Results

CHARACTERISTICS OF CALLERS

Table 1 shows the sociodemographic profile for a representative sample of adult smokers in England and callers to the Quitline. Compared with all smokers, callers were more likely to be women, to be in the age groups 25–34 or 35–44 years, to come from households with children under the age of 16 years, and to be heavy smokers (smoke 20 or more cigarettes a day). At one year the social class profile of callers to the helpline reflected the social class profile of all adult smokers; 63% of the sample were of manual occupations or unemployed compared with 61% of the adult smoker population.

Some differences emerged between the caller profiles at baseline and one year recall. Compared with callers at baseline, women, those aged 35 and over, and those with moderate consumption levels (10–19 cigarettes a day) were overrepresented in the one year recall sample. The one year recall sample also included more long term smokers.

CHANGE IN SMOKING STATUS

Among smokers at baseline, 23% reported having stopped smoking at the two month follow up while 22% reported having stopped at one year. A further 35% of smokers reported that they had reduced their consumption at the two month recall compared with 29% at the one year recall. Among ex-smokers, 57% reported still not smoking at two months while 41% reported still not smoking at one year (table 2). Of those who had resumed smoking at one year 28% were smoking less than they had been initially.

To estimate conservatively the impact of the helpline it is assumed: first, that all those who refuse to take part in the one year follow up are smokers at one year; and second, that 20% of all those who report not smoking at one year fail biochemical validation.11 Of those who refused to participate in the one year recall 61 were smokers at baseline and 28 were ex-smokers at baseline. The adjustments yield a cessation rate of 15.6% (95% CI 12.7% to 18.5%) at one year among callers who were smokers at baseline and 29% (95% CI 23.2% to 34.8%) among ex-smokers at baseline.

Of those not smoking at the one year recall stage, 31% had not smoked since their initial call and a further 30% had not smoked for over nine months of that time. Of the group who had smoked since their initial call but were not smoking at one year the average number of quit attempts was 2.0 (61% giving up just once and 7% over four times).

Of those smoking at one year, 67% reported having stopped at some point over the past year. The vast majority (97%) of those making a quit attempt had been able to give up for over 24 hours on at least one occasion, with the average period of the last cessation attempt being around 22 weeks. A total of 3% claimed not to have smoked for over nine months of that time, and 41% claimed to have stopped for up to a month. Among those quitting at all the average number of quit attempts was 2.2 (51% had tried once and 10% over four times).

The average length of time non-smokers at one year claimed to have not been smoking since their initial call was 39.6 weeks compared to 6.9 weeks for those who were smoking at one year recall.

Looking in more detail at those who were not smoking at one year, among those who were smokers at their initial call 21% had not smoked since that call, a further 34% had not smoked for over nine months and 10% claimed to have stopped for up to a month. Among ex-smokers initially 43% had not smoked since their initial call, 24% had not smoked for over nine months, and 6% claimed to have stopped for up to a month.

Table 3 shows some of the characteristics associated with smoking behaviour changes among callers who were smokers at the initial call. The variables included in this table were predetermined by the published literature on factors predictive of smoking cessation. In this study an increase in the likelihood of quitting appeared to be associated with an increase in age and a decrease in consumption level.

Among ex-smokers initially, the main characteristics associated with not smoking at one year were social classes ABC1, light smokers, and being male.

The behaviour change reported at one year appeared to be influenced by whether the caller received a two month recall interview. Thus among smokers at baseline, 24% of those who received the two month recall reported not smoking at one year compared with 18% who had not received the two month recall (χ² = 3.123, narrowly missed significance at p < 0.05). Among ex-smokers at baseline, the difference between those who did and those who did not receive a two month recall was not significant (43% and 38%, respectively) but the base was very small. A similar effect was also observed for length of quit attempt, number of quit attempts, and confidence in remaining a non-smoker. Thus, among those recalled at two months, the average length of a quit attempt was 17.5 weeks compared with 13 weeks for those not recalled. The average number of quit attempts was 2.3 for those recalled at two months and 1.5 for those not recalled. Similarly, callers who received the two month recall were more confident about remaining non-smokers than callers who had not received the call back (66% and 50%, respectively).

Among those not smoking at one year, levels of confidence in remaining a non-smoker were similar to those reported at two months, with 61% reporting that they felt very confident. However, motivation to stop smoking altogether among callers who were smoking at one year was somewhat lower than that reported at the two month stage (68% want to stop very much indeed at two months compared with 57% at one year). Not surprisingly, a major difference was observed between those smoking and not smoking at one year in terms of their expectation to be smoking in three months' time (61% of smokers felt they would compared with 2% of those not smoking).

Currently, Quitline receives around 500 000 calls a year, of which just over half (268 626) are transferred to counsellors. Of these counselling calls, 71.4% are received during the three months of the advertising campaign. During this period 93% of callers are phoning for themselves, of whom 73% are smokers and 27% are ex-smokers at their initial call.15 Applying the unadjusted figure for stopping smoking at one year to the total number of smokers calling the service for themselves and who speak to a counsellor yields a total of 40 121 (33 556–46 614) ex-smokers at one year. The figure for ex-smokers who stay stopped is 27 655 (23 136–32 175). When adjusted for refusals and an estimated 20% failed validation, the conservative estimate for smokers who stop at one year is 28 450 (23 161–33 738) and for ex-smokers who stay stopped is 19 561 (15 649–23 473).

Discussion

The results indicate that the Health Education Authority's advertising campaign has been extremely successful in generating calls to Quitline and thus encouraging activity directed towards cessation. More than two fifths of all the calls made to Quitline in one year were received during the three month advertising campaign, and of these more than four out of five were answered by counsellors. With an annual rate of half a million callers, 93% of whom are calling for themselves, the Quitline is currently reaching more than 4% of the total population of adult smokers in England, an impressive achievement for a single intervention. Given the diverse backgrounds of the callers these data provide strong support for the feasibility and potential of the Quitline as a model for public health intervention programs.

The Quitline was first advertised as part of a national anti-smoking campaign in England in December 1994. In the present study around 60% of callers claimed advertising as the source of Quitline awareness. The content of the advertising (for example, supportive or hardhitting) and the advertising mix (for example, television, or television, radio and magazine) appeared not to have influenced call volumes. Compared with previous campaigns overall call volumes have remained more or less the same despite substantial changes to the content and mix of advertising. The constant media spend across different advertising mixes and changes to content may explain this.15 A crucial factor in achieving the reach observed in this study appears to be the advertising of a freephone number. When Quitline was relaunched with a freephone number in November 1995, call volumes increased nearly nine fold.15 The use of a freephone number may also be an important factor in the diverse reach of the Quitline.

Compared with the population of adult smokers in England, callers were more likely to be women, older, more heavily addicted, and have children under the age of 16 years living in the household. The social class distribution of callers to the Quitline reflected the social class distribution of smoking in the population, with nearly two thirds of callers being in manual occupations or unemployed. The profile of callers suggests that the service is successful in attracting some of the groups highlighted in the UK government's recent White Paper on tobacco, “Smoking kills”.2 Given that one fifth of the smokers who called Quitline who were in manual occupations or unemployed reported having stopped at one year, it seems likely that such a service can make a major contribution to achieving smoking reductions among these priority groups.

Currently, 54% of callers are transferred to counsellors with an associated success rate of 15.6% (adjusted estimate) among smokers and 29% among ex-smokers (adjusted estimate). A fourfold increase in calls to counsellors from smokers and ex-smokers (from 3.1% to 12.4%) would mean that the UK government's target to reduce adult smoking to 26% by the year 2005 could be reached sooner, by the year 2002. This assumes a similar success rate could be achieved with all potential callers to the Quitline. This assumption could prove invalid if the self selected nature of the Quitline callers means that they differ from smokers in the general population in ways that affect success in smoking cessation. Alternatively, the findings might be further improved by efforts to recruit those more likely to succeed in quitting, such as less addicted smokers and those with a shorter history of smoking. As noted above, the profile of callers at one year appears, if anything, to biased away from achieving successful quitting. Additional improvements may also be observed if those who phoned the service and received an information pack, but did not get the opportunity to speak to a counsellor, also stop smoking. Separate studies would be needed to explore these issues. Research would also be needed to determine whether similar rates could be obtained with smokers who phone during periods without paid advertising. The cessation rates reported in this study were obtained with callers who phoned during the Health Education Authority's advertising campaign.

The lack of an appropriate comparison group limits the conclusions that can be drawn about the efficacy of the helpline. Nevertheless, the cessation rate in this study is very similar to that reported by Platt and colleagues8 and Borland and Hill.13 It also compares very favourably with estimates of cessation rates obtained in the absence of interventions,16 17 In a study by Baillie and colleagues a cessation rate of 7.33% was observed without any apparent intervention. Even the most cautious estimate in the present study (15.6%) is double that reported by Baillie and colleagues.16

Lack of biochemical validation is another concern. Several studies have documented that self reported smoking status leads to an overestimate of successful quit attempts.9 11 12 However, when adjusted for an estimated deception rate of 20% and refusals the present cessation rate is still comparable with studies that incorporated both control groups and biochemical validation.9 11 12

Another consideration is the bias that may have resulted from attrition. Compared with baseline callers, those followed up at one year were more likely to be women, aged 35 and over and smokers with light (< 10 cigarettes a day) or moderate cigarette consumption (10–19 cigarettes a day). Various factors have been identified with successful quitting such as being a lighter smoker and being older (aged 40+).1 18 19 The evidence concerning sex differences is conflicting. Although sex differences have not always been found, when differences have emerged men have been more successful.7 The effect of sex may also be influenced by age. In one study young women were significantly more likely to give up than young men, while middle aged women were less likely to give up than middle aged men.19 Given the interaction between age and sex, the change in age and sex profile observed in the present study appears to offer an unlikely explanation of the findings. At one year, 44% of the sample were women aged 35+, the group least likely of all to quit. Only 17% of the sample were men aged 35+. Changes in consumption are also unlikely to account for the findings because the greatest increase was among moderate smokers from 31% of the sample at baseline to 40% at one year. Among light smokers the increase was from 11% at baseline to 15% at one year. On balance, the profile of callers at one year appears, if anything, to be biased away from achieving successful quitting.

Information on social class was not available at baseline and so it is not possible to assess any bias attributable to this factor. However, as reported earlier, the social class profile at one year matched that of the adult population of smokers in England.(survey of knowledge, attitudes and behaviour 1994-1997, unpublished data). Two thirds of the callers at one year were from manual and unemployed households. Since successful quitting is associated with higher rather than lower socioeconomic status,20-22 social class bias in the Quitline sample appears an unlikely explanation of the present findings.

The results of this study also point to the need to consider reducing prevalence by providing relapse prevention to smokers who have already achieved initial cessation. In the present study 27% of initial callers were ex-smokers of whom 59% relapsed at one year. One approach to preventing relapse might be to offer proactive call backs to initial callers since research has shown that the provision of more than one telephone session may be beneficial.9 10 The timing of call backs is arguably best determined by the shape of the relapse curve.23 Typically the probability of relapse is much greater in the initial stages than in the later stages of quitting and so smokers' need for help is strongest soon after they attempt to quit. Indeed the relapse curve for smoking cessation indicates that maximum relapse (60%) occurs within one week after a quit attempt.23 The optimum timing and the number of call backs would need to be tested in a future study. Additionally, future research should include an assessment of the content of counselling, including any relapse prevention component, in order to determine which aspects of the counselling process are important in generating positive smoking behaviour change.

Overall, the results indicate that telephone counselling can produce positive outcomes in terms of smoking cessation. The high accessibility and convenience of the telephone format suggests that this approach to smoking cessation provides a promising addition to the interventions currently available.