The findings of this subanalysis of CHOICES indicate that a high smoking abstinence was achieved in the Greek population of this study, despite participants' heavy dependence on nicotine. This study has shown that treatment with varenicline as provided by a smoking cessation clinic outside a clinical trial environment can be very effective as a smoking cessation treatment.
Overall, 70.4% of Greek participants successfully quit smoking at the end of a 12-week course of varenicline, as assessed by self-reported 7-day PPA as a measure of abstinence. Neither the total FTND score nor FTND Question 1 response (ie, time to first cigarette on waking) significantly predicted a successful 7-day PPA at Week 12. The most frequently observed adverse events were nausea (13.3%), insomnia (4.6%) and upper abdominal pain (4.1%), mostly of mild to moderate severity.
The efficacy and tolerability of varenicline has been shown in various placebo-controlled, double-blind, randomized clinical trials [11, 12]. Stringent inclusion and exclusion criteria applied in these trials meant that participants with chronic comorbidities and medication were excluded. The value of observational studies, such as CHOICES, lies in the heterogeneity of participants who are enrolled, making the findings more representative of the general population. For example, in the analysis presented here, a high proportion of participants had a medical history of hypertension, chronic obstructive pulmonary disease or depression and had effective treatment for smoking cessation with varenicline. These findings support varenicline use in populations with these comorbidities. This has also been shown in double-blind clinical trials .
In the main CHOICES study  the 7-day PPA by Week 12 was 64.6% (95% C.I. 60.1-68.3), while 61.1% was reported for a subanalysis of the Belgian population enrolled in the CHOICES study. This compares with a 7-day PPA of 70.4% (95% C.I. 64.0-76.7) for the Greek population at Week 12 which is reported here. The higher rate of abstinence in our population may be due to a number of factors including differences between the Greek study centres versus the other countries in the CHOICES trial (Belgium, Hungary and Slovenia), such as, differences in behavioural support and/or reimbursement. Although these factors were not compared in the CHOICES trial and a further trial would be required in order to evaluate the impact of any differences on outcomes, it is important to note that in Greece this observational study was carried out in respiratory departments and smokers were treated and monitored by the same medical doctor throughout the trial. Doctors had received training in smoking cessation strategies and were voluntarily referred to the smoking cessation clinics, where behavioural support was provided. High abstinence rates have been previously reported in these clinics using other smoking cessation treatments, such as behavioural support, nicotine replacement products and bupropion SR, in Greek smokers [20, 21].
It is also important to note that at the time the CHOICES study was conducted, there was no reimbursement policy for smoking cessation medications (including varenicline) in Greece, although there are current proposals to partially reimburse Greek smokers for smoking cessation medications in the future.
One additional important finding of the study is the higher rate of abstinence with better compliance. We considered as completers those participants who had good treatment compliance, defined as taking study medication for 80% of the planned number of treatment days (more than 8 weeks). 86% of smokers who were completers had quit smoking by Week 12. This shows the greater effectiveness of varenicline treatment when it is taken for more than a month, and indicates that good compliance to treatment is important and that treatment for more than 8 weeks is associated with higher success rates. Close follow-up by specialized smoking cessation clinics increases the compliance to treatment and further the abstinence rates, as other published studies in Greek smokers have also reported [20, 21].
In the Belgian subanalysis of CHOICES, the FTND total score was not predictive of successful smoking cessation at Week 12 (P = 0.12; odds ratio [O.R.] 0.65 [95% C.I. 0.38-1.11]),  similar to what is reported here for the Greek population (P = 0.057; O.R. 0.55 [95% C.I. 0.29-1.02]). The latter was anticipated, as traditional measures of nicotine dependence, such as the FTND, are not consistently predictive of successful smoking cessation, as reported in the literature .
Time to first cigarette on waking was predictive of the 7-day PPA at Week 12 in the Belgian subpopulation . Although there was a similar trend in the Greek subpopulation, this was not statistically significant (O.R. 0.73, 95% C.I. 0.51-1.06, P = 0.096). This may be so because the majority of the Greek smokers in this study belong to the first two categories, lighting up their first cigarette within the first 30 minutes. In a previously published study in Greek smokers, multivariate regression analysis indicated that the time to first cigarette on waking, and use of bupropion, independently predicted abstinence .
In a recent observational study based on records stored in The Health Improvement Network (THIN) database in the United Kingdom, the 7-day PPA 6 months after the initiation of varenicline therapy was 49.5%, based on self-reporting . A longer follow-up of the participants of the CHOICES study would be of interest and may give comparable results. A strong and positive message of this subanalysis of CHOICES is the high effectiveness of varenicline treatment in Greek smokers. Unfortunately, in Greece, the smoking ban laws are not implemented broadly in public places, and there has been no substantial media campaign supporting smoking bans .