The Scottish Government

« Previous | Contents | Next »

Listen

EFFECTIVE AND COST-EFFECTIVE MEASURES TO REDUCE ALCOHOL MISUSE IN SCOTLAND: AN UPDATE TO THE LITERATURE REVIEW

SECTION FIVE: PREVENTION

SUMMARY

This section reviews the effectiveness and cost-effectiveness evidence relating to the prevention of problem drinking and alcohol abuse. The main findings of the previous review were;

• there is only weak evidence of effectiveness for school-based interventions;

• evaluations tend to show changes in knowledge rather than behaviour;

• characteristics which contribute to success are interactive design, parental, community and peer involvement;

• health promoting schools have some positive impacts but less on alcohol;

• mass media campaigns affect knowledge and attitudes rather than behaviour;

• screening and counselling adolescents was not cost-effective in one US study; and

• thiamine supplementation of beer was shown to be cost-effective in preventing Wernicke-Korsakoff encephalopathy in an Australian study.

The additional studies reviewed confirm the general findings relating to school based interventions and mass media campaigns. In addition:

• there is some evidence that booster sessions may increase effectiveness;

• the impact of specific components of programmes varies with other characteristics of the programmes;

• community involvement may increase effectiveness by restricting access to alcohol for young people;

• some studies have identified negative impacts of prevention programmes; and

• one study demonstrates the cost-effectiveness of family focussed prevention programmes.

Better research designs are needed to provide more robust evidence in the area of prevention.

INTRODUCTION

5.1 This section focuses on interventions to prevent alcohol misuse. The majority of the literature remains concerned with school-based interventions, although there is some increasing overlap with community interventions. The limited evidence relating to other interventions, including mass media campaigns, is then presented. Similar issues have been covered by a Health Development Agency Evidence Briefing (Waller et al 2002)

SCHOOL-BASED INTERVENTIONS

Types of intervention

5.2 Interventions are based around teaching social skills that will help in resisting social and peer pressure. As well as the content of the programme, the style of delivery can also be important. The main distinction drawn is between:

• interactive programmes, which include social influence and comprehensive life skills;

• interactive delivery, which requires elements such as participation, student to student communication, small group activities and corrective feedback; and

• non-interactive programmes, which are based on delivery of knowledge through mainly didactic teaching.

Some of the more recent reviews have focussed on the effectiveness of individual programme components. Most of the literature concerns substance abuse more generally, rather than alcohol misuse specifically.

Effectiveness

5.3 An updated meta-analysis of drug prevention programmes confirms some previous findings (Tobler et al 2000; Tobler 2000). Considering all areas of substance abuse, interactive programmes are more effective than non-interactive (see table 5.1). For programmes targeting alcohol, however, the difference is not significant. When the analysis is restricted to a subset of high quality programmes (features include random assignment, interventions of 4 hours or more, follow up of 3 months or more, control for pre-existing differences), the difference becomes highly significant (p=0.001).

Table 5.1 Comparison of non-interactive and interactive programmes.

Source: Tobler et al (2000) Table 14.

5.4 The studies reported in Tobler et al (2000) had follow up periods of at least 3 months, with between a quarter and a third of studies having longer term follow up (beyond one year). However, the results were not presented by length of follow up. Skara and Sussman (2003) reviewed adolescent programmes with at least 2 years follow up but only 7 of 25 studies provided results for alcohol (2 of which were also included by Tobler et al 2000). Five of these reported positive outcomes, of which 4 were designed to provide booster sessions.

5.5 Foxcroft et al (2002; 2003) have also reviewed programmes targeted at young people and with longer term follow up but with specific interest in the results relating to alcohol. They included 56 studies in their review, less than half of which were included in the review by Tobler et al (2000). Twenty of the studies showed evidence of ineffectiveness and interventions with short term or medium term results were only partially effective at best. Three longer-term studies reported effective interventions, based on life skills training, a school and community intervention and a family based intervention. Five other studies with longer term follow up, including the 4 year follow up of Project Northland, showed no effect.

5.6 Cuijpers (2002a) reviews evidence relating to specific components of school based drug prevention programmes. Evidence relating to the added effectiveness of booster sessions suggests that the effect depends on other characteristics of the programme. The only study specifically on alcohol misuse showed a positive effect (Shope et al 1992 cited in Cuijpers 2002a). By contrast, community interventions added to school programmes significantly increase the effect. Again, only one of these studies was concerned with alcohol misuse. Studies that directly compare peer leadership and adult leadership for the same drug prevention programmes are reviewed in more detail in a second paper (Cuijpers 2002b). Overall, peer leadership was more effective but with wide variations in the results. In some cases, the adult led programmes were more effective but there were too few studies to identify the other characteristics of programmes or settings which result in peer led programmes being more effective. Only 4 of the 12 studies included alcohol.

5.7 Babor et al (2003) conclude that resistance skills may produce small, short-lived changes in behaviour unless booster sessions are provided. Studies of normative education report small to moderate behavioural changes. Combining school and community interventions appears to have more impact, not least because the community intervention may be successful in restricting access to alcohol by young people.

Population groups

5.8 Most interventions target school age children but the age groups vary and results for particular interventions may not translate to other age groups. The majority of school-based interventions are aimed at the 11-14 age range. There are few studies of programmes aimed at college or university students and no convincing evidence of impact on behaviour (Babor et al 2003).

Settings

5.9 No studies were found comparing alternative settings for the delivery of programmes to the target group.

Providers

5.10 There were no new studies in this area other than comparisons of peers and non peers (see 5.6 above).

OTHER HEALTH PROMOTION INTERVENTIONS

Types of intervention

5.11 Mass media campaigns aim to communicate health promotion messages to the general population through a variety of media. Media campaigns can also form part of enforcement programmes. Community prevention initiatives combine education with community action related to under-age drinking. Given the type of initiatives pursued, there is some overlap with the legislation and enforcement literature but the focus of interest here is the community setting.

Effectiveness

5.12 Derzon and Lipsey (2002) reviewed mass media campaigns targeted at youth. Their findings confirmed the positive effects on attitudes and knowledge but also suggest some positive effect on behaviour with respect to alcohol. Babor et al (2003) review counter-advertising but conclude that there has been too little and the quality has been too poor to draw conclusions about its effectiveness. US warning labels on alcohol have raised awareness of specific health messages but there is little evidence of change in behaviour (Babor et al 2003). Labels are often small and obscure.

5.13 Hingson and Howland (2002) review of community interventions covering studies included previously. These are found to be generally effective. Treno and Lee (2002) provide an overview of community action, which combines media advertising, legislative action and enforcement. The target of such community action is drink related problems rather than problem drinkers, per se. Although problem drinkers have a higher rate of drink related problems, most drink related problems are not caused by problem drinkers.

Population groups

5.14 Mass media campaigns and community actions may be aimed at a general population or aimed at teenagers and minors.

Settings

5.15 No studies were found comparing alternative settings for the delivery of programmes.

Providers

5.16 No studies were found comparing alternative providers for the delivery of programmes.

COST-EFFECTIVENESS

5.17 Spoth et al (2002) carried out an evaluation of two family focussed adolescent substance-use prevention programmes in terms of cost per case of alcohol-use disorder prevented, benefit cost ratio and net benefits per participating family. Schools were randomised to either the Iowa Strengthening Families Progam (ISFP) or the Preparing for the Drug-Free Years Progam (PDFY) or to a minimal contact control group. Both interventions included instruction on parenting skills but the ISFP includes the target child in all sessions and has seven sessions compared with 5 for PDFY.

5.18 Cases of alcohol-use disorder in each group were projected from data relating to age of starting alcohol use. Costs of alcohol-use disorder were estimated from a societal perspective using a human capital approach. This provides a conservative estimate of benefit, as it does not value the health gain per se but only the economic consequences of the gain. Direct and indirect costs of the intervention were included but time costs to participants were not estimated. It was assumed that voluntary participation implied that benefits to the participants were perceived to outweigh costs. Travel and childcare costs were included.

5.19 Analysis was on an intention to treat basis, so included data on those who failed to attend. Students who had not started alcohol use at the end of follow up (age 17) were all assigned an onset age of 18+, which may underestimate differences between groups. Both programmes had positive benefit:cost ratios (9.6 for ISFP and 5.85 for PDFY) and positive net benefits per family ($5923 for ISFP and $2697 for PDFY).

QUALITY AND RELEVANCE OF EVIDENCE

5.20 Concerns remain about the quality of the evaluations and the lack of longer-term follow up. Werch and Owen (2002) provide a review of negative outcomes of substance use prevention problems; i.e. programmes that result in increases in substance use, including alcohol. They identify three basic types of negative effects; studies indicating mixed positive and negative programme effects, where negative outcomes may relate to a subgroup of the young people studied, such as existing users versus non users; studies indicating negative effects together with mainly non significant programme effects, where the problems relate to both theory failure and implementation failure; and finally, but rarely, an isolated negative effect in a predominantly positive programme, which may be an artefact. The relative frequency of such negative effects in the overall literature is not provided but this study serves to emphasise the importance of programme populations, contexts and design in the transference of evidence into practice.

FURTHER RESEARCH

5.21 The research needs identified in the previous review were:

• better designed studies, particularly with respect to the comparison groups;

• translation of effect sizes into outcomes;

• identifying the potential benefit of targeting high-risk groups for health promotion;

• evaluation of the costs and benefits of community interventions;

• establishing the generalisability and sustainability of community programmes;

• development and evaluation of novel interventions in health promotion; and

• consideration of whether there is a cumulative effect of exposure to health promotion messages.

These research requirements remain unchanged.

« Previous | Contents | Next »