Send email for updates
Bischof G., Grothues J.M., Reinhardt S. et al.
Drug and Alcohol Dependence: 2008, 93(3), p. 244–251
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 Bischof at firstname.lastname@example.org. You could also try this alternative source.
This German study saved valuable counselling time by only offering further advice to primary care patients who had not yet responded to brief computerised feedback on their risky drinking.
Summary Brief interventions for problem drinking in medical settings are effective but rarely conducted, mainly due to insufficient time. A stepped care approach (starting with a very brief intervention and intensifying efforts in case of no success) could save resources and improve effectiveness. However, research is lacking. The present study compared a full care brief intervention for patients with at-risk drinking, alcohol abuse or dependence, against a stepped care approach in a randomised controlled trial.
Participants were proactively recruited from general practices in two northern German cities. In total, 10,803 screenings were conducted (refusal rate: 5%). Alcohol use disorders according to DSM-IV The Diagnostic and Statistical Manual of Mental Disorders (the 'IV' signifies the fourth edition) is published by the American Psychiatric Association and provides diagnostic criteria for mental disorders. It is used particularly in the USA. were assessed with the Munich-Composite International Diagnostic Interview (M-CIDI).
The 408 eligible participants who joined the study were randomly assigned to one of three conditions:
Stepped care – a computerised intervention plus up to three 40-minute telephone-based interventions after one, three and six months depending on the success of the previous intervention and the patient’s confidence that they could sustain this;
Full care – a computerised intervention plus at the same time a counselling session and then another three 30-minute telephone-based interventions, equalling the maximum counselling time of the stepped care intervention;
Untreated control group.
To evaluate these interventions, time spent in counselling in the intervention conditions and quantity/frequency of drinking were assessed 12 months after baseline assessments.
At follow-up the two intervention groups combined were not drinking significantly less than the untreated control group patients, though there was a non-significant trend in this direction among the 42% of patients who at the start of the trial had been assessed as at-risk drinkers or alcohol abusers rather than dependent or drinking relatively rarely but heavily when they did. Among this same set of patients – the typical targets for brief interventions – significantly fewer offered intervention (25% v. 41%) met the study’s criterion for ‘binge’ drinking at the end of the trial.
However, especially among abuse/at-risk drinkers, the intervention groups had been drinking more at the start of the trial. This meant that though they ended up drinking about the same, the reduction in average consumption was significantly greater among patients offered an intervention. When the sample was divided by severity level, this reduction remained significant only among the abuse/at-risk drinkers. After intervention they were drinking 18g less alcohol per day (over two UK units) compared to just 4g (half a UK unit) among the controls.
Next the analysts compared to the two types of interventions – stepping up if needed versus offering the full programme to all patients. Overall, stepped care patients received roughly half the amount of intervention in minutes compared to full-care patients, yet at 12–13g per day the average reduction in their drinking since baseline assessments was almost identical. This was also the case among abuse/at-risk drinkers, but on average once-in-a-while ‘bingers’ slightly reduced their average consumption after stepped care but slightly increased it after full care, resulting in a statistically significant difference in favour of stepped care.
The authors concluded that a stepped care approach can be expected to improve the cost effectiveness of brief interventions for patients with at-risk drinking. A substantial proportion of patients offered stepped care responded well to computerised assessment feedback only, indicating that very brief, economic interventions may be sufficient for a subgroup of at-risk drinkers. Findings also showed that the intervention programmes were effective mainly for non-dependent at-risk drinkers or alcohol abusers.
Only half of the patients who screened positive for risky drinking agreed to join the study, so its results might not generalise to all risky drinkers at the primary care practices in the study. Also, screening and interventions were conducted by study staff and exceeded the duration of typical brief interventions offered in primary care.
commentary Generally the duration of a single-session brief intervention makes little or no difference to drinking outcomes. However, several But not all. See for example: ● WHO Brief Intervention Study Group. "A cross-national trial of brief interventions with heavy drinkers." American Journal of Public Health: 1996, 86(7), p. 948–955. http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1380435 ● Burge S.K. et al. "An evaluation of two primary care interventions for alcohol abuse among Mexican-American patients." Addiction: 1997, 92(12), p. 1705–1716. http://dx.doi.org/10.1111/j.1360-0443.1997.tb02891.x previous studies have indicated that follow-on advice sessions do improve outcomes. For the first time in a randomised trial, the issue addressed by the featured study is whether it is best to offer these sessions as a set programme, or whether time and money can be saved by offering them only to patients who have not yet responded to the initial interventions(s).
In the context of the study, the verdict was clear: the 'as needed' strategy cut counselling time in half, saving Assuming the doctor was doing the counselling. 22 Euros per patient, yet with minor exceptions Puzzlingly, these favoured the 'as needed' strategy, the briefer of the two interventions. This difference was seen only among the least heavy drinkers (the episodic 'bingers'), raising the possibility that making these patients go through four advice sessions in addition to the brief intervention was counterproductively disproportionate to their need. the outcomes were not significantly different. It was, however, important to at least monitor progress and offer further help if needed: about 30% of patients seemed to have securely resolved their drinking problems after brief intervention alone, but after another two advice sessions this proportion had doubled.
In line with earlier findings that dependent or long-term very heavy drinkers do not benefit from brief interventions, the other major finding in the study was that neither of the interventions seemed to have benefited dependent drinkers. At the start of the study these averaged about 80gm of alcohol a day, 10 UK units. Neither did they benefit patients who on average drank quite moderately At the start of the study averaging about 14gm of alcohol a day, under two UK units. but sometimes to excess. It was the in-between patients At the start of the study they averaged about 37–49gm of alcohol a day, about five or six UK units. who benefited – patients whose health was at risk from regular excessive drinking and/or who were experiencing adverse consequences, but in both cases short of dependence. Yet even among these patients, classic targets for brief intervention, this on its own At least in the form studied, ie, computerised assessment feedback with tailored written advice rather than face-to-face consultation. was insufficient for many.
The findings have clear implications for primary care alcohol interventions, where the nature of the practice-patient relationship makes it feasible to follow brief advice with periodic check-ups and further intervention if needed. It suggests this potential is worth exploiting. However, as the authors caution, it cannot simply be assumed that such a strategy would be feasible and effective in normal primary care practice. Though patients were identified in GP practices, both the identification and the interventions were conducted by specialist staff. Nearly half the patients who screened positive for possible risky drinking refused to participate in the study, raising doubts about the representativeness of those who did. Also it is unfortunate that by chance the risky drinkers in the study's intervention groups were from the start drinking more A not quite statistically significant extra 10gm a day. than the non-intervention control group, raising the possibility that the clearest benefit from the intervention was partly It did not account for all the extra drop in consumption among the intervention groups. due to catching these patients at an atypical high in their drinking, from which they might in any event have descended.
Last revised 05 February 2009. First uploaded 05 February 2009
Give us your feedback on the site (two-minute survey)
Open Effectiveness Bank home page
Add your name to the mailing list to be alerted to new studies and other site updates