Send email for updates
Williams E.C., Rubinsky A.D., Chavez L.J. et al.
Addiction: 2014, 109(9), p. 1472–1481.
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 Williams at firstname.lastname@example.org. You could also try this alternative source.
Evaluated across an entire region, a determined effort to implement alcohol screening and brief intervention in the US health system for ex-military personnel led to no significant reductions in drinking – results seen as a prime example of the disappointing impacts of alcohol brief interventions in real-world conditions.
Summary Routine alcohol screening and brief intervention for outpatients who screen positive is a top prevention priority but has been extremely challenging to implement. Without successful implementation, providers offer brief interventions primarily to patients with medical conditions related to alcohol or to those with severe drink problems, for whom brief intervention may be inadequate.
The US Veterans Affairs (‘VA’) health service for ex-military personnel has mounted probably the world’s most successful effort to implement widespread alcohol screening and brief intervention.
From 1 October 2007 this included a national performance measure incentivising brief intervention for patients who screened positive, aided from January 2008 by an automated prompt to clinicians.
But in this study across a VA region, patients recorded as having been given brief advice were no more likely than others later to have stopped drinking at risky levels.
VA studies have yet to show that its major implementation effort has reduced drinking, results seen as a prime example of the disappointing impacts of alcohol brief interventions in real-world conditions.
The US Veterans Affairs (‘VA’) health care service for ex-military personnel is the largest integrated healthcare system in the United States. Commonly its clinics use reminders triggered by electronic recording ssytems to prompt clinicians to take recommended measures. Whether clinicians act accordingly can be fed into national performance measures linked to financial incentives. Using such systems can increase provision of recommended preventive care, including brief alcohol counselling for patients who screen positive for risky drinking.
Nationally the service implemented annual alcohol screening in 2003, and from 1 October 2007 implemented a national performance measure incentivising brief intervention for patients who screened positive for risky drinking. In January 2008 the service made available an electronic reminder to prompt clinicians to offer brief alcohol counselling when patients screened positive on the three questions which constitute the AUDIT-C screening questionnaire. Look here for more on the reminder.
In response to these measures, documented provision of brief interventions increased, but it remained unclear whether as a result risky drinking had been reduced. To assess this the featured study analysed records of patients seen at 30 VA medical centres in the northern and western United States during the first six months of the brief intervention performance measure – from 1 October 2007 to 4 April 2008. Of the 269,937 patients, 22,214 or just over 8% scored five or more, meaning that according to the measure they should have been offered a brief intervention. Of these 22,214 patients, 6210 or 28% were screened again nine to 15 months later. Nine months was chosen because the VA’s requirement for annual screening means most sites prompt clinicians to re-screen nine months after a prior screen. On average the time between the two screens was just under a year. Patients were nearly all men, 90% were aged 50 or more, and 37% scored on AUDIT-C as drinking at a level posing severe risks.
Of the 6210 repeat-screen patients, records for 1751 or 28% indicated that between the two screens their VA clinicians had implemented the required brief intervention or had advised them to reduce or abstain from drinking, considered a good proxy for the required intervention. Advised patients differed from those not advised in several ways, including being more likely to be physically and mentally ill, to have drug use problems in addition to their drinking, and to be relatively heavy drinkers. Though it averaged 28%, across the 30 centres the intervention rate varied from 0% to 68%; at seven fewer than 1 in 10 risky drinkers were recorded as having been advised about their drinking.
The key question asked by the study was whether relative to non-advised patients, patients recorded as having received brief advice were at their second screen more likely to have appreciably reduced their drinking to the point where they no longer met the VA’s criterion for risky drinking – operationalised as scoring under the five-point cut-off after having reduced their AUDIT-C scores by at least two points. To account for differences between advised and non-advised patients, the analysis took into account factors including initial severity of drinking, other substance use, and physical and mental health.
Across all 6210 repeat-screen patients, and when the sample was segmented by alcohol use severity, documented brief intervention was not associated with a greater likelihood that patients had stopped drinking at risky levels chart. For the whole sample remission proportions were virtually identical – adjusted for other factors, 47% with advice, 48% without. Among the less severe drinkers, the remission rate was 51% with advice, 48% without, and at 45% was identical among patients with more severe drinking patterns. The analysis was re-run dropping the requirement for a two-point reduction in AUDIT-C scores; still there was no greater remission after brief advice.
These results indicate that during early implementation stages, the reach and effectiveness of alcohol brief interventions may be limited, and they may still tend to be reserved for more severe drinkers with greater health problems. Only slightly over a quarter of patients who screened as drinking in an unhealthy manner were recorded as having received a brief intervention, and intervention was not associated with significantly more patients remitting from unhealthy drinking. The latter finding held also among patients in the lower severity range, for whom brief interventions are thought most suitable.
It should be remembered that this study was confined to patients seen initially during the first six months of the VA’s national programme to incentivise brief intervention through a performance measure. By 2010, rather than just over a quarter, the intervention rate nationally had reached 77%. The findings also contrast with those from a pilot study pre-dating the performance measure which found that 3% more patients with documented brief interventions had resolved their unhealthy drinking. With more repeat screening and therefore a larger sample, the featured study too might have found some extra drinking reductions after brief advice. Other differences between the studies were that in the pilot the intervention rate was 71%, and intervention was no more likely to be recorded for patients with severe versus less severe drinking patterns.
If replicated, the featured study’s findings could reflect the need for improvements to the quality of brief interventions in the VA system. Electronic clinical reminders do increase documented provision of recommended care, but documentation sometimes does not match what actually happens. Moreover, impacts of brief interventions demonstrated in randomised trials have been found by some studies not to transfer to more routine implementation. It could be that the VA’s top–down quality improvement initiatives raise documented compliance, but in reality are not enough to overcome barriers to implementing brief interventions.
This study may not have been able to fully account for differences between patients for whom brief interventions had versus has not been documented, or between clinics where intervention was more versus less common. Given the way patients were sampled, the results may not apply to less frequent users of VA health services, nor to those less likely to be married or to have significant mental health problems. Also the study could not account for the possible impacts of alcohol-related advice not recorded within the clinical reminder system.
commentary Results from this early phase of the VA’s national system offered no encouragement to its continuation, though results may change as the system beds in and is developed. Meantime the featured study has been seen as the definitive demonstration that studies truly close to real-world conditions have not found brief interventions work. It is the latest in a series of VA studies which have as yet produced no convincing evidence that what seems the world’s most successful effort to implement widespread alcohol screening and brief intervention has had the intended impacts on drinking; details below.
Like the Veterans Affairs system, the UK health service has placed considerable emphasis on alcohol screening and brief intervention. As yet there are no system-wide studies like the featured study which can show whether these have curbed risky drinking. These US results show that even a determined and sophisticated implementation drive cannot be assumed to reduce drinking.
The study’s authors contrasted their findings with the significant extra reduction in drinking found in an earlier pilot study at a single multi-clinic VA facility where there were management expectations on clinicians not to dismiss electronic reminders. However, the interventions stimulated by these expectations only slightly reduced drinking, and without randomising clinics or patients to be counselled or not, the results were vulnerable to bias. At a less promising facility no significant extra reductions in drinking were found. To address possible quality deficits, the service tried automating brief intervention via a web-based program, but it had no demonstrable impacts on drinking over and above the service’s mandated but patchily delivered usual alcohol advice requirements.
Though quantity was there, screening quality was also called in to question when it was found that 61% of patients who screened positive to a confidential postal survey did not do so when the same questions were asked as part of their routine care. These results mean that even if brief interventions reduced drinking, the impact across the VA caseload would be less than expected because many risky drinkers will have been missed. Look here for more on this study.
Other examples of disappointing results from brief alcohol interventions mounted under relatively real-world conditions derive from the English SIPS trials funded by the Department of Health in 2006. Results from primary care surgeries, probation offices and emergency departments were essentially the same: whether the intervention truly was a brief intervention as usually conceived, or a minimal warning intended as a control procedure, made no appreciable difference to drinking reductions. Numbers screened also seem to have been small, and achieving them often required specialist support.
Real-world trials of web-based brief intervention among college students in Sweden and New Zealand have also either found no extra reductions in drinking, or very small effects which might not have been due to the interventions.
For more on brief interventions and related UK policy see this Effectiveness Bank hot topic.
Last revised 26 October 2015. First uploaded 18 October 2015
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