Background: Five nicotine replacement therapies and two non–nicotine replacement pharmacotherapies have been endorsed by the Public Health Service as reliably increasing abstinence rates in smoking cessation compared with placebo. Comparisons among the different therapies are difficult because of the lack of head-to-head studies and differences in the methodologies of studies comparing individual agents to placebo. Piper and colleagues studied the relative effectiveness of five pharmacotherapies in a placebo-controlled clinical trial.

The Study: Public advertising campaigns were used to recruit adults who smoked more than nine cigarettes daily on average and were motivated to quit. Exclusions included low alveolar carbon monoxide level (less than 9 ppm or less), use of any other form of tobacco, and current use of or contraindication to any of the study agents. After screening for eligibility, participants completed three baseline sessions that included collection of extensive demographic, medical, and tobacco-related history plus physical examination and targeted physiologic assessment (carotid intima media thickness, brachial artery reactivity, and small-particle lipoprotein testing).

The 1,504 participants were randomized to one of the following six treatment groups: bupropion SR (Zyban), nicotine lozenge, nicotine patch, nicotine patch plus nicotine lozenge, bupropion SR plus nicotine lozenge, or placebo. All participants received six individual counseling sessions. Follow-up visits were scheduled for one, two, four, and eight weeks after the quit date. Medication compliance, adverse effects, vital signs and smoking status were reviewed at each visit. Smoking status was assessed as continuous abstinence after the quit date, or seven-day abstinence (any use of tobacco in the previous seven days). Self-reported status was confirmed by expired carbon monoxide levels.

Results: The treatment groups did not differ notably in any important variables after randomization. All treatments showed higher rates of cessation during the study and six months after the quit date than did placebo (see accompanying table). After statistical correction, only the patch and the two combination therapies were effective at the end of treatment (eight weeks) and only the patch plus lozenge was effective at six months after the quit date. All treatments improved the time to relapse better than placebo, but after statistical correction the gain in time to relapse remained significant only for combination therapies.

Participants used an average of 77 percent of the study medications, with the lowest usage rates for lozenge. Four participants withdrew because of side effects, and one serious adverse event (seizures) was recorded. The most common adverse events were headache, nausea, sleep disturbances, and skin irritation. Participants using combination therapies reported more side effects than other participants.

Conclusion: The authors conclude that the nicotine patch plus lozenge had the greatest effect in this study and was the only intervention notably better than placebo at six months after the quit date. They speculate that the high six-month quit rates achieved by the placebo group may be attributed to the motivation of participants and effective counseling. They also note that adherence rates were lowest for the lozenge and suggest this may be because of “as needed” dosing.