Table 3.

Results of Multivariable Logistic Regression for Continuous Smoking Abstinence From 4 to 12 Months

	Odds ratio (95% CI)	P
Specialist services vs. nonspecialist services	1.477 (1.092, 1.997)	.011
Age (y)	1.013 (1.002, 1.023)	.016
Female vs. male	0.933 (0.723, 1.205)	.596
Married or living with a partner vs. all other	1.520 (1.158, 1.995)	.003
Low education (up to GCSE vs. A-level or above)	0.940 (0.726, 1.218)	.641
Unemployed vs. all other	0.634 (0.398, 1.012)	.056
Free prescription vs. no free prescription	0.898 (0.675, 1.195)	.461
Any previous quit attempts vs. no previous quit attempt	0.702 (0.439, 1.123)	.140
Longest time managed to quit before: >4 weeks vs. ≤4 weeks	1.770 (1.245, 2.517)	.001
Cigarettes per day before quitting: <10 vs. ≥10	0.785 (0.605, 1.018)	.068
First cigarette within 5 minutes after waking vs. >5 minutes	0.817 (0.586, 1.140)	.235
Living with a smoking partner vs. not living with a smoking partner	0.888 (0.639, 1.236)	.480
Use of NRT vs. no use of NRT during 2–3 months	1.652 (1.214, 2.248)	.001
Use of varenicline vs. no use of varenicline during 2–3 months	1.335 (0.987, 1.805)	.060

CI = confidence interval; GCSE = General Certificate of Secondary Education; NRT = nicotine replacement therapy. The multivariable analysis used data from 1154 participants, and data from 250 participants were excluded from the analysis because of missing data from some independent variables. After including the treatment condition as an independent variable, the results of the logistic regression analysis remain materially unchanged.