Effectiveness of tobacco cessation interventions and risk factors for tobacco-use relapse: A cohort study

INTRODUCTION

Smoking is defined as a habit in which a substance (most commonly tobacco) is burnt, and the resulting smoke is breathed in to be tasted and absorbed in the bloodstream.¹ Although smoking is the most common form of ingesting tobacco, there are other smokeless forms available. Nicotine, the main addictive chemical in tobacco, causes a rush of adrenaline when absorbed in the bloodstream or inhaled via cigarette smoke. It also triggers an increase in dopamine—the brain’s ‘happy’ chemical. This stimulates the area of the brain associated with pleasure and reward,² the reason which makes people smoke.

Smoking is one of the largest preventable causes of human diseases. Tobacco use kills nearly 8 million people worldwide each year. More than 7 million of those deaths are the result of direct tobacco use, while around 1.2 million are the result of non-smokers being exposed to second-hand smoke.³ The Global Adult Tobacco Survey (GATS)-2 revealed that 28.6% (266.8 million) of adults in India, aged 15 and above, use tobacco in some form.⁴

Tobacco cessation in general means the process of discontinuing tobacco use⁵ Tobacco cessation interventions are a cost-effective way of reducing ill health as they provide both immediate and long-term health benefits, irrespective of the age of quitting.⁶ Most people who try to quit unaided are unsuccessful in the long term because smoking is a complex addiction, with physical and psychological components.⁷ The GATS studies have reported that there was a significant increase in the proportion of tobacco users between GATS-1 and GATS-2 who visited a healthcare provider in the preceding 12 months and, subsequently, an increased number (46.3% in GATS-1 to 48.8% in GATS-2) were advised by a healthcare provider to stop tobacco use.³ However, there is still uncertainty regarding the relative effectiveness of intense tobacco cessation therapies.⁸ For instance, a study by Baker et al. showed that there was no significant difference in biochemically confirmed rates of smoking abstinence at 26 weeks following 12 weeks of open-label treatment with nicotine patch, varenicline, or combination nicotine replacement therapy.⁸ Similarly, a study done at our centre assessing patients who reported to the tobacco cessation clinic (TCC) between August 2007 and July 2009 showed that only 15% (28/189) attended follow-up clinics, and the rate of 1-month abstinence was only 10% (11/106).⁹ Since a decade has passed since our previous study and Indian data on the effectiveness of tobacco cessation interventions are limited, we assessed the effectiveness of tobacco cessation strategies in our TCC in quitting or reducing tobacco use and the risk factors for tobacco-use relapse.

METHODS

RESULTS

We screened a total of 135 participants, of whom 127 were recruited. Five participants refused consent, and 3 participants were excluded due to severe illness. At the 6th month follow-up, 6/127 participants were lost to follow-up, all of whom used to smoke tobacco and did not use the smokeless forms of tobacco. Among the remaining 121 participants, only 6 were on a regular follow-up to the TCC, and the rest were contacted through telephone. The demographic characteristics are depicted in Table 1.

The mean (standard deviation [SD]) age of the participants was 51.7 (14.4) years. There were 2/127 (1.57%) women among the study participants. Those using smokeless forms of tobacco were 9/127 (7.09%), and of these, 3/9 also smoked tobacco. Nearly 90% of them attended the TCC as their doctor had referred them. Of all these participants, 77/127 (60.63%) had tried previously to quit using tobacco but failed, and 84/127 (66.14%) had a positive family history of using at least one form of tobacco. Further, 41/127 (32.28%) of all the participants reported using alcohol along with tobacco. The most common comorbidity was asthma, seen in 67/127 (52.76%) of the participants, and a high-resolution CT scan of the lungs had been obtained by the treating physician for 16/127 (12.60%) participants, out of which 13/127 (81.25%) participants had features suggestive of chronic obstructive pulmonary disease (COPD). The details pertaining to their motivational stage and treatment are given in Table 2.

Approximately two-thirds of the participants were prescribed only behavioural counselling, while the remaining 44/127 (34.65%) were prescribed some form of pharmacotherapy. Of these, 43 (97.73%) were prescribed nicotine replacement therapy (NRT) and 1/44 (2.27%) were prescribed bupropion and varenicline. ADRs occurred in 4/43 (9.30%) of the participants on NRT, where 3/43 (6.98%) participants had dyspepsia, 2/43 (4.65%) suffered a headache, and 1/43 (2.33%) participants had nausea. There were no ADRs reported by the participant who was taking bupropion and varenicline.

At the 6th month follow-up, 28/121 (23.14%) participants reported quitting using tobacco of any form. This included 3/6 participants who used only smokeless forms of tobacco. Twenty-eight participants reported not being on any treatment during the follow-up. Eighteen of these 28 participants (16 smokers and 2 users of smokeless forms) reported having quit using tobacco. The details of participants at follow-up are summarized in Table 3.

The regression analysis of hypothesized predictors is listed in Table 4. Younger age showed a trend towards being a significant factor for not quitting tobacco use in the univariate analysis, but failed to achieve statistical significance in the multivariate analysis.

DISCUSSION

We found that with the use of tobacco cessation interventions, 23% of the participants quit tobacco use and 74% reduced use at 6 months. The most common intervention prescribed was behavioural counselling.

Compared to our previous study⁹, there was no significant change in the socio-demographic characteristics of patients visiting the TCC. Most patients were middle-aged male smokers with comorbid respiratory diseases. This is in line with other published studies from India. An analysis of 23 320 cases registered across 18 TCCs between 2002 and 2007 reported that 92.2% were males, and the majority were middle-aged.¹⁴ This clearly indicates that there has not been any change over time in the characteristics of people who seek professional help from a TCC. In India, men use tobacco throughout their lives, whereas women tend to become tobacco users at an older age.¹⁵ This could be explained by the fact that smoking among women is not socially accepted in India.¹⁶ Although the number of women smokers in India has increased and India has the second highest number of women smokers after the USA,¹⁷ under-reporting is quite prevalent among Indian women.¹⁸

The most common reason for attending the TCC continued to be a referral from the treating physician of a primary ailment.⁹ Despite many awareness campaigns against the use of tobacco, not many were motivated to attend TCC on their own, suggesting that tobacco cessation is still not perceived as a felt need by the public. However, caution is required while interpreting these numbers, as ours was a TCC set-up within a tertiary referral centre, and the patients who visited our TCC were those who would otherwise be accessing healthcare services from our institute.

We observed that the quit rate was 23.14% and 74.38% reported a reduction in tobacco usage. Although quitting is the ideal outcome with unequivocal health benefits, reduction in tobacco use, especially smoking, also improves respiratory symptoms and lung function.¹⁹ The quit rate observed in our study was higher than the median unassisted quit rate of India (14.2%) as reported in the GATS-2 study.²⁰ Other studies from India also found a similar quit rate (Kumar et al.²¹ 12.5%; Varghese et al.¹⁴ 14%; Jindal et al.²² 10%; and Thresia et al.²³ 16%). This high quit rate seen in our study is probably because it was institution-based with a small sample size. Most of the participants were primarily patients suffering from other illnesses, and quitting tobacco would have been a health mandate required by their treating physician. Hence, they are likely to be more motivated than the normal population to quit tobacco use due to the morbidity of their primary illness. This also explains the higher proportion of patients in the action stage of motivation among those who mentioned that they had previously thought about quitting tobacco use. Further, the majority of our participants were well educated (58% graduate and above), and higher levels of education and occupation are known to be positively associated with cessation, probably because of the increased awareness about the harmful effects of tobacco.²⁴ However, compared to our previous study,⁹ the quit rate is high (10% v. 23.1%), which suggests a possible healthy attitude change in our population towards quitting tobacco use and/or better awareness regarding the ill-effects of tobacco use.

The most common interventions prescribed were behavioural counselling only (65.35%), followed by a combination of behavioural counselling and nicotine replacement therapy (NRT; 33.07%). This finding was similar to the one reported by Varghese et al. among 34 741 participants, wherein 68.9% received behavioural counselling and 31% were prescribed an adjunct pharmacotherapy.¹⁴ Use of medication was very low at just 1.6% in our study. Although there is high-quality evidence from a meta-analysis of 52 studies (19 488 participants) proving the superiority of combined behavioural counselling and pharmacotherapy (including NRT) over usual care, brief advice, or less intensive behavioural support (relative risk 1.83, 95% CI 1.68 to 1.98), use of medicines as well as NRT is still limited.²⁵ This trend has still not changed despite the availability of newer medicines, probably due to the associated cost burden.²⁶

We were, however, unable to identify any statistically significant risk factor for relapse or treatment failure at 6 months. Younger age, although significant in univariate analysis, did not reach statistical significance in multivariable analysis, probably due to the small sample size. Young age may be assumed as a possible predictor. A similar trend was seen in a secondary analysis of the GATS-2 study by Srivastava et al., which found higher odds of quitting attempts among younger age groups but the lowest odds of successful quitting.²⁷ This is concerning because if not addressed adequately at an early stage, India could become a country with the highest burden of tobacco-attributable diseases in the near future.

Our study has a few limitations. Objective evidence of tobacco cessation by biochemical methods was not evaluated. There was no clear risk factor identified for relapse, probably due to the small sample size. The results may not be generalizable to a normal general population as all our participants were patients accessing healthcare services with some illness, mostly tobacco-related, and were likely to be more motivated than the normal population to quit tobacco use. Having said that, the main strength of our study is that follow-up was done in approximately 95% of the study participants at 6 months, as opposed to our previous study, where 6-month data were available only for 56% of patients.⁹ Further, our study is one of the few hospital-based studies from India evaluating the use of a TCC in a hospital setting, as most studies from India are community-based.

In conclusion, we report that the quit rate following TCC interventions was low at 23.14%, but 74.38% reported a reduction in tobacco use, thus demonstrating that TCC could play a pivotal role in preventing tobacco-related mortality and morbidity. Furthermore, the quit rate has improved since our previous study approximately a decade ago. Hence, better awareness among physicians regarding the availability of a dedicated TCC would help in the optimal utilization of services rendered by the TCC. Behavioural counselling alone, followed by a combination of NRT and behavioural counselling, were the most common interventions. However, combination therapy is known to provide greater benefits, and the cost of NRT, or medications, is likely to influence the decision. A pharmacoeconomic analysis would thus be helpful. No predictors of relapse/treatment failure were identified in our study, probably due to the small sample size.