Effectiveness of antismoking telephone helpline: follow up surveyBMJ 1997; 314 doi: https://doi.org/10.1136/bmj.314.7091.1371 (Published 10 May 1997) Cite this as: BMJ 1997;314:1371
- Stephen Platt, directora,
- Andrew Tannahill, chief executiveb,
- Jonathan Watson, director of research and evaluationb,
- Elizabeth Fraser, data analystb
- a Research Unit in Health and Behavioural Change Department of Public Health Sciences University of Edinburgh Medical School Edinburgh EH8 9AG
- b Health Education Board for Scotland Woodburn House Edinburgh EH10 4SG
- Correspondence to: Dr Platt
- Accepted 24 January 1997
Objective: To evaluate the effectiveness of an antismoking campaign conducted by the Health Education Board for Scotland.
Design: Descriptive survey of adult callers to a telephone helpline (Smokeline) for stopping smoking; panel study of a random sample of adult callers; assessment of changes in prevalence of smoking in Scotland before and after introduction of the helpline
Setting: Telephone helpline.
Subjects: Callers to Smokeline over the initial one year period. Detailed information was collected on a 10% sample (n=8547). A cohort of adult smokers who called Smokeline (total n=848) was followed up by telephone interview three weeks, six months, and one year after the initial call.
Main outcome measures: Numbers of adult smokers calling helpline; changes in smoking behaviour, especially stopping smoking among cohort members; and changes in prevalence of smoking in the general population.
Results: An estimated 82 782 regular adult smokers made genuine contact with Smokeline over the year, representing about 5.9% of all adult smokers in Scotland. At one year 143 of the cohort of 848 callers (23.6%; 95% confidence interval 20.2% to 27.0%) reported that they had stopped smoking, and 534 (88.0%; 85.4% to 90.6%) reported having made some change. About 19 500 (16 700 to 22 350) adult smokers, equivalent to 1.4% (1.2% to 1.6%) of the mean adult smoking population, stopped smoking with direct help from Smokeline. During the second year of the campaign (1994) smoking prevalence among 25-65 year olds in Scotland was 6% (2.0% to 10.0%) lower than it had been before the start of the campaign.
Conclusion: The Health Education Board for Scotland's antismoking campaign reached a high number of adult smokers, was associated with a highly acceptable quit rate among adults given direct help through Smokeline, and contributed considerably to an accelerated decline in smoking prevalence in Scotland.
There was an unprecedented response to the antismoking campaign of the Health Education Board for Scotland, with an estimated 5.9% of adult smokers in Scotland responding to the invitation to call Smokeline, a free telephone helpline, in its first year of operation
A panel study of callers to Smokeline, with telephone interviewing, obtained an acceptable response rate of 71.6% at one year follow up
Nearly a quarter (23.6%) of smokers who called the helpline were not smoking at the one year follow up, a success rate that exceeds a proposed standard for comparable health education and promotion interventions
The Smokeline campaign contributed considerably to an accelerated (6%) reduction in smoking among people aged 25-65 in Scotland during 1992-4
The study findings provide further evidence of the efficacy of mass media antismoking initiatives with a social support component
Smoking remains the largest single cause of preventable death and serious ill health in Scotland.1 It is estimated that some 10 600 people in Scotland die each year as a result of their smoking,2 and the extent of damage to health related to smoking has been acknowledged in recent national policy statements,3 4 in which smoking is identified as a first order priority for health education. Targets for reduction of smoking in Scotland have been set at 30% (from 30% to 21%) among 12-24 year olds and at 20% (from 40% to 32%) among 25-65 year olds between 1986 and 2000.
Paid advertising has been identified as a leading intervention to promote stopping smoking and as a comparatively cost effective way of reaching most smokers.1 5 Available empirical evidence suggests that mass media campaigns, using either paid or unpaid advertising to reach the general smoking population, can be effective in encouraging and helping people to change their smoking behaviour in some way–for example, to reduce consumption 6 7 8 or stop smoking altogether.5 9 10 11
We report on the evaluation of the Smokeline campaign launched in October 1992 by the Health Education Board for Scotland. We examined the effectiveness of the campaign during its first year of operation in three respects: adult (age ≥16 years) uptake of the service provided; change in smoking behaviour among adult users of the service; and impact on overall prevalence of smoking among adults in Scotland.
The overall aim of the campaign is to make an important contribution to the national effort to reduce the prevalence of cigarette smoking. The more specific campaign objectives are set out in box.
Box 1 Campaign objectives
To remind smokers and their families and friends of the negative consequences of smoking
To challenge complacency about smoking
To motivate and support smokers and their families and friends towards stopping smoking or encouraging and helping others to do so
To provide direct advice, help, and support to smokers and families and friends to enable them to stop smoking
The campaign has three intertwined strands: Smokeline, a telephone service; You Can Stop Smoking, a self help guide to stopping smoking, available exclusively through Smokeline; and mass media advertising, including television, outdoor posters, and press. Further details of all these elements are given in box.
Box 2 Campaign elements
Smokeline is a free interactive telephone service, available from noon to midnight each day. It is staffed by trained counsellors and provides advice and information, including tips on how to give up and stay off cigarettes; general support and encouragement when people want it; and, if required, a free copy of You Can Stop Smoking, for the caller or an adult friend or relative.
You Can Stop Smoking was designed to be a realistic, eye catching, step by step guide that reinforces the smoker's fundamental belief that responsibility for stopping lies squarely with him or her and offers practical advice to increase the likelihood of successfully stopping.
Mass media advertising has been widely used to promote the campaign. Smokeline was launched on 23 October 1992 through two television advertisements. The first of these was a hard hitting, realistic portrayal of a mother, seriously ill in a hospital bed, imploring her young son never to smoke. The second was shown at the end of the same commercial breaks as the first. It was highly positive in tone, featuring a friendly and sympathetic female “counsellor” encouraging smokers and their families to contact Smokeline. The advertisements were shown regularly for four weeks (first burst) and again from weeks 16-23 (second burst).
A further television advertisement, featuring a man who has been told he is dying of a disease related to smoking who expresses his regrets that he will miss seeing his family growing up, was shown intermittently from weeks 46-52 (third burst). Again, it was paired with the positive advertisement.
The television advertising was combined with three striking outdoor posters: two billboard sized, the other smaller. All were displayed intermittently from week 7. In addition, four press advertisements were published intermittently from week 12.
Calls to Smokeline–Information on sex, smoking, nature of call (first or repeat), reason for calling, and source of obtaining Smokeline telephone number was recorded by the counsellor for each interactive call. (A call is recorded as interactive when a counsellor has offered information to a caller or has begun a process of encouraging the caller to clarify his or her request for information.) At the inception of the line it was not anticipated that children would call, and age data were not at first systematically recorded. It became apparent, however, that considerable numbers of callers were aged under 16 years, and all interactive callers were asked to give their ages from the ninth week onwards. Information on smoking habits and history, intentions to stop smoking, and sociodemographic characteristics was sought from a 10% systematic random sample of all calls from adults over the whole 12 month period (n=8547).
Panel study–Behavioural outcomes were assessed by means of a panel study of adult callers. From the 10% sample of adult callers to Smokeline (see above) a group of 970, of whom 848 (87.4%) were current smokers, was randomly selected for follow up at three points in time (three weeks, six months, and 12 months after the initial call). All had consented to participate at the initial interview. Follow up interviews were conducted over the telephone by an independent research team.
Smoking in the general population–Data on prevalence of smoking in Scotland are issued by the Scottish Office Department of Health, drawing on information collected by the Office for National Statistics for the general household survey and survey on smoking among school aged children.
Adult smokers using Smokeline
The best estimate of the total number of interactive calls from adults during the first to the eighth week was 36 060 (out of 37 370). During the 44 remaining weeks of the year there were 59 664 interactive calls from adults. The calculated total for the year was thus 95 724 (actual total for all ages was 129 717). To this total, we adjusted for repeat calls (6%) and calls by non-smokers (8%) to obtain the number of adult callers who were smokers–82 782, representing about 5.9% of all adult smokers in Scotland (estimated at 1.4 million).
Characteristics of callers
We calculated that over the period under review 49 660 (60%) of the 82 782 adult callers to Smokeline were women, compared with 52% of the general adult population of Scotland.12 Callers to Smokeline were younger than the Scottish adult (16-44 years) general population as a whole13 (72% v 54%, respectively), more likely to be unemployed (16% v 10%), and less likely to be home owners (41% v 52%). Over two thirds of callers had tried to stop in the past, while 58% were desperate to stop at the time of their call, and one third claimed that they would do so immediately. The consumption of cigarettes among callers was particularly high, with 56% smoking 20 or more a day (compared with 42% of smokers in the adult Scottish population12).
Change in smoking behaviour
Among the 848 adult smokers in the combined initial panel sample, attrition was evident over the course of the follow up period, although it was not exceptionally high. At one year follow up 607 (71.6%) of the original sample were interviewed again. The group successfully followed up was found to be representative of the original sample in respect of motivational factors (for example, intention to stop smoking), previous successful attempts at stopping and (baseline) cigarette consumption. The groups differed, however, with respect to sociodemographic features: those followed up were significantly more likely to be women, older, and owner occupiers and they were less likely to be unemployed.
Table 1) shows point prevalence measures of smoking status, based on respondents' self reports at each interview. The non-smoking rate among those followed up at one year was 23.6% (95% confidence interval 20.2% to 27.0%). Of those smoking at one year, 42.5% reported having stopped at some point during the preceding six months, 19.5% were smoking less, and 21.6% had switched to a lower tar brand compared with that smoked six months earlier. If we consider a broader measure of any positive action taken with respect to smoking (defined to include non-smoking at follow up, stopping smoking at any time during follow up, reduced cigarette consumption, or switching to lower tar cigarettes) 88.0% (85.4% to 90.6%) had made some behavioural change in a positive direction by the time of the one year follow up (table 2).
Table 3) provides an analysis of time spent during the year as a non-smoker by respondents' smoking status at 12 months. Just over a third (34.3%; 26.4% to 42.2%) of those who were non-smokers at 12 months had stopped smoking for at least 80% of the period. This figure equates to 8.2% (6.0% to 10.4%) of the total sample (data missing for 20) and is taken as the period prevalence measure of stopping smoking.
We estimated the number of adult Scottish smokers who stopped smoking with direct help from Smokeline. Given that 82 782 adult smokers called Smokeline over the one year period, the point prevalence measure of stopping smoking (23.6% (SD 3.4%)) gives a rough total of 19 500 adult smokers (range 16 700-22 350), equivalent to 1.4% (range 1.2%-1.6%) of the adult smoking population. When we applied the period prevalence figure (8.2% (SD 2.2%)) the rough total was 6800 adult smokers (range 5000-8600), equivalent to 0.5% (range 0.4%-0.6%) of the adult smoking population.
Change in smoking prevalence in the general population
Table 4) gives the trends in smoking prevalence among Scottish adults during 1984-94. Among all adults aged ≥16 there was a decline in the percentage of smokers, from 39% in 1984 to 30% in 1994. The decline of 4% (0.7% to 7.3%) between 1992 and 1994 is particularly noteworthy. Among those aged 25-65 the decline was even more pronounced, from 38% in 1992 to 32% in 1994, a reduction of 6% (2% to 10%), thus reaching the national target for this age group six years ahead of the year 2000 deadline.
The Health Education Board for Scotland's campaign seems to have been highly effective in encouraging use of the telephone helpline and facilitating positive behavioural change among adult smokers. At follow up one year after first calling Smokeline, nearly nine in 10 smokers using the service had taken some form of positive action on their smoking; just under a quarter were non-smokers. If we generalise from these findings to the total sample of adult callers to Smokeline we get a one year point prevalence estimate of 19 500 adult smokers (1.4% of the adult smoking population) who stopped smoking with direct help from Smokeline.
Possibility of bias
Confidence in these estimates depends largely on dealing satisfactorily with the problems of bias. The first consideration is possible attrition and other bias in the panel sample. There are no relevant data on the whole Smokeline sample with which to compare the successfully followed up group. The one year sample was found to be representative of the original panel sample in respect of motivational factors (for example, intention to give up), previous successful attempts at stopping, and cigarette consumption, all of which have been shown to be key predictors of stopping smoking in these cohorts.14 There was bias in respect of certain sociodemographic characteristics, but of these only socioeconomic status seems to be a predictor of successfully stopping smoking and by no means one of the strongest.14 On the basis of this evidence we conclude that biased sample attrition is unlikely to be large enough to invalidate our findings, and we defend the use of actual denominators (rather than the original denominator) in calculating rates of stopping smoking and other measures of behavioural change.
Another potential source of bias is the use of uncorroborated self reports to measure current smoking behaviour. Biochemical data on stopping smoking are difficult to obtain in large scale campaigns, without considerably inflating the costs of the intervention. Even with such investment the level of confirmation may be unsatisfactory.15 In any case, there is evidence that self reports of smoking status are fairly accurate.16 17 18
Direct impact of the campaign
The underlying trend in the prevalence of adult smoking during 1975-92 was a decline of about 0.8% a year (47% to 34%).12 19 The reduction in adult smoking between 1992 (before campaign) and 1994 (during campaign) was 4% (table 4)–much more than the underlying reduction in smoking prevalence that is not attributable to an intervention (1.6%). The cumulative direct impact of Smokeline (between 1.4% (all change achieved within the first year and no effect in the second year) and 2.8% (rate of change in the first year continues into the second year)), together with the underlying reduction in smoking prevalence (1.6%) yields an overall estimated reduction of 3.0% to 4.4% (over two years). The actual decrease of 4% lies within this range. There is thus no evidence that the effect of Smokeline on overall smoking prevalence among Scottish adults arose otherwise than through its direct impact on service users.
This discussion assumes that all of the excess reduction in prevalence of smoking can be attributed (directly or indirectly) to the campaign. A well known difficulty in assessing the success of a mass media health education campaign, however, is how to separate out the true impact of the campaign from other extraneous factors. Naturalistic designs (particularly before and after studies without control groups, such as ours) are subject to confounding period effects, which hamper interpretation of the findings. Three such possible confounders have been identified: cigarette advertising, pricing of tobacco, and advertising of products to help people stop smoking.
Possible confounding effects
In comparison with the Health Education Board for Scotland's expenditure of about £550 000 on television, print, and poster advertising during the first year of the adult Smokeline campaign,21 the tobacco companies are estimated to have spent about 10 times this amount (£5.6m) on above the line advertising in Scotland.22 Such advertising is unlikely to have exerted a positive impact on rates of stopping smoking from October 1992 to October 1993. Indeed, it would probably have had the opposite effect.
The cost of a packet of 20 cigarettes rose by 6.5% in March 1993 as a result of increased taxation. This may have had a positive bearing on the rates of stopping smoking at six months and one year.
The marketing of products to stop smoking–for example, transdermal patches–has grown in recent years in response to the demand for practical support from smokers who want to stop. There is no evidence to suggest, however, that their advertising presence correlates with use of Smokeline. Nicorette was the only product advertised during the launch of Smokeline, while total advertising for products to stop smoking remained at a comparatively constant level before and after Smokeline came into existence.22
Comparison with other antismoking interventions
The success of the campaign needs to be assessed against other antismoking interventions, as well as in terms of its overall impact on smoking trends among adults in Scotland. Table 5 presents the baseline, standard, and ultimate standard measures against which, as Flay suggests,9 the results of interventions should be compared; the table also shows the achieved rates of stopping smoking for face to face interventions26 27 and nicotine replacement programmes.28 The outcome of the campaign compares well with these other types of intervention.
The Scottish antismoking campaign provided direct help to an exceptionally high number of adult smokers; use of Smokeline by adults was associated with a one year rate of stopping smoking that compares highly favourably with those for other interventions; and the campaign made a significant contribution to an accelerated reduction in the prevalence of smoking in Scotland. The study provides further evidence of the efficacy of mass media campaigns with a social support component in motivating and enabling positive behavioural change among adult smokers.
Funding: Health Education Board for Scotland.
Conflict of interest: None.