ORIGINAL INVESTIGATION Randomized Comparative Trial of Nicotine Polacrilex, a Transdermal Patch, Nasal Spray, and an Peter Hajek, PhD; Robert West, PhD; Jonathan Foulds, PhD; Fredrik Nilsson, MSc; Sylvia Burrows, BSc; Anna Meadow, BSc Background: There are several nicotine replacement products on the market, and physicians are likely to be asked with increasing frequency about which of these products their patients should use. Objective: To provide a basis for rational advice by comparing nicotine polacrilex (gum), a transdermal patch, nasal spray, and an inhaler. Design: Randomized trial with assessments at the quit date and 1, 4, and 12 weeks later. Setting: Hospital smokers clinic. Patients: Male and female community volunteers (N = 504) smoking 10 or more cigarettes per day and seeking help to stop smoking. Interventions: Patients were given brief advice, and purchased their nicotine replacement treatment at approximately half the regular retail price. Main Outcome Measures: Nicotine replacement treatment use, ratings of withdrawal symptoms, ratings of product characteristics and helpfulness, and biochemically validated continuous lapse-free abstinence. Results: The products did not differ in their effects on withdrawal discomfort, urges to smoke, or rates of abstinence. The continuous validated 12-week abstinence rates were 20%, 21%, 24%, and 24% in the gum, patch, spray, and inhaler groups, respectively. Compliance with recommended nicotine replacement treatment use was high for the patch, low for gum, and very low for the spray and the inhaler. The spray was underused because of adverse effects more often than the other products. In the subjects using the spray, the level of use among abstainers at week 1 predicted outcome at week 12. The inhaler was rated as more embarrassing to use than the other products, but provided at least as much nicotine as the gum. Conclusion: When asked about nicotine replacement treatment products available, physicians should note that, despite low compliance with the recommended dose of the spray and inhaler and differences in product ratings, overall, there are no notable differences between the products in their effects on withdrawal discomfort, perceived helpfulness, or general efficacy. Arch Intern Med. 1999;159:2033-2038 From the Department of Human Science, St Bartholomew s and the Royal London School of Medicine and Dentistry (Dr Hajek and Ms Meadow) and Department of Psychology, St George s Hospital Medical School (Drs West and Foulds and Ms Burrows), University of London, London, England; Department of Psychology, University of Surrey, Guilford, England (Dr Foulds); and Pharmacia & Upjohn, Helsingborg, Sweden (Mr Nilsson). THERE ARE 4 nicotine replacement treatments (s) available (nicotine polacrilex [gum], nicotine transdermal patch, nicotine nasal spray, and nicotine inhaler). They have all been shown to increase smokers chances of stopping smoking for at least a year. 1 In 2 withinsubject short-term (2-3 days) studies, 2,3 the gum and the patch had similar effects on withdrawal; in another within-subject study, 4 the patch and the spray were shown to produce similar nicotine and cotinine plasma levels during 5 days of use. Otherwise, no direct comparisons have been made between the different products for their ease of use, suitability for different patient groups, nicotine delivery or efficacy in relief of withdrawal symptoms, reduction in urges to smoke, or ability to maintain abstinence. Such direct comparisons are required between the 4 products to make rational decisions concerning their use. As the public becomes aware of various options, health professionals are likely to be asked with increasing frequency about which of these products their patients should use. Faster delivery products, such as nicotine nasal spray, may be more effective in relieving urges to smoke. 4 However, they have been reported to cause greater local irritation, which may translate into lower compliance and efficacy. The inhaler, which mimics some of the cigarette manipulation and puffing behaviors, may help to relieve discomfort 2033
SUBJECTS AND METHODS SAMPLE This randomized, controlled study was performed at 2 sites in London, England (Royal London Hospital and St George s Hospital). The project was approved by the appropriate ethics committees. Recruitment was conducted through advertisements and referrals. A total of 506 smokers met the study criteria and underwent randomization. Two became pregnant during the study, and their participation was discontinued, leaving 504 subjects. The subjects were eligible for inclusion if they were at least 18 years old, had smoked an average of 10 or more cigarettes per day, were motivated to stop smoking, were generally in good health, were not receiving treatment for a psychiatric disorder, had not tried to stop smoking using an within the past 3 months, and for whom none of the 4 products was contraindicated. Table 1 gives the sample characteristics. There were no significant differences between the groups (P.05 for all comparisons). PROCEDURE Smokers volunteering for the study were screened by telephone interview. Those that fulfilled the entry criteria were invited to attend a hospital smokers clinic. They were also sent a patient information sheet and a questionnaire to complete and bring with them at their first attendance. The questionnaire included questions concerning smoking history, previous use, and demographic details. The subjects were instructed to smoke normally up to the first session and that the session would mark the beginning of the cessation attempt (ie, they should be prepared not to smoke thereafter). At their first individual attendance at the clinic (week 0), smokers were reassessed for eligibility and invited to provide written consent for participation in the study. Subjects viewed an 8-minute videotape describing the 4 products being tested, and provided ratings of their opinion of the products. They were then randomized, by a computer-generated schedule in sealed envelopes, to receive 1 of 4 products as follows: (1) nicotine polacrilex (gum), 2 or 4 mg (subjects using 12 pieces of 2-mg gum per day were offered 4-mg gum); (2) a nicotine transdermal patch, 15 mg, for 16 hours (n = 124); (3) a nicotine nasal spray (n = 126); and a nicotine inhaler. The subjects agreed to participate in the study knowing that they might be allocated to use any one of these products. They were also aware that they would have to purchase the products at approximately half the regular retail price ( 7; ie, $12 for 105 pieces of gum, 7 patches, 1 bottle of nasal spray solution giving 200 0.5-mg doses, or 42 inhaler cartridges). The reason for this was to mimic as far as possible the conditions in countries such as England, where is not reimbursable. The subjects then viewed a 7-minute videotape describing in greater detail how to use the product to which they had been assigned. The videotapes standardized the presentation of information about the products as far as possible. The subjects rated their mood and physical symptoms to provide a baseline for later measurements of withdrawal symptoms. Subjects tried 1Uoftheirproduct at the first session under supervision and then were dispensed their supply of to last them for 1 week. Subjects were given brief encouragement to remain abstinent and were Table 1. Sample Characteristics of 504 Subjects Nicotine Replacement Treatment Nicotine Polacrilex (Gum) Transdermal Patch Nasal Spray Characteristics* (n = 124) (n = 126) Men, % 35 33 42 32 Age, mean (SD), y 40 (10.0) 40 (10.8) 41 (10.3) 39 (10.2) No. of cigarettes smoked 24 (8.0) 25 (10.0) 25 (9.0) 24 (9.1) per day, mean (SD) FTND score, mean (SD) 5.9 (2.5) 5.6 (2.3) 5.9 (2.4) 5.9 (2.1) CO, mean (SD), ppm 29 (11.3) 29 (10.8) 29 (12.3) 30 (12.5) No. of prior quit attempts, mean (SD) 2.8 (2.9) 2.5 (2.4) 2.7 (3.2) 2.9 (3.0) White collar, % 50 41 50 50 *FTND indicates Fagerstrom Test of Nicotine Dependence; CO, end-expired carbon monoxide. related to behavioral components of smoking. 5 Another hypothesis assumes that the gum, spray, and inhaler, which all provide sensory stimulation, should have bigger impact than the patch on initial craving. 6 The patch, because it is only put on once a day, is much easier to use than the other products and may, therefore, be associated with better compliance and outcome. None of these issues has been addressed empirically within a single randomized trial. To our knowledge, we performed the first study to directly compare the 4 products on a range of variables. Results from a comparison of the 4 products in their use, acceptability, cotinine levels, rated helpfulness, withdrawal symptom relief, and abstinence rates up to 3 months are reported. Findings regarding the products dependence potential, subject preferences, and product effects in different patient subgroups will be reported separately. 2034
instructed to use the product to which they had been assigned according to the manufacturer s instructions for up to 12 weeks. For the gum, this involved at least 10 to 15 pieces per day, approximately hourly; for the patch, this involved 1 patch per day applied to the skin on waking; for the nasal spray, this involved at least 30 single shots per day, approximately hourly; and for the inhaler, this involved using the inhaler hourly for some 20 minutes and using at least 6 cartridges per day, with each cartridge being used 3 to 4 times. Subjects were also given a sheet containing written guidance on use of their product. Subsequent individual visits to the clinic were scheduled as follows: session 2, 1 week after the quit date; session 3, 4 weeks after the quit date; and session 4, 12 weeks after the quit date (only those abstinent at session 3 were invited). Subjects received 10 for expenses per visit, excluding session 1. Those who were abstinent and using at week 12 were invited back 3 weeks later to assess dependence on their product. The following measures were taken: 1. In sessions 2 through 4, data regarding use since the last session, reasons for nonuse (subjects using less than the recommended dose were asked to choose 1 of the following: forgot, it is not helping, adverse effects, did not need it, to save money, or another reason), and ratings concerning ease of use (ranging from 1 [very difficult] to 9 [very easy]) and embarrassment using (1 indicates not at all; 2, somewhat; and 3, very much so) were obtained. 2. In sessions 1 through 4, ratings of depression, irritability, restlessness, hunger, and poor concentration (using the Mood and Physical Symptoms Scale) 7 were obtained. 3. In sessions 2 through 4, ratings of strength of urges to smoke, time spent with urges to smoke, and difficulty not smoking 7 were obtained. 4. In sessions 2 through 4, data regarding complete, lapse-free abstinence from cigarettes, verified by expiredair carbon monoxide concentrations (cutoff point 10 ppm), were obtained. If biochemical verification was not possible because of nonattendance at a session, subjects were counted as smokers. 5. In session 2, the saliva cotinine concentration was obtained. STATISTICAL ANALYSES Between-product comparisons on categorical variables used 2 analyses with pairwise comparisons made using Bonferroni adjustment for multiple comparisons. Comparisons involving continuous variables used analysis of variance with the Scheffé test for post hoc pairwise comparisons. The withdrawal score at week 1 was calculated by subtracting the week 1 ratings of irritability, depression, restlessness, poor concentration, and hunger from the ratings at week 0. The total withdrawal score was calculated by adding the previously described ratings. The compliance and abstinence rates were calculated on an intention-to-treat basis, while other analyses were based on subjects who provided the required measures, as stated in the Results section. The sample size of 504 subjects provided 80% statistical power to detect a difference between any 2 groups in total withdrawal scores of 1.4 scale points (scores can range from 20 to 20) (2-tailed tests, P.001). It provided 80% power to detect differences between any 2 groups of 15% in abstinence rates (2-tailed tests, P.001). Logistic regression was used to assess whether cotinine measures in smokers abstaining at 1 week predicted success at 12 weeks. Table 2. Compliance With the Minimum Recommended Use of During the First Week and Mean Use* Variable Nicotine Polacrilex (Gum) (1) Transdermal Patch (2) Nasal Spray (3) (4) Difference (P ) Compliance, % All subjects 38 82 15 11 2 1 3and4(.01) Abstinent subjects 48 96 24 17 2 1 3and4(.01) use per day in week 1 abstainers, mean (SD) 9.5 (3.0) Pieces 1.0 (0.2) Patches 24.5 (11.8) Single shots 4.3 (1.9) Cartridges *Nonattendees were assumed to be noncomplying. indicates nicotine replacement treatment; ellipses, data not applicable.... RESULTS Table 2 shows that compliance was high for the patch and low for the other products: only 17% to 48% of even those abstaining in the first week used the recommended amount. It also shows the mean use among those abstinent at week 1. Subjects using the spray and the inhaler complied with recommended use less often than subjects using the gum, and all 3 groups had lower compliance than subjects using the patch. In the gum group, 6 subjects used 4-mg gum. Table 3 shows that self-reported reasons for nonuse were different for the different products. The most common reason for not using the patch was forgetting, while for the other products it was believing that subjects did not need it. Adverse effects were cited more often as a reason for nonuse with the spray than with the other products. The other category included items such as embarrassing to use, being ill, and avoiding dependence. Table 4 shows that, during the first week, the patch was rated as easier to use than all the other products, which did not differ significantly from each other. The inhaler 2035
was rated as more embarrassing to use than the other products, followed by the spray. The products did not differ in perceived helpfulness and in stress relief. Table 5 shows that subjects allocated to use the spray and abstaining at week 1 had lower saliva cotinine concentrations after 1 week of abstinence than those using the gum. Among subjects who complied with the minimum recommended use, those using the inhaler had higher cotinine concentrations than those using the patch. Table 6 shows that the products did not differ significantly for effects on withdrawal symptoms, difficulty not smoking, urges to smoke at week 1, or weight gain at week 12. There were no differences between subjects abstinent for 12 weeks at any point. Table 6 only shows the total withdrawal scores, but there were no significant differences in any of the constituent ratings. Table 3. Subjects Attending the Smokers Clinic at Week 1 Who Underused Giving Reasons for Not Using the Recommended Amount* Reason Nicotine Polacrilex (Gum) (n = 67) Transdermal Patch (n = 23) Nasal Spray (n = 72) (n = 80) Forgot 10 65 21 14 Do not need it 43 0 39 41 Adverse effects 16 4 25 8 Impractical 3 0 3 18 Other 27 30 13 20 *Values are given as percentage of subjects. 2 = 68.9 for table, df = 12, P.001. indicates nicotine replacement treatment. Other indicates embarrassing to use, being ill, avoiding dependence, etc. Table 7 shows that the products did not differ for abstinence rates. We compared the relapse rates in nonsmoking subjects who used with those in nonsmoking subjects who failed to use at least the minimum recommended dose during the first week of abstinence. The differences were not significant in the gum (P =.57), patch (P =.72), and inhaler (P =.90) groups. However, subjects using at least the minimum recommended dose of the nasal spray at week 1 did significantly better at week 12 (69% vs 39% still abstinent, 2 = 4.5, P =.02). There was also a significant positive association between amount of nasal spray used in week 1 and abstinence at week 12 in the nasal spray group (logistic regression, Wald score = 5.2, P =.02, 2-tailed test). COMMENT In naturalistic conditions, the 4 nicotine replacement products did not differ in their effects on withdrawal discomfort, and there was no difference in 3-month sustained abstinence rates. However, there were significant differences in compliance with recommended use and in some of the product ratings. In the nasal spray group only, level of initial use was positively related to subsequent abstinence. It may be argued that with only 504 subjects participating in the study, the failure to find a difference between the products for abstinence rates may have been due to low statistical power. However, the abstinence rates in the 4 groups were similar, and there was little suggestion of a clinically significant difference. Lack of power to detect a difference is unlikely to explain the failure to find a difference in relief of withdrawal discomfort, where the outcome variable was a quantitative index. Thus, based on these results, we could not recommend one product vs the others for overall efficacy. On the other hand, Table 4. Ratings for in Subjects Attending the Smokers Clinic at Week 1* Nicotine Polacrilex (Gum) (1) Transdermal Patch (2) Nasal Spray (3) (4) Difference Variable (n = 104) (n = 108) (n = 105) (n = 97) (P ) Rating of ease of use (1-10), mean (SD) 6.2 (2.6) 8.6 (1.1) 5.6 (2.7) 5.7 (2.7) 2 1,3and4(.01) Subjects reporting at least some 10 3 37 58 4 3 2and1(.05) embarrassment, % Helpfulness (1-5), mean (SD) 4.4 (0.7) 4.4 (0.8) 4.4 (0.8) 4.5 (0.7) None (.10) Stress relief (1-5), mean (SD) 3.1 (0.9) 3.2 (1.0) 3.5 (1.0) 3.3 (1.0) None (.05) * indicates nicotine replacement treatment. No post hoc comparisons were significant. Table 5. Saliva Cotinine Concentrations in Subjects Abstinent From Smoking at Week 1 Nicotine Replacement Treatment* Transdermal Patch (2) Nasal Spray (3) (4) Difference Nicotine Polacrilex (Gum) Subjects (1) All those abstaining 949 (943) (n = 64) Abstaining subjects using at least 1238 (1216) the minimum recommended dose (n = 31) 818 (312) 829 (307) (n = 63) 562 (432) 909 (511) (n = 16) 863 (778) (n = 54) 1624 (1164) (n=9) 1 3 4 2 *Values are given as mean (SD) cotinine concentrations, measured in nanomoles per liter. 2036
Table 6. Ratings of Urges to Smoke and Withdrawal Symptoms in Subjects Abstinent at Week 1 and Weight Gain and Use in Subjects Abstinent for 12 Weeks* Variable Nicotine Polacrilex (Gum) (n = 64) Transdermal Patch Nasal Spray (n = 54) Difficulty not smoking 2.0 (1.4) 2.1 (1.3) 2.2 (1.2) 1.9 (1.4) Time spent with urges 2.0 (1.1) 2.0 (1.0) 1.9 (1.0) 1.7 (0.8) Strength of urges 2.6 (1.3) 2.4 (1.2) 2.6 (1.1) 2.3 (1.3) Total withdrawal score 1.6 (3.9) 1.7 (4.2) 2.3 (4.2) 1.9 (4.1) use at week 12, % (N) 76.0 (19) 53.8 (14) 90.0 (27) 74.2 (23) Weight gain at week 12, kg 3.6 (3.0) 3.6 (3.5) 2.5 (1.8) 2.3 (2.8) *Values are given as mean (SD), unless otherwise indicated. There were no significant differences between the groups, except for use at week 12. The scale was from 0 to 15. The scale was from 20 to 20, and is the sum of the differences of 5 ratings from baseline (each rating being a 5-point scale). Overall 2 = 9.6, P =.02. The nicotine polacrilex (gum), nasal spray, and inhaler groups were significantly (P.05) different from the patch group by post hoc comparison, after Bonferroni adjustment. N signifies number of subjects. Table 7. Abstinence From Smoking in Each Group (n = 504)* Abstinence Nicotine Polacrilex (Gum) Transdermal Patch (n = 124) Nasal Spray (n = 126) Up to 1 wk 64 (50.4) 68 (54.8) 68 (54.0) 54 (42.5) Up to 4 wk 39 (30.7) 45 (36.3) 48 (38.1) 44 (34.6) Up to 12 wk 25 (19.7) 26 (21.0) 30 (23.8) 31 (24.4) *Values are given as number (percentage) of total in each group. There were no significant differences between the groups ( P =.19,.65, and.77 for weeks 1, 4, and 12, respectively). indicates nicotine replacement treatment. this still leaves the possibility that the products are differentially effective in different types of smokers. It is also possible that products that might be considered on a priori grounds to be more effective had lower compliance, which undermined their potential efficacy. Thus, the nasal spray, which has the fastest nicotine delivery but is initially particularly unpleasant to use, yielded lower cotinine concentrations than the other products, and level of use at week 1 was positively associated with outcome at week 12. The fact that the inhaler was no more effective than the other products and that the actions of puffing as a person would on a cigarette and the irritation in the throat do not seem to enhance efficacy can be interpreted as an indication that efficacy is determined primarily by the extent of nicotine rather than behavioral substitution. However, participants typically used less than the recommended amount, which could have diluted its potential behavioral impact. The study included only 1 type of nicotine patch. There are several other types of patches on the market with 24-hour delivery, 8 and some caution is needed in generalizing the study results to the other patch types. The study differed from most previous trials in that subjects paid for their. This was to mimic as far as possible the real-life situation in which use of different products is to a different degree codetermined by their cost. In particular, products other than the patch readily allow economizing, and clinical experience and research data suggest that this is 1 of the reasons for their underuse. In a study involving accompanied by brief advice, Hughes et al 9 found that when the gum was provided free, its rate of use and its efficacy were higher than when it had to be purchased. Despite having to pay for the products, the self-reported use of the inhaler and the nasal spray was only slightly lower than that reported in studies 10,11 providing free, while the use of the gum and the patch was at least as high as in previous trials. 2 The comparatively good product use could be related to carefully briefing subjects on use and to recommending hourly vs ad lib use. 12 Although the products were used as much as in other trials, the gum and the spray were used less than recommended (ie, subjects used an average of 25 rather than a minimum of 30 single shots of nicotine nasal spray per day and 9 rather than 10-15 pieces of gum per day), and the inhaler was used much less than the recommended dose (4 per day rather than at least 6 per day). Equivalent use of the inhaler and the gum can be expected to provide similar saliva cotinine levels, because absorption in both cases is primarily buccal. In this study, subjects in the gum group using 90% of the minimum recommended dose had cotinine levels similar to those in the inhaler group, who only used 67% of the recommended dose, while in those using the recommended dose, cotinine levels in subjects using the inhaler were 25% higher than in subjects using the gum. Taking the gum as a benchmark, the recommendation for inhaler use may be too high. Another point that arose with the inhaler was that we instructed subjects to use it hourly for about 20 minutes but that each cartridge could be used 3 to 4 times. The instructions on the (now commercially available) packet state that each cartridge should be used once intensively for 20 minutes. This is based on laboratory results with forceful inhalations every few seconds. In real life, subjects are unlikely to puff on the inhaler with the required intensity for long enough to use its nicotine contents at 1 use. As would be expected, the patch was generally not considered embarrassing to use, while the spray and especially the inhaler were considered embarrass- 2037
ing by many subjects. Since the study, the inhaler has been launched in several countries; it can be expected that as its use becomes more widespread, embarrassment will decrease. The study was not set up to evaluate efficacy compared with a placebo, as there is extensive literature 1 showing the treatment to be efficacious. The follow-up period did not extend beyond 3 months, as there was little reason to expect any new differences between products to emerge after the cessation of their use, and the study had sufficient statistical power for short-term comparisons. About two thirds of the sample were women, which is typical for smoking cessation clinics. 13 The absolute cessation rates achieved in this naturalistic setting with minimal behavioral support seem encouraging, and are in accord with other studies. 1 In conclusion, it appears that, in a naturalistic setting, the type of has little bearing on overall efficacy despite the differences in the sensory and behavioral aspects of s and in the speed of nicotine absorption. The nasal spray, which has the fastest nicotine delivery, may be more helpful to those who manage to use it properly, but it causes more adverse effects than the other products and tends to be underused. The inhaler was more embarrassing to use than the other products, although this may subside when its use becomes more widespread. The products may have different appeal and efficacy in different subgroups of smokers, but the results indicate that overall, the 4 therapies have similar perceived helpfulness and similar effects on craving, withdrawal symptoms, and abstinence rates. Accepted for publication January 25, 1999. We thank Pharmacia & Upjohn, Helsingborg, Sweden, for funding the study. Reprints: Peter Hajek, PhD, Psychology Section, St Bartholomew s and the Royal London School of Medicine and Dentistry, Turner Street, London E1 2AD, England. REFERENCES 1. Silagy C, Mant D, Fowler G, Lancaster T. Nicotine Replacement Therapy for Smoking Cessation [serial on CD-ROM]. Oxford, England: Update Software; 1999. 2. Fagerstrom KO, Schneider N, Lunell E. Effectiveness of nicotine patch and nicotine gum as individual versus combined treatments for tobacco withdrawal symptoms. Psychopharmacology. 1993;111:271-277. 3. Leischow S, Valente S, Hill A, et al. Effects of nicotine dose and administration method on withdrawal symptoms and side effects during short-term smoking abstinence. Exp Clin Psychopharmacol. 1997;5:54-64. 4. Benowitz N, Zevin S, Jacob P. Sources of variability in nicotine and cotinine levels with use of nicotine nasal spray, transdermal nicotine, and cigarette smoking. Br J Clin Pharmacol. 1997;43:259-267. 5. Hajek P, Jarvis M, Belcher M, Sutherland G, Feyerabend C. Effect of smoke-free cigarettes on 24 h cigarette withdrawal: a double-blind placebo-controlled study. Psychopharmacology. 1989;97:99-102. 6. Rose J, Levin E. Inter-relationships between conditioned and primary reinforcement in the maintenance of cigarette smoking. Br J Addict. 1991;86:605-609. 7. West R, Russell M. Pre-abstinence smoke intake and smoking motivation as predictors of severity of cigarette withdrawal symptoms. Psychopharmacology. 1985; 87:334-336. 8. Gorsline J. Nicotine pharmacokinetics of four nicotine transdermal systems. Health Values. 1993;17:20-24. 9. Hughes J, Wadland W, Fenwick J, Lewis J, Bickel W. Effect of cost on the selfadministration and efficacy of nicotine gum: a preliminary study. Prev Med. 1991; 20:489-496. 10. Sutherland G, Stapleton J, Russell M, et al. Randomised controlled trial of nasal nicotine spray in smoking cessation. Lancet. 1992;340:324-329. 11. Tonnesen P, Norregaard J, Mikkelsen K, Jorgensen S, Nilsson F. A double-blind trial of nicotine inhaler in smoking cessation. JAMA. 1993;269:1268-1271. 12. Killen J, Fortmann S, Newman B, Varady A. Evaluation of a treatment approach combining nicotine gum with self-guided behavioral treatments for smoking relapse prevention. J Consult Clin Psychol. 1990;58:85-92. 13. Hajek P. Helping smokers to overcome tobacco withdrawal: background and practice of withdrawal-oriented therapy. In: Richmond R, ed. Interventions for Smokers: An International Perspective. Baltimore, Md: Williams & Wilkins; 1994:29-46. 2038