To our knowledge, this is the first study of BI in the psychiatric setting using clinical psychiatric staff. The major finding is that BI has a positive (although small) 12-month effect on alcohol habits in psychiatric outpatients. The minimal intervention produced similar results as the more intense BI: i.e., assessment, feedback, and a leaflet, with or without brief advice, reduced AUDIT score in risky-drinking patients. The reported major strategy among patients for reduced drinking was cutting down on number of drinks per drinking day. Gender or psychiatric severity as measured by number of psychiatric consultations had no impact on results.
Investigations of BI in a psychiatric outpatient setting are rare. In one such study, by Eberhard et al., the design was similar to ours but had some important differences. The intervention was telephone-based and delivered by the research team. In our study, clinical psychiatric staff performed the interventions within their normal schedule. The intervention in the Eberhard study was evaluated at six months only and not at 12 months, and intervention effects were measured on a group rather than an individual level. These differences may have contributed to their result in which 43.8% of participants in the intervention group lowered their AUDIT scores to nonhazardous levels, as compared with only 21% in our study. It is notable that a large percentage of the patients in the control group in the Eberhard study (27.7%) reduced their AUDIT score to nonhazardous levels also, simply after assessment and after declaring interest to take part in the study.
In one other study conducted targeting inpatients in a psychiatric setting, Hulse et al. compared outcomes between patients receiving a 45-minute BI and those receiving an information package only. As in our study, psychotic patients were excluded. Also similar to our study, the clinical setting demanded reaction to patients with risky drinking, and there was no group that received assessment only. A reduction in weekly consumption was found after six months in both intervention groups, with a median weekly consumption of 36 standard drinks in the BI group and 33 standard drinks in the information package group, respectively. The median weekly consumption at six months was 6.8 standard drinks in the BI group and 10 in the information package group. A greater proportion of those in the BI group improved their drinking category as classified by national Australian criteria. The same researchers later studied five-year outcomes in terms of general hospital and mental health morbidity and mortality. They found that patients who had received a BI or an information package had longer intervals to both first general hospital and mental health inpatient event than matched controls. Importantly, no differences in outcome were identified between the intervention groups.
Patients with psychiatric disorders were included in a recent review of BI for substance use. Except for the Hulse et al. study, only one investigation targeting alcohol use was included in this review, a pilot study investigating the effects of three MI sessions versus a control condition among patients with schizophrenia. A significant reduction in drinking days and an increase in abstinence rate was found in the MI group.
The results of our study are in line with those of Hulse et al. and Eberhard et al., suggesting that BI may be of some value also in the psychiatric setting. Our results imply that the intervention may be very brief and still be effective. Addressing the issue through assessment, feedback, and information leaflet may be sufficient. Intervention effects are relatively small, which indicates that regression to the mean as well as natural recovery processes cannot be ruled out.
In this study we preferred the label “minimal intervention group” rather than control group. Patients in the minimal intervention group underwent assessment, were given feedback on drinking habits, gave consent, and received a leaflet. Intrinsic as they are to intervention studies, such factors are known to reduce drinking significantly over time[30, 31].
In general, the mechanisms that lead the risky-drinking person to change drinking habits after BI are unclear. It is plausible that BI works simply by pointing out a potential problem, which would stimulate risky-drinking persons to contemplate change. Most excessive drinkers who change, even those with more advanced alcohol problems or comorbid psychiatric disorders, change their drinking patterns without treatment[32, 33]. The BI may serve as a trigger mechanism for such processes of natural recovery.
Psychiatric outpatient units are busy; therefore, secondary alcohol assessment and prevention is not commonly given high priority. Consequently, such measures need to be short and easy to use but still efficient. In this study, BI was administered in a clinical psychiatric setting with caregivers of all categories performing both the screening and the intervention. In previous studies conducted in psychiatric settings, interventions were conducted by research staff. Physicians in primary care commonly are considered the most suitable professional group to perform BI, although nurses may assist[34, 35]. Psychiatric care, however, is not as physician-centered as primary care; i.e. psychiatric patients are likely to have the major part of their contact with a nonphysician caregiver. Psychiatric staff may even be more prone to raise the subject of alcohol than other medical staff. Anecdotal reports from staff within the present study inform that BI had a strong impact on specific patients. Still, the most efficient form of BI in the psychiatric setting needs elaboration. Judging from the small effect of very brief interventions, more profound interventions may be more suitable.
Methodological limitations of this study need to be addressed. One limitation is the reliance on self-reported data. To engage all categories of psychiatric staff, we chose to use questionnaires only and not biological markers. However, biomedical markers have disadvantages: their use requires medical staff, and they are not reliable enough on their own. Furthermore, alcohol intake must be substantial to be biologically measurable.
Another limitation of our study is the small sample size. If a larger sample was used, conclusions could be more strongly supported. The dropout rate is another limitation. Women dropouts (n = 28) had higher AUDIT scores at baseline than women whom completed follow-up (mean, 11.8 ± 3.5 points versus 10.1 ± 3.7 points [t = 2.1, p < 0.05]). No other systematic attrition could be identified. With the use of a conservative ITT analysis, we eliminated the risk of overestimating intervention effects. Also, follow-up data were collected by the research team only and not by the patient’s psychiatric caregiver, thus lowering the risk that the patient underestimated his or her drinking to please the caregiver.
The major strength of this study is that it adds knowledge to the field of secondary alcohol prevention in a group of patients who are in urgent need of such strategies. By addressing alcohol habits in a simple, time-efficient manner, psychiatric staff of any profession may initiate small but measurable improvements in risky drinking behavior.