen publications (13 studies; n=6,031) published between 1988 and 2007 were included. Follow up ranged from three to 12 months. Attrition resulting in completion rates at 12 months follow up ranged from 58 per cent to 88 per cent. The sample sizes of most of the included studies were greater than 100 participants.

Most studies reported improved outcomes for intervention and control groups. There was a general trend in favour of treatment with BI, with 11 of 12 studies that compared pre- and post-BI results reporting a significant effect of BI on at least some of the outcomes. More intensive interventions showed more favourable results. Five studies did not find significant differences in alcohol intake between treatment groups. Computer-based BI studies reported inconsistent results.

Although favourable trends towards BI were observed, specific conclusions could not be drawn because of variations in study protocols, recruitment criteria for participants, screening and assessment methods, and injury severity.

Inclusion criteria for the review were defined very broadly with respect to participants, interventions and outcomes. This seemed appropriate for the topic area, although a lack of specific criteria for participants, interventions and outcomes together with an inability to stratify presented difficulties when interpreting the narrative synthesis. A range of databases and journals were searched. As no attempts were made to locate unpublished studies and studies published languages other than English, relevant studies may have been missed. There was potential for bias and error in the selection of studies, because the methods were not reported and study validity was not assessed. The heterogeneity in the included studies in terms of participants, interventions and outcomes made it difficult to draw conclusions. The process of screening to recruit participants and the experience of being injured, the severity of injury and the emergency setting were all factors that may have influenced the results.

The authors' statement that specific conclusions could not be reached seemed appropriate, but given the above concerns and the potential for bias in the reporting of the review, the reliability of their conclusions was unclear and should be interpreted with caution.

Practice: The authors did not state any implications for practice.

Research: The authors stated that more research was needed for the interaction between methods of assessment and behaviour change, the nature of changes (how and why) in alcohol related outcomes, the optimal setting in which changes can occur and assessment of changes with computer-based BIs.

The Swedish National Rescue Services Agency, Center for Disease Control (Grants R49/CCR1232280-01 and R49CE000544-01), and the National Institute of Alcohol Abuse and Alcoholism (Grant 1R01AA13709-01A1).

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.