This study was a cross-sectional questionnaire survey conducted on a sample of 130 Narikuravars residing in Chennai, Tamil Nadu. The study was conducted from November 2018 – December 2018. Ethical approval for the study was obtained from the Institutional Review Board of SRM Dental College and Hospital. The sample size was estimated using the highest prevalence percentage of tobacco use for a covariate in a study done on tribal people of Madhya Pradesh by Verma P et al., [11]. The highest prevalence was noted among tribal people who had completed primary schooling which was 66.7%. Using this prevalence percentage, and assuming a finite population of 500 tribal people, 10% absolute precision, 99% confidence level and design effect of 1 for cluster surveys, the sample size was estimated to be 114 using Open Epi sample size calculator for estimation of sample size. Assuming a 20% non-response rate, the sample size was increased to 130. Participants were recruited into the study after explaining the purpose of the study and obtaining informed consent. Post-hoc power analysis with a prevalence percentage of 56.2% observed in the present study, 66.7% true proportion, 130 study subjects, and precision of 10% was found to be 80.2%. For the purpose of this study, three clusters of Narikuravars residing in three different areas of Chennai namely Avadi, Thirumullaiavoyal and Red Hills were identified. Following this, a cluster randomised sampling procedure was followed for selection of clusters. Two clusters out of three were selected randomly using lottery method.

The clusters that were selected were Avadi and Thirumullaivoyal. All individuals aged 15 years or above in both the clusters were included in the study. This age cut-off was based on GATS surveys conducted by WHO where prevalence of tobacco use is estimated for adolescent and adults above 15 years of age. In the TNTS (2015-2016), it was found that 25.7% of surveyed users of tobacco had initiated the practice before 18 years of age [8]. Hence, an age cut-off of 15 years or above was used in the present study. On such inclusion, a final sample size of 130 study subjects were obtained with complete responses to all the questions giving a 100% response rate. The study instrument was a self-styled questionnaire consisting of 14 questions that was framed and administered by the principal investigator to all the study participants through a face-to-face interview in the local vernacular language Tamil. The English version of the questionnaire was translated to Tamil version. The Tamil version was then back translated into English by a linguistic expert fluent in both Tamil and English. The back translated version was then compared with the English version to check that the questions were properly translated. The questionnaire consisted of four parts. The first part consisted of three items to assess the awareness of health hazards of tobacco use, the second part consisted of two items to assess current tobacco use practices, the third part consisted of six items to assess quitting practices and factors that were challenging or led to relapse, the last item in this part of the questionnaire was partly open ended and assessed whether participants had motivated any other person to quit using tobacco and what were the reasons they cited in asking others to quit. The fourth part of the questionnaire consisted of three items to assess the awareness on the availability of professional help to aid in quitting tobacco, awareness on cessation aids and awareness on tobacco control laws in India. All the questions except for the eleventh question were closed ended with dichotomous or multiple-choice answers. Skip patterns were used where appropriate. A skip pattern is a question or series of questions associated with a conditional response, that is some questions pertain only to certain subjects. For example, in this study subjects who answered no for the first question were not asked for a response to the second question. Likewise, question six to ten was administered only to participants who used tobacco in any form. [Annexure]

In addition to this, the questionnaire contained questions to assess the demographic characteristics of the study participants including age, gender, education status and occupation.

A pilot study was conducted among twenty-five members in the cluster that was excluded from the main study to assess the reliability of the questionnaire. These questionnaires were not included in the final analysis. The Cronbach’s alpha for the tested 14-item questionnaire was found to be 0.88 showing a good degree of internal consistency and homogeneity between items. The face validity of the questionnaire was assessed by distributing the questionnaires to 20 dentists who were familiar with concepts related to tobacco use. The dentists were required to assess four characteristics namely ambiguity, comprehensibility, use of proper terms and relevance using a 5-point Likert scale ranging from one to 5 (1-not important, 2-somewhat important, 3-important, 4-very important, 5-highly important). Following this all questionnaires were collected and impact score of each item was calculated with a cut-off of >1.5 being considered as acceptable.

Descriptive statistics was used to summarise the data. The data was entered into MS Excel and descriptive statistics including frequency distributions were calculated and tabulated for all the 14-items in the questionnaire. The open-ended question was analysed using frequency distribution of various responses given by the participants.

In this study, 60% (n=78) participants were females; the mean age range of the study participants was 29.7 years. Most of the participants were uneducated (53.1%) or had completed middle school (21.5%). Likewise, a majority of participants 46.2% (n=60) had no occupation and only 21.5% (n=28) of participants had a full-time occupation. [Table/Fig-1]. The frequency distributions for all the 14-items in the questionnaire are presented in [Table/Fig-2]. In the present study, an overwhelming majority of 86.2% (n=112) of Narikuravars were aware of the health hazards of tobacco use. About 75% (n=84) felt that using tobacco would result in lung cancer, 16.9% (n=19), 2.7% (n=3) and 5.4% (n=6) felt that using tobacco would result in oral cancer, heart attack or all the three diseases respectively. Eighty six percent of Narikuravars felt that smokeless tobacco was as harmful as smoking tobacco and 56.2% (n=73) were using tobacco in any form during the time of the study. In this study, 57.5% (n=42) used tobacco in a natural preparation of tobacco leaf, betel leaf, areca nut and slaked lime.

Among users of tobacco, 49.3% (n= 36) answered that they have thought of quitting tobacco use in the past. Out of this, 58.4% (n=21) answered that they had tried abstaining from tobacco use. Among these participants who tried abstaining from using tobacco, only 14.4% (n=3) were successful in quitting for more than six months’ time and did not relapse back into using. A vast majority of participants (57.1%, n=12) felt that withdrawal symptoms were the biggest challenge encountered when they attempted to quit tobacco and also identified it to be the most important factor that led to relapse.

The eleventh question was partly open ended, participants who answered yes to having motivated others to quit using tobacco were asked about the advice/reasons they gave to motivate others to quit tobacco use. Many participants (76.9%, n=100) stated that they motivated others to quit citing the health hazards of tobacco use while some others stated that they did not like the smell of tobacco products. A small number of participants (9.2%, n=12) stated that they motivated people to quit because using tobacco would result in harm to the children or family members of the user.

A majority of 97.6% (n=127) of study participants were unaware that professional help can be sought to assist tobacco users to quit, and 91.5% (n=119) were unaware about pharmacological or non-pharmacological aids to cessation. Eighty percent (n=104) participants were aware about the existence of regulations against tobacco products in India.

[Table/Fig-3] shows the comparison of studies conducted on tobacco consumption in tribal people in India [11-14].

Tobacco epidemic is known to be one of the major contributors of preventable mortality in the world. Despite two decade long efforts to curtail tobacco consumption, tobacco use is still prevalent in many adolescents and adults all over the world. Among tobacco users, tribal people have been found to have high rates of using smokeless and smoking tobacco in many studies done in India [9,10]. However, studies that assess concomitant factors such as attitudes to quitting, success in quitting tobacco, factors leading to relapse and knowledge regarding cessation aids are rarely reported in context of tribal people. Hence, this study was done to bridge the gap that exists in these areas for a more holistic understanding to tobacco menace in tribal population.

In the present study, the prevalence of tobacco use in Narikuravars in any form was found to be 56.2%. This is somewhat similar to results obtained by Zahiruddin QS et al., who found the prevalence of tobacco use in any form to be 54.45% in adolescents residing in six tribal villages of Wardha district, Maharashtra [15]. The same study also found the prevalence of using smokeless tobacco to be 23.14% whereas in the present study the prevalence of using smokeless tobacco was found to be 57.5% which is comparatively higher. The discrepancy can be explained based on the fact that Zahiruddin QS et al., study was conducted among adolescents aged 11 to 19 years and among young adolescents using cigarettes or any other form of smoking tobacco products is considered much more fashionable than using smokeless forms to tobacco.

In the present study, 86.2% of Narikuravars knew about the health hazards of tobacco use. This percentage is similar to that found in a study conducted by Imtiaz D et al., who found that 87.4% of current tobacco users in rural areas of Dehradun district knew about the health hazards of tobacco use [16]. Likewise, Nicholson AK et al., study on Aboriginal smokers in Australia found that 94% of smokers were aware that tobacco use causes lung cancer, and 89% were aware that smoking causes heart disease [17]. In the present study, most of the Narikuravars (75%) were aware that smoking tobacco causes lung cancer, 16.9% were aware that chewing tobacco products can cause oral cancer. However, only 2.7% of participants knew that tobacco use can result in heart disease and 13.8% felt that there were no health risks associated with using tobacco.

The relatively high awareness on health hazards of tobacco usage even among indigenous people such as the Narikuravars can be attributed to mass media campaigns and stringent legislative actions taken by the Government of India to tackle tobacco epidemic in the country. In the present study, even the proportion of patients who answered that they felt there were no health risks in using tobacco made the statement in a lethargic fashion stating that there were health risks in everything, fate determines one’s chances of falling ill and not due to genuine unawareness about the health risks of using tobacco.

In the present study, nearly half (49.3%) of current users of tobacco stated that they had thought of quitting in the past. Among these participants, 58.4% had tried abstaining from tobacco use. The questions on tobacco cessation in the present study were structured based on the Prochaska and Di Clemente’s stages of change model where individuals pass through various stages while attempting to change behaviour [18]. A person is said to be in the action stage if they have been able to change behaviour for a month. In the maintenance stage, the new behaviour has been adopted and continued for a period of six months. A person may relapse to the previous behaviour at any stage in the stages of change model.

In the present study, a participant in the maintenance stage who had successfully quit tobacco for six months and did not relapse to use again at the time of the study period was considered to be successful in quitting. In using such criteria, only 14.4% participants were successful in quitting tobacco in the present study, whereas, a majority of 66.6% of Narikuravars failed in their attempts at quitting and could not quit tobacco successfully even for a month.

A previous review done by Di Giacomo M et al., on smoking cessation in indigenous populations of Australia, New Zealand, Canada and United States have revealed that smoking cessation interventions were successful when they were tailored according to an indigenous community after identifying enabling factors and barriers to quitting [19].

In the present study, 97.6% of study participants were unaware that professional help can be sought in assisting a person to quit tobacco. In addition, 91.5% of study participants were unaware of any medications that could help a person to quit tobacco. In the present study, the term medications were used broadly to include all pharmacological and non-pharmacological aids to tobacco cessation. Cowie N et al., study on Maori and Pacific Islander parents of pre-adolescent children in Auckland, New Zealand revealed that 78% of parents had awareness about nicotine gum and 75% of parents were aware about nicotine patch as aids to tobacco cessation [20]. In the present study, only 8.5% of Narikuravars knew about the existence of pharmacological or non-pharmacological aids to tobacco cessation, and even in this 8.5% awareness regarding the specific forms of Nicotine Replacement Therapy (NRT) was vague and inadequate.

Eighty percent of Narikuravars in this study knew about the existence of legislation against tobacco products in the country. Again, the participants were unaware of the specific conditions or prohibitions existing in regards to tobacco in the country.

However, as far as we know this is the first study that has tried to understand the tobacco use practices along with other factors such as; whether the indigenous users had thought of quitting tobacco in the past, whether they had made any actual efforts to quit, whether they were successful if they did make a quit attempt. In addition, challenges to quitting were identified along with factors that led to relapse among users who made a quit attempt and relapsed back to using. Di Giacomo M et al., review clearly states the need for identifying factors that act as facilitators and barriers to quitting tobacco for any cessation intervention to be successful in an indigenous population [19]. The awareness on seeking professional help to assist in quit attempts, pharmacological and non-pharmacological aids to cessation and tobacco control laws among these indigenous people was also assessed in this survey. This study can be used as a foundation to conduct more intense research on tobacco epidemic in this sub-set of indigenous population in Tamil Nadu.

The limitations of this study were that this study was a cross-sectional survey and did not use any interventions either in the form of cessation counselling, NRT or other pharmacological aids to assess whether these interventions could prove to be successful in helping Narikuravars who used tobacco to quit. Assessment of level of nicotine dependence among users and exhaled carbon monoxide measurements to confirm the status of non-users were not done in this study. As this study relied completely on the answers of the participants to determine whether they were tobacco users or not, social desirability bias may have prompted participants to answer that they were non-users. Likewise, social desirability bias may have made participants to project themselves as successful in their quit attempts. Future studies could include larger samples of tribal people in addition to incorporating objective measures to assess status of tobacco use.

From the findings of this study, it is clear that tobacco consumption is higher among the indigenous population of Narikuravars residing in Chennai, Tamil Nadu when compared to national prevalence estimates from the second GATS survey. A majority of study participants were aware of the health risks of tobacco and nearly half of current users of tobacco had thought of quitting in the past. Only a very few participants had success in quitting tobacco. The awareness on seeking professional assistance to quit and cessation aids to quit was poor. This study reveals that targeted and community-based approaches must be designed to improve knowledge on tobacco cessation, so that indigenous people may benefit from tobacco cessation interventions tailored specifically to them.