Table 2.

Multivariable logistic regression models: partial and fully adjusted odds ratios and 95% confidence intervals for the association between varenicline versus NRT and smoking cessation at 3, 6 and 9 months and 1, 2 and 4 years after exposure, N = 220136^a

Model	3 months	6 months	9 months	1 year	2 years	4 years
	Odds ratio (95% confidence interval)
Partial adjustedb	1.47	1.50	1.44	1.38	1.30	1.23
	(1.42 to 1.52)	(1.46 to 1.55)	(1.40 to 1.48)	(1.35 to 1.42)	(1.27 to 1.33)	(1.21 to 1.26)
Fully adjustedc	1.42	1.46	1.40	1.34	1.26	1.19
	(1.38 to 1.47)	(1.42 to 1.50)	(1.36 to 1.44)	(1.31 to 1.38)	(1.23 to 1.29)	(1.16 to 1.21)

^aMissing BMI and IMD values were imputed using multiple imputation.²⁹

^bPartial adjusted models were adjusted for: age, sex and year of prescription.

^cFully adjusted models were adjusted for: age, sex, days in history, IMD, number of GP visits 1 year preceding first prescription, BMI, year of first prescription, history of major physical morbidity (Charlson Index), alcohol misuse, drug misuse, bipolar, depression, neurotic disorder, self-harm, other mental disorder, antidepressant prescription ever, antipsychotic prescription ever, hypnotics/anxiolytics prescription ever, other psychotropic medication.