Send email for updates
Eberhard S, Nordström G., Höglund P. et al.
Social Psychiatry and Psychiatric Epidemiology: 2009, 44(12), p. 1013–1021.
Unable to obtain a copy by clicking title? Try asking the author for a reprint by adapting this prepared e-mail or by writing to Dr Eberhard at firstname.lastname@example.org. You could also try this alternative source.
Set in Sweden, the first study among psychiatric outpatients to test brief alcohol interventions against screening alone found worthwhile extra drinking reductions after brief motivational advice. Use of a telephone-based intervention was another innovation.
Summary Excessive or unhealthy drinking is unusually common among people being treated for mental health problems, associated with a poorer mental health prognosis and a diminished quality of life. This Swedish study investigated whether hazardous drinking among non-psychotic psychiatric outpatients could be reduced to non-hazardous levels by a brief telephone intervention.
Over a 10-week period in 2004, non-psychotic patients The only other exclusions were patients not proficient in Swedish, too ill to participate, or already known to have a substance use disorder. visiting adult psychiatric outpatient units in Lund and Uppsala were invited to be screened for alcohol and drug problems. Patients were given two self-completion screening forms: the AUDIT (Alcohol Use Disorders Identification Test) questionnaire for alcohol and another for drugs, though only hazardous drinkers and drink-related outcomes were investigated There were too few problem drug users to make a viable study. by the study. All but 3% of the patients seen at the clinics were given the forms. Of these, 74% (amounting to 1746 patients) filled them in and agreed to participate in the study. Of these, 344 met the study's AUDIT score criteria Patients scoring under 6 for women or 8 for men were considered non-hazardous drinkers; those scoring 19 or over were considered probably dependent. for hazardous but probably non-dependent drinking and were randomly allocated to a brief intervention or to no intervention Within the six-month follow-up period. After this they were offered intervention if still drinking to hazardous levels. other than the screening.
After screening and randomisation, patients in the intervention group were phoned as soon as possible to be given brief advice featuring feedback from their AUDIT screening results. The roughly 15 minutes of advice was given by experienced Experienced in both mental health and substance use treatment. nurses trained to use their discretion to implement the manualised protocol based on motivational interviewing principles, tailored according to each patient's readiness Assessed according to Prochaska and DiClemente's cycle of change model. to curb their drinking.
Six months later an attempt was made to phone and screen the patients again. AUDIT scores were obtained from 291 patients (85% of the baseline sample) and compared against their initial scores. At both time points the questions related to drink problems over the past six months. When re-contacted patients missed a few questions, their scores on these were allocated the average score; scores for missing patients were assumed to have changed similarly to known patients. On these assumptions, 44% of patients who had been allocated to brief advice no longer met criteria for hazardous drinking but just 28% of those simply screened, a statistically significant difference. The gap was similar for women (42% v. 23%) and men (49% v. 34%), but probably because there were fewer, the difference for men was not statistically significant. This primary outcome measure was complemented by analyses showing that typical and average AUDIT scores had also declined more sharply after brief advice.
The authors argued that a near-halving in the proportion of patients risking their physical and mental health through risky drinking meant that the intervention should be considered for routine psychiatric outpatient care. In the Swedish context, the study had confirmed that hazardous drinking is a common concern among such patients and that (contrary to expectations) most will accept and complete screening procedures. Screening itself seemed to have led to many to reduce their drinking, but significantly more did so when brief advice was also offered.
commentary This seems the first study among psychiatric outpatients (in this study, non-psychotic patients mainly suffering from mood or anxiety disorders) to test brief alcohol interventions against screening alone. Use of a telephone-based intervention was also an innovation. The researchers found extra drinking reductions attributable to the 15 minutes on the phone which seem worthwhile gains in curbing excessive drinking. The results are important in two ways: first in demonstrating that brief alcohol interventions can work with non-psychotic mentally ill populations as well as the general run of patients; second, in pointing towards a low-cost way of perhaps also reducing the extent to which drinking impedes progress in mental health care. These messages are as much of interest to GPs as to psychiatric units; primary care is where many such patients will be treated because neither their mental illness nor their drinking is sufficiently severe to attract specialist care. The role of brief alcohol interventions in primary care was recognised in guidance for England on the treatment of mentally ill problem substance users.
An unusually all-embracing patient inclusion policy and a high follow-up rate instil confidence that the results would apply to all similar patients where the trial was conducted. There seems no a priori reason why the results should not also apply elsewhere in Sweden and to other countries with similar health systems and drinking cultures. Telephone delivery is, as the authors commented, convenient for all concerned. It means expert therapists can be deployed without the expense of them having to wait at the clinic for potential brief intervention candidates, and takes the load off the clinic's psychiatric staff. The alternative of these staff delivering the intervention themselves is likely to run in to resistance typical of attempts elsewhere, during which medical staff have complained that they did not have the time or skills. Commonly they also feared that addressing drinking without this having been asked for, and when there were no obvious reasons to do so, might undermine their relationship with the patient.
As the authors acknowledged, the main methodological concern is that the nurses who offered the brief advice also conducted follow-up interviews. Patients advised by this person to cut down and/or who gave a commitment to do so may have felt more obliged than screening-only patients to create a good impression. Also the nurses could have been aware which patients had been advised and which not, introducing the possibility of perhaps unintentional bias.
Though this seems the first study with this precise combination of interventions and patients, similar studies have been done. Rather than psychiatric unit patients, some have concerned general medical patients with mental health problems, others patients recruited in non-medical settings, and others patients not just drinking at hazardous levels, but dependent on alcohol. As in the featured study, generally patients were non-psychotic and not severely mentally ill. Generally too, offering them a brief motivational intervention focused on drinking helped them cut back more than usual medical or mental health care. Reviewers for the Cochrane collaboration came to a similar conclusion. One study did also suggest that at the severer end both of depression and of drinking, yet more patients would make substantial drinking cutbacks if brief interventions were followed by more extended therapy. Details of these studies can be found in the background notes.
Identifying heavy drinking among mental health caseloads is of course a crucial first step in addressing it. A set of quick but reliable and valid screening questions is the ideal tool for identifying risky drinking not (or not yet) apparent during routine medical care. Reviewers have usefully assessed the evidence for different screening methods among mental health caseloads. Also a recent US study set in an outpatient psychiatry clinic compared two popular alcohol screening surveys against a routine clinical intake interview and found both identified many more patients, with T-ACE preferable to AUDIT. Finally, a British study tested a community implementation of the Dartmouth Assessment of Lifestyle Instrument (DALI), the only screening tool specifically intended to identify substance misuse among people with severe mental health problems. Only a few of the questions in the tool usefully predicted drinking or drug problems. These were incorporated in a much briefer set of questions called the Simple Substance Use Screening Scale.
Last revised 27 July 2010
Background notes to this entry
Give us your feedback on the site (one-minute survey)
Open Effectiveness Bank home page
Add your name to the mailing list to be alerted to new studies and other site updates