Re: Effectiveness of screening and brief alcohol intervention in primary care (SIPS trial): pragmatic cluster randomised controlled trial
SIPS trial authors conclude that “simple feedback and written information may be the most appropriate strategy to reduce hazardous and harmful drinking…” I don’t think that conclusion is the one most consistent with their findings. The best evidence for efficacy is still for brief counseling (1) -- which has efficacy above screening, simple feedback and consent in randomized trials--and SIPS results are not inconsistent with the findings of those efficacy trials. The SIPS study (not an efficacy trial but a pragmatic/effectiveness study) found that efficacy of brief intervention is lost in clinical practice, even under the best of circumstances.
First let’s be clear about the data. The SIPS trial findings were that brief advice and brief counseling as implemented in real world practices (the latter, 57% of the time) had, if anything, less effect (though not significantly so; odds ratios 0.78 and 0.85) than a leaflet, in patients, 15-20% of whom did not meet the investigators’ chosen criterion for unhealthy alcohol use (an AUDIT score of 8 or greater) to begin with, and 17% of whom were lost to follow-up (meaning that in the counseling group, 0.57*0.85*0.83, or at most 40% with unhealthy use received the intervention and were followed).
We should not conclude that a leaflet had any effect. Decreases in use across the three groups may have been self-change related to choosing to have a medical visit, or a response to screening or informed consent (no need to invoke statistical regression to the mean). Particularly given the robust findings from systematic reviews that favor brief intervention (advice, counsel) when compared to no brief intervention in efficacy trials (1), the conclusion most consistent with these data is that even when great efforts are made to implement screening and brief intervention (SBI) in real world clinical care (e.g. with less external researcher support), the effects seen in efficacy studies do not translate into effective interventions in practice.
The research and policy implications are enormous since alcohol SBI is among the most potentially effective preventive service in health care (2). But SBI is a skilled service that needs to be done well; done poorly (e.g. without fidelity to health behavior change counseling skills, with a confrontational approach or with a breach of confidentiality) it won’t work (or could even be harmful). And the effect sizes in efficacy studies, while large from a public health perspective, are small enough (e.g. 3 fewer drinks per week) that they could easily be erased when SBI is not implemented in practice exactly like it was in those studies. Implementation of SBI in practice as well as it is done in efficacy studies, is unlikely. Yet alcohol SBI can only reach its potential if the effects seen in efficacy studies can be achieved in real world practice. Kaner et al’s systematic review suggested that the practice was similarly effective in trials in which SBI implementation looked more like it would in clinical practice and less like research implementation, but none of those studies came close to being pragmatic trials like SIPS, so they couldn’t really inform that question (1).
The SIPS trial is one of very few pragmatic implementation studies of alcohol SBI. One other had an even more disappointing result, finding that a huge implementation effort resulted in little implementation, and worse patient outcome (i.e. more hazardous drinking) in practices randomly assigned to implement screening and brief intervention (3,4)(a finding consistent with SIPS study results).
The implications seem clear (and they are not about leaflets--feedback and leaflets might be effective but that isnt what was shown in the SIPS study). Policymakers should be leery of widespread implementation unless it is done well. And it will take a lot to do it well—saying we are doing it well without assuring high quality implementation (which reflects current practice, at best) will give us false reassurance that we have taken care of unhealthy alcohol use and will waste time and money. And researchers and educators should turn their attention to how to implement alcohol screening and brief intervention in clinical practice in a way that retains the efficacy seen in clinical trials.
Kaner EF, Dickinson HO, Beyer FR, Campbell F, Schlesinger C, Heather N, Saunders JB, Burnand B, Pienaar ED. Effectiveness of brief alcohol interventions in primary care populations. Cochrane Database of Systematic Reviews 2007, Issue 2. Art. No.: CD004148. DOI: 10.1002/14651858.CD004148.pub3.
Solberg LI, Maciosek MV, Edwards NM. Primary care intervention to reduce alcohol misuse: ranking its health impact and cost effectiveness. Am J Preventive Med 2008;34:143-152.
van Beurden I, Anderson P, Akkermans, R P, Grol RPTM, Wensing M, Laurant MGH. Involvement of general practitioners in managing alcohol problems: a randomized controlled trial of a tailored improvement programme. Addiction 2012;107:1601–1611.
Hilbink M, Voerman G, van Beurden I, Penninx B, Laurant M. A randomized controlled trial of a tailored primary care program to reverse excessive alcohol consumption. J Am Board Fam Med 2012;25:712–22.
Competing interests: A clinical and academic career devoted to improving care for people with unhealthy alcohol and other drug use; honoraria for lectures on these topics at academic and other institutions and professional meetings; US National Institutes of Health research grants on related topics. No related commercial interests.