Abstract
BACKGROUND: Smoking cessation services are available in England to provide assistance to those wishing to quit smoking. Data from one such service were analysed in order to investigate differences in quit rate between males and females prescribed with different treatments. METHODS: A logistic regression model was fitted to the data using the binary response of self-reported quit (failed attempt = 0, successful attempt = 1), validated by Carbon Monoxide (CO) monitoring, 4 weeks after commencing programme. Main effects fitted were: client gender; age; region; the type of advisory sessions; and pharmacotherapy, Nicotine Replacement Therapy (NRT) or Varenicline. A second model was fitted including all main effects plus two-way interactions except region. These models were repeated using 12-week self-reported quit as the outcome. RESULTS: At 4 weeks, all main effects were statistically significant, with males more likely (odds ratio and 95 % CI, females v males = 0.88 [0.79-0.97]), older smokers more likely (adjusted odds ratios [OR] and 95 % confidence interval [CI] respectively for groups 20-29, 30-49, 50-69 and 70+ vs 12-19 age group: 1.79 [1.39-2.31], 2.12 [1.68-2.68], 2.30 [1.80-2.92] and 2.47 [1.81-3.37] and for overall difference between groups, χ(2)(4) = 53.5, p < 0.001) and clients being treated with Varenicline more likely to have successfully quit than those on NRT (adjusted OR and 95 % CI for Varenicline vs NRT = 1.41 [1.21-1.64]). Statistically significant interactions were observed between (i) gender and type of counselling, and (ii) age and type of counselling. Similar results were seen in relation to main effects at 12 weeks except that type of counselling was non-significant. The only significant interaction at this stage was between gender and pharmacotherapy (adjusted OR and 95 % CI for females using Varenicline versus all other groups = 1.43 [1.06-1.94]). CONCLUSION: Gender and treatment options were identified as predictors of abstinence at both 4 and 12 weeks after quitting smoking. Furthermore, interactions were observed between gender and (i) type of counselling received (ii) pharmacotherapy. In particular, the quit rate in women at 12 weeks was significantly improved in conjunction with Varenicline use. These findings have implications for service delivery.