Randomised controlled trial of cognitive behavioural therapy for smokers living in a deprived area of London: outcome at one-year follow-up Professor David F Marks and Catherine M Sykes Centre for Health & Counselling, City University, Northampton Square, London EC1 V 0HB Email:
[email protected]
Abstract In order to reach the smoking prevalence targets set by UK government, it is necessary to increase the adult quit rates, especially among smokers on low incomes. One way of increasing quit rates is to provide cost-effective smoking cessation therapy within the National Health Service (NHS). This paper compares the efficacy of smoking cessation advice in a health promotion campaign with that of an alternative intervention consisting of a cognitive behavioural therapy for smokers, the QUIT FOR LIFE Programme. Two hundred and sixty self- and GP-referred smokers took part in a randomised controlled trial with follow up at 12 months. All lived in an economically deprived area of north London. At a 12-month follow up twenty-three of 116 participants (19.8%) receiving therapy were abstinent and 10 (8.6%) had reduced cigarette consumption by at least 25% of pre-treatment level. Six (5.8%) of 104 participants in the control group were abstinent and none had reduced consumption. Cognitive behavioural therapy can provide a cost-effective intervention for low income smokers and should be available within the NHS in addition to nicotine replacement therapy.
Introduction The Smoking Kills White Paper sets out a target for adult smokers. It aims to reduce adult smoking in all social classes so that the overall rate falls from 28% to 24% or less by the year 2010; with a fall to 26% by 2005 (Department of Health, 1998a). The main methods to reach this target are nicotine replacement therapy (NRT), advice from health professionals and counselling. The White paper promised that £60m would be spent over three years on setting up smoking cessation services. These services were launched with the ‘Don’t give up giving up’ advertising campaign launched in December 1999. One of the features of the ‘Don’t give up giving up’ campaign is a helpline. Smokers are offered literature on smoking cessation and they can also be transferred to a counsellor. However the Department of Health’s analysis of calls shows that only 9% of callers were transferred to counsellors and 82% were sent literature (Foulds, 2000). The literature offered to smokers is based the Stopping Smoking Made Easier (SSME) leaflet published in 1992 by the Health Education Authority. SSME recommends a staged approach: preparing to stop, stopping, and staying stopped. It is based on
Randomised controlled trial of cognitive behavioural therapy for smokers Prochaska and DiClemente’s (1983) model of stages of change. The preparation stage consists of clarifying reasons for stopping, becoming ready to stop, and making an action plan. Smokers are recommended to stop suddenly, using their will power, rather than to gradually reduce their consumption. Smokers are advised to choose a day, obtain support from family and friends, review the action plan the day before, plan a suitably healthy reward for the end of the first day, and another for the end of the first week, and the first month, plan ahead, and to telephone the national ‘Quitline’, GP or health centre for further help. Once smokers have stopped they are exhorted to think positively, take care, not to play games, keep busy, avoid alcohol, refuse or break up cigarettes offered by friends and to learn to relax. No techniques are actually offered to implement this advice. In this journal, Marks (1998) argued that in addition to being willing and ready to quit, there are at least four other requirements before addictive behavioural change is likely to occur. The smoker must believe that the change will bring about the desired outcome, believe that he/she will be able to make the change, possess the skills that are necessary to make and cope with the change, and possess the skills to recover and to prevent relapse. Marks (1998) also pointed out that the use of NRT only has an efficacy of 2-6% in family practice settings and 716% in self-referred smokers, that NRT lacks cost-effectiveness and the cost makes it inaccessible to many smokers and that NRT only deals with physiological dependency. It completely neglects smoker’s psychological dependency on tobacco. Psychological dependency, especially in the context of stressful life events, is the major cause of continuing the habit in spite of having all of the necessary advice, health information and motivation to make the change (Gilbert, 1997; Marks, 1998; US Department of Health & Human Sciences, 1996). Cognitive-behavioural therapy (CBT) enables addictive behaviours to be changed by facilitating the control of the psychological processes responsible for, or associated with, the acquisition of the habit, e.g. conditioning, reinforcement, arousal, attention to cues, moods, feelings, suggestion, and imagery. CBT has the potential to deliver costeffective interventions for the reduction of the psychological dependency that the majority of smokers experience and which is left intact by NRT.
The purpose of this paper is to describe a randomised controlled trial (RCT) of the efficacy of CBT for smokers, the QUIT FOR LIFE (QFL) Programme (Marks, 1993) in comparison to a control treatment. The control treatment consisted of SSME. This paper presents analyses of efficacy and cost-effectiveness of QFL. The QFL Programme uses a spectrum of thirty CBT and other relevant methods in a self-help package consisting of a handbook, reduction cards, a progress chart, and other necessary materials. A cassette tape of the handbook recorded by the BBC’s Nick Ross is also given to participants. This is useful for participants who do not like reading or are illiterate. The reverse side of the tape includes relaxation music. The self-help format requires attendance at one introductory session, a few days before the participants initiate a programme of systematic reduction. The methods are listed in Table 1, which appears at the end of the paper. There is a section to make smokers more aware of the trickery of tobacco advertising which 2
Randomised controlled trial of cognitive behavioural therapy for smokers reveals to smokers the multiple causes of addictive behaviour (see Whitehead, 1995). This removes the emphasis from blaming the smoker. QFL aims at a gradual reduction of cigarette consumption over a period of 7-10 days. A set of targets guides the smoker towards a daily reduction of 50%. Of particular importance is the objective of increasing the smokers’ self-efficacy, a belief that they can and will be able to make the change from smoker to nonsmoker. The therapy encourages smokers to be constantly vigilant concerning the process of rationalisation that gains more force as they approach complete cessation. This occurs on “D-Day”, a planned 24-hour period when smoking is discontinued. The programme also trains participants to employ mental imagery, suggestion, relaxation and meditation (Marks, 1993). The QFL Programme is in two stages, reduction and relapse-prevention. The Programme is further sub-divided into ten sections spaced across a period of three months. The reduction stage begins on a Tuesday and finishes the following week with ‘D-Day’. Relapse-prevention starts on ‘D-Day’ and continues thereafter. In contrast to the control intervention, it is emphasised that will power is not needed in the QFL Programme. NRT is an optional adjunct during the initial period of 10-30 days immediately following D-Day. It is recommended that smokers use only the lowest dosage level gum or patch for this purpose. In practice NRT is used by a small minority of QFL participants owing to its high over-the-counter price. The control intervention (SSME) is available in many GP surgeries throughout the UK and consists of a small booklet providing smoking cessation advice. Normally the SSME does not include a group introductory session. Smokers are usually handed the SSME or sent it through the post. Methods The planned study population was adult smokers living in a deprived area of London. All participants telephoned a smoking cessation clinic and booked to attend a smoking cessation trial. Posters and pamphlets advertising free quit smoking assistance were displayed in GP waiting rooms, pharmacies and a library noticeboard. The national Quitline was also informed about the trial. Dates and times of a one-hour introductory session to the interventions were given to a receptionist. These sessions consisted of treatment or control in a random order. The receptionist was unaware this order. When smokers telephoned the receptionist gave the smokers a choice of dates and times and booked them into the first convenient group session. There were no criteria for exclusion; all smokers who applied were admitted to the trial subject only to their informed consent. The primary outcome measure was point prevalence of smoking validated by a breath carbon monoxide (CO) reading. Power analysis suggested a sample of 100 would provide a 90% chance of detecting a difference of 10% between control and treatment conditions. 131 individuals were allocated to the CBT programme and 129 to the control condition. Sessions were arranged by booking smokers into mixed groups of 3 to 12 people. Upon arrival at the clinic, participants were given an explanation about the trial and a consent form that informed them that two interventions were being evaluated and that they had been allocated to one of them by random selection. None refused to give informed consent.
3
Randomised controlled trial of cognitive behavioural therapy for smokers Each session lasted 60 minutes and was run interactively, inviting questions and comments, and urging the smokers to comply with the instructions in the booklets provided. The two interventions were delivered with similar amounts of energy and enthusiasm. Every effort was made to motivate the smokers to follow the procedures recommended in the booklets provided. A breath CO reading was taken during the session using a Bedfont Smokerlyzer. The smokers were invited to call the clinic’s help-line if they had any further questions or problems. Finally, a 3-month follow-up meeting was arranged to monitor the participants’ progress. The meeting reinforced the approach that was delivered at the original session and so its content differed for the two treatment conditions. The attendance rate at this meeting was approximately 60%. Irrespective of their attendance at the meeting, all participants were contacted by telephone at the 12-week point and their progress discussed and, in the majority of cases, a Smokerlyzer reading taken. At the 12 month follow up each participant’s smoking status was ascertained. If they failed to attend the follow up for a meeting, smoking status was ascertained by telephone or letter and arrangements were made to take a CO reading in cases where reduced or abstinence was reported. All participants were asked if they had used NRT in addition to CBT or the control treatment. Eleven had done so. Twenty nine participants left the trial before the 12-month data-point: one died; four moved house, and twenty-four became non-contactable by telephone or mail. Results As is typical of smoking cessation interventions, the attendance rates for this trial were low; approximately one in four people who telephoned did not book into a session and one in two who booked into a session failed to appear. This sample, as with all samples of smokers in cessation trials, is representative of smokers at the action stage. Table 2. Socio-economic status of the participants SES QFL I 0 II 20 IIIN 21 IIIM 14 IV 20 V 6 Unempl. 21 Retired 9 Students 10 Unknown 13 Totals 131
Contro 0 23 19 14 15 4 16 11 7 17 129
Totals 0 43 40 28 35 10 37 20 17 30 260
The SES distributions for the QFL and control groups were highly similar (Table 2). The sample contained a significant proportion of smokers on lower incomes including smokers from manual occupations (73/260 = 28%) and unemployed smokers (37/260 = 14%). The QFL sample contained 45 males and 86 females and reported an average cigarette consumption of 24.5 (SD, 6.18) cigarettes per day at baseline. The control sample contained 49 males and 80 females and reported an average cigarette consumption of 25.5 (SD, 6.42) cigarettes per day at baseline.
The participants who could be contacted at follow up totalled 231 at 12 months. Twenty-nine participants were unavailable for follow-up at 12 months, 12 (9%) from the CBT group and 17 (13%) from the control group. This difference was not statistically significant.
4
Randomised controlled trial of cognitive behavioural therapy for smokers
At the three-month point, a few participants reported using NRT as a complement to CBT (3 = 2%) or SSME (8 = 6%). Of these, in the CBT group, two proved to be abstinent at 12 months, and one had relapsed; in the SSME group, 4 were abstinent and 4 had relapsed. There were insufficient numbers to perform separate statistical analyses on participants using NRT. Had they been more numerous it would have been interesting to analyse their abstinence rates, but this must await the results of a separate study. Adding those who were eliminated because of drop-out to those eliminated because of their use of NRT gave combined “drop-out rates” of 15 (11%) from the CBT group and 25 (19%) from the control group. Statistical analysis showed that this difference was not significant (chi-squared = 2.04; 1 d.f.; p>.10). Smoking status of the remaining 220 participants at follow-up was classified into three categories: abstinence (zero cigarette consumption per day continuously for at least seven days immediately prior to the follow-up test), reduction (cigarette consumption at least 25% lower than baseline for at least four continuous weeks immediately prior to the follow-up test), or no change. This information was obtained by self-report using a standard set of questions worded using the above definitions. These reports were validated using the breath CO tests. In the majority of cases (77/78 or 98.7%) participants' verbal reports of their smoking status were consistent with their breath carbon monoxide readings (Table 3). Only one participant (in the QFL condition) reported a smoking status (abstinence) inconsistent with her breath CO reading. Her data were removed from the analysis. Table 3. Breath carbon monoxide readings (parts per million) for participants reporting abstinence, reduced cigarette consumption, or no change CO level Totals (ppm)
Abstinence
Reduced consumption
No change
Totals
