Appendix C: The evidence

This appendix lists evidence statements from one review and a focus group research report provided by external contractors (see appendix A) and links them to the relevant recommendations (see appendix B for the key to study types and quality assessments). The evidence statements are presented here without references – these can be found in the full review (see appendix E for details).It also sets out a brief summary of findings from the economic appraisal and the fieldwork.

Evidence statement number 1 indicates that the linked statement is numbered 1 in the review 'Interventions to prevent the uptake of smoking in children and young people'. Evidence statement FG1 indicates that the linked statement is numbered 1 in the focus group research report 'The prevention of uptake of smoking by children and young people, with reference to the areas of mass media and the sale of tobacco products: findings from a multi-method primary research study'.

The review, focus group research report and economic appraisal are available on the NICE website. Where a recommendation is not directly taken from the evidence statements, but is inferred from the evidence, this is indicated by IDE (inference derived from the evidence) below.

Recommendation 1: evidence statements 1, 1.1, 1.1.1, 1.3.1, 1.3.3, 1.3.4, 1.4.1, 1.7, 1.7.1, 1.7.2, 1.7.3, 1.7.4, FG 2

Recommendation 2: evidence statements 1, 1.3.1, 1.3.3, 1.3.4, 1.3.5, 1.6.1, 1.6.2, 1.7, 1.7.1, 1.7.2, 1.7.3, FG 2, FG 5–8

Recommendation 3: evidence statements 1, 1.3.2, 1.3.3, 1.4, 1.4.1, 1.6, 1.6.1, 1.6.2, 1.8, 1.8.1, FG1–4

Recommendation 4: evidence statements 2, 2.1, 2.1.1, 2.2.1, 2.3, 2.3.1, 2.3.2, 2.4, 2.4.1, 2.5, 2.5.1, 2.6, 2.6.1, 2.7, 2.7.1, 2.7.2, 2.7.3, 2.8, 2.8.1, FG 9

Recommendation 5: evidence statements 2, 2.1, 2.1.1, 2.3, 2.3.1, 2.3.2, 2.4, 2.4.1, 2.5, 2.5.1, 2.6, 2.6.1, 2.7, 2.7.1, 2.7.2, 2.7.3, FG 9

Evidence statements

Evidence statement 1

There is evidence that mass-media campaigns can prevent the uptake of smoking and also influence knowledge, attitudes and intentions of children and young people. Factors that have been shown to influence effectiveness in terms of attitudes, perceptions, beliefs and intentions include message source, message content, message format, message framing, duration, target audience, demographics of the audience, and the site/setting of the campaign. Factors that have been shown to influence effectiveness in terms of smoking behaviour (that is, smoking in the past 30 days, decreased initiation of smoking, quitting, number of cigarettes smoked) include message content, target audience, duration of the mass-media campaign, demographics of the audience, the number of anti-tobacco message sources and the Truth campaign. Overall, the factors outlined above work best when combined with broader tobacco control initiatives produced by tobacco control bodies. Furthermore, campaigns are most effective when they are long in duration and greater in intensity of exposure.

Evidence statement 1.1

Some mass-media interventions are more effective than others. Comparing interventions, prevention campaigns produced by the tobacco industry are less effective than anti-tobacco campaigns produced by tobacco control bodies. Young people perceive industry campaigns to be less effective, less interesting and less engaging. Industry campaigns also 'appeared to move youth attitudes in a pro-tobacco direction'.

Evidence statement 1.1.1

Evidence from one cluster RCT (++) suggests that adolescents perceive tobacco industry sponsored advertisements less favourably and as less effective (that is, participants rated these advertisements as less convincing and less helpful in keeping friends from smoking and starting smoking) in reducing smoking (specifically, fewer people taking up smoking based on the following outcome measures: intention to smoke, curiosity of tobacco use, tobacco industry sympathy) than other smoking prevention advertisements, but also express greater sympathy with the tobacco companies after viewing their advertisements. Yet, neither the industry sponsored nor other prevention advertisements changed adolescent's intention to smoke.

One cross-sectional (+) study found that an American tobacco control campaign did increase anti-tobacco attitudes and beliefs, while an industry-sponsored campaign 'appeared to move youth attitudes in a pro-tobacco direction'. Similarly, one cross-sectional study (++) found that exposure to tobacco industry youth-targeted smoking prevention advertising generally had no beneficial outcomes (measured by young people's attitudes, beliefs and intentions regarding the tobacco industry, and tobacco use 10 months into the Truth campaign). Exposure to tobacco company parent-targeted advertising was associated with lower perceived harm of smoking, stronger approval of smoking, stronger intentions to smoke in the future and greater likelihood of having smoked in the past 30 days. Another (+) US-based cross-sectional study found that tobacco industry advertisements were less interesting, less cognitively engaging, and held less negative emotional appeal for teenagers than advertisements created by tobacco control programmes.

Evidence statement 1.3.1

How an intervention is delivered does influence the attitudes, perceptions and behaviours of young people. Evidence from two (+) reviews found that message content does influence the effectiveness of an intervention (see below), though the impact is not consistent, and also depends on the duration of delivery. One (++) RCT study found that message content could change perceptions of health risk severity and intentions not to smoke, though none of the message themes resulted in: increased self-efficacy for refusing cigarette offers or resisting tobacco marketing, or improved health risk vulnerability. Another (++) RCT study found that using tobacco-related disease messaging was more effective for increasing anti-tobacco attitudes and perceptions of social disapproval risks associated with smoking, whereas anti-industry advertisements did not decrease young people's intention to smoke.

Evidence from a US cross-sectional (+) study found that 'truth' messages were effective in decreasing and preventing smoking in young people (Florida teens were less likely to smoke in the past 30 days, to have ever tried smoking, or to indicate that they could not rule out the possibility of smoking in the future).

A UK-based (++) qualitative study found that social norms messages were more effective than fear messages at encouraging more committed smokers to consider their smoking behaviours and reinforcing awareness of the dangers of smoking in less committed smokers. 'Industry manipulation advertisements' were aesthetically appealing but ineffective for preventing the uptake of smoking. Similarly, one (+) review and one RCT (+) study concludes that anti-smoking advertisements can improve smoking prevention and cessation in the young (by making them less likely to smoke, have lower intentions to smoke, and have greater intentions to quit smoking), but the specific outcomes of any message type depends on the context and the values that the audience associates with smoking.

Applicability: most of the studies were conducted in the USA. It is not clear if these findings are directly applicable to the UK since the mass-media campaigns under investigation are specific to the USA. Furthermore, demographics of participants are different from those in the UK. International review data may be broadly applicable to the UK since the review is international in scope.

Evidence statement 1.3.2

Studies analysed the effectiveness of a variety of mass-media formats. One cross-sectional (-) study found that television advertisements were recalled more often than other formats and that viewing the advertisements increased intention to quit, though did not affect actual quit attempts. Evidence from one qualitative (+) study indicates that young people deemed websites as effective in obtaining information on smoking, if they incorporated: interactivity, expert-trusted guidance, and appealing graphics. One (+) cross-sectional study reveals that youth-led tobacco prevention movements and intensive counter-marketing media campaigns can be effective in preventing the uptake of smoking and 'generating negative attitudes about the [tobacco] industry'.

Applicability: all three studies were conducted in the USA. Given that the findings are in response to specific USA interventions, it is not clear if findings are applicable to the UK.

Evidence statement 1.3.3

Evidence from one cross-sectional (+) study and one (+) review suggest that adult-focused or general population campaigns are successful in reducing smoking (cutting down the number of cigarettes smoked, increasing the numbers attempting to quit, making it easier to stay a non-smoker) in young people. Yet, one (+) review contends that both messages aimed at young people and general messages can be effective in developing awareness, and changing attitudes and behaviours associated with tobacco use, as long as messages are not deemed patronising by the young.

Applicability: no studies were conducted in the UK. It is not clear if the findings are directly relevant to the UK.

Evidence statement 1.3.4

One RCT (+) found that message framing impacts the effectiveness of an intervention by lowering intentions to smoke, lowering the perceived pharamacological benefits of smoking, and lowering the perceived psychological benefits of smoking. In particular, it is important that the message framing is consistent with the desired outcome.

Applicability: given the broad cultural differences between South Korea and the UK, the findings of this study are less relevant to the UK.

Evidence statement 1.3.5

One (+) review contends that effective messaging should attend to all elements (such as content, format and tone). Specifically, evidence from one cross-sectional (+) study suggests that message processing in older teens improves when messages incorporate unrelated cuts and use suspenseful images. One cross-sectional study (+) found sources were evaluated more positively for implicit rather than explicit messages, and for anti-smoking rather than pro-smoking messages. Evidence from an RCT (++) study reveals that exposure to cigarette advertisements depicting young people can decrease negative stereotypic beliefs about smoking and increase an intention to smoke in the young.

Applicability: the demographics of study participants and the mass-media interventions under investigation are specific to the USA. It is not clear if findings are applicable to the UK.

Evidence statement 1.4

There was a lack of information regarding whether the effectiveness of a mass-media intervention depends on the status of the person delivering it. However, evidence indicates that young people who receive anti-smoking messages from a variety of sources (for example, family, friends, internet, sporting events), as opposed to only a few, are more likely to refuse tobacco.

Evidence statement 1.4.1

No studies specifically discussed how the status of a person delivering an intervention can have an impact on its effectiveness. Yet, one cross-sectional study (+) and one (+) review reveal that young people who are exposed to a large variety of anti-tobacco sources are more likely to refuse tobacco, and that social interactions can support anti-tobacco messaging. Evidence from two cross-sectional studies (+) indicates that the tobacco industry is not a trusted source of anti-tobacco information among young people.

Applicability: it is not clear if the findings are directly applicable to the UK as they are USA-based. However, international review data may be broadly applicable, since multiple studies have produced similar results. Given the differences in demographics of study participants and the interventions under investigation it is not clear if findings are directly applicable to the UK.

Evidence statement 1.6

The duration of a mass-media intervention influences its effect. Increased exposure to anti-tobacco messages over time decreases intent to smoke and smoking initiation, meanwhile, increasing negative attitudes towards the tobacco industry.

Evidence statement 1.6.1

Evidence from one (++) Cochrane review suggests that the duration of an intervention will have the greatest bearing on health behaviours. In support of this, evidence from three cross-sectional studies (one [++] and two [+]) identified by the literature search, reveals that increased exposure to anti-smoking advertisements over time results in a decrease in: young people smoking in the past 30 days (compared to those in markets with no exposure to state-sponsored anti-tobacco laws), intent to smoke, initiation of smoking, enhanced perception of risk, and negative attitudes about smoking.

Similarly, two cross-sectional (+)US studies demonstrate that young people living in states with aggressive counter-industry media campaigns are more likely to have 'negative beliefs about tobacco industry practices', are less likely to smoke, and are more informed about the dangers of second-hand smoke. As well, one (+) cohort study found that pro-tobacco media increased susceptibility to smoking, while anti-tobacco media decreased susceptibility. Conversely, one (++) US-based cross-sectional study did not find a relationship between exposure to anti-smoking campaigns and improved ideas about smoking or health behaviours. They argue that in order to be effective, exposure must be supported by other tobacco control initiatives. A cross-sectional (++) study found increased exposure to anti-tobacco mass-media messages in the absence of school-based tobacco prevention measures was not successful in reducing tobacco use among adolescents.

Applicability: none of the studies were conducted in the UK. However, given the nature of exposure to mass-media campaigns, findings may be applicable to the UK.

Evidence statement 1.6.2

Results from four cross-sectional studies (two [++] and two [+]) indicate that the Truth campaign was successful in improving the prevention of youth smoking over time. Studies show that the campaign resulted in: decreased prevalence rates of smoking in young people (through reduced uptake and/or increased quitting), greater agreement with anti-smoking statements by young people, and stronger anti-tobacco attitudes and beliefs.

Applicability: The Truth campaign is a USA anti-tobacco mass-media campaign. Due to the nature of the campaign and the demographics of US young people, results are not directly relevant to the UK.

Evidence statement 1.7

Effectiveness may vary according to a variety of demographic factors. Mass-media campaigns appear to benefit younger children more than their older counterparts. However, findings regarding the impact of sex and ethnicity are inconclusive. Mass-media messages and themes are received differently depending on age, sex, and ethnicity. There was a lack of information regarding the impact of socioeconomic status. A variety of other individual characteristics can also impact effectiveness.

Evidence statement 1.7.1

Several studies discuss sex and gender-based differences in the effectiveness of media interventions. One RCT (+) found that for girls, cosmetic advertisements had a greater impact on smoking behaviour (including how often they smoked, how long they have been smoking for and the number of cigarettes smoked) and intentions to quit; while health ads had a greater impact on the smoking behaviour of boys (including how often they smoked, how long they have been smoking for and the number of cigarettes smoked) and intentions to quit. Health advertisements were also most useful for reducing girls and boys intention to start smoking. Evidence from one (+) cohort study found that over time, boys were more susceptible (expressed greater interest in smoking uptake) to smoking than girls. One (3 +) cross-sectional study found no gender differences in the effectiveness of an anti-smoking campaign. A cross-sectional (-) study found that while awareness was similar for girls and boys, girls had a greater recall of anti-tobacco messaging. In a (+) cross-sectional study based in Norway, girls demonstrated a stronger behavioural response (reporting that the campaign had affected their beliefs or decisions concerning smoking) to an anti-smoking media campaign that was targeted at girls.

Applicability: none of these studies were conducted in the UK. It is not clear if the findings are directly relevant, as gender is culturally defined and prescribed.

Evidence statement 1.7.2

Evidence from one review (+), one US-based cohort study (+), and four cross-sectional (two [++], one [+], and one [−]) studies reveals that for younger children, media campaigns are more likely to decrease intentions to smoke and improve smoking behaviour by decreasing initiation rates and continuation of current smoking. Similarly, one review (+) suggests that those close to the minimum legal age (older youth) are less affected by anti-tobacco industry campaigns since they have the least awareness of, and receptivity to, mass-media messages. In order to target this group, they suggest using campaigns that appeal to the general population, rather than just young people.

Conversely, one cross-sectional study (+) found that older youth demonstrated greater change in behavioural intentions after exposure to a media campaign. Also, one cross-sectional (+) study testing emotional reactions to smoking advertisements, found only a weak relationship between age and response.

Evidence from one RCT study (+) found that message content differentially impacts the outcomes of the campaign (how often young people smoke, number of cigarettes smoked, intentions to start smoking, and intentions to quit), depending on the age of the students. In general, health messages were more effective in changing smoking behaviour (how often young people smoke, how long they have been smoking, and the number of cigarettes smoked), intention to start smoking and intention to quit smoking for older students. Cosmetic messages were more effective in changing smoking behaviour (how often young people smoke and the number of cigarettes smoked) for younger students. In another RCT (+) study, the investigators also concluded that age and message types have a statistically significant impact on the interpretation of tobacco-related messages. Older youth were less likely to positively accept explicit anti- or pro-tobacco messages that limited their internalised decision making, compared to younger children.

Applicability: none of these studies were conducted in the UK. It is not clear if findings are directly relevant.

Evidence statement 1.7.3

A variety of studies explored the impact of ethnicity on the effectiveness of youth interventions. One (++) cross-sectional study revealed that African Americans and Hispanics were more affected (defined as the level to which young people reported advertising had made them less likely to smoke cigarettes) by anti-smoking messaging than white young people. Evidence from one cross-sectional (+) study found no relationship between ethnicity and emotional reaction to anti-smoking messages. Finally, one (+) cross-sectional study found that a web-based tobacco prevention programme had a greater impact on intentions not to smoke among Hispanic and white students than black students.

Applicability: as these studies deal with specific populations in the USA, it is unclear how applicable these findings are to a UK setting.

Evidence statement 1.7.4

One cross-sectional (+) study found that a number of variables were associated with a greater intention to smoke, including: brand recognition, willingness to use or wear products with tobacco brands, stress and having friends who smoke. Having a live-in father who smoked, and agreeing with anti-tobacco ads were both associated with a lesser intention to smoke. Evidence from one cross-sectional (+) study found that young people who smoked demonstrated a greater awareness of the pervasiveness of anti-smoking campaigns than among young people who had never smoked or who were susceptible to smoking.

Applicability: as neither of the studies were conducted in the UK it is not clear if findings are directly relevant.

Evidence statement 1.8

Lack of exposure and longevity are barriers to effective mass-media interventions.

Evidence statement 1.8.1

No studies specifically examined facilitators or barriers to the implementation of mass-media interventions. Yet, two (+) reviews suggest that mass-media interventions are most effective when they are longer in duration and greater in intensity of exposure. One review cites the guidelines developed by the Centre for Disease Control which recommend that advertisements should be aired for a minimum of 6 months to affect awareness and up to 24 months to have an impact on behaviours; advertisements should also be aired as frequently as possible, particularly within the first 6 months of a campaign. The other review contends that mass-media interventions should be large, intense and of 'sufficient duration' but the duration or intensity have not been explicitly defined.

Applicability: both studies were conducted in the USA. However, given the nature of exposure to mass-media campaigns findings may be applicable to the UK.

Evidence statement 2

There is evidence that access restriction interventions impact effectiveness in terms of the number of sales to young people, young people's ability to access cigarettes and store clerk compliance. There was a lack of information regarding whether interventions impact behaviours, attitudes, beliefs, intentions or perceptions. Only two studies addressed the impact of interventions on smoking behaviour. Factors that have been shown to influence number of sales, young people's ability to access cigarettes and store clerk compliance include active enforcement, comprehensive interventions, interventions produced by tobacco control bodies, requesting age/proof of ID, demographics of the vendor/store clerk, site/setting of the access intervention, and the demographics of the target audience. Overall, the factors outlined above work best when combined with requesting proof of age/ID, active enforcement (in relation to both retailer-youth purchaser and trading standards-retailers) and other youth prevention strategies.

Evidence statement 2.1

Some access restrictions appear to be more effective than others. Compared to interventions created by tobacco control bodies, interventions produced by the tobacco industry do not decrease the sale of tobacco to young people. Store clerks participating in the tobacco industry intervention were still willing to illegally sell tobacco to children even after state mandated warnings were issued.

Evidence Statement 2.1.1

One cross-sectional (−) article found that a tobacco industry sponsored campaign in the US did not significantly reduce the sale of tobacco to minors, yet state mandated warnings were only slightly more successful in reducing young people's ability to purchase tobacco. Tobacco industry interventions may not prevent the illegal sale of tobacco to children and young people; active enforcement of tobacco sales laws by health officials may be more effective.

Applicability: findings are not applicable to the UK since the findings are specific to a US-based tobacco industry campaign.

Evidence Statement 2.2.1

No studies in the review examined whether interventions were delaying rather than preventing the onset of smoking. For the most part, studies examined the effect of access restrictions on illegal sales (for example number of sales to youth, merchant compliance) not the effect on behaviour or prevention of uptake. One US-based cross-sectional study (+) did find that interventions impacted youth's stage of smoking uptake. Stage of smoking uptake was rated on a continuum of 1 to 5, with stage 1 being someone who has never smoked and has no intention to smoke, and stage 5 being someone who currently smokes, has smoked at least 100 cigarettes and has no intention to quit. Evidence from this study suggests that compliance with youth access laws reduces the probability of being in higher stages of smoking. Youth who are in earlier stages of smoking depend more on social sources for acquiring tobacco. Interestingly, evidence from one American review (+) shows no difference in youth smoking rates between communities with and without greater merchant compliance with sales restrictions.

Applicability: the findings are in relation to two US-specific interventions. It is not clear if findings are directly applicable to the UK.

Evidence statement 2.3

The way in which an intervention is delivered does influence effectiveness. There is strong evidence that comprehensive interventions are more effective than individual restrictions alone. Furthermore, active enforcement and requesting age/ID can also decrease sales of tobacco. Similar findings were highlighted from English survey data.

Evidence statement 2.3.1

One (++) Cochrane review and one US-based cross-sectional study (+) found that multi-faceted interventions (active enforcement, multi-component educational strategies, and increased taxing and restrictions on smoking in public places respectively) are most effective for reducing youth's ability to access tobacco, particularly when combined with ongoing and active enforcement of minimum age restrictions. Similarly, English survey data indicates that a broad set of actions is the key to successfully increasing compliance with minimum age laws. Active law enforcement has been identified by one review (+) and two cross-sectional studies (−) as an important part of multi-component interventions. Evidence from one review (+) suggests that vending machine policies are most effective at reducing youth access to tobacco when combined with locking devices or complete vending machine bans.

Applicability: the majority of the studies took place outside of the UK in a wide range of countries, including Australia, the USA and New Zealand. However, it is likely that their findings are applicable to the UK, given the broad similarities in the impact of enforcement.

Evidence statement 2.3.2

Two cross-sectional (+) US-based studies found that when store clerks requested proof of age, illegal sales decreased. There is some evidence that asking for identification decreases illegal sales more than asking for age. Yet evidence from a non-RCT study (+) in the US suggests that minors who present ID are more successful when purchasing tobacco than those who do not. Therefore, while cashier compliance with enforcing age restrictions can decrease young people's ability to purchase tobacco, evidence suggests that this will be most effective when stringent verification of ID occurs.

Applicability: as none of these studies were conducted in the UK it is not clear if findings are directly applicable.

Evidence statement 2.4

The status of the person delivering an access restriction does impact on effectiveness. The age, gender and ethnicity of shop assistants selling tobacco appear to influence sales to young people.

Evidence statement 2.4.1

In one cross-sectional study (+), store clerks participating in a compliance programme were as likely to make illegal sales of tobacco to young people as store clerks who were not participating in the programme. However, US-based evidence from one (+) non-RCT and two cross-sectional (+) studies suggests that the age, gender and ethnicity of the person delivering an intervention influences the outcomes. Overall, younger store clerks are more likely to sell tobacco illegally to a minor, identification is less likely to be requested and an illegal sale is more likely to occur when the store clerk is a man. Some evidence also suggests that ethnicity may influence intervention outcomes; Asian clerks were found more likely to request age, with white store clerks most often requesting identification.

Applicability: all four studies were conducted in the USA. It is not clear if findings are applicable to the UK.

Evidence statement 2.5

Evidence shows that the site/setting does influence effectiveness. Based on English survey data, young people are successful at buying tobacco in a variety of locations including newsagents, tobacconists or sweet shops. Similar findings were highlighted by US studies which found that young people buy cigarettes from convenience stores, gas stations and food stores. One Tasmanian study also found that youth are successful in purchasing cigarettes from a variety of locations, including: service stations, supermarkets and corner stores.

Evidence statement 2.5.1

Evidence shows that site/setting does influence the effectiveness of the intervention, and youth's ability to purchase tobacco. Evidence from one cross-sectional (+) study in Sweden indicates that younger looking adolescents were most successful when purchasing tobacco from newsstands, tobacconists and service stations (compared to department stores, grocery stores, cafes, restaurants, and video rental shops). Survey data from England indicates that young people close to the legal purchase age (older young people) are more successful at purchasing cigarettes than their younger counterparts. Another cross-sectional study (++) in the US found that minors were most successful at purchasing tobacco in convenience stores, followed by gas stations and food stores.

One Tasmanian cross-sectional (+) study also found that youth are successful in purchasing cigarettes from a variety of locations, including: service stations, supermarkets and corner stores. Survey data from England similarly indicates that young people often buy cigarettes from newsagents, tobacconists or sweet shops. The availability of tobacco vending machines also influences access to tobacco. Two (+) cross-sectional studies based in the US, found that young people were more successful when purchasing tobacco from unlocked vending machines or self-service displays than from locked vending machines or over-the-counter outlets.

Applicability: all five studies took place outside of the UK. However, it is likely that their findings are applicable to the UK given the broad similarities in the locations where young people purchase cigarettes.

Evidence statement 2.6

The duration of access restrictions may impact effectiveness. There is some evidence that compliance with access restrictions increases over time. However, effectiveness may not be self-sustainable and may be impacted by social sources of tobacco.

Evidence statement 2.6.1

No studies in the review directly studied the intensity of interventions, although some did examine the impact of an intervention over time. Evidence from two (+) cross-sectional studies indicate that over time (between 2001 and 2003, and between 1996 and 2005 respectively) factors such as successive retail inspections, public prosecutions, awareness of campaigns and implementing a minimum-age law, result in decreased illegal sales of tobacco. Yet, evidence from one (+) review demonstrates that the effectiveness of access restrictions on purchasing tobacco may depend on the level of implementation (level of fines, rate of compliance, community involvement). Lastly, according to evidence from a (+) empirical review, interventions may not produce a sustained decrease in the illegal sale of tobacco. The authors do not specify the impact of the interventions on duration of effect; they only state that interventions without compliance checks, significant penalties, and store clerk awareness have limited long-term effects. Similarly, findings from one (+) cross-sectional study in Tasmania showed a decrease in non-compliance over time.

Applicability: all five studies took place outside of the UK. As a result, it is not clear if findings are directly applicable.

Evidence statement 2.7

The effectiveness of access restrictions is affected by a variety of demographic variables. Those close to the legal minimum age (older youth) and more established smokers (who are also likely to be older) are more successful at purchasing tobacco. Although there were mixed findings regarding the impact of sex, findings from a strong piece of evidence indicate that boys are more successful than girls at purchasing tobacco. However, English survey data indicates that girls are more likely to try and buy cigarettes. However, refusal rates, and therefore purchasing success rates, are similar for boys and girls. The ethnicity of the young person influenced whether or not age/ID was requested. There was a lack of information regarding the impact of socioeconomic status.

Evidence statement 2.7.1

Access restrictions on the sale of tobacco have an impact on people who smoke in different ways, depending on their age and smoking status. Evidence from one (++) Cochrane review reveals that regular smokers encounter access restrictions on the sale of tobacco more frequently, but also employ more techniques to obtain cigarettes – such as presenting fake ID or lying about their age. One Australian-based cross-sectional (−) study found that retailer compliance resulted in the greatest decrease in smoking behaviour for younger and less experienced smokers. For example, the number of regular smokers decreased, the number of young people reporting at least monthly smoking decreased and the frequency of smoking decreased. Similarly, there is some US-based evidence from one (+) cross-sectional study, one (4 +) non-randomised controlled trial study, and one (++) cross-sectional study that young people close to the legal minimum age (older youth) are more successful in purchasing tobacco. Some evidence also suggests that the youth's age or appearance affects their ability to purchase tobacco. Two (+) cross-sectional studies and survey data from England found that young people who appear older are more successful in purchasing tobacco than those who look younger.

Applicability: although all of these studies took place outside of the UK, it is likely that their findings are applicable to the UK, given the outcomes being measured.

Evidence statement 2.7.2

Evidence from one US cross-sectional study (++) found that males had greater purchasing success rates. English survey data indicates that girls try to purchase cigarettes more than boys, however, refusal rates and therefore purchasing success rates, are similar. Evidence from two (+) Swedish cross-sectional studies indicate that boys were more successful in purchasing tobacco, both before and after minimum age restrictions were applied. Conversely, one US (+) cross-sectional study suggests girls are more successful in buying tobacco and one (+) cross-sectional study found that girls were more frequently asked to present ID when attempting to buy cigarettes. Some evidence also suggests that requesting ID results in the greatest reduction of girl's access to purchasing cigarettes.

Applicability: all five studies took place outside the UK. Furthermore, some evidence is not consistent with English survey data. Findings may not be directly relevant to the UK.

Evidence statement 2.7.3

Evidence indicates that ethnicity influences the ability to buy tobacco among young people. One US (+) cross-sectional study found that African American children, followed by Latino and white children respectively, were more likely to be asked for ID when attempting to purchase cigarettes. ID requests resulted in the greatest reduction of African American children's success in purchasing cigarettes. The authors do not indicate whether or not ID requests resulted in a reduction of purchasing success for Hispanic or white youths. One US-based (+) cross-sectional study found that tobacco policies impact young people differently. Evidence shows that smoking rates for white male young people are more responsive to anti-tobacco activities and clean indoor restrictions, while young black males are more influenced by smoking protection and youth access laws (that is, purchasing restrictions).

Applicability: as these studies deal with specific populations in the USA, it is unclear how applicable these findings are to a UK setting.

Evidence statement 2.8

Acquiring tobacco from social sources and lack of enforcement are barriers to the effective implementation of access restrictions.

Evidence statement 2.8.1

Two key barriers to the implementation of access restrictions on purchasing tobacco were identified. Evidence from three (+) reviews and one (++) review indicates that access restrictions are impeded by a young person's ability to access tobacco products from social sources including friends, family, and strangers. English survey data reveals similar findings. Furthermore, evidence from one (+) cross-sectional study based in the USA shows that weak enforcement of laws and policies creates a barrier to the effective reduction of the number of young people who smoke. In particular, minimum age restrictions are not well enforced.

Applicability: although the studies were conducted in the USA, their results are likely to be broadly applicable to the UK setting.

Evidence statement FG1

On the basis of young people's recognition of the format, television campaigns should be continued to be funded as part of comprehensive prevention and cessation campaigns.

Evidence statement FG2

There was evidence to suggest that national smoking prevention campaigns with both adult and young person-oriented messages would be successful approaches for reducing smoking.

Evidence statement FG3

Health promotion campaigns using the Internet will benefit from cutting-edge design and programming.

Evidence statement FG4

Social networking and communication sites may be useful hosts of electronic smoking prevention interventions. However, these should be well designed 'click-through adverts' with clear NHS branding, rather than dedicated pages within the sites.

Evidence statement FG5

Despite similar levels of smoking knowledge, current smokers had more positive smoking attitudes, and were less likely to believe that prevention campaigns could be effective. Smoking cessation and prevention campaigns are therefore likely to have differential effects, depending upon current smoking status. Content should be altered depending upon whether the aim of the intervention is to prevent uptake, delay uptake, or promote cessation.

Evidence statement FG6

From the results obtained in this sample, male smokers may be most resistant to attempts to persuade them to change their smoking behaviours.

Evidence statement FG7

If asked to express a preference, young people tend to value 'socially desirable' traditional intervention techniques (that is, fear arousal/'shock tactics') rather than evidence-based approaches. Some campaign elements should therefore proceed in opposition to young people's preferences.

Evidence statement FG8

Young people would prefer campaigns to be delivered by well-known individuals with personal smoking stories.

Evidence statement FG9

Young people aged under 18 are able to obtain cigarettes from a wide variety of sources that circumvent legal controls. Proof of age schemes will not be effective for young people who obtain contraband or illegally imported cigarettes. Furthermore, young people are able to purchase cigarettes online with minimum information checking by retailers. Proof of age schemes need to be supported by test purchasing and enforcement.

Cost-effectiveness evidence

Overall, mass-media campaigns and age restrictions on the sale of tobacco were found to be cost effective.

The review of economic evaluations identified one study (raising the legal age of smoking) which was estimated to be cost-saving. Another four studies were deemed to be cost effective (that is, they were estimated to be well below £20,000 to £30,000 per quality-adjusted life year [QALY]). The latter four studies comprised: two mass-media campaigns, an age-enforcement programme and a multi-component tobacco control programme involving a school, the media and the community.

A cost-effective modelling analysis for both mass-media and point-of-sale interventions came to similar conclusions: both were estimated to be a cost effective or very cost effective use of resources. (For further details, see 'Cost-effectiveness of a mass media campaign and a point-of-sale intervention to prevent the uptake of smoking in children and young people: economic modelling report'.)

The main limitations of the modelling analyses concerned uncertainty about how many children and young people were prevented from taking up smoking – and how long the effect of the interventions last.

Fieldwork findings

Fieldwork aimed to test the relevance, usefulness and the feasibility of implementing the recommendations and the findings were considered by PHIAC in developing the final recommendations. For details, go to the fieldwork section in appendix B and online.

Fieldwork participants were generally positive about the recommendations and their potential to help prevent the uptake of smoking among children and young people. Many stated that the recommendations would lead to a more concerted effort to prevent under-age smoking. This, they said, was not generally considered a high priority locally, regionally, or nationally (for example, compared to activity to combat under-age drinking). Some primary care trust (PCT) staff felt they needed the support of other local strategic partners, including local authority trading standards teams, the police and local magistrates; they welcomed their inclusion in the recommendations.

Several participants thought the recommendations should become part of PCT or local authority regulatory assessments, or should be enshrined in national standards and targets. This, they felt, would ensure statutory bodies take the action needed to prevent illegal sales and to discourage children and young people from smoking.

However, some participants believed adults who smoke should remain the top priority. They believed that a reduction in smoking prevalence among adults would lead to a similar reduction in smoking prevalence among children and young people.

Some retailers expressed concerns about some of the recommendations relating to them. They acknowledged that they should not sell tobacco products to young people under 18 and should ask for proof-of-age to prevent illegal sales. However, they also pointed out that staff may be reluctant to ask for proof-of-age due to the risk of abuse from customers. Trading standards officers said they could and do work with retailers to help them ask for proof-of-age in a way that reduces this risk.

Overall, participants thought the draft recommendations offered a useful checklist for preventing illegal sales and how to set up and run mass-media campaigns to discourage children and young people from smoking. Many thought they would help to prevent smoking among children and young people under 18 (albeit as part of a wider programme).