Aim and objective:
In 2007 Balcova Heart Study cohort was started to determine cardiovascular disease risk in 35 years and older people in Balcova district, Izmir, Turkey. Among participants who were willing to quit smoking were invited to smoking cessation clinic. This study aims to analyze the effect of inverse probability treatment weighting (IPTW) against no weighting to predict relapse in smoking cessation.

Methods:
Smokers who were screened in Balçova Heart Study for ischemic heart disease risk factors and declared the will to quit smoking were invited to the outpatient smoking cessation clinic. All patients were given behavioural counseling and a proportion of the smokers received pharmacotherapy (varenicline or bupropion) by clinical decision. After one year all patients were called to estimate the success rate of the intervention (n=359). The results of pharmacotherapy and behavioural counseling versus only behavioural counseling were compared in generalized estimating of effects model with and without IPTW weighting. The fully adjusted model included age and education category, gender, marital and working status, Fagerström Test for Nicotine Dependence category.

Results:
Relapse rate was 51.3% (n=184). Additional to the behavioural counselling pharmacotherapy lowered relapse rate by 20.0% [relative risk=0.80, 95% confidence interval (CI): 0.64-0.99] relative to only behavioural counselling in crude analysis. In the fully adjusted model without IPTW beta regression coefficient of pharmacotherapy was -0.548 [odds ratio (OR)=0.58, 95% CI: 0.37-0.90, p=0.016). After IPTW weighting the coefficient was -0.545 (OR=0.58, 95% CI: 0.37-0.90)p=0.016).

Conclusions:
Additional to behavioural counseling pharmacotherapy is effective to prevent relapse in smoking cessation follow-up and IPTW weighting reveals only a slight change in beta regression coefficient.

Funding:
Balçova Municipality funded Balçova Heart Study.