Skip to main content Skip to left navigation Skip to footer
U.S. Department of Health and Human Services
Skip Left Navigation

Browse by Topic

Tobacco Smoking Cessation in Adults and Pregnant Women, Behavioral and Pharmacotherapy Interventions, 2015

* Indicates an old grade definition


Recommendations: Behavioral and Pharmacotherapy Interventions for Tobacco Smoking Cessation in Adults and Pregnant Women

Discussion:

Burden of Disease

Tobacco use is the leading preventable cause of disease, disability, and death in the United States. Smoking increases the risk for all-cause mortality and has been found to increase risk for various types of cancer (such as lung, liver, and colorectal), respiratory diseases (such as chronic obstructive pulmonary disease and tuberculosis), cardiovascular disease (such as stroke from secondhand smoke), diabetes, impaired immune function and autoimmune disease (such as rheumatoid arthritis), eye disease (such as age-related macular degeneration), and erectile dysfunction in men. Smoking early in pregnancy has been found to cause ectopic pregnancy and orofacial clefts in infants.According to National Health Interview Survey data from 2013, a total of 42.1 million adults in the United States (17.8%) smokes cigarettes. Rates are higher among men; adults aged 24 to 44 years; lesbian, gay, bisexual, or transgender adults; multiracial groups and American Indians or Alaska Natives; persons whose highest education level attained is a GED; and persons living below the poverty level. Adults with mental health conditions have higher smoking rates and tend to smoke a higher average number of cigarettes than adults without mental health conditions. Approximately 27% of persons with a mental health or substance use disorder smoke.Based on data from the Pregnancy Risk Assessment and Monitoring System, 23.2% of women smoked during the 3 months before conception. Data from 2011 showed that 55% of women quit smoking during pregnancy, roughly 10% of pregnant women smoked during the last 3 months of pregnancy, and 40% of women relapsed within 6 months of delivery. Smoking during pregnancy is associated with fetal growth restriction, preterm delivery, and sudden infant death syndrome. Prenatal smoking contributed to an estimated 5.3% to 7.7% of preterm deliveries, 13.1% to 19.0% of low-birthweight term deliveries, 23.2% to 33.6% of cases of sudden infant death syndrome, and 5.0% to 7.3% of preterm-related deaths in 2002.Quitting smoking is one of the most important and yet challenging preventive health measures a person can take to improve his or her health. Most smokers make several serious attempts to quit before achieving permanent abstinence.

Scope of Review

The current review focused on appraising evidence on interventions for smoking cessation that are relevant to primary care (behavioral interventions, pharmacotherapy, and complementary or alternative therapy) in adults, including pregnant women. Only pharmacotherapy interventions that were approved by the FDA as first-line agents for smoking cessation were included (NRT, bupropion SR, and varenicline). Health outcomes included all-cause mortality, tobacco-related mortality and morbidity, and perinatal mortality and morbidity. Smoking-cessation outcomes had to be reported at 6 months or later to be included; findings on smoking reduction (based on frequency or quantity only), intention to quit, and cessation at less than 6 months were excluded. The USPSTF also evaluated adverse events reported at any time point.

Effectiveness of Interventions

Nonpregnant Adults

Behavioral Interventions
The USPSTF reviewed 11 good- or fair-quality systematic reviews on behavioral interventions, including complementary or alternative therapies, and smoking-cessation outcomes. Evidence on increasing smoking abstinence was strongest for physician and nurse advice, tailored self-help materials, and telephone counseling. Based on a 2013 systematic review of 28 studies (n = 22,239), rates of smoking abstinence at 6 months or more were 8.0% in groups that received physician advice compared with 4.8% in groups that received no advice or usual care (risk ratio [RR], 1.76 [95% CI, 1.56 to 1.96]). Based on pooled analyses of 35 studies from a 2013 systematic review that evaluated advice delivered by nurses, 13.3% of participants who received interventions from nurses achieved smoking abstinence at 6 months or more compared with 11.3% of those who received usual care or minimal intervention (RR, 1.29 [CI, 1.20 to 1.39]). When stratified by intensity level, both minimal advice (defined as a single session lasting <20 minutes with ≤1 follow-up session) and intensive advice (defined as a single session lasting ≥20 minutes or >1 follow-up session) from a physician significantly increased cessation rates compared with no advice. In a subset of 15 trials that directly compared intensive versus minimal advice from physicians, intensive advice significantly increased smoking abstinence compared with minimal advice (RR, 1.37 [CI, 1.20 to 1.56]).Based on pooled analyses of 32 studies from a 2014 review of print-based self-help materials (n = 40,890), participants who received tailored self-help materials showed a significantly higher rate of smoking abstinence at 6 months or more than control participants (7.1% vs. 5.8%; RR, 1.28 [CI, 1.18 to 1.37]). However, no significant improvement was seen when nontailored materials were compared with no self-help materials (RR, 1.06 [CI, 0.98 to 1.16]), based on pooled analyses of 33 studies (n = 29,495). A 2013 review on telephone counseling interventions found that recruiter-initiated telephone support improved rates of smoking cessation at 6 months or more.Based on pooled analyses of 12 studies (n = 30,182), abstinence rates were significantly higher in participants who received several counseling sessions via a telephone quit line than in those who received only a single session or self-help materials (10.7% vs. 7.6%; RR, 1.41 [CI, 1.20 to 1.66]). Pooled analyses of 51 studies (n = 30,246) found that recruiter-initiated telephone counseling that did not result from patients’ calls to help lines also showed a benefit on smoking abstinence rates at 6 months or more compared with controls (13.1% vs. 9.7%; RR, 1.27 [CI, 1.20 to 1.36]). The USPSTF also reviewed the 2008 Public Health Service guidelines on treating tobacco use and dependence for additional details on intensity and content of behavioral interventions, as well as type of staff providing counseling for smoking cessation. The USPSTF also considered evidence from additional reviews on mobile phone and Internet-based interventions. Although findings suggested a benefit, there were too few studies and the studies were too heterogeneous to draw definitive conclusions. The USPSTF also considered evidence from reviews on biomedical risk assessment interventions, exercise, and hypnotherapy for smoking cessation; the studies were too limited to draw conclusions. A 2014 review on the use of acupuncture for smoking cessation (9 studies; n = 1892) did not find increased rates of smoking cessation at 6 to 12 months.
Pharmacotherapy
The USPSTF reviewed 3 good-quality systematic reviews on pharmacotherapy for smoking cessation in nonpregnant adults. A 2012 systematic review on NRT (117 studies; n = 51,265) found that 17.3% of participants taking any form of NRT achieved abstinence at 6 months or more compared with 10.3% of participants receiving placebo or taking no NRT (RR, 1.60 [CI, 1.53 to 1.68]). Those who used 2 forms of NRT achieved higher abstinence rates than those who used just 1 (20.6% vs. 15.6%; RR, 1.34 [CI, 1.18 to 1.51]). All studies of dual NRT used the patch plus another type of NRT (such as gum, nasal spray, or inhaler), although which form served as the control varied. All single forms of NRT significantly improved cessation rates at 6 months or more (gum: RR, 1.49 [CI, 1.40 to 1.60]; patch: RR, 1.64 [CI, 1.52 to 1.78]; tablets or lozenges: RR, 1.95 [CI, 1.61 to 2.36]). A 2014 systematic review on the use of antidepressants for smoking cessation (44 studies; n = 13,728) found that bupropion SR was associated with a significantly higher rate of smoking abstinence at 6 months or more than placebo or no bupropion SR (19.7% vs. 11.5%; RR, 1.62 [CI, 1.49 to 1.76]). Based on a smaller number of studies (14 studies; n = 6166), a 2012 systematic review found that varenicline was associated with a higher cessation rate than placebo (28.0% vs. 12.0%; RR, 2.27 [CI, 2.02 to 2.55]). Pooled analyses found that NRT plus bupropion SR was more effective than bupropion SR alone (RR, 1.24 [CI, 1.06 to 1.45]; 4 studies; n= 1991); however, it was not more effective than NRT alone (RR, 1.19 [CI, 0.94 to 1.51]; 12 studies). Smaller subsets of studies from these reviews directly compared types of pharmacotherapy for smoking cessation. Abstinence rates among participants using NRT versus bupropion SR at 6 months or more did not significantly differ (8 studies; n = 4086). Two studies (n = 778) compared NRT and varenicline and found no significant difference between groups. Four studies evaluated bupropion SR versus varenicline; although not all of the studies found a significant difference, a pooled estimate (n = 1810) found a lower cessation rate with bupropion SR than varenicline (RR, 0.68 [CI, 0.56 to 0.83]).
Combinations of Behavioral Interventions and Pharmacotherapy
Combinations of behavioral counseling and pharmacotherapy for smoking cessation were also effective. A 2012 good-quality systematic review (40 studies; n = 15,021) found that participants who received combination pharmacotherapy and intensive behavioral counseling had a higher abstinence rate at 6 months or more compared with control participants who received usual care, self-help materials, or brief advice on quitting (which was less intensive than the counseling or support given to the intervention groups) (14.5% vs. 8.3%; RR, 1.82 [CI, 1.66 to 2.00]). Most studies (27 of 40) used NRT as the pharmacotherapy and offered at least 4 behavioral counseling sessions (33 of 41). Another good-quality systematic review found that abstinence rates at 6 months or more were higher in participants who received behavioral support as an adjunct to pharmacotherapy than in those who received pharmacotherapy alone (21.4% vs. 18.3%; RR, 1.16 [CI, 1.09 to 1.24]). Most studies (27 of 38) offered NRT alone as the pharmacotherapy. Participants in the control group may have also received some counseling or support, but it was less intensive than in the intervention group. Most studies (28 of 38) offered at least 91 minutes of total contact time, and many (16 of 38) offered greater than 300 minutes.
ENDS
The USPSTF identified only 2 fair-quality RCTs that reported on the effects of e-cigarettes for stopping conventional cigarette smoking. One study of persons not attempting or wishing to quit found a higher abstinence rate at 12 months in participants using e-cigarettes than in control participants (11.0% vs. 4%; P = 0.04), whereas a larger study found no significant difference in abstinence rates at 6 months. Neither trial was conducted in the United States.

Pregnant Women

Behavioral Interventions
Based on a good-quality systematic review of 86 studies done in 2013, the USPSTF found that behavioral interventions in pregnant women (including behavioral counseling, feedback, health education, incentives, and social support) are effective at improving rates of smoking cessation as well as perinatal health outcomes. Compared with control participants (most often defined as usual care or less intensive interventions), pregnant women who received any type of behavioral intervention before the third trimester had higher cessation rates late in pregnancy (15.2% vs. 11.2%; RR, 1.45 [CI, 1.27 to 1.64]). Their children also had improved mean birthweight (mean difference, 40.78 g [CI, 18.45 to 63.10 g]), rates of low birthweight (RR, 0.82 [CI, 0.71 to 0.94]), and rates of preterm birth (RR, 0.82 [CI, 0.70 to 0.96]). When analyzed separately, behavioral counseling (the behavioral intervention in most studies) was effective in increasing abstinence rates in late pregnancy and mean birthweight in infants.
Pharmacotherapy
The USPSTF found a much smaller body of evidence on pharmacotherapy for smoking cessation in pregnant women. Only 5 RCTs (n = 11,922) evaluating NRT use were found; there were no published trials on the use of bupropion SR or varenicline as an intervention for smoking cessation in pregnant women. The NRT studies reported very low (as low as <25%) adherence rates, which limited the interpretability of findings. Meta-analysis of these 5 trials showed no significant improvement in rates of smoking abstinence late in pregnancy (10.8% vs. 8.5%; RR, 1.24 [CI, 0.95 to 1.64]). Given the small number of studies that reported on perinatal health outcomes, meta-analysis was not done. Although results on preterm birth were mostly in the direction of benefit, only 1 study showed a significant benefit. Reported results on mean birthweight and stillbirth were mixed. The largest trial (n = 1050) reported a higher rate of survival with no impairment at 2 years in children whose mothers received NRT during pregnancy.

Potential Harms of Interventions

Nonpregnant Adults

Behavioral Interventions
Evidence on harms of behavioral interventions for smoking cessation is limited. Based on the evidence reviewed by the USPSTF, only minor adverse events related to ear acupuncture, ear acupressure, and other auriculotherapy were identified. Adverse events related to other forms of behavioral interventions were not reported.
Pharmacotherapy
A 2014 fair-quality systematic review of 21 studies (n = 11,647) found that NRT use was associated with a higher rate of all cardiovascular adverse events (RR, 1.81 [CI, 1.35 to 2.43]) compared with placebo. This result seems to be driven by minor events, such as tachycardia and arrhythmia. No significant increase in major cardiovascular adverse events was found (RR, 1.38 [CI, 0.58 to 3.26]). An older fair-quality review from 2010 also found higher rates of heart palpitation and chest pain, nausea and vomiting, gastrointestinal symptoms, and insomnia in NRT users than control participants. The patch was associated with an increase in skin irritation, and oral NRT was associated with throat soreness, mouth ulcers, and hiccups. Based on a subset of 8 studies, this older review also found no difference in mortality rates in participants using NRT versus control participants (odds ratio, 0.74 [CI, 0.33 to 1.67]). The same systematic review also evaluated the potential harms of bupropion SR for smoking cessation (27 studies; n = 10,02). No significant difference was found between groups using bupropion SR versus placebo in all cardiovascular adverse events (RR, 1.03 [CI, 0.71 to 1.50]). A trend toward reduction in major cardiovascular adverse events was found in the bupropion SR group (RR, 0.57 [CI, 0.31 to 1.04]). A good-quality review (33 studies; n = 9631) found an increased rate of serious adverse events in bupropion SR users versus those using placebo or controls that was of borderline significance (RR, 1.30 [CI, 1.00 to 1.69]). On the basis of 19 studies, the current USPSTF review found no significant increase in serious psychiatric events with bupropion SR (RR, 0.60 [CI, 0.28 to 1.28]). It found 10 reported cases of seizures among 13,000 study participants. Bupropion SR labels include a boxed warning about serious neuropsychiatric events in persons who take bupropion SR for smoking cessation. Concerns have been raised about varenicline and potential cardiovascular and neuropsychiatric harms. In 2012, the FDA issued a safety communication indicating a nonsignificant increase in major adverse cardiovascular events. Based on a fair-quality systematic review done in 2014, a meta-analysis of 18 studies (n = 9072) showed no significant increase in all cardiovascular adverse events (RR, 1.24 [CI, 0.85 to 1.81]) or major cardiovascular adverse events (RR, 1.44 [CI, 0.73 to 2.83]). Varenicline labels contain a boxed warning about serious neuropsychiatric events; however, none of the evidence reviewed by the USPSTF reported on this outcome. A good-quality systematic review done in 2012 found an increased risk for 1 or more serious adverse events in participants using varenicline compared with those receiving placebo (RR, 1.36 [CI, 1.03 to 1.81]), but the type of serious adverse event was not specified. Most recently, the FDA issued a warning and approved updated labeling to describe a potential but rare occurrence of seizures with varenicline use and a potential interaction between varenicline and alcohol.
Combinations of Behavioral Interventions and Pharmacotherapy
The USPSTF did not identify any reports of adverse events related to combinations of behavioral interventions and pharmacotherapy.
ENDS
Neither of the 2 trials on e-cigarette use for conventional smoking cessation reported any serious adverse events related to product use or any difference in the frequency of adverse events. One study reported a higher number of serious adverse events in the nicotine e-cigarette group (27 [19.7%]) than in the nicotine patch group (14 [11.8%]). However, the authors stated that they did not find any evidence on an association with the study product, and there was no difference in overall incidence of adverse events (serious and nonserious) between groups (incidence rate ratio, 1.05 [CI, 0.82 to 1.34]). Although it was not reported in the 2 trials, concerns have been raised about potentially harmful ingredients in e-cigarettes. Because of the large number and variation of available devices and cartridge fluids and the lack of regulation, the toxicity and safety of e-cigarettes and ENDS are extremely difficult to determine uniformly.Inadvertent poisoning in children who mishandle nicotine cartridges is also a concern, and the number of calls to poison centers involving nicotine-containing e-cigarette liquids has increased from 1 per month in 2010 to 215 per month in 2014. Based on estimates from a recent small survey, dual use of e-cigarettes and conventional cigarettes may be high: 9.4% of persons who smoke conventional cigarettes reported concurrently using e-cigarettes, and 76.8% of persons who use e-cigarettes reported concurrently smoking conventional cigarettes. Further research is needed to better understand how dual use of e-cigarettes and conventional cigarettes may affect attempts at cessation and initiation of smoking.

Pregnant Women

Behavioral Interventions
The USPSTF did not find any serious adverse events related to behavioral interventions for smoking cessation in pregnant women. According to 1 systematic review, an increase in smoking was reported in 4 out of 86 studies.
Pharmacotherapy
Evidence on harms from pharmacotherapy interventions for smoking cessation in pregnant women is limited. The USPSTF found no evidence of perinatal harms, although the number of studies was low and they were underpowered for rare outcomes. Similarly, there were too few studies to make any definitive determination on maternal harms of NRT. Maternal harms of NRT that have been reported in some studies include increased rates of cesarean delivery, slightly increased blood pressure (1 study reported an increase of 0.02 mm Hg per day in diastolic blood pressure over time, and skin reaction to the NRT patch.

Estimate of Magnitude of Net Benefit

Nonpregnant Adults

Behavioral Interventions
The USPSTF found convincing evidence that behavioral interventions (such as physician and nurse advice, tailored self-help materials, and telephone counseling) provided or referred to by primary care providers substantially improve achievement of smoking cessation in nonpregnant adults. The USPSTF found adequate evidence that there are no harms of behavioral interventions. The USPSTF concludes with high certainty that the net benefit of providing behavioral interventions for smoking cessation in nonpregnant adults is substantial.
Pharmacotherapy
The USPSTF found convincing evidence that pharmacotherapy with NRT, bupropion SR, or varenicline substantially improves achievement of smoking cessation in nonpregnant adults. Evidence is also convincing that using 2 types of NRT moderately improves achievement of smoking cessation over single forms and that the addition of NRT to treatment with bupropion SR provides additional benefit over use of bupropion SR alone. The USPSTF found adequate evidence that the harms of pharmacotherapy, including serious cardiovascular adverse events and neuropsychiatric events, are small. It concludes with high certainty that the net benefit of providing pharmacotherapy for smoking cessation in nonpregnant adults is substantial.
Combinations of Behavioral and Pharmacotherapy Interventions
The USPSTF found convincing evidence that combinations of behavioral plus pharmacotherapy interventions substantially increase achievement of smoking cessation in nonpregnant adults. The USPSTF found adequate evidence that the harms of combined interventions are small. The USPSTF concludes with high certainty that the net benefit of providing combined interventions for smoking cessation in nonpregnant adults is substantial.
ENDS
The USPSTF concludes that the current evidence on the use of ENDS for conventional smoking cessation is insufficient. Evidence is lacking and conflicting, and the balance of benefits and harms cannot be determined. Given the established safety and effectiveness of behavioral and pharmacotherapy interventions, the USPSTF recommends that primary care providers direct patients who smoke to these other interventions.

Pregnant Women

Behavioral Interventions
The USPSTF found convincing evidence that behavioral interventions (such as behavioral counseling, feedback, health education, incentives, and social support) substantially improve achievement of smoking cessation in pregnant women, increase birthweight, and decrease preterm birth in their children. The USPSTF found convincing evidence that there are no harms of behavioral interventions and concludes with high certainty that the net benefit of providing behavioral interventions for smoking cessation in pregnant women is substantial.
Pharmacotherapy
The USPSTF concludes that the current evidence on pharmacotherapy for smoking cessation in pregnant women is insufficient. Evidence on NRT is limited and conflicting, evidence on other forms of pharmacotherapy is lacking, and the balance of benefits and harms cannot be determined.

Response to Public Comment

A draft version of this recommendation statement was posted for public comment on the USPSTF Web site from 5 May to 1 June 2015. Several comments requested clarification about which types of interventions are being recommended—pharmacotherapy, behavioral interventions, or combinations of both. Several comments also expressed concern that the draft language for the recommendation may cause clinicians to offer only 1 type of intervention. The USPSTF clarified that both intervention types (pharmacotherapy and behavioral interventions) are effective and recommended; combinations of interventions are most effective, and all should be offered. The best and most effective combinations are those that are acceptable to and feasible for an individual patient. Clinicians should consider the patient’s specific medical history and preferences and offer and provide the combination that works best for that patient. Comments also sought clarification about which populations are included in the I statement for ENDS; the USPSTF clarified that the I statement for ENDS includes pregnant women. A few comments requested additional implementation resources. The USPSTF revised the recommendation to include a table that highlights effective components of behavioral interventions and provided links to additional resources on pharmacotherapy options and resources for pregnant women. The USPSTF revised its language on ENDS to reflect current terminology.

How Does Evidence Fit With Biological Understanding?

Because of the well-established health benefits of smoking cessation, most of the research on interventions for smoking cessation focuses on cessation (rather than health outcomes) as a primary outcome. The current review identified 1 study of middle-aged men at high risk for cardiorespiratory disease that found a significantly smaller number of deaths from respiratory illness at 33 years of follow-up in participants who received advice from medical practitioners. The study also found favorable effects on all-cause mortality, coronary disease mortality, and lung cancer incidence and mortality at 20 years of follow-up, although these effects were not significant.

Updates of Previous USPSTF Recommendation

In 2009, the USPSTF recommended that clinicians ask all adults about tobacco use and provide interventions for smoking cessation for those who use tobacco products (A recommendation). It also recommended that clinicians ask all pregnant women about tobacco use and provide augmented, pregnancy-tailored counseling for those who smoke (A recommendation). The current recommendation updates and is consistent with the 2009 recommendation. In addition, the USPSTF reviewed evidence on ENDS, an emerging tobacco product, but found insufficient evidence to recommend for or against its use for smoking cessation. The USPSTF also reviewed updated evidence on pharmacotherapy interventions for smoking cessation in pregnant women and found that the evidence is still insufficient to make a recommendation.
Other Sections:
Numerous professional societies and health organizations, including the American College of Physicians, American College of Preventive Medicine, American Heart Association, and American Congress of Obstetricians and Gynecologists recommend that clinicians screen for tobacco use and provide interventions to patients who smoke. More recently, with the emergence and increased use of e-cigarettes, some organizations have incorporated statements about e-cigarettes into their guidelines. Both the American Congress of Obstetricians and Gynecologists and the American Heart Association recommend that e-cigarettes be included in smoking screening questions. The American Heart Association also concludes that evidence is insufficient to counsel patients on the use of e-cigarettes as a primary cessation aid. The American Academy of Family Physicians has updated its recommendations, which are consistent with those of the USPSTF.
Agency for Healthcare Research and Quality - Advancing Excellence in Health Care
AHRQ ePSS

You are now leaving ePSS Electronic Preventive Services Selector website and going to
This website contains links to other federal and state agencies and private organizations. The Agency for Healthcare Research and Quality (AHRQ) cannot attest to the accuracy of information provided by these nonfederal website links. Linking to nonfederal sites does not constitute an endorsement by AHRQ or any of its employees of the sponsors or the information and products presented on the sites. You will be subject to the destination site's privacy policy when you leave the AHRQ website. Thank you for visiting epss.ahrq.gov.

Continue Cancel