Parkes T, Atherton I., Evans J. et al.
Edinburgh: NHS Health Scotland, 2011.
In three years from 2008 Scottish national policy drove delivery of nearly 175,000 brief alcohol interventions, testament to what can be done when policy is backed by funding and infrastructure and incentive payments contingent on implementation. Leverage and acceptance were greatest in primary care, where the vast majority of the work took place.
Summary Brief interventions in healthcare settings are a significant component of the Scottish government's alcohol strategy. The aim is to use quick screening questionnaires to identify people drinking above sensible drinking guidelines and to offer brief advice to hazardous and harmful drinkers to reduce their consumption and risk of alcohol-related harm. The government established a health improvement target (HEAT H4) for NHS health boards in 2008 supported by a substantial increase in funding for alcohol treatment and support services. Between April 2008 and March 2011 this required the Scottish national health service to deliver 149,449 brief interventions across three priority settings: primary care, emergency departments, and antenatal care. Later it was extended to the delivery of 61,081 alcohol brief interventions during April 2011 to March 2012, with the aim of embedding these interventions in health service practice.
The featured evaluation focused on implementation in primary care but also reported on some dimensions of implementation in the other two settings. Data was collected mainly in 2010 at national, health board, practitioner and patient levels. The next section is based on the report's own summary.
The initial three-year target was reached nationally ahead of schedule in March 2011, and examples of successful delivery in all priority settings clearly indicate the ability of health boards to do this work well. Most of this activity has been in line with Scottish guidance on delivery of brief alcohol interventions, evidence-based practice, and the international literature.
Importantly, healthcare staff see brief alcohol interventions as a worthwhile activity for NHS staff and a valid use of NHS resources. The vast majority of primary care patients interviewed for the study accepted that conversations about alcohol are part of a GP's or healthcare worker's role and, in the main, appeared to have no problem with being offered alcohol screening or brief interventions. The programme was expected to result in increased referrals to specialist alcohol services. Though referrals have not necessarily increased as much as expected, some health board leads felt more of the 'right people' are now being seen in specialist alcohol services – more motivated to seek help and create change in their lives.
Interviews suggested that the reach and impact of the initiative was mixed across Scotland. Some health board leads were clear there was much more work to be done, while others felt that two years in to the HEAT H4 target, already they had outperformed. The population-wide approach was viewed as an effective and essential mechanism to avoid patients feeling stigmatised by being advised on their drinking. However, gaps in coverage were noted, especially in rural and remote areas and in relation to age/gender groups who less frequently attend mainstream medical services.
The extension year for HEAT H4 provides further opportunity and further support to build the capacity of NHS staff in Scotland to provide effective and evidence-based brief alcohol interventions. However, the featured study's findings suggest funding and infrastructure support are likely to be needed for some time yet before mainstreaming into routine practice is guaranteed. As well as developing skills and knowledge, the investment in staff training has also been responsible for building grassroots support for brief alcohol interventions; practitioners who had taken part in more in-depth training expressed greater enthusiasm for the programme and motivation to become involved. Due to staff turnover, the need to update and renew skills, and to reach those who remain untrained, the training infrastructure will need to continue to be adequately resourced to effectively 'bed in' more advanced levels of skill, confidence and competence.
Findings on data collection and monitoring suggest the importance of developing more universal systems and standards for recording activity if enhanced public health surveillance is required. However, practitioners were clear that recording needed to be simplified if brief alcohol interventions are to be effectively mainstreamed. This represents one of the key tensions of this work: low demands on practitioners may make it easier to embed new activities into practice and therefore enhance adoption, but also mean that their impact becomes more difficult to monitor.
One of the most significant findings of the evaluation is the variation across Scotland in how brief alcohol interventions have been implemented. For example, different payment structures in primary care enhanced service contracts have resulted in practices emphasising different aspects of interventions and developing different approaches and infrastructures for implementation. Health boards have found numerous ways to localise the programme to fit their social and geographic contexts, indicating a willingness to be flexible and accommodating to the different issues each faces. However, the many commonalities and shared experiences enable substantial generalised learning for the mainstreaming of brief alcohol interventions into routine health service practice in Scotland and more widely.
The following have been selected from among the implications of the evaluation listed in the featured report.
• Continued prioritisation of investment in brief alcohol interventions is essential to ensure that the benefits of the investment are fully realised and sustained.
• Specialist roles for coordination and facilitation of this work are considered to be essential to continued implementation.
• It should be emphasised that brief alcohol interventions are only effective with harmful and hazardous drinkers and should not be used as a routine method of intervention with dependent drinkers.
• Models of delivery appropriate to one healthcare setting, or even one geographic area, may not be appropriate for other settings, so localised and tailored models and approaches should be explored, in line with the evidence base.
• Maintain financial incentives in primary care to help establish delivery as part of routine practice. Ensure payments are sufficient to cover perceived costs and to incentivise delivery and recording.
• Strategic buy-in at the most senior levels of health boards seems essential for effective and successful delivery. Relevant board sectors such as public health, health promotion and addictions, should be represented and involved.
• It may be helpful to rely less on the 'cascade' approach to training, as practitioners who have received more intensive training have limited scope to disseminate its more complex aspects, such as motivational interviewing, to colleagues and peers.
• Ensure staff understand that the impact on many different levels (national, local/community, individual) may not be detected or realised for a number of years; this does not mean brief alcohol interventions are not working. Related to this, acknowledge 'target fatigue' and try to find ways to make the work meaningful to all involved. Training is a key mechanism for achieving this.
• Delivery of screening and brief intervention together as part of the same consultation would appear to be more effective in most settings. However, in emergency departments, 'light touch', context-appropriate and evidence-informed models should be used if they enable effective implementation and are acceptable to patients. In such settings there may be advantages in restricting staff involvement to screening, with those who screen positive being referred to other services and departments for brief intervention. However, the sustainability of services where the delivery model is designed to minimise or circumvent involvement of frontline staff is likely to be limited.
• Avoid setting ambitious local screening targets based on proportions of practice populations. This may result in repeat screenings and less efficient use of resources. Local targets and contracts should encourage focused screening and discourage the practice of giving interventions to patients who screen negative or are dependent drinkers, and encourage referral of dependent drinkers to other treatment services.
• Simpler screening tools, such as FAST, are easier to use/remember and are therefore more likely to facilitate accurate recording. Adoption of effective screening tools that are sensitive to the needs of patients in different health care settings is important to the identification of those likely to benefit.
• Use of information materials by practitioners with patients should be encouraged as this helps to underline key messages, and is particularly valuable where consultation times are limited.
commentary In relation at least to primary care, the Scottish experience shows what can be done when a determined national policy drive is backed by funding and organisational infrastructure and incentive payments contingent on the work being done – findings in line with reviews (1 2) of the implementation of brief interventions, which have emphasised organisational factors such as adequate financial and managerial support, training opportunities, workload management, incentives, and clear direction from the top prioritising this work not just in theory but also in practice via the levers available to management.
In Scotland national policy clearly prioritised screening and brief intervention in primary care, antenatal care, and accident and emergency departments, realised in a health service target for 2008/09–2010/11 to deliver 149,449 brief interventions across the three years supported by dedicated funding. The target was set on the basis that 19% of adult patients would present to these services with conditions possibly related to drinking and would be screened for excessive drinking, of whom a fifth would screen positive and three quarters of these at-need patients would actually be counselled. As the featured report notes, in the event the target was exceeded; over the three-year period, 174,205 brief alcohol interventions were recorded across the three priority settings.
The Scottish programme's intended exclusion of dependent drinkers was made on the basis of evidence available to that date that brief interventions do not affect these patients, but from later studies we now know that this is not a universal finding. Contrary findings include a study in the north of England where just a few (and often just one) counselling sessions by a specialist nurse had a remarkable impact on dependent drinkers seeking medical care at an accident and emergency department.
The featured evaluation was not set up to assess impacts on drinking and interviewed just 25 primary care patients. Nevertheless their responses do not suggest a major impact: "Patients largely believed that discussion on alcohol was an incidental part of the consultation and, for most, of limited consequence ... many ... indicated that the intervention had not made any difference to them". It is however in the nature of such a programme that only a fraction of patients may respond by cutting down on drinking and then not always very much; the thinking is that across an entire population these small changes will cumulate to a worthwhile public health impact. The two patients who said they stopped drinking as a result of the advice and the others who tried to cut back may be as much as can be expected.
As documented in the featured report, the Scottish drive to implement brief alcohol interventions seems to have been most successful in GPs' surgeries. In the three health board areas where these figures were known, 83%, 84% and 92% of brief interventions were recorded as delivered in primary care. In contrast, 17%, 8% and less than 8% of interventions were delivered in emergency departments. If these kind of figures applied nationally, perhaps 11% of the 174,205 brief alcohol interventions recorded over three years were in these departments, or about 6400 a year – just over 6% of the estimated 100,000 alcohol-related attendances per year to Scottish emergency departments. Why this might be was revealed by staff interviews. Resistance from staff (feeling that this was not their business and detracted from core activities and objectives) and time pressures sometimes led (contrary to the preferred option in guidance) to intervention by appointment some time after screening rather than immediately, and this in turn reduced attendance. Screening rates probably too suffered from inadequate buy-in by staff.
At national level too, general acceptance of primary care screening and intervention contrasted with views that emergency departments were suitable for screening and referral only, and mixed views on the appropriateness if intervening in antenatal care. Again contrasting with emergency departments, in primary care practices screening and brief intervention were generally delivered together; when they were not, intervention attendance suffered. Another key factor was the specific funding available through local enhanced contracts to incentivise primary care practices. These were probably among the factors which in one of the featured report's case study areas led to an estimated 41% of hazardous or harmful drinkers visiting the practices during a year being screened, and in two others areas, about 30%. Also indicative of the leverage exerted by payment-for-performance contracts is the way the payment criteria were able to influence not just the number but the nature of primary care interventions.
Greater leverage and greater acceptance in primary care were also evident in the SIPS trial in England. Preliminary findings released by the SIPS project in the form of factsheets and conference presentations suggest that implementation was more difficult in emergency departments than primary care, though in neither setting were as many patients screened as might have been expected; weekly averages were 12 per emergency department and less than two per GP practice.
The key finding from SIPS was that across these settings and in probation offices, extended brief alcohol interventions were no more effective at reducing drinking than an alcohol advice booklet plus a sentence or two of feedback alerting someone to their risky drinking. Among the reasons may have been that these interventions came after patients and offenders had been through research and screening assessments of their drinking and related problems and their readiness to do something about these, possibly thought-provoking interventions in themselves. Also, while what was intended in the interventions is clear, what was actually done has as yet not been reported.
However, it may simply be that longer (but still brief) and more sophisticated interventions are no more effective than shorter and simpler ones. For example, a comprehensive review of primary care brief alcohol intervention study found no extra benefit from extended interventions, while in a Scottish hospital, handing heavy drinking medical inpatients a guide to sensible drinking led to declines in consumption as great as more extended advice. The evidence on this issue is mixed, but if it is the case that a straightforward warning based on screening results achieves all that can be achieved, then the training and resource load documented in the featured report might be substantially reduced.
Thanks for their comments on this entry in draft to Tessa Parkes of the University of Stirling. Commentators bear no responsibility for the text including the interpretations and any remaining errors.
Last revised 10 December 2012. First uploaded 24 November 2012
STUDY 2016 Improving the delivery of brief interventions for heavy drinking in primary health care: outcome results of the Optimizing Delivery of Health Care Intervention (ODHIN) five-country cluster randomized factorial trial