Effects on abstinence of nicotine patch treatment prior to quitting smoking: a parallel, two-arm, pragmatic randomised trial

Objectives: Standard smoking cessation pharmacotherapy is recommended in the post-quit period, but may also facilitate abstinence if used prior to quitting. The objective was to examine the effectiveness and cost-effectiveness of nicotine patch worn for four weeks before a quit attempt. Design: Randomised controlled open-label trial. Setting: Primary care and smoking cessation clinics in England enrolled patients between 13/08/2012 and 10/03/2015. Participants: 1792 people, mainly middle-aged and of lower socioeconomic status, who were daily smokers with tobacco dependence. Interventions: Participants were randomised 1:1, using concealed randomly permuted blocks stratified by centre, to either standard smoking cessation pharmacotherapy and behavioural support or the same treatment supplemented by four weeks of 21mg nicotine patch use prior to quitting: ‘preloading’. Main outcome measures: The primary outcome was biochemically confirmed six-month prolonged abstinence and secondary outcomes were prolonged abstinence at four weeks and 12 months. Results: 899 people were allocated to the preloading arm and 893 to the control arm. Biochemically validated six-month abstinence, was achieved by 157/899 (17.5%) of the intervention arm and 129/893 (14.4%) of the control arm; difference (95% confidence interval) 3.0% (-0.4% to 6.4%), odds ratio (OR) 1.25 (0.97 to 1.62), p=0.08 in the primary analysis. There was an imbalance between arms in the frequency of use of varenicline for post-cessation medication and planned adjustment for this gave an odds ratio for the effect of pre-loading of 1.34 (1.03, 1.73), p=0.03, difference 3.8% (0.4% to 7.2%). At four weeks, the difference in prolonged abstinence unadjusted for varenicline use was OR 1.21 (1.00 to 1.48), difference 4.3% (0.0% to 8.7%), p=0.05 and adjusted for varenicline use it was OR 1.32 (1.08 to 1.62) p=0.007. At 12 months the OR was 1.28 (0.97 to 1.69), difference 2.7% (-0.4% to 5.8%), p=0.09 unadjusted for varenicline use and after adjustment was OR 1.36 (1.02 to 1.80) p=0.04. 5.9% of participants discontinued preloading because they could not tolerate it. Gastrointestinal symptoms, chiefly nausea, occurred in 4.0% (2.2 to 5.9) more people in the preloading than control arm. There were eight serious adverse events in the preloading arm and eight in the control arm, odds ratio OR of 0.99 (0.36 to 2.75). Conclusions: In the primary analysis there was insufficient evidence to confidently demonstrate that nicotine preloading increases subsequent smoking abstinence. The beneficial effect appears to have been masked by a concurrent reduction in the use of varenicline in people using nicotine preloading and future studies should explore ways to mitigate this unintended effect. Registration: Current Controlled Trials, ISRCTN33031001