Does alcohol screening work in general practice?

tmr-coverThis clinical article on the evidence surrounding alcohol screening and brief interventions was published in the medical periodical, The Medical Republic.

The bottom line

“Yes”, alcohol screening and brief interventions (ASBI) in general practice do work. However, it’s not a blanket yes as the devil is in the detail. Understanding our patient’s needs is key.

The background

Risky alcohol use is a social and health problem in Australia.  The general population has consistently perceived “excessive alcohol consumption” as the “drug of most serious concern”.  Thirty percent of people aged 14 or older were risky drinkers in the 2013 National Drug Strategy Household Survey.  In the ongoing BEACH study, 26% of adult patients in general practice have reliably been risky drinkers.

The current paradigm of a preventive health approach to alcohol-related harms has been established since the 1980s.  Operationally, this has been through the use of alcohol screening and brief interventions (ASBI) in primary care.  The rationale is that if risky drinking is detected early before the onset of harm, brief interventions can be delivered and the burden of alcohol-related disease and injury in the population reduced.

General practitioners (GPs) have been seen as instrumental in delivering ASBI.  Statements that declare general practice as the “ideal” setting for ASBI are cliché in the academic literature – I have been guilty of writing this myself [1].  With ASBI recommended in routine practice, a basic empirical question must be answered – do alcohol screening and brief interventions work in general practice settings?

The relevant Cochrane systematic review is often cited as the definitive reference [2].  At face value, its findings are supportive of ASBI – participants who received ASBI consumed on average 4 standard drinks less per week than the control group at one year follow up [2].

However, digging a little deeper reveals clinically meaningful uncertainties.  There was a major difference in the effect of ASBI depending on sex.  Men reduced their mean alcohol intake by almost 6 standard drinks per week, while it was only 1 drink per week for women [2].  It is arguably unreasonable to pool the results from men and women together.

Moreover, there was at least moderate heterogeneity, that is, lack of consistency in the results of the individual studies in the meta-analysis, even within the male and female subgroups [2]. The implication is that there are clinically important contexts that influence the effectiveness of ASBI.

The evidence

If we step back and take a broad overview of the evidence, there are curious anomalies that deserve inspection.  If ASBI are as effective and acceptable as commonly claimed, then why have they not already been successfully implemented broadly in primary care [3]?

Research over decades has consistently found GP scepticism on the utility of ASBI tools, even though GPs agree that risky drinking is a problem, and that they have an important role to play.  This scepticism and reluctance has traditionally been dismissed and even criticised as evidence of “low levels of therapeutic commitment” (p. 249) [4].

It was just a decade ago that even the suggestion that ASBI may not be effective in general practice, was seen as heresy.  In 2003, Beich and colleagues published a systematic review in the BMJ, reporting that for every 1000 patients receiving ASBI in general practice, only 2 or 3 can be expected to have reduced their alcohol consumption to below the recommended maximum level after 12 months [5].  This generated irate responses from others in the field – some of which can be read online at the BMJ website: http://www.bmj.com/content/327/7414/536/rapid-responses.

Stephen Rollnick (the Rollnick of motivational interviewing fame), one of the co-authors of that paper, wrote:

“The ferocity of many responses to our paper astonished me. Clearly, I have misjudged the momentum that has built up around alcohol screening. I thought that it was still in need of considerable debate and refinement, yet others were clearly (and sincerely) working towards widespread dissemination in general practice. This came home to me when I was invited to attend a “friendly” debate soon after the publication of our paper, at a meeting of European alcohol researchers. At one point in the debate, it was suggested that an analysis be conducted of the number of people likely to lose their lives as a result of GPs abandoning alcohol screening after reading our paper.”

Efficacy versus effectiveness

Thankfully, recent commentary has recognised the issue of efficacy versus effectiveness – that is, even if ASBI are efficacious in controlled settings, they might not be effective in real-world practice [3].  There remains disagreement as to whether ASBI are effective in pragmatic practice – some interpret the evidence in the affirmative [6], and others in the negative [3].

From a clinical perspective, the contemporaneous, larger, and pragmatic trials in general practice have relatively consistent results.  These trials conducted in Europe and the US (Box 1), all demonstrated no or minimal effect of ASBI as compared to the control intervention.

However, this doesn’t mean we should abandon the preventive approach to risky drinking.  All researchers agree that ASBI are efficacious.  Fundamentally, this is an implementation problem.

An interesting observation from the aforementioned pragmatic trials in general practice is that participants in the control groups, who received screening and usual care, had important reductions in risky drinking.  The null result in these studies was due to the fact that the control participants improved just as much as those who received the formal ASBI package.  One encouraging hypothesis from these findings is that the alcohol assessment process itself may be the “active ingredient”, encouraging individuals who are ready to reduce their drinking to do so [6].  This could explain the previously observed phenomena that longer brief interventions are no better than shorter [2], and that ASBI are ineffective in people with alcohol abuse and dependence.

Thus, rather than “do they work”, the better question is in what contexts are ASBI effective, and how do we implement?

The research agenda has been committed to the universal application of ASBI, and has tended to ignore discrepant clinical viewpoints from GPs [1].  Dismissing this collective wisdom may have been a missed opportunity – the average GP has many years of lived experiences interacting with patients in real clinical situations.  While universal ASBI is seen and experienced by GPs as impractical, targeted screening and pragmatic case finding are potentially acceptable.

Valuing patient perspectives

Undervalued too are patient perspectives – a strategic misstep given that they are the recipients of ASBI [1].  Patient acceptance of alcohol assessment is influenced by the context of the consultation [7].  Patients are more positive towards this assessment when it is framed within a SNAP (smoking, nutrition, alcohol, physical activity) framework [7].  Alcohol assessment is also rated as highly acceptable by patients in a number of common reasons for presentation (diabetes, hypertension, GORD, mental health) [7], and also during new patient registration.  These are opportunities for targeted screening.  Other strategies that improve the acceptability of alcohol discussions include setting the context for assessment, linking it to patient agendas, a collaborative consultation style, and respecting patient sensitivity to the moral dimension of alcohol consumption [8].

The overall message for clinical GPs is quite positive.  Much of the benefit of formal ASBI in routine practice is likely the result of simply engaging patients in a discussion about their alcohol use, within the context of the therapeutic patient-doctor relationship.  Patients see alcohol counselling as indicative of higher quality primary care.  Within a practice, focussing clinical behaviour change to capitalise on targeted screening opportunities (e.g., for new patients, chronic disease and mental health care planning) may be the most effective initial activity.  More intensive interventions in general practice might work, but may need to be supported by broader community and policy interventions to meet their theoretical potential [6].

Box 1 – Recent pragmatic trials of ASBI in general practice

Beich et al. 2007, Alcohol and alcoholism, 42(6): 593-603

A Danish randomised controlled trial (RCT).  There were no differences between the brief intervention and control groups at 1-year follow-up.  There was a small increase in alcohol consumption in groups at the end of the study.

Hilbink et al. 2012, Journal of the American Board of Family Medicine, 25(5): 712-22

A Dutch RCT.  The brief intervention appeared less effective than the control.  Of the risky drinkers in the study, 47% in the control group, and 36% of the intervention group, were low-risk drinkers at 2-years follow-up.

Kaner et al. 2013, BMJ, 346: e8501

An English RCT.  Brief (5 min) advice, and brief (20 min) lifestyle counselling, was no better than a patient information leaflet at 12-months follow-up.  All groups demonstrated a modest decrease in alcohol consumption.

Butler et al. 2013, BMJ, 346: f1191

A Welsh RCT.  The brief intervention was no better than the control at 12-months follow-up.  Both groups demonstrated a modest decrease in alcohol consumption.

Williams et al. 2014, Addiction, 109(9): 1472-81

A US Veteran Affairs retrospective cohort study.  Of veterans who screened positive for risky drinking, those who received documented brief intervention did no better than those who did not, at 9-15 months follow-up, in a brief intervention implementation study.  Both groups demonstrated a substantial resolution in risky drinking on screening.

References

  1. Tam CWM, Leong LH, Zwar N. Let’s listen to patients’ and GPs’ perspectives on alcohol-screening research. Australian family physician 2015;44(6):427-8.
  2. Kaner EF, Beyer F, Dickinson HO, et al. Effectiveness of brief alcohol interventions in primary care populations. Cochrane database of systematic reviews 2007(2):CD004148.
  3. Saitz R. The best evidence for alcohol screening and brief intervention in primary care supports efficacy, at best, not effectiveness: you say tomato, I say tomato? That’s not all it’s about. Addiction science & clinical practice 2014;9:14.
  4. Deehan A, Templeton L, Taylor C, Drummond C, Strang J. Low detection rates, negative attitudes and the failure to meet the “Health of the Nation” alcohol targets: findings from a national survey of GPs in England and Wales. Drug Alcohol Rev 1998 Sep;17(3):249-58.
  5. Beich A, Thorsen T, Rollnick S. Screening in brief intervention trials targeting excessive drinkers in general practice: systematic review and meta-analysis. Bmj 2003 Sep 6;327(7414):536-42.
  6. O’Donnell A, Wallace PG, Kaner E. From efficacy to effectiveness and beyond: what next for brief interventions in primary care? Frontiers in psychiatry 2014;5:113.
  7. Tam CWM, Leong L, Zwar N, Hespe C. Consultation contexts and the acceptability of alcohol enquiry from general practitioners – a survey experiment. Australian family physician 2015;44(7):490-6.
  8. Tam CWM, Leong L, Zwar N, Hespe C. Alcohol enquiry by GPs – understanding patient perspectives: a qualitative study. Australian family physician 2015;44(11):833-838.

 

Leave a Reply

%d bloggers like this: