Dr Neil Walker Statistician, Oxford Biomedical Research Centre, Oxford
Abstract
Smoking cessation services in England provide both material and psychological support for smokers wishing to quit. The standard pharmacological treatment offered is Nicotine replacement Therapy (NRT), although the main alternative (Champix/varenicline) has been demonstrated to be more efficacious in Randomised Controlled Trials (RCT). We analysed observational data collected by one of the primary smoking cessation services in England to assess whether the effectiveness of the treatments are moderated by the type of counselling service provided and individual—level variables, i.e. age, gender, ethnicity, cigarette dependency (Fagerstrom score) and GP ward deprivation score.
Generalised Linear Mixed Model (GLMM) analysis, incorporating 4-week Carbon Monoxide (CO)- validated quit as a response, revealed the presence of statistically significant interactions between treatment and (i) dependency (odds ratio [OR] and 95% confidence interval [CI] for Champix+Fagerstrom: 1.08 [1.04, 1.13]) and (ii) service use (OR and CI for Champix+GP: 0.68 [0.54, 0.86], Champix+pharmacy: 0.57 [0.43, 0.76] and Champix+other: 0.36 [0.12, 1.07] with Champix+one-to-one and all NRT combinations fixed at unity). These results imply that whilst Champix is associated with a higher overall quit rate than NRT, the relative effectiveness of NRT is enhanced when the smoker has a high level of dependence. Furthermore, Champix is found to be especially effective in conjunction with one-to-one support whereas NRT appears to be more effective in combination with other forms of support.
Source of funding: Neil Walker funded in-post by NIHR.
Declaration of interest: The authors declare no conflict of interest with respect to the presented research.
