Crawford M.J., Sanatinia R., Barrett B. et al.
Health Technology Assessment: 2014, 18(30).
A major study conducted in London did not find clinically important reductions in drinking among excessive drinkers offered a brief intervention while attending sexual health clinics, nor did brief intervention seem a cost-effective use of health service resources.
Summary Surveys have repeatedly demonstrated that a high proportion of people attending sexual health clinics are drinking above recommended levels. Those who drink excessively are more likely to be diagnosed with a sexually transmitted infection. Brief interventions for risky drinking have been shown to be effective across a range of medical settings, but there is very little evidence from sexual health clinics.
Conducted in London, the featured study examined the (cost-)effectiveness of a brief intervention for people attending sexual health clinics identified by a screening test as drinking excessively.
The trial did not find clinically important reductions in drinking, nor did brief intervention seem a cost-effective use of health service resources.
This major trial bolsters the impression that real-world brief interventions are not effective enough to justify widespread implementation, but their potential may yet be realised, and/or targeting screening to new patients and people who seem at risk may be more cost-effective.
To help fill this evidence gap, the featured study examined the effectiveness and cost-effectiveness (in respect of both drinking and on sexual behaviour) of a brief intervention for people attending sexual health clinics whom screening tests identified as drinking excessively.
Between 2010 and 2012, 802 participants were recruited to the study from three hospital-based sexual health clinics in London, England. Participants had to be aged 19 or above and drinking excessively according to the Modified-Single Alcohol Screening Question. This asks men, “How often do you drink more than 8 units of alcohol on one occasion?”; women are asked a corresponding question but stipulating six units. Excessive drinking was defined as drinking these amounts at least once a month. Participants were typically aged 24–30 years and heterosexual, and just over half were women. Nearly two-thirds drank heavily at least weekly but very few did so daily. Most had attended the clinics because they were experiencing genitourinary symptoms or for a health check.
On days when recruitment took place, clinic staff gave all those attending the clinic a postcard with information about the study (presented as about sexual health and lifestyle, not specifically drinking) and asked if they would meet a researcher. Participants who agreed to join the study (of 1640 assessed, 447 refused) were assessed for eligibility, including the question about their drinking. Those who qualified were then allocated at random to the evaluated intervention or to a control procedure which involved clinic staff offering them a health and lifestyle information leaflet including information on alcohol.
Patients randomly allocated to the intervention were instead offered brief (two or three minutes) face-to-face advice on drinking from the clinician treating them at the clinic, based on that previously found effective in an emergency department setting. It consisted of feedback on the possible health consequences of their drinking, highlighting any links between their attendance and their drinking, and written information about alcohol and health. Patients were also asked if they would be willing to attend an appointment with an alcohol health worker lasting up to half an hour for further advice, which (if the case warranted it) might include offering another appointment or referral for treatment. Participants unable to attend an appointment on the day were offered a later appointment or telephone-based information and advice.
Whether (relative to just being given a leaflet) the intervention had led to further improvements in drinking and sexual risk behaviour was assessed six months later, when around three-quarters of the participants could be contacted by phone and answered questions about their behaviour in the previous three months.
All but five of the 402 patients allocated to the intervention received brief advice from their sexual health clinician but just 81 received follow-on advice from an alcohol health worker, mostly over the phone. At the start of the study these 81 were as far was known no more likely to be drinking heavily than other patients. How many more patients accepted the offer of the appointment but failed to turn up and did not answer calls is not reported.
During the period from three to six months after intervention, on average participants allocated to the intervention indicated they had been drinking 18.1 UK units (each about 8g alcohol) a week, compared to 20.3 units among control patients just offered a leaflet chart. Adjusted for other factors, the difference of 2.3 units favouring intervention patients just failed to meet conventional criteria for statistical significance, meaning chance variation could not be ruled out. The difference in weekly amounts was due to intervention patients drinking just over a unit (1.1 or 8.8g alcohol) less on each day they drank, a gap which was statistically significant. When assessed in the same way as at the start of the trial, about 24% of intervention patients no longer scored as drinking excessively compared to just 18% of control patients, but this difference was not statistically significant.
Unprotected sex during the period from three to six months after intervention was reported by 53% of intervention patients and 59% of control patients, a difference which though it favoured intervention patients, fell well short of the conventional criterion for statistical significance, so may have been a chance finding.
Including the interventions assessed by the study, over the six-month follow-up period intervention patients used health services which cost on average £319, slightly and non-significantly more than the £311 for control patients. If a service were willing to pay £5 for a per-patient reduction of one unit a week in drinking, the brief intervention would have over a 50% chance of being cost-effective. Though they absorbed slightly more health service resources, intervention patients on average scored as having experienced a slightly lower increase in their quality of life relative to control patients, meaning that on these grounds the intervention could not be considered cost-effective, despite costing just £12.57 extra per patient.
The researchers concluded that the trial had not produced evidence that brief intervention for excessive alcohol use among people attending sexual health clinics is associated with clinically important reductions in drinking, nor that it is a cost-effective use of health service resources. Although there was a statistically significant difference in the number of units drank per drinking day, the scale of the difference (1.1 units) is unlikely to be clinically important. No statistically significant improvements in sexual health were found, though the possibility could not be ruled out that brief intervention might be associated with small, but clinically important changes.
Assessing drinking as part of wider efforts to tackle poor sexual health among some patients makes sense, but screening all sexual health patients and offering brief interventions to all risky drinkers does not appear clinically effective or cost-effective. It is, however, possible that some patients would benefit from brief alcohol intervention, such as those who both drink heavily and who frequently acquire sexually transmitted infections, and those seeking emergency contraception or post-exposure treatment to prevent HIV disease. Impacts might be greater among heavier drinkers who more often exceed recommended limits.
Confidence in the practical relevance of the findings is bolstered by an intervention which was delivered by and manageable by front-line clinicians rather than specially trained staff, and by recruitment and baseline assessment procedures which minimised the degree to which control patients might in response have reduced their drinking. Also, though only about a fifth of intervention patients took up the offer of further advice, making this available over the telephone seems to have meant this proportion was far higher than in a pilot study.
Findings were similar to those from a corresponding study in Australia which assessed the effects of brief intervention for risky drinking delivered by a trained nurse in a sexual health clinic. Compared to screening and research assessments only, the intervention resulted in a small, non-statistically significant extra reduction in the proportion of people drinking harmfully.
In contrast to the featured study, the many trials in other health-care settings have generally found brief intervention associated with self-reported reductions in alcohol consumption over the next six to 12 months. It is unclear why the featured study too did not find appreciable impacts on drinking. Possible reasons explored below include the level of alcohol misuse among people attending sexual health clinics, their readiness to change, and the intervention actually received.
Though typically when they drank they drank heavily, overall levels of alcohol consumption among the patients were far lower than those seen in other brief intervention studies. There is some evidence that brief intervention is less effective among lower-level consumers.
While acknowledging a link between alcohol use and sexual behaviour, patients attending sexual health clinics do not seem to see drinking as leading to sexually transmitted infections or other negative health consequences, nor do they see their drinking as excessive even if they are drinking above recommended daily units. Instead, drinking is seen as a normal part of their social life and a means of having fun. In the featured study just 19% of patients thought their attendance could be related to their drinking, compared to over 69% in a previous study in an emergency department. People attending sexual health clinics may want better sexual health, but reducing drinking is generally not seen as essential to achieving this aim.
The quality of the intervention itself is unlikely to have been a factor. It was modelled on one previously found effective, and nearly all those allocated to it received each of the intended four components of brief advice. A survey of the clinicians who delivered the brief intervention found most believed there was a link between drinking and sexual health, and that the intervention could help improve sexual health.
commentary Following results from the English SIPS trials (1 2 3), this major trial may be seen as bolstering the impression that in the UK, real-world brief interventions delivered across the board to all risky drinkers identified by screening are not effective enough to be considered a good use of health service resources. Given the general finding (1 2) that many control groups in alcohol brief intervention studies who received no or minimal intervention substantially reduced their drinking, it seems entirely possible that in the featured study neither screening and research assessments on their own, nor these plus brief intervention, had any impact on drinking.
The way such studies are designed means that all that can be said on strict scientific grounds is that so far generally they have not found real-world brief interventions to have greater impacts than doing almost nothing (or not great enough to warrant substantial investment) – an absence of evidence of impact, not positive evidence of no impact. Sampling variations may have obscured a real effect, and other trials show there is a potential for effectiveness, which might yet emerge in more real-world conditions. But the more consistently and the more often real-world implementations fail to deliver, the stronger grows the impression that this reflects the reality which research trials dip into and sample/manipulate.
As the authors suspected, the intervention’s impact may have foundered on the mismatch between levels of drinking seen by national guidelines as excessive and the views of this mainly young, sexually active group who saw their ‘excessive’ drinking, probably usually confined to weekends, as a normal feature of their social life, and generally unrelated to the problems which brought them to the sexual health clinics. The drinking amounts which triggered intervention were the standard set for the screening test, but at a monthly drinking session of over eight units for men and six for women were low for this caseload; just 369 of 1690 patients approached were found not to be drinking above these levels. In emergency departments and trauma centres where there is some evidence (1 2) that brief interventions modestly reduce drinking, excessive drinkers are likely to have suffered an unintended repercussion of their drinking in the form of an injury, but having fun and perhaps too having sex will often be a desired and intended consequence of drinking, which in the latter case entails some risks which may result in a visit to a sexual health clinic. Persuading these drinkers that their consumption has led to a problem rather than on balance enhanced their life would seem a more difficult task.
Though brief interventions are assumed applicable only to risky but pre-problematic drinkers, findings from trials are consistent with impacts being as great among heavier drinkers, and there is some evidence that they respond best to brief interventions. The presumed tension is between drinking heavily enough for this to seem even to the drinker to be a valid concern, and being so reliant on drinking that it will take more than a few minutes of advice to shift the habit. But in certain circumstances, even dependent drinkers respond well to brief advice.
These plausible explanations for the intervention’s minor impact must however be set alongside equally or more negative findings from the English SIPS brief intervention trials in primary care surgeries, probation offices and emergency departments. Similar findings among very different populations with different reasons to consider or not consider changing their drinking suggest that these factors are not determining influences – that it is simply that real-world brief interventions are ineffective or at best only minimally and inconsistently effective.
In the SIPS trials, a year later the proportion of risky drinkers had fallen by about the same proportion whatever the intervention. Most basic was a simple warning that the patient or offender was drinking “above safe levels, which may be harmful to you”, plus an instruction to read the alcohol information booklet they were handed. Supplementing this with an individualised brief intervention much like that trialled in the featured study – including referral to an alcohol health worker – made no difference. As in the featured trial, generally drinking problems and amounts were modest but still most participants had at some time thought of cutting down. These intentions were most often solidified into decisions and actions among offenders seen in probation offices whose drinking was also on average most problematic, and it was in this trial that (especially among the heaviest drinkers) there were some signs that brief counselling did have an extra impact.
Given these findings, it seems possible that brief interventions when delivered in ways approximating routine practice are simply not effective enough to justify the considerable resources it would take to ensure and maintain widespread implementation. Across the spectrum of populations targeted, settings and intervention methods, truly real-world trials of brief interventions are few (1 2 3 4 5) and they tend to find that the interventions are often not delivered and do not affect drinking to a statistically significant degree. More promising results from trials more selective about their participants and in which there is non-routine support to promote implementation seem not to survive the loss of these controls and supports.
Nevertheless, findings from trials in general have been encouraging enough for alcohol screening and brief intervention to feature as important strands in public health policies across the UK, though not in the universal form trialled in the featured study. Instead, ambition has been scaled back to screening new patients and/or those thought in advance to possibly be at risk – so-called ‘targeted’ screening.
Thanks for their comments on this entry to research author Michael Crawford of Imperial College London in England, Adrian Brown of Northwick Park and Central Middlesex Hospitals in England, and Richard Saitz of the Boston University School of Public Health in the USA. Commentators bear no responsibility for the text including the interpretations and any remaining errors.
Last revised 22 March 2016. First uploaded 29 January 2016