bs_bs_banner
TRIAL PROTOCOL
doi:10.1111/add.12420
A cluster randomized controlled trial of a brief tobacco cessation intervention for low-income communities in India: study protocol Bidyut K. Sarkar1,2, Lion Shahab2, Monika Arora1, Fabiana Lorencatto3, K. Srinath Reddy1 & Robert West2 Public Health Foundation of India, New Delhi, India,1 Department of Epidemiology and Public Health, University College London, London, UK2 and NCSCT, University College London, London, UK3
ABSTRACT Background India has 275 million adult tobacco users and tobacco use is estimated to contribute to more than a million deaths in the country each year. There is an urgent need to develop and evaluate affordable, practicable and scalable interventions to promote cessation of tobacco use. Because tobacco use is so harmful, an increase of as little as 1 percentage point in long-term quit success rates can have an important public health impact. This protocol paper describes the rationale and methods of a large randomized controlled trial which aims to evaluate the effectiveness of a brief scalable smoking cessation intervention delivered by trained health professionals as an outreach programme in poor urban communities in India. Methods/design This is a pragmatic, two-arm, community-based cluster randomized controlled trial focused on tobacco users in low-income communities. The treatment arm is a brief intervention comprising brief advice including training in craving control using simple yogic breathing exercises (BA-YBA) and the control arm is very brief advice (VBA). Of a total of 32 clusters, 16 will be allocated to the intervention arm and 16 to the control arm. Each cluster will have 31 participants, making a total of 992 participants. The primary outcome measure will follow the Russell Standard: self-report of sustained abstinence for at least 6 months following the intervention confirmed at the final follow-up by salivary cotinine. Discussion This trial will inform national and international policy on delivery of scalable and affordable brief outreach interventions to promote tobacco use cessation in low resource settings where tobacco users have limited access to physicians and medications. Keywords India, intervention, non-pharmacological, outreach, single session, smokeless, tobacco cessation, trial, validated quit rate, yogic breathing. Correspondence to: Bidyut K. Sarkar, Public Health Foundation of India, New Delhi 110019, India. E-mail: [email protected] Submitted 12 February 2013; initial review completed 1 May 2013; final version accepted 11 November 2013
INTRODUCTION Tobacco use is a risk factor for six of the eight leading causes of death in the world [1]. More than 5 million deaths are attributed to tobacco use globally each year [2] and an estimated 1 million deaths in India alone [3]. India has 275 million tobacco users, of whom 164 million are smokeless tobacco users, 69 million are smokers and 42 million use both smoked and smokeless tobacco [4]. With regard to smoking, it has been shown that an increase of as little as 1 percentage point in longterm quit success rates can have an important public health impact [5]. A 1% decrease in prevalence of © 2014 Society for the Study of Addiction
tobacco use in India would represent 7.8 million fewer tobacco users. The prevalence of tobacco use is much higher among lower socio-economic groups in many countries. Moreover, quit rates are typically lower among those with lower education level and socio-economic status [6]. In India, 70% of the health-care services are obtained from the private sector and very few people are covered by health insurance. Therefore, lower-income groups in India would have much less access to healthcare services and would be prone to higher morbidity and mortality due to tobacco use. This trial focuses upon the most vulnerable strata to which the majority Addiction, 109, 371–378
372
Bidyut K. Sarkar et al.
of tobacco users in India belong, those with severe economic disadvantage. Brief advice is a low-cost intervention which can be delivered en masse by trained health workers [7]. Systematic reviews of randomized controlled trials in western samples have found that brief advice to quit smoking is better than no advice, increasing 12-month quit rates by 1–3% [7–10], with offer of support leading to a larger effect than simple advice to quit [11]. However, the effectiveness of brief advice for smoking cessation has yet to be evaluated in a randomized trial in low- or middle-income countries; nor is it known in any setting whether brief advice is effective in reducing smokeless tobacco use. Unlike developed countries such as the United Kingdom, where behavioural therapy and/or pharmacotherapy is provided at little or no cost to patients, support services are very limited for quitting tobacco use in India [12–13]. Hence, there is an urgent need to identify and evaluate low-cost, affordable, non-resource-intensive, easily scalable interventions to promote cessation of tobacco use in India. A promising, affordable and culturally appropriate option to enhance brief advice and aid smoking cessation in the Indian context is the use of breathing exercises, known as ‘Pranayama’ within yoga. Yoga is a traditional method of achieving good physical and mental health. Therefore, yogic breathing exercises are expected to have a high degree of acceptability as a tobacco cessation aid.
Recently, Pranayama has been widely popularized in India by guru Ramdev of Haridwar, who has provided mass training through television and mass yoga camps attended by tens of thousands. He, along with other yoga gurus, have attributed the benefits of breathing exercises specifically to two key breathing exercises called ‘Kapalbhati-KB’ (normal deep inhalation—forceful exhalation) and ‘Anulom vilom-AV’ (alternate nostril breathing), which are also the easiest to learn and practice. These two breathing exercises were selected for the present intervention (see Table 1). Although relatively little research has been undertaken to test the hypothesis that yogic breathing exercises may improve quitting rates, the existing research is promising. McClernon et al. [14] observed that controlled deep breathing significantly reduced smoking withdrawal symptoms, including craving for cigarettes and negative affect. More recently, Shahab et al. have reported a significant reduction in craving immediately after a session of yogic breathing exercises in yoga-naive smokers abstinent for at least 10 hours who were randomized to an intervention arm and a video control arm [15]. Given that there is good evidence that the experience of severe cravings early on in quit attempts is an important mediator of relapse [16–17], yogic breathing exercises may increase smoking cessation rates. Indeed, Kochupillai et al. [18] reported a positive impact of yogic breathing exercises on tobacco abstinence and found the 12-month
Table 1 Breathing exercise instructionsa. Kapalbhati Normal inhalation–forceful exhalation (1–2 minutes on empty stomach once in morning)
Anulom vilom Alternate nostril inhalation–exhalation (3 minutes in morning and then for cravings during the day)
• Start by breathing deeply: inhale slowly and deeply filling up your abdomen, then fill up your rib cage up to your collar bone and then exhale slowly, emptying your rib cage and then pull in your abdomen emptying it fully. Repeat a few times. • Now, inhale your breath naturally and exhale forcefully (moderate force) with drawing in/pulling in of the abdomen • Do this natural inhale–exhale cycle forcefully repeatedly for a period of 1 minute and then rest by breathing normally • Do this for 1 minute at a time, rest by normal breathing in between to a total duration of 2–3 minutes for the first few days and then you can gradually increase the duration • Increase the duration to 5 minutes over a period of 2 months • Do not strain yourself and stop at any point in time if you feel exhausted • Rest by breathing normally for half a minute before next exercise
• Sit in a comfortable position with the back straight and upright • Fold the fore and middle fingers of your hand into the palm and position your hand on the bridge of your nose to enable use of your thumb and the ring finger to close the nostrils alternatively • Close the right nostril with the thumb of your right hand and breathe in deeply inhaling through the left nostril, then close the left nostril with your ring finger and release the thumb from the right nostril and now exhale through the right nostril • Then inhale through the right nostril, close the right nostril and exhale through the left nostril • Now inhale through the left nostril, close the left nostril and exhale through right nostril
a In this study a set of two simple, beginner-level breathing exercises will be taught to participants. These breathing exercises are based on yogic ‘Pranayama’ recommended by Yoga Guru Swami Ramdev, Haridwar, India, for use for the general population. The participants will be shown a short video demonstration of the breathing exercises and will then be trained to undertake the exercises on their own for about 5 minutes. The breathing exercises should be performed in an upright sitting position. This can then be repeated during the day based on need and feasibility.
© 2014 Society for the Study of Addiction
Addiction, 109, 371–378
Brief tobacco cessation intervention trial
abstinence rates of the yoga breathing exercises group to be 21%, similar to that reported for pharmacological agent bupropion [19]. Given the above findings and the gap in the literature, this pragmatic trial was designed to evaluate the effectiveness of a brief tobacco use cessation intervention. The intervention comprises a single session of face-to-face quit advice and training in yogic breathing exercises aimed at tobacco users in low-income communities in India.
373
of the selected low-income administrative blocks can be one of two types of community: an authorized government rehabilitation known as a ‘resettlement colony’ with slightly better facilities, and an unauthorized settlement known as ‘J.J. cluster’ (J.J. stands for ‘Jhuggi-Jhopri’, meaning huts in Hindi, the local language). Equal numbers of both types of community was selected for inclusion in the study. Recruitment into the study started in January 2012 and it is currently ongoing. The study is anticipated to conclude in November 2013. Trial interventions and delivery
METHODS AND DESIGN Study design and timing This is a cluster randomized trial with two arms: a treatment arm consisting of a brief face-to-face session of quit advice including training in yogic breathing exercises (BA-YBA) versus a control arm of very brief advice (VBA) alone. The participants will be followed-up after the intervention to assess tobacco use status (see Fig. 1). The proposed study area has a total population of approximately 36 000 adults living in randomly selected administrative blocks of 32 large slum areas of lowincome residents spread over a wide geographical area in the city of Delhi, India. Each of these blocks will be considered as a cluster for randomization in the study. Each
The treatment arm is a single session of face-to-face tobacco quit advice including training in BA-YBA to control cravings. The control arm is VBA. A written standard operating procedure (SOP) will be followed by the research team, and a written script for quit advice and a standard video for training on yogic breathing exercises will be used for intervention delivery. For both groups, the baseline interview will be conducted at the doorstep of the tobacco-user’s residence. The intervention will be delivered by a member of the delivery team. Delivery teams will consist of one trained health professional and two field health workers who will be trained and supervised by the health professional. All content is delivered in Hindi, the local language. The
32 Clusters (Low-income urban slum areas) EsƟmated Adult PopulaƟon: 30,655 Randomized straƟfied allocaƟon by computer
16 clusters allocated to intervenƟon
16 clusters allocated to control
(8 JJ clusters & 8 ReseƩlement)
(8 JJ clusters & 8 ReseƩlement)
16 clusters given BA-YBA intervenƟon (Brief Advice and yogic breathing exercises)
Figure 1 Brief Advice and Breathing EXercises (BABEX): cluster randomized trial flowchart © 2014 Society for the Study of Addiction
16 clusters given VBA intervenƟon (Very Brief Advice)
Follow-up at 4 weeks
Follow-up at 4 weeks
Self-reported absƟnence rate
Self-reported absƟnence rate
Follow-up at 7 months
Follow-up at 7 months
Self-reported absƟnence rate for last 6 months (RS6)
Self-reported absƟnence rate for last 6 months (RS6)
Validated by salivary coƟnine
Validated by salivary coƟnine
Addiction, 109, 371–378
374
Bidyut K. Sarkar et al.
intervention will be delivered face to face to participants in both arms, either singly or in groups of two to four participants. The content of the BA-YBA, a brief intervention comprising advice to quit and breathing exercises, will: 1 Explain the important groups of chemicals present and released on smoking and chewing tobacco and the specific diseases caused by tobacco use including cancer, respiratory and cardiovascular diseases. 2 Identify the personal health risks associated with the individual’s specific form(s) of tobacco use and also the economic burden and implications of tobacco use, including cost of hospitalization and treatment of tobacco use-related diseases. 3 Explain the benefits of early and complete abstinence. 4 Explain that the chances of success are higher with the use of some support, as most people experience temporary withdrawal symptoms, as well as the importance of avoiding places, situations and company which could precipitate relapse for a while; and advice to quit within the next month. 5 Give advice to seek support of family members and close friends in the quitting effort. 6 Explain the support options available for quitting, including medications such as nicotine replacement therapy and non-pharmacological aids such as yogic breathing exercises. 7 Perform training in two yogic breathing exercises (Kapalbhati and Anulom-vilom) with the help of a standard training video (see Table 1 for details on instructions). 8 Recommendation to use these breathing exercises once in the morning (both exercises) and Anulomvilom whenever they experience cravings for tobacco during the day. 9 Check understanding with a brief practice session of breathing exercises by the tobacco user for 2–3 minutes. 10 Promote monitoring of adherence by participants by providing a blank calendar and instructions to place a tick by days when the breathing exercises have been undertaken. 11 Provide a leaflet of breathing exercise instructions. 12 State that the research team will follow-up after 1 and 7 months. The VBA, the control condition, will: 1 Explain that tobacco use is very harmful and causes major diseases such as cancer. 2 Explain that quitting is the most important action they can take for their health. 3 State that the research team will follow-up after 1 month and 7 months. To assess the full effect of the intervention, the control condition would ideally involve no advice on tobacco use, © 2014 Society for the Study of Addiction
but this was considered unethical by relevant review boards. Therefore, any effect-size estimate produced by this trial is likely to be conservative. We will conduct a survey to assess background quit rates among randomly selected non-participants from each cluster, using the same measures as are used in the trial. This will provide an indication of the full effect size. The contents of the intervention and control conditions have been coded into component behaviour change techniques (BCTs) using an established taxonomy of smoking cessation BCTs [20]. BCTs identified at least once within each condition are listed in Table 2. BCTs marked with an asterisk (*) have been shown in other research to be associated significantly with improved short-term quit outcomes [21]. Fidelity assessment Fidelity will be promoted by following standard guidelines [22]. To assess fidelity, the first 10 episodes of quit advice will be tape-recorded and the recording checked against the script by the health professional. If discrepancies are discovered at this stage, retraining will be undertaken. Adherence by the participant will be assessed by requesting participants to record practice of the breathing exercises on a blank calendar provided and checked at the 4-week follow-up. Randomization As low-income blocks in the study area are stratified into two community types (‘J.J cluster’ and ‘resettlement colony’), an equal number of each was identified for inclusion prior to randomization. Further, a list of all tobacco users who met the eligibility criteria for recruitment (being aged 23 years or above, a current tobacco user and residing in the low-income community) was also finalized within each cluster prior to randomization of clusters to the intervention or control arm. The random sequence for allocation of clusters to the intervention and control arms was generated independently by computer using a random number generator in SPSS software with a uniform distribution and blocked so that an equal number of each community type are allocated to the intervention and control conditions. Therefore, 16 clusters (eight from each community type) will be allocated randomly to the BA-YBA arm and 16 clusters (eight from each community type) to the VBA arm (see Fig. 1). Sample size For a two-group comparison of proportions of a binary outcome with a standard alpha of 0.05, 390 individuals per group, or a total of 780 individuals, would provide Addiction, 109, 371–378
Brief tobacco cessation intervention trial
375
Table 2 Behaviour change techniques (BCT) analysis of intervention and control conditions of the Brief Advice and Breathing EXercises (BABEX) trial. Intervention condition (brief advice + yogic breathing)
Control (very brief advice)
BM1: provide information on the consequences of smoking and smoking cessation
BM1: provide information on the consequences of smoking and smoking cessation RC4: provide expectations regarding the treatment programme
BM5: provide normative information on others’ behaviours and experiences BM9: help to identify reasons for wanting and not wanting to stop smoking *BS2: relapse prevention and coping BS3: facilitate action planning/develop treatment plan BS6: prompt self-recording BM10: emphasize the importance of abrupt cessation BS11: advise on avoidance of cues for smoking BS14: teach relaxation techniques *A1: advise on stop smoking medications *A2: advise on/facilitate use of social support RC4: provide expectations regarding the treatment programme RC5: offer/direct towards appropriate written materials RC6: provide information on withdrawal symptoms *RC9: summarize information/confirm client decisions
*Indicates BCT shown to be associated with improved 4-week carbon monoxide-validated quit outcomes in English smoking cessation services [20].
90% power to detect a difference of 5% between the intervention (BA-YBA) group (7%) and the control (VBA) group (2%) in two-tailed analysis [23]. As this sample size is based on individual randomization, this sample size has to be inflated for the design effect (Deff) of clustering. Taking into account the clustering designated as mcluster, the sample size is obtained by multiplying the individual randomization sample size by the design effect [24,25]. Therefore, the sample size required is calculated as mcluster = Deff × mindividual. The design effect is calculated as = 1 + (n* − 1) × ICC, where n* = average cluster size and ICC = intracluster correlation co-efficient. The ICC is defined as the ratio of the between-cluster variance to the total variance. As there are no directly applicable values of ICC for our study, it is advised to calculate the sample size for a range of design effects. The power analysis is geared towards the main outcome variable and analysis using logistic regression adjusted for clustering [24]. Intracluster correlation coefficient values in the public health context are typically very small, at approximately 0.001 [23]. Given the fixed number of clusters and estimates from the uncorrected power analysis, the average cluster size is approximately 24.4 (= 780/32), resulting in a design effect of 1.024 [= 1 + (25-1) × 0.001] or, using an ICC a magnitude larger of 0.01, a design effect of 1.24. The required total sample size to detect an effect with 90% power is therefore 967 (= 780 × 1.24), i.e. 484 participants in each arm, or a minimum of 30.2 (= 967/ 32) participants from each cluster, assuming equal cluster size. Thus it is proposed to recruit a total of 992 © 2014 Society for the Study of Addiction
(= 31 × 32) participants into this study, maximizing the number of clusters [26]. This sample size provides sufficient power at 90% to detect significant differences for an expected design effect of 1.24 and at 80% power for correspondingly higher design effects of up to 1.84.
Participant recruitment A census survey of the randomly selected blocks identified as clusters for the study was undertaken from January to August 2012. This included 30 655 adults, of whom a total of 7467 tobacco users were identified to prepare the sampling frame. This census survey also provided the baseline prevalence of tobacco use in each cluster. In order to select the 31 tobacco users per cluster, each cluster is divided into three to four approximately equal geographical subclusters (based on streets/lanes) for selection of participants. From each geographical cluster, tobacco users will be invited to participate by visiting houses consecutively until the target number of eight or 10 participants per subcluster is met. Those meeting the inclusion criteria (see below) will be asked to provide consent to participate in the trial. In the case of multiple eligible participants in the same household, only one participant from each household will be recruited. In case the target number of eight to 10 tobacco users is not met in a particular geographic subcluster, the remaining target number will be added to the next subcluster. Recruitment will continue in the cluster until the target number of 31 is achieved. Addiction, 109, 371–378
376
Bidyut K. Sarkar et al.
The inclusion criteria for participation will be any adult of age 23 years or above, who is a current self-reported daily user of any tobacco product, and residing in the study community. There will be no exclusion criteria except pregnancy in women, as the safety of yogic breathing exercises in pregnant women is not established. As a previous research study conducted in the same area undertook interventions for tobacco cessation in the age group 10–19 years—who would be 22 years old by the time recruitment for the current study is completed—an older age cohort is the target for inclusion to avoid contamination with the previous research study. In addition, as the practice of yogic breathing exercises involves abdominal movements, the effect of which has not been evaluated in pregnant women, this group will not be advised to use them. Consent Informed written consent in the local language, Hindi, will be obtained from all participants individually. Those who cannot read or write will be read out the patient information sheet and will be requested to affix their thumb impression in the presence of a literate witness who will countersign the consent form. Due to the nature of the intervention, blinding the participant to the intervention will not be feasible. Measures
slightly higher levels than gas chromatography–mass spectrometry (GC-MS) because of inclusion of metabolites [29], the usual cut-off for GC-MS of 13 ng/ml will be increased by the ratio of the typical ELISA to GC-MS value. Secondary outcome measures Major secondary outcome measures will be 1-week point prevalence abstinence assessed 4 weeks and 7 months after intervention delivery. Attempts to stop tobacco use will also be documented. Further, immediately before and after the intervention, tobacco-related cognitions (intention to stop tobacco use and self-efficacy regarding tobacco use cessation) will be assessed [31]. In addition, satisfaction with the yogic breathing exercises will be assessed immediately after delivery of intervention in the BA-YBA group only. Participants will also be asked about the usefulness of yogic breathing and the frequency of use of these exercises at the 4-week follow-up. Data analysis and management Case report forms (CRF) will be prepared. All data will be entered directly into the CRF; there will be no other source documents. Double data entry will be undertaken into an electronic database. Data analysis will be undertaken by the investigators using the SPSS statistical analysis package.
Baseline measures
Main analysis
At the first visit, tobacco users will be asked to provide socio-demographic information (age, gender, occupation, education and income), information about their tobacco use history, any previous quit attempts, knowledge, attitudes and quitting intentions, nicotine dependence and screening for current depression. Nicotine dependence will be assessed with the Fagerström test for nicotine dependence (FTND) [27] modified for smokeless users as recommended by Fagerström [28]. Saliva samples will be collected for salivary cotinine level assessment in a randomly selected 40% of participants in both arms at baseline, and from all self-reported quitters at the follow-up at 7 months for validation of quit status
An intention-to-treat analysis in which all participants including those lost to follow-up will be counted, as continuing tobacco use will be undertaken [32]. The ‘complex samples’ procedure within the SPSS software will be used allowing logistic regression of the outcome measures on to the treatment group, taking account of the hierarchical organization of the data set, with confidence intervals and P-values adjusted to take this into account. It is possible that a bias will be introduced if different proportions of eligible participants agree to take part in the intervention and control conditions. In that event we will undertake a sensitivity analysis in which the denominator, instead of being enrolled participants, is participants who were eligible to be enrolled whether or not they agreed to receive the intervention and be followed-up. Those who decline to take part will be considered as continuing tobacco users.
Primary outcome measure The primary outcome measure will follow the Russell Standard: self-report of a maximum of five cigarettes/ bidis or five occasions of use of chewing tobacco (or any other form of tobacco) in the 6 months preceding the follow-up at 7 months after intervention delivery, validated by salivary cotinine concentration of less than 20 ng/ml. Cotinine will be assessed by enzyme-linked immunosorbent assay (ELISA) [29]. As this tends to yield © 2014 Society for the Study of Addiction
Subgroup analyses Planned subgroup analyses will involve separate estimation of effect-sizes for users of smokeless and smoked tobacco products. Similarly, estimation of any differences Addiction, 109, 371–378
Brief tobacco cessation intervention trial
in the intervention delivery by the trained health professional versus trained field health workers or delivery of intervention singly or in groups will also be analysed.
DISCUSSION This study is set up to be a pragmatic trial evaluating a low-cost, easily scalable intervention to promote tobacco use cessation, designed specifically for application in low- to middle-income countries such as India. It follows the Medical Research Council (MRC) guidelines for developing and evaluating complex interventions which states that: ‘complex interventions may work best if tailored to local circumstances rather than being completely standardised’ [33]. Ideally, a no-intervention control arm would be included to test the full effect of the experimental intervention. However, on advice that this would be considered unethical, a very brief advice condition is used as the control. Thus, the effect of the experimental intervention may be underestimated. To address this, we will conduct a follow-up survey of randomly selected samples of tobacco users who were not approached to take part in the trial in each of the participating communities to assess the background quit rate. It is recognized that the design will not permit dissociation of the effect of the breathing exercises from quit advice. However, a three-arm trial that would enable us to do this would be prohibitively expensive. As the priority at this stage is to identify a cost-effective, scalable intervention that is effective, this multi-component intervention (quit advice and yogic breathing exercises) is evaluated with a two-arm cluster randomized study design. Quit advice has been found to be slightly more effective if there are follow-up sessions [7], but as follow-up sessions would not be an easily scalable intervention, only a single quit advice session without any follow-up intervention visits was selected for evaluation. Following the same rationale, only a single session of training for simple beginner-level, yogic breathing exercises instead of multiple training sessions for yogic breathing exercises was adopted for this study. If the intervention is found to be effective, it could inform tobacco cessation support policy at the national and international levels by providing a low-cost, nonphysician-dependent, easily scalable option for the government to deliver to a substantial proportion of the 275 million tobacco users in India. It is important that this does not detract from other core tobacco control interventions, such as raising tobacco duty, but it could benefit millions of tobacco users in this and other countries with limited low resources. © 2014 Society for the Study of Addiction
377
Governance and ethical considerations The trial will conform to the UCL ethical procedures and subject to oversight by the UCL Ethics Committee, London, UK and PHFI Ethics Committee, New Delhi, India. UCL Ethics Registration ID: 3051/002; Data Protection Registration: Ref No Z6364106/2011/05/37. The study has received approval from UCL Research Ethics (3051/002 dated 20 October 2011) and Institutional Ethics Committee, Public Health Foundation of India, New Delhi, India (TRC-IEC-122/11 dated 23 February 2012). The protocol has also been registered at ISRCTN with trial registration no: ISRCTN23362894 and acronym BABEX. Clinical trial registration Trial registration: ISRCTN23362894 Acronym BABEX. Declaration of interests B.K.S. has a UKC–Wellcome Trust capacitystrengthening strategic award to the Public Health Foundation of India and has no conflicts of interest to declare. M.A. and K.S.R. have no conflicts of interest to declare. L.S. has received honoraria for talk and travel expenses from manufacturers of medications for smoking cessation to attend meetings and workshops. F.L. has received travel funds and hospitality from manufacturers of medications for smoking cessation. R.W. undertakes consultancy and research for and receives travel funds and hospitality from manufacturers of medications for smoking cessation. Acknowledgements This work is supported by a Wellcome Trust Capacity Strengthening Strategic Award to the Public Health Foundation of India and a consortium of UK Universities (WTP grant reference no: 6936), with additional support from Cancer Research UK and NCSCT UK. References 1. Mathers C. D., Loncar D. Projections of global mortality and burden of disease from 2002 to 2030. PLOS Med 2006; 3: e442. 2. World Health Organization (WHO). WHO Report on the Global Tobacco Epidemic, 2008: The MPOWER package. Geneva: WHO, 2008. 3. Jha P., Jacob B., Gajalakhshmi V., Gupta P. C., Dhingra N., Kumar R. et al. A nationally representative case–control study of smoking and death in India. N Engl J Med 2008; 358: 1137–47. 4. International Institute for Population Sciences (IIPS) Mumbai. Global Adult Tobacco Survey (GATS INDIA), 2009–2010. New Delhi: Ministry of Health and Family Welfare, Government of India; 2010. Addiction, 109, 371–378
378
Bidyut K. Sarkar et al.
5. West R. The clinical significance of ‘small’ effects of smoking cessation treatements. Addiction 2007; 104: 506–9. 6. Kotz D., West R. Explaining the social gradient in smoking cessation: it’s not in the trying, but in the succeeding. Tob Control 2009; 18: 43–6. 7. Stead L. F., Bergson G., Lancaster T. Physician advice for smoking cessation (Review). Cochrane Database Syst Rev 2008; (2): CD000165. 1–55. 8. Lancaster T., Stead L. Physician advice for smoking cessation. Cochrane Database Syst Rev 2004; (4): CD000165. 9. Lancaster T., Stead L. F. Individual behavioural counselling for smoking cessation. Cochrane Database Syst Rev 2005; (2): CD001292. doi: 10.1002/14651858.CD001292.pub2. 10. Stead L., Lancaster T. Group behaviour therapy programmes for smoking cessation— 2005. Cochrane Database Syst Rev 2009; (2): CD001007. doi: 10.1002/ 14651858.CD001007.pub2. 11. Aveyard P., Begh R., Parsons A., Robert W. Brief opportunistic smoking cessation interventions: a systematic review and meta-analysis to compare advice to quit and offer of assistance. Addiction 2011; 107: 1066–73. 12. Murthy P., Saddichha S. Tobacco cessation services in India: recent developments and the need for expansion. Indian J Cancer 2010; 47: S69–74. 13. West R., May S., West M., Croghan E., McEwen A. Performance of English stop smoking services in first 10 years: analysis of service monitoring data. BMJ 2013; 347: f4921. 14. McClernon F. J., Westman E. C., Rose J. E. The effects of controlled deep breathing on smoking withdrawal symptoms in dependent smokers. Addict Behav 2004; 29: 765– 72. 15. Shahab L., Sarkar B. K., West R. The acute effects of yogic breathing exercises on craving and withdrawal symptoms in abstaining smokers. Psychopharmacology (Berl) 2013; 225: 875–82. 16. West R., McNeill A., Raw M. Smoking cessation guidelines for health professionals: an update. Thorax 2000; 55: 987– 99. 17. Kober H., Kross E. F., Mischel W., Hart C. L., Ochsner K. N. Regulation of craving by cognitive strategies in cigarette smokers. Drug Alcohol Depend 2010; 106: 52–5. 18. Kochupillai V., Kumar P., Singh D., Aggarwal D., Bhardwaj N., Bhutani M. et al. Effect of rhythmic breathing (Sudarshan Kriya and Pranayam) on immune functions and tobacco addiction. Ann N Y Acad Sci 2005; 1056: 242– 52. 19. Ahluwalia J. S., Harris J. K., Catley D., Okuyemi K. S. Sustained—release bupropion for smoking cessation in
© 2014 Society for the Study of Addiction
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
African Americans—a randomized controlled trial. JAMA 2002; 288: 468–74. Michie S., Hyder N., Walia A., West R. Development of a taxonomy of behavioural support for smoking cessation. Addict Behav 2011; 36: 315–9. West R., Walia A., Hyder N., Shahab L., Michie S. Behaviour change techniques used by the English Stop Smoking Services and their association with short term outcomes. Nicotine Tob Res 2010; 12: 742–747. Advance access. Borrelli B. The assessment, monitoring, and enhancement of treatment fidelity in public health clinical trials. J Public Health Dent 2011; 71: 552–63. Machin D., Campbell M. J., Tan S. B., Tan S. H. Sample Size Tables for Clinical Studies (with CD Rom), 3rd edn. Singapore: Wiley Blackwell; 2009. Donner A., Klar N., editors. Design and Analysis of Cluster Randomized Trials in Health Research. Sussex: John Wiley & Sons; 2000. Hayes R. J., Moulton L. H., editors. Cluster Randomised Trials. Boca Raton, FL: Chapman & Hall/CRC Press/Taylor & Francis Group; 2009. Campbell M. K., Piaggio G., Elbourne D. R., Altman D. G. Consort 2010 statement: extension to cluster randomized trials. BMJ 2012; 345: e5661. 1–21. Heatherton T. F., Kozlowski L. T., Frecker R. C., Fagerstrom K. O. The Fagerstrom test for nicotine dependence: a revision of the Fagerstrom Tolerance Questionnaire. Br J Addict 1991; 86: 1119–27. Ebbert J. O., Patten C. A., Schroeder D. R. The Fagerstrom Test for Nicotine Dependence–Smokeless Tobacco (FTNDST). Addict Behav 2006; 31: 1716–21. Watts R. R., Langone J. J., Knight G. J., Lewtas J. Cotinine analytical workshop report: consideration of analytical methods for determining cotinine in human body fluids as a measure of passive exposure to tobacco smoke. Environ Health Perspect 1990; 84: 173–82. Alterman A. I., Gariti P., Niedbala R. S. Varying results for immunoassay screening kits for cotinine level. Psychol Addict Behav 2002; 16: 256–9. Hall S., Bishop A. J., Marteau T. M. Increasing readiness to stop smoking in women undergoing cervical screening: evaluation of two leaflets. Nicotine Tob Res 2003; 5: 821–6. Newell J. D. Intention to treat analysis: implications for quantitative and qualitative research. Int J Epidemiol 1992; 21: 837–41. Craig P., Dieppe P., Macintyre S., Mitchie S., Nazareth I., Petticrew M. Developing and evaluating complex interventions: the new MRC guidelines. BMJ 2008; 337: a1655.
Addiction, 109, 371–378
TRIAL PROTOCOL
doi:10.1111/add.12420
A cluster randomized controlled trial of a brief tobacco cessation intervention for low-income communities in India: study protocol Bidyut K. Sarkar1,2, Lion Shahab2, Monika Arora1, Fabiana Lorencatto3, K. Srinath Reddy1 & Robert West2 Public Health Foundation of India, New Delhi, India,1 Department of Epidemiology and Public Health, University College London, London, UK2 and NCSCT, University College London, London, UK3
ABSTRACT Background India has 275 million adult tobacco users and tobacco use is estimated to contribute to more than a million deaths in the country each year. There is an urgent need to develop and evaluate affordable, practicable and scalable interventions to promote cessation of tobacco use. Because tobacco use is so harmful, an increase of as little as 1 percentage point in long-term quit success rates can have an important public health impact. This protocol paper describes the rationale and methods of a large randomized controlled trial which aims to evaluate the effectiveness of a brief scalable smoking cessation intervention delivered by trained health professionals as an outreach programme in poor urban communities in India. Methods/design This is a pragmatic, two-arm, community-based cluster randomized controlled trial focused on tobacco users in low-income communities. The treatment arm is a brief intervention comprising brief advice including training in craving control using simple yogic breathing exercises (BA-YBA) and the control arm is very brief advice (VBA). Of a total of 32 clusters, 16 will be allocated to the intervention arm and 16 to the control arm. Each cluster will have 31 participants, making a total of 992 participants. The primary outcome measure will follow the Russell Standard: self-report of sustained abstinence for at least 6 months following the intervention confirmed at the final follow-up by salivary cotinine. Discussion This trial will inform national and international policy on delivery of scalable and affordable brief outreach interventions to promote tobacco use cessation in low resource settings where tobacco users have limited access to physicians and medications. Keywords India, intervention, non-pharmacological, outreach, single session, smokeless, tobacco cessation, trial, validated quit rate, yogic breathing. Correspondence to: Bidyut K. Sarkar, Public Health Foundation of India, New Delhi 110019, India. E-mail: [email protected] Submitted 12 February 2013; initial review completed 1 May 2013; final version accepted 11 November 2013
INTRODUCTION Tobacco use is a risk factor for six of the eight leading causes of death in the world [1]. More than 5 million deaths are attributed to tobacco use globally each year [2] and an estimated 1 million deaths in India alone [3]. India has 275 million tobacco users, of whom 164 million are smokeless tobacco users, 69 million are smokers and 42 million use both smoked and smokeless tobacco [4]. With regard to smoking, it has been shown that an increase of as little as 1 percentage point in longterm quit success rates can have an important public health impact [5]. A 1% decrease in prevalence of © 2014 Society for the Study of Addiction
tobacco use in India would represent 7.8 million fewer tobacco users. The prevalence of tobacco use is much higher among lower socio-economic groups in many countries. Moreover, quit rates are typically lower among those with lower education level and socio-economic status [6]. In India, 70% of the health-care services are obtained from the private sector and very few people are covered by health insurance. Therefore, lower-income groups in India would have much less access to healthcare services and would be prone to higher morbidity and mortality due to tobacco use. This trial focuses upon the most vulnerable strata to which the majority Addiction, 109, 371–378
372
Bidyut K. Sarkar et al.
of tobacco users in India belong, those with severe economic disadvantage. Brief advice is a low-cost intervention which can be delivered en masse by trained health workers [7]. Systematic reviews of randomized controlled trials in western samples have found that brief advice to quit smoking is better than no advice, increasing 12-month quit rates by 1–3% [7–10], with offer of support leading to a larger effect than simple advice to quit [11]. However, the effectiveness of brief advice for smoking cessation has yet to be evaluated in a randomized trial in low- or middle-income countries; nor is it known in any setting whether brief advice is effective in reducing smokeless tobacco use. Unlike developed countries such as the United Kingdom, where behavioural therapy and/or pharmacotherapy is provided at little or no cost to patients, support services are very limited for quitting tobacco use in India [12–13]. Hence, there is an urgent need to identify and evaluate low-cost, affordable, non-resource-intensive, easily scalable interventions to promote cessation of tobacco use in India. A promising, affordable and culturally appropriate option to enhance brief advice and aid smoking cessation in the Indian context is the use of breathing exercises, known as ‘Pranayama’ within yoga. Yoga is a traditional method of achieving good physical and mental health. Therefore, yogic breathing exercises are expected to have a high degree of acceptability as a tobacco cessation aid.
Recently, Pranayama has been widely popularized in India by guru Ramdev of Haridwar, who has provided mass training through television and mass yoga camps attended by tens of thousands. He, along with other yoga gurus, have attributed the benefits of breathing exercises specifically to two key breathing exercises called ‘Kapalbhati-KB’ (normal deep inhalation—forceful exhalation) and ‘Anulom vilom-AV’ (alternate nostril breathing), which are also the easiest to learn and practice. These two breathing exercises were selected for the present intervention (see Table 1). Although relatively little research has been undertaken to test the hypothesis that yogic breathing exercises may improve quitting rates, the existing research is promising. McClernon et al. [14] observed that controlled deep breathing significantly reduced smoking withdrawal symptoms, including craving for cigarettes and negative affect. More recently, Shahab et al. have reported a significant reduction in craving immediately after a session of yogic breathing exercises in yoga-naive smokers abstinent for at least 10 hours who were randomized to an intervention arm and a video control arm [15]. Given that there is good evidence that the experience of severe cravings early on in quit attempts is an important mediator of relapse [16–17], yogic breathing exercises may increase smoking cessation rates. Indeed, Kochupillai et al. [18] reported a positive impact of yogic breathing exercises on tobacco abstinence and found the 12-month
Table 1 Breathing exercise instructionsa. Kapalbhati Normal inhalation–forceful exhalation (1–2 minutes on empty stomach once in morning)
Anulom vilom Alternate nostril inhalation–exhalation (3 minutes in morning and then for cravings during the day)
• Start by breathing deeply: inhale slowly and deeply filling up your abdomen, then fill up your rib cage up to your collar bone and then exhale slowly, emptying your rib cage and then pull in your abdomen emptying it fully. Repeat a few times. • Now, inhale your breath naturally and exhale forcefully (moderate force) with drawing in/pulling in of the abdomen • Do this natural inhale–exhale cycle forcefully repeatedly for a period of 1 minute and then rest by breathing normally • Do this for 1 minute at a time, rest by normal breathing in between to a total duration of 2–3 minutes for the first few days and then you can gradually increase the duration • Increase the duration to 5 minutes over a period of 2 months • Do not strain yourself and stop at any point in time if you feel exhausted • Rest by breathing normally for half a minute before next exercise
• Sit in a comfortable position with the back straight and upright • Fold the fore and middle fingers of your hand into the palm and position your hand on the bridge of your nose to enable use of your thumb and the ring finger to close the nostrils alternatively • Close the right nostril with the thumb of your right hand and breathe in deeply inhaling through the left nostril, then close the left nostril with your ring finger and release the thumb from the right nostril and now exhale through the right nostril • Then inhale through the right nostril, close the right nostril and exhale through the left nostril • Now inhale through the left nostril, close the left nostril and exhale through right nostril
a In this study a set of two simple, beginner-level breathing exercises will be taught to participants. These breathing exercises are based on yogic ‘Pranayama’ recommended by Yoga Guru Swami Ramdev, Haridwar, India, for use for the general population. The participants will be shown a short video demonstration of the breathing exercises and will then be trained to undertake the exercises on their own for about 5 minutes. The breathing exercises should be performed in an upright sitting position. This can then be repeated during the day based on need and feasibility.
© 2014 Society for the Study of Addiction
Addiction, 109, 371–378
Brief tobacco cessation intervention trial
abstinence rates of the yoga breathing exercises group to be 21%, similar to that reported for pharmacological agent bupropion [19]. Given the above findings and the gap in the literature, this pragmatic trial was designed to evaluate the effectiveness of a brief tobacco use cessation intervention. The intervention comprises a single session of face-to-face quit advice and training in yogic breathing exercises aimed at tobacco users in low-income communities in India.
373
of the selected low-income administrative blocks can be one of two types of community: an authorized government rehabilitation known as a ‘resettlement colony’ with slightly better facilities, and an unauthorized settlement known as ‘J.J. cluster’ (J.J. stands for ‘Jhuggi-Jhopri’, meaning huts in Hindi, the local language). Equal numbers of both types of community was selected for inclusion in the study. Recruitment into the study started in January 2012 and it is currently ongoing. The study is anticipated to conclude in November 2013. Trial interventions and delivery
METHODS AND DESIGN Study design and timing This is a cluster randomized trial with two arms: a treatment arm consisting of a brief face-to-face session of quit advice including training in yogic breathing exercises (BA-YBA) versus a control arm of very brief advice (VBA) alone. The participants will be followed-up after the intervention to assess tobacco use status (see Fig. 1). The proposed study area has a total population of approximately 36 000 adults living in randomly selected administrative blocks of 32 large slum areas of lowincome residents spread over a wide geographical area in the city of Delhi, India. Each of these blocks will be considered as a cluster for randomization in the study. Each
The treatment arm is a single session of face-to-face tobacco quit advice including training in BA-YBA to control cravings. The control arm is VBA. A written standard operating procedure (SOP) will be followed by the research team, and a written script for quit advice and a standard video for training on yogic breathing exercises will be used for intervention delivery. For both groups, the baseline interview will be conducted at the doorstep of the tobacco-user’s residence. The intervention will be delivered by a member of the delivery team. Delivery teams will consist of one trained health professional and two field health workers who will be trained and supervised by the health professional. All content is delivered in Hindi, the local language. The
32 Clusters (Low-income urban slum areas) EsƟmated Adult PopulaƟon: 30,655 Randomized straƟfied allocaƟon by computer
16 clusters allocated to intervenƟon
16 clusters allocated to control
(8 JJ clusters & 8 ReseƩlement)
(8 JJ clusters & 8 ReseƩlement)
16 clusters given BA-YBA intervenƟon (Brief Advice and yogic breathing exercises)
Figure 1 Brief Advice and Breathing EXercises (BABEX): cluster randomized trial flowchart © 2014 Society for the Study of Addiction
16 clusters given VBA intervenƟon (Very Brief Advice)
Follow-up at 4 weeks
Follow-up at 4 weeks
Self-reported absƟnence rate
Self-reported absƟnence rate
Follow-up at 7 months
Follow-up at 7 months
Self-reported absƟnence rate for last 6 months (RS6)
Self-reported absƟnence rate for last 6 months (RS6)
Validated by salivary coƟnine
Validated by salivary coƟnine
Addiction, 109, 371–378
374
Bidyut K. Sarkar et al.
intervention will be delivered face to face to participants in both arms, either singly or in groups of two to four participants. The content of the BA-YBA, a brief intervention comprising advice to quit and breathing exercises, will: 1 Explain the important groups of chemicals present and released on smoking and chewing tobacco and the specific diseases caused by tobacco use including cancer, respiratory and cardiovascular diseases. 2 Identify the personal health risks associated with the individual’s specific form(s) of tobacco use and also the economic burden and implications of tobacco use, including cost of hospitalization and treatment of tobacco use-related diseases. 3 Explain the benefits of early and complete abstinence. 4 Explain that the chances of success are higher with the use of some support, as most people experience temporary withdrawal symptoms, as well as the importance of avoiding places, situations and company which could precipitate relapse for a while; and advice to quit within the next month. 5 Give advice to seek support of family members and close friends in the quitting effort. 6 Explain the support options available for quitting, including medications such as nicotine replacement therapy and non-pharmacological aids such as yogic breathing exercises. 7 Perform training in two yogic breathing exercises (Kapalbhati and Anulom-vilom) with the help of a standard training video (see Table 1 for details on instructions). 8 Recommendation to use these breathing exercises once in the morning (both exercises) and Anulomvilom whenever they experience cravings for tobacco during the day. 9 Check understanding with a brief practice session of breathing exercises by the tobacco user for 2–3 minutes. 10 Promote monitoring of adherence by participants by providing a blank calendar and instructions to place a tick by days when the breathing exercises have been undertaken. 11 Provide a leaflet of breathing exercise instructions. 12 State that the research team will follow-up after 1 and 7 months. The VBA, the control condition, will: 1 Explain that tobacco use is very harmful and causes major diseases such as cancer. 2 Explain that quitting is the most important action they can take for their health. 3 State that the research team will follow-up after 1 month and 7 months. To assess the full effect of the intervention, the control condition would ideally involve no advice on tobacco use, © 2014 Society for the Study of Addiction
but this was considered unethical by relevant review boards. Therefore, any effect-size estimate produced by this trial is likely to be conservative. We will conduct a survey to assess background quit rates among randomly selected non-participants from each cluster, using the same measures as are used in the trial. This will provide an indication of the full effect size. The contents of the intervention and control conditions have been coded into component behaviour change techniques (BCTs) using an established taxonomy of smoking cessation BCTs [20]. BCTs identified at least once within each condition are listed in Table 2. BCTs marked with an asterisk (*) have been shown in other research to be associated significantly with improved short-term quit outcomes [21]. Fidelity assessment Fidelity will be promoted by following standard guidelines [22]. To assess fidelity, the first 10 episodes of quit advice will be tape-recorded and the recording checked against the script by the health professional. If discrepancies are discovered at this stage, retraining will be undertaken. Adherence by the participant will be assessed by requesting participants to record practice of the breathing exercises on a blank calendar provided and checked at the 4-week follow-up. Randomization As low-income blocks in the study area are stratified into two community types (‘J.J cluster’ and ‘resettlement colony’), an equal number of each was identified for inclusion prior to randomization. Further, a list of all tobacco users who met the eligibility criteria for recruitment (being aged 23 years or above, a current tobacco user and residing in the low-income community) was also finalized within each cluster prior to randomization of clusters to the intervention or control arm. The random sequence for allocation of clusters to the intervention and control arms was generated independently by computer using a random number generator in SPSS software with a uniform distribution and blocked so that an equal number of each community type are allocated to the intervention and control conditions. Therefore, 16 clusters (eight from each community type) will be allocated randomly to the BA-YBA arm and 16 clusters (eight from each community type) to the VBA arm (see Fig. 1). Sample size For a two-group comparison of proportions of a binary outcome with a standard alpha of 0.05, 390 individuals per group, or a total of 780 individuals, would provide Addiction, 109, 371–378
Brief tobacco cessation intervention trial
375
Table 2 Behaviour change techniques (BCT) analysis of intervention and control conditions of the Brief Advice and Breathing EXercises (BABEX) trial. Intervention condition (brief advice + yogic breathing)
Control (very brief advice)
BM1: provide information on the consequences of smoking and smoking cessation
BM1: provide information on the consequences of smoking and smoking cessation RC4: provide expectations regarding the treatment programme
BM5: provide normative information on others’ behaviours and experiences BM9: help to identify reasons for wanting and not wanting to stop smoking *BS2: relapse prevention and coping BS3: facilitate action planning/develop treatment plan BS6: prompt self-recording BM10: emphasize the importance of abrupt cessation BS11: advise on avoidance of cues for smoking BS14: teach relaxation techniques *A1: advise on stop smoking medications *A2: advise on/facilitate use of social support RC4: provide expectations regarding the treatment programme RC5: offer/direct towards appropriate written materials RC6: provide information on withdrawal symptoms *RC9: summarize information/confirm client decisions
*Indicates BCT shown to be associated with improved 4-week carbon monoxide-validated quit outcomes in English smoking cessation services [20].
90% power to detect a difference of 5% between the intervention (BA-YBA) group (7%) and the control (VBA) group (2%) in two-tailed analysis [23]. As this sample size is based on individual randomization, this sample size has to be inflated for the design effect (Deff) of clustering. Taking into account the clustering designated as mcluster, the sample size is obtained by multiplying the individual randomization sample size by the design effect [24,25]. Therefore, the sample size required is calculated as mcluster = Deff × mindividual. The design effect is calculated as = 1 + (n* − 1) × ICC, where n* = average cluster size and ICC = intracluster correlation co-efficient. The ICC is defined as the ratio of the between-cluster variance to the total variance. As there are no directly applicable values of ICC for our study, it is advised to calculate the sample size for a range of design effects. The power analysis is geared towards the main outcome variable and analysis using logistic regression adjusted for clustering [24]. Intracluster correlation coefficient values in the public health context are typically very small, at approximately 0.001 [23]. Given the fixed number of clusters and estimates from the uncorrected power analysis, the average cluster size is approximately 24.4 (= 780/32), resulting in a design effect of 1.024 [= 1 + (25-1) × 0.001] or, using an ICC a magnitude larger of 0.01, a design effect of 1.24. The required total sample size to detect an effect with 90% power is therefore 967 (= 780 × 1.24), i.e. 484 participants in each arm, or a minimum of 30.2 (= 967/ 32) participants from each cluster, assuming equal cluster size. Thus it is proposed to recruit a total of 992 © 2014 Society for the Study of Addiction
(= 31 × 32) participants into this study, maximizing the number of clusters [26]. This sample size provides sufficient power at 90% to detect significant differences for an expected design effect of 1.24 and at 80% power for correspondingly higher design effects of up to 1.84.
Participant recruitment A census survey of the randomly selected blocks identified as clusters for the study was undertaken from January to August 2012. This included 30 655 adults, of whom a total of 7467 tobacco users were identified to prepare the sampling frame. This census survey also provided the baseline prevalence of tobacco use in each cluster. In order to select the 31 tobacco users per cluster, each cluster is divided into three to four approximately equal geographical subclusters (based on streets/lanes) for selection of participants. From each geographical cluster, tobacco users will be invited to participate by visiting houses consecutively until the target number of eight or 10 participants per subcluster is met. Those meeting the inclusion criteria (see below) will be asked to provide consent to participate in the trial. In the case of multiple eligible participants in the same household, only one participant from each household will be recruited. In case the target number of eight to 10 tobacco users is not met in a particular geographic subcluster, the remaining target number will be added to the next subcluster. Recruitment will continue in the cluster until the target number of 31 is achieved. Addiction, 109, 371–378
376
Bidyut K. Sarkar et al.
The inclusion criteria for participation will be any adult of age 23 years or above, who is a current self-reported daily user of any tobacco product, and residing in the study community. There will be no exclusion criteria except pregnancy in women, as the safety of yogic breathing exercises in pregnant women is not established. As a previous research study conducted in the same area undertook interventions for tobacco cessation in the age group 10–19 years—who would be 22 years old by the time recruitment for the current study is completed—an older age cohort is the target for inclusion to avoid contamination with the previous research study. In addition, as the practice of yogic breathing exercises involves abdominal movements, the effect of which has not been evaluated in pregnant women, this group will not be advised to use them. Consent Informed written consent in the local language, Hindi, will be obtained from all participants individually. Those who cannot read or write will be read out the patient information sheet and will be requested to affix their thumb impression in the presence of a literate witness who will countersign the consent form. Due to the nature of the intervention, blinding the participant to the intervention will not be feasible. Measures
slightly higher levels than gas chromatography–mass spectrometry (GC-MS) because of inclusion of metabolites [29], the usual cut-off for GC-MS of 13 ng/ml will be increased by the ratio of the typical ELISA to GC-MS value. Secondary outcome measures Major secondary outcome measures will be 1-week point prevalence abstinence assessed 4 weeks and 7 months after intervention delivery. Attempts to stop tobacco use will also be documented. Further, immediately before and after the intervention, tobacco-related cognitions (intention to stop tobacco use and self-efficacy regarding tobacco use cessation) will be assessed [31]. In addition, satisfaction with the yogic breathing exercises will be assessed immediately after delivery of intervention in the BA-YBA group only. Participants will also be asked about the usefulness of yogic breathing and the frequency of use of these exercises at the 4-week follow-up. Data analysis and management Case report forms (CRF) will be prepared. All data will be entered directly into the CRF; there will be no other source documents. Double data entry will be undertaken into an electronic database. Data analysis will be undertaken by the investigators using the SPSS statistical analysis package.
Baseline measures
Main analysis
At the first visit, tobacco users will be asked to provide socio-demographic information (age, gender, occupation, education and income), information about their tobacco use history, any previous quit attempts, knowledge, attitudes and quitting intentions, nicotine dependence and screening for current depression. Nicotine dependence will be assessed with the Fagerström test for nicotine dependence (FTND) [27] modified for smokeless users as recommended by Fagerström [28]. Saliva samples will be collected for salivary cotinine level assessment in a randomly selected 40% of participants in both arms at baseline, and from all self-reported quitters at the follow-up at 7 months for validation of quit status
An intention-to-treat analysis in which all participants including those lost to follow-up will be counted, as continuing tobacco use will be undertaken [32]. The ‘complex samples’ procedure within the SPSS software will be used allowing logistic regression of the outcome measures on to the treatment group, taking account of the hierarchical organization of the data set, with confidence intervals and P-values adjusted to take this into account. It is possible that a bias will be introduced if different proportions of eligible participants agree to take part in the intervention and control conditions. In that event we will undertake a sensitivity analysis in which the denominator, instead of being enrolled participants, is participants who were eligible to be enrolled whether or not they agreed to receive the intervention and be followed-up. Those who decline to take part will be considered as continuing tobacco users.
Primary outcome measure The primary outcome measure will follow the Russell Standard: self-report of a maximum of five cigarettes/ bidis or five occasions of use of chewing tobacco (or any other form of tobacco) in the 6 months preceding the follow-up at 7 months after intervention delivery, validated by salivary cotinine concentration of less than 20 ng/ml. Cotinine will be assessed by enzyme-linked immunosorbent assay (ELISA) [29]. As this tends to yield © 2014 Society for the Study of Addiction
Subgroup analyses Planned subgroup analyses will involve separate estimation of effect-sizes for users of smokeless and smoked tobacco products. Similarly, estimation of any differences Addiction, 109, 371–378
Brief tobacco cessation intervention trial
in the intervention delivery by the trained health professional versus trained field health workers or delivery of intervention singly or in groups will also be analysed.
DISCUSSION This study is set up to be a pragmatic trial evaluating a low-cost, easily scalable intervention to promote tobacco use cessation, designed specifically for application in low- to middle-income countries such as India. It follows the Medical Research Council (MRC) guidelines for developing and evaluating complex interventions which states that: ‘complex interventions may work best if tailored to local circumstances rather than being completely standardised’ [33]. Ideally, a no-intervention control arm would be included to test the full effect of the experimental intervention. However, on advice that this would be considered unethical, a very brief advice condition is used as the control. Thus, the effect of the experimental intervention may be underestimated. To address this, we will conduct a follow-up survey of randomly selected samples of tobacco users who were not approached to take part in the trial in each of the participating communities to assess the background quit rate. It is recognized that the design will not permit dissociation of the effect of the breathing exercises from quit advice. However, a three-arm trial that would enable us to do this would be prohibitively expensive. As the priority at this stage is to identify a cost-effective, scalable intervention that is effective, this multi-component intervention (quit advice and yogic breathing exercises) is evaluated with a two-arm cluster randomized study design. Quit advice has been found to be slightly more effective if there are follow-up sessions [7], but as follow-up sessions would not be an easily scalable intervention, only a single quit advice session without any follow-up intervention visits was selected for evaluation. Following the same rationale, only a single session of training for simple beginner-level, yogic breathing exercises instead of multiple training sessions for yogic breathing exercises was adopted for this study. If the intervention is found to be effective, it could inform tobacco cessation support policy at the national and international levels by providing a low-cost, nonphysician-dependent, easily scalable option for the government to deliver to a substantial proportion of the 275 million tobacco users in India. It is important that this does not detract from other core tobacco control interventions, such as raising tobacco duty, but it could benefit millions of tobacco users in this and other countries with limited low resources. © 2014 Society for the Study of Addiction
377
Governance and ethical considerations The trial will conform to the UCL ethical procedures and subject to oversight by the UCL Ethics Committee, London, UK and PHFI Ethics Committee, New Delhi, India. UCL Ethics Registration ID: 3051/002; Data Protection Registration: Ref No Z6364106/2011/05/37. The study has received approval from UCL Research Ethics (3051/002 dated 20 October 2011) and Institutional Ethics Committee, Public Health Foundation of India, New Delhi, India (TRC-IEC-122/11 dated 23 February 2012). The protocol has also been registered at ISRCTN with trial registration no: ISRCTN23362894 and acronym BABEX. Clinical trial registration Trial registration: ISRCTN23362894 Acronym BABEX. Declaration of interests B.K.S. has a UKC–Wellcome Trust capacitystrengthening strategic award to the Public Health Foundation of India and has no conflicts of interest to declare. M.A. and K.S.R. have no conflicts of interest to declare. L.S. has received honoraria for talk and travel expenses from manufacturers of medications for smoking cessation to attend meetings and workshops. F.L. has received travel funds and hospitality from manufacturers of medications for smoking cessation. R.W. undertakes consultancy and research for and receives travel funds and hospitality from manufacturers of medications for smoking cessation. Acknowledgements This work is supported by a Wellcome Trust Capacity Strengthening Strategic Award to the Public Health Foundation of India and a consortium of UK Universities (WTP grant reference no: 6936), with additional support from Cancer Research UK and NCSCT UK. References 1. Mathers C. D., Loncar D. Projections of global mortality and burden of disease from 2002 to 2030. PLOS Med 2006; 3: e442. 2. World Health Organization (WHO). WHO Report on the Global Tobacco Epidemic, 2008: The MPOWER package. Geneva: WHO, 2008. 3. Jha P., Jacob B., Gajalakhshmi V., Gupta P. C., Dhingra N., Kumar R. et al. A nationally representative case–control study of smoking and death in India. N Engl J Med 2008; 358: 1137–47. 4. International Institute for Population Sciences (IIPS) Mumbai. Global Adult Tobacco Survey (GATS INDIA), 2009–2010. New Delhi: Ministry of Health and Family Welfare, Government of India; 2010. Addiction, 109, 371–378
378
Bidyut K. Sarkar et al.
5. West R. The clinical significance of ‘small’ effects of smoking cessation treatements. Addiction 2007; 104: 506–9. 6. Kotz D., West R. Explaining the social gradient in smoking cessation: it’s not in the trying, but in the succeeding. Tob Control 2009; 18: 43–6. 7. Stead L. F., Bergson G., Lancaster T. Physician advice for smoking cessation (Review). Cochrane Database Syst Rev 2008; (2): CD000165. 1–55. 8. Lancaster T., Stead L. Physician advice for smoking cessation. Cochrane Database Syst Rev 2004; (4): CD000165. 9. Lancaster T., Stead L. F. Individual behavioural counselling for smoking cessation. Cochrane Database Syst Rev 2005; (2): CD001292. doi: 10.1002/14651858.CD001292.pub2. 10. Stead L., Lancaster T. Group behaviour therapy programmes for smoking cessation— 2005. Cochrane Database Syst Rev 2009; (2): CD001007. doi: 10.1002/ 14651858.CD001007.pub2. 11. Aveyard P., Begh R., Parsons A., Robert W. Brief opportunistic smoking cessation interventions: a systematic review and meta-analysis to compare advice to quit and offer of assistance. Addiction 2011; 107: 1066–73. 12. Murthy P., Saddichha S. Tobacco cessation services in India: recent developments and the need for expansion. Indian J Cancer 2010; 47: S69–74. 13. West R., May S., West M., Croghan E., McEwen A. Performance of English stop smoking services in first 10 years: analysis of service monitoring data. BMJ 2013; 347: f4921. 14. McClernon F. J., Westman E. C., Rose J. E. The effects of controlled deep breathing on smoking withdrawal symptoms in dependent smokers. Addict Behav 2004; 29: 765– 72. 15. Shahab L., Sarkar B. K., West R. The acute effects of yogic breathing exercises on craving and withdrawal symptoms in abstaining smokers. Psychopharmacology (Berl) 2013; 225: 875–82. 16. West R., McNeill A., Raw M. Smoking cessation guidelines for health professionals: an update. Thorax 2000; 55: 987– 99. 17. Kober H., Kross E. F., Mischel W., Hart C. L., Ochsner K. N. Regulation of craving by cognitive strategies in cigarette smokers. Drug Alcohol Depend 2010; 106: 52–5. 18. Kochupillai V., Kumar P., Singh D., Aggarwal D., Bhardwaj N., Bhutani M. et al. Effect of rhythmic breathing (Sudarshan Kriya and Pranayam) on immune functions and tobacco addiction. Ann N Y Acad Sci 2005; 1056: 242– 52. 19. Ahluwalia J. S., Harris J. K., Catley D., Okuyemi K. S. Sustained—release bupropion for smoking cessation in
© 2014 Society for the Study of Addiction
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
African Americans—a randomized controlled trial. JAMA 2002; 288: 468–74. Michie S., Hyder N., Walia A., West R. Development of a taxonomy of behavioural support for smoking cessation. Addict Behav 2011; 36: 315–9. West R., Walia A., Hyder N., Shahab L., Michie S. Behaviour change techniques used by the English Stop Smoking Services and their association with short term outcomes. Nicotine Tob Res 2010; 12: 742–747. Advance access. Borrelli B. The assessment, monitoring, and enhancement of treatment fidelity in public health clinical trials. J Public Health Dent 2011; 71: 552–63. Machin D., Campbell M. J., Tan S. B., Tan S. H. Sample Size Tables for Clinical Studies (with CD Rom), 3rd edn. Singapore: Wiley Blackwell; 2009. Donner A., Klar N., editors. Design and Analysis of Cluster Randomized Trials in Health Research. Sussex: John Wiley & Sons; 2000. Hayes R. J., Moulton L. H., editors. Cluster Randomised Trials. Boca Raton, FL: Chapman & Hall/CRC Press/Taylor & Francis Group; 2009. Campbell M. K., Piaggio G., Elbourne D. R., Altman D. G. Consort 2010 statement: extension to cluster randomized trials. BMJ 2012; 345: e5661. 1–21. Heatherton T. F., Kozlowski L. T., Frecker R. C., Fagerstrom K. O. The Fagerstrom test for nicotine dependence: a revision of the Fagerstrom Tolerance Questionnaire. Br J Addict 1991; 86: 1119–27. Ebbert J. O., Patten C. A., Schroeder D. R. The Fagerstrom Test for Nicotine Dependence–Smokeless Tobacco (FTNDST). Addict Behav 2006; 31: 1716–21. Watts R. R., Langone J. J., Knight G. J., Lewtas J. Cotinine analytical workshop report: consideration of analytical methods for determining cotinine in human body fluids as a measure of passive exposure to tobacco smoke. Environ Health Perspect 1990; 84: 173–82. Alterman A. I., Gariti P., Niedbala R. S. Varying results for immunoassay screening kits for cotinine level. Psychol Addict Behav 2002; 16: 256–9. Hall S., Bishop A. J., Marteau T. M. Increasing readiness to stop smoking in women undergoing cervical screening: evaluation of two leaflets. Nicotine Tob Res 2003; 5: 821–6. Newell J. D. Intention to treat analysis: implications for quantitative and qualitative research. Int J Epidemiol 1992; 21: 837–41. Craig P., Dieppe P., Macintyre S., Mitchie S., Nazareth I., Petticrew M. Developing and evaluating complex interventions: the new MRC guidelines. BMJ 2008; 337: a1655.
Addiction, 109, 371–378
