EVIDENCE-BASED ANSWER
NICOTINE REPLACEMENT THERAPY (NRT), bupropion, nortriptyline, clonidine, and varenicline are all effective, although insufficient evidence exists to recommend one intervention over another (SOR: A, systematic reviews).
Effective nonpharmacologic interventions include brief physician advice and more intensive counseling, such as proactive telephone counseling, group and individual counseling, and use of quit lines (SOR: A, systematic reviews).
Evidence summary
NRT. A Cochrane review of 111 randomized controlled trials (RCTs) with a total of >40,000 subjects evaluated abstinence rates after 6 months of NRT and placebo or no treatment.1 All forms of NRT increased abstinence vs placebo or no treatment, independent of setting, duration of treatment, and intensity of nonpharmacologic therapies. Overlapping confidence intervals suggested that no one form of NRT was superior. (The TABLE summarizes all the studies discussed here.)
Bupropion. A Cochrane review of 36 RCTs (N=11,140) showed higher abstinence rates with bupropion than placebo after ≥6 months of follow-up (average quit rate 17% vs 9%). Duration (6 vs 12 months) and intensity (150 vs 300 mg) of therapy didn’t influence the results.2 Six separate RCTs comparing bupropion plus NRT with NRT alone showed significant heterogeneity, but found no significant differences using a mixed-effects model.2
Nortriptyline. A Cochrane review that pooled results from 6 RCTs (N=975) showed superior 6-month abstinence rates for nortriptyline compared with placebo.2 Adding nicotine patches in other RCTs (N=1219) didn’t change abstinence rates.2 No long-term studies have examined other tricyclic antidepressants.
Clonidine. A pooled analysis of 6 RCTs found clonidine superior to placebo after ≥12 weeks of follow-up.3 Results were heavily influenced by one trial limited to heavy smokers and poor tolerability due to adverse effects of therapy, especially sedation and dry mouth.
Nicotine receptor partial agonists and antagonists. Standard dose varenicline was more than twice as likely as placebo to produce abstinence at 6 months in a Cochrane review of 10 RCTs.4 Lower doses were slightly less effective, but had fewer side effects. Adverse effects included mild to moderate nausea and sleep disorders; causation has not been established between varenicline and rare postmarketing reports of severe psychiatric disturbances.4,5
The pooled results of 3 RCTs suggested that varenicline was superior to bupropion, but different abstinence rates for bupropion users in other placebo-controlled trials necessitate caution in interpreting these results.4 Varenicline was not superior to NRT.4
One RCT (N=48) comparing nicotine patches plus the nicotine antagonist mecamylamine with patches plus placebo found improved abstinence rates at 6 and 12 months; a larger RCT didn’t support these findings.6
Table
How effective are smoking cessation interventions?
| Intervention | No. of studies | Effect size* (95% confidence interval) | Total N |
|---|---|---|---|
| NRT vs placebo or no treatment1 | 111 | 1.58 (1.50-1.66) | >40,000 |
| Bupropion vs placebo2 | 36 | 1.69 (1.53-1.85) | 11,140 |
| Bupropion 300 mg/d vs 150 mg/d2 | 3 | 1.08 (0.93-1.26) | 2042 |
| Bupropion + NRT vs NRT2 | 6 | 1.23 (0.67-2.26) | 1106 |
| Nortriptyline vs placebo2 | 6 | 2.03 (1.48-2.78) | 975 |
| Nortriptyline + NRT vs NRT2 | 4 | 1.29 (0.97-1.72) | 1219 |
| Clonidine vs placebo3 | 6 | 1.63 (1.22-2.18) | 776 |
| Varenicline vs placebo, standard dose4 Varenicline vs placebo, low dose4 Varenicline vs bupropion4 Varenicline vs NRT4 | 10 4 3 2 | 2.31 (2.01-2.66) 2.09 (1.56-2.78) 1.52 (1.22-1.88) 1.13 (0.94-1.35) | 4443 1272 1622 778 |
| Mecamylamine + NRT vs NRT+ placebo6 | 1 | 37.5% vs 12.5% | 48 |
| Simple advice vs usual care10-13 | 17 | 1.66 (1.42-1.94) | 15,930 |
| Patient-initiated telephone quit line vs usual care14 | 9 | 1.37 (1.26-1.50) | 24,000 |
| NRT, nicotine replacement therapy. *An effect size >1.0 means that patients using this intervention are more likely not to smoke at 6 to 12 months; larger numbers correlate with greater effectiveness | |||
These interventions are not supported
A review of placebo-controlled RCTs found no evidence of improved abstinence at 6 to 12 months with fluoxetine, paroxetine, sertraline, venlafaxine, citalopram, or monoamine oxidase inhibitors, alone or as adjuncts to NRT.2
No good evidence supports using anxiolytics, silver acetate, Nicobrevin (a nicotine-free smoking cessation aid), lobeline, or naltrexone for smoking cessation.7-9
Simple advice and quit lines help
A Cochrane review of 17 RCTs found that simple advice improved quit rates and maintenance of abstinence at 12 months.10-13
A review of 9 RCTs (N>24,000). found that telephone quit lines increased abstinence, particularly after more than 2 sessions.14
No high-quality studies demonstrate the effectiveness of acupuncture, hypnotherapy, or acupressure for smoking cessation.15,16
Recommendations
The Agency for Health Care Research and Quality recommends counseling (including individual, group, and telephone sessions and brief physician advice) in addition to sustained-release bupropion, NRT, and varenicline as first-line agents. It considers clonidine and nortriptyline second-line therapies.17