CADTH is committed to supporting Canada’s health care decision-makers through this challenging and uncertain time.
For evidence, tools, and resources related to COVID-19, visit our COVID-19 Evidence Portal.


Begin main content

Pharmacologic-Based Strategies for Smoking Cessation: Clinical and Cost-Effectiveness Analyses

Last updated: August 21, 2012
Issue: Volume 2, Issue 3
Result type: Report

Adapted from Tran K, Asakawa K, Cimon K, Moulton K, Kaunelis D, Pipe A, Selby P. Pharmacologic-based Strategies for Smoking Cessation: Clinical and Cost-Effectiveness Analyses [Technology Report; No. 130]. Ottawa: Canadian Agency for Drugs and Technologies in Health; 2010.

For more information on this project, visit


Tobacco smoking is a risk factor for cancer, respiratory disease, and cardiovascular disease. It is estimated that 19% of Canadians (approximately 5.2 million) aged 15 years and older were smokers in 2007.1 Each year, approximately 45,000 Canadians die from smoking.2 Significantly, one-third of Canadian smokers aged 15 years or older express an intention to quit in the next 30 days.3

Nicotine, the addictive chemical component of tobacco products that attracts smokers, has consequences for personal and community health. Smoking is an addiction. Although many smokers report that they quit unaided, most smokers who try to quit without smoking cessation aids are unsuccessful in the long term. In the US and UK, 70% of smokers intend to quit every year, 45% try to quit, and less than 5% are successful.4,5 The rate of relapse is high among smokers who quit without treatment. The proportion of smokers who can achieve abstinence for at least one week is 25% to 51%, 10% to 20% for those able to do it for at least three months, and only 3% to 5% for those able to abstain for six months.6

There are many prescribed pharmacotherapies for smoking cessation, including nicotine replacement products, bupropion, and varenicline. Smoking cessation programs often combine drug treatment and behavioural support (such as psychological interventions, telephone support, and self-help).


The objective of the report was to:

  • Compare the pharmacological agents (drug-based smoking cessation aids) to determine which of these works best (clinical effectiveness) at six months and 12 months after patients attempt to quit smoking.
  • Perform an economic analysis to determine which smoking cessation therapy was the most cost-effective for patients, drug plans, and the Canadian health care system.


To compare pharmacological agents with or without behavioural support programs for smoking cessation in adults, a systematic review with meta-analyses and an economic evaluation were carried out. A budget impact analysis and an examination of current public funding and planning issues were also performed.

For the clinical review, a total of 155 articles (from August 1982 to June 2009) describing 143 trials were selected for inclusion in the clinical analyses. To meet the inclusion criteria, randomized controlled trials (RCTs) must have reported biochemically validated measures of smoking abstinence from the study start date or from the target quit date of at least six months. Analyses were conducted using a Bayesian random effects model for mixed treatment comparisons meta-analysis. Continuous abstinence rates (patients have not smoked at all since quitting) and point prevalence abstinence rates (patients have not smoked in the past seven days) were analyzed separately.

For the economic analysis, a decision-analytic model was constructed to assess the cost-effectiveness (cost per additional quitter and cost per life-year gained; cost per quality-adjusted life-year [QALY] gained) of nicotine replacement therapy (NRT), bupropion, and varenicline compared with no pharmacologic treatment, from the perspective of a publicly funded health care system. For the base case, the general population (stratified by age and gender) motivated to quit smoking was considered. Treatment strategies included one-time cessation interventions of NRT, bupropion, varenicline, and no pharmacologic treatment; it was assumed that all patients received the same level of behavioural support. Budget impact analyses were conducted by comparing current utilization patterns of NRT, bupropion, and varenicline, with alternative scenarios reflecting the prescription patterns of these therapies.


Clinical Efficacy in Healthy Smokers

Eighty-one trials (from August 1982 to June 2009) were identified, involving a total of 40,317 participants. Varenicline, bupropion, and all six NRTs (patch, gum, lozenge, inhaler, spray, and sublingual) were found to be more efficacious than placebo after six months and one year of follow-up. Use of NRTs (odds ratio [OR] 1.89, credible interval [CrI] 95% 1.63 to 2.18) or use of bupropion (OR 1.95, CrI 95% 1.58 to 2.41) was found to approximately double the odds of quitting successfully and maintaining abstinence for at least 12 months compared with placebo. Use of varenicline (OR 2.78, CrI 95% 2.17 to 3.57) also improved the odds of success between two- and three-fold. The use of mixed treatment comparisons in the review found varenicline to be superior to conventional NRT use and bupropion. There were no differences in efficacy between bupropion and NRT or between the various forms of NRT alternatives.

Clinical Efficacy in Specific Populations of Smokers

Nicotine gum, nicotine patch, and bupropion were shown to be superior compared with placebo at 12 months of follow-up in five trials (N = 1,673) assessing five different interventions for smoking cessation in smokers with cardiovascular or smoking-related diseases. For patients with chronic obstructive pulmonary disease, nicotine sublingual (one trial, N = 370) and bupropion (two trials, N = 579) were superior to placebo.

Within this review's strict selection criteria, evidence on the long-term effectiveness of pharmacotherapy is limited and/or mixed on the efficacy of pharmacotherapy for hospitalized patients, adolescents, pregnant women, those with mental disorders, those with substance abuse, low-income smokers, and cancer patients.

Combination Therapy

There were few RCTs of combined therapy versus monotherapy. The addition of nicotine gum to nicotine patch was superior compared with nicotine patch alone for the six-month continuous abstinence rate, but not at one year of follow-up (one RCT, N = 374). The combined therapies of nicotine patch and inhaler or bupropion and NRT did not provide an added effect compared with monotherapy (four RCTs). Thus, combined therapy should be prescribed with caution because the evidence is scarce and the adverse effects unknown.

Economic Results

A review of existing economic analyses showed that varenicline (with or without behavioural interventions) seemed to be the most cost-effective therapy, followed by bupropion and NRT. However, the results should be interpreted with caution because most of the studies comparing varenicline with bupropion or NRT were funded by the manufacturer of varenicline.

Our economic analyses targeting a general population showed that, for all ages and gender, bupropion and varenicline dominated (i.e., cost less and was more effective) nicotine gum, patch, lozenge, and inhaler. If providers' willingness to pay was greater than $10,000 per QALY, then varenicline was preferred.


Few studies examining the long-term benefit of combining pharmacotherapies were found.

Evidence of the long-term benefits and safety of combination therapy remains uncertain.


Given the available evidence, all pharmacotherapies under review are efficacious in helping the general population (relatively healthy smokers) quit smoking for at least six to 12 months. Thus, NRT, bupropion, and varenicline could all be used as aids for smoking cessation in the general population of smokers. Evidence on the long-term effectiveness of pharmacotherapy is limited for hospitalized patients, adolescents, pregnant women, those with mental disorders, those with substance abuse, low-income smokers, and cancer patients. Bupropion or NRT were found to be effective for those with cardiovascular disease or chronic obstructive pulmonary disease who wanted to quit smoking.

For pharmacological interventions targeting a general population, bupropion and varenicline were dominating options (cost less and were more effective) over nicotine gum, patch, lozenge, and inhaler. If a provider's willingness to pay was greater than $10,000 per QALY gained, varenicline was the optimal treatment choice compared with NRT and bupropion.


  1. Supplementary tables: annual (February-December 2007). In: Canadian Tobacco Use Monitoring Survey (CTUMS) 2007. Ottawa: Health Canada; 2008 [cited 2008 Nov 25]. Available from:
  2. Tobacco and the health of Canadians. Ottawa: Physicians for a Smoke-Free Canada; 2008.
  3. Leatherdale ST, Shields M. Smoking cessation: intentions, attempts and techniques. Health Rep. 2009 [cited 2009 Jul 31];20(3):31-9. Available from:
  4. Cigarette smoking among adults--United States, 2000. MMWR Morb Mortal Wkly Rep. 2002 Jul 26;51(29):642-5.
  5. Lader D. Smoking-related behaviour and attitudes, 2007. A report on research using the National Statistics Omnibus Survey produced on behalf of the NHS Information Centre for health and social care. Newport (UK): Office for National Statistics; 2008. (Omnibus survey report no. 36).
  6. Hughes JR, Keely J, Naud S. Shape of the relapse curve and long-term abstinence among untreated smokers. Addiction. 2004 Jan;99(1):29-38.