Adapting and Evaluating a Brief Advice Tobacco Intervention in High-Reach, Low-Resource Settings in India

This study aims to promote tobacco cessation among adults in high-reach, low-resource community settings in Mumbai, India. Tobacco use is a major driver of cancer deaths and as of 2017, about 267 million individuals use smokeless and/or smoked tobacco in India. One of the WHO-endorsed evidence-based practices for tobacco cessation is brief advice interventions, which involve screening for tobacco use, advising patients to quit, and referring them to treatment. While these interventions often include medication for tobacco cessation in higher-income countries, such treatments can be an expensive and impractical solution in low- and middle-income countries. The team proposes a simplified brief advice intervention without the use of pharmacotherapy, to be implemented in community-based healthcare settings in Mumbai (TB treatment clinics, NGO-run health centers, and dental practices serving populations of lower-socioeconomic status). A task-shifting model will be used, moving program delivery responsibilities from clinicians to community health practitioners. The team also proposes to use a mobile app and a WhatsApp group to support ongoing training and engagement of practitioners. The central questions are: Does a brief advice intervention adapted for use in low-resource settings in India through task-shifting and technology-based training support result in higher cessation rates than usual care? What are the key barriers to and facilitators of program implementation? The study has three aims: Aim 1: Adapt and pilot-test a tobacco cessation evidence-based program in three types of low-resource community-based healthcare settings in Mumbai. Aim 2: The clinical trial itself involves assessing whether the adapted brief advice program results in increased quit rates among tobacco users (compared to usual care) in three types of healthcare settings. The hypothesis is that those assigned to the brief advice program will be more likely to have maintained tobacco cessation after 6 months compared to those who received usual care. Aim 3: Evaluate the use of communication technologies, such as social media and apps, to support ongoing training and networking among practitioners who are implementing the intervention. The long-term goal is to support adaptation and scale-up of tobacco control EBPs from high-resource to low-resource settings. Towards that goal, the overall objective is to develop a scalable, resource-appropriate brief advice EBP for use in India.

Background and Significance: About 1 million deaths are attributed to tobacco use in India each year. In 2017 there were roughly 267 million tobacco users including 199 million smokeless tobacco users, 100 million smokers, and 32 million dual users. Tobacco use is higher in rural versus urban populations (32% vs. 21%) and among men versus women (24% vs. 14%). In Mumbai, there are disparities in tobacco use based on education and occupation. Cessation is of interest to tobacco users in India. In a national survey, about 39% of smokers and 33% of smokeless tobacco users reported a quit attempt in the year preceding the survey. The challenge in India, given the magnitude of tobacco use and limited resources, is in delivering cessation supports at scale and with low cost. The problem of tobacco is multifaceted and widespread, with a lack of coordination among those engaged in tobacco cessation: community centers, NGOs and clinics. India is a signatory to the WHO Framework Convention on Tobacco Control (FCTC) and launched the National Tobacco Control Program in 2007, though the impact is limited by inconsistent enforcement across states and regions. Greater attention to and support for cessation of smokeless tobacco is also needed. Low-resource settings require evidence-based cessation programs that can be delivered, sustained, and scaled-up at low cost. However, there are important barriers to scaling up evidence-based tobacco cessation in such settings. First, research syntheses, such as systematic reviews and meta-analyses mainly come from high-income settings and the interventions and strategies are difficult to implement in limited-resource settings such as India. Second, tobacco cessation EBPs typically include pharmacotherapy, a resource beyond the reach of most tobacco users and many healthcare systems in India. Also, most tobacco cessation EBPs focus solely on smoking, which is a mismatch for India given the high rates of smokeless, smoked, and dual tobacco use and the fact that the smokeless tobacco market in India is the world's largest. Third, EBPs are more likely to be scaled up in low- and middle-income countries (LMICs) if they are simple and inexpensive and can be used in alternative channels, such as NGOs or commercial operations, rather than relying entirely on government infrastructure and highly skilled personnel. Fourth, there is a gap in the literature regarding interventions to influence practice change among practitioners in LMICs. Together these gaps highlight a need to advance the science of adapting and implementing low-cost, scalable cessation EBPs for use in India and other LMICs. One type of low-cost, scalable cessation support is brief advice interventions, or short engagement by providers to screen for tobacco use, offer cessation advice, and refer users to treatment resources. These interventions can typically be administered in 30 seconds to 2 minutes. Brief advice interventions are effective for both smoked and smokeless tobacco. Simple forms of these programs have been shown to have an impact of 1-3% increased cessation rates over no advice, which could translate to 7.8-23.4 million fewer tobacco users in India. Programs that have a higher level of counseling (as in this proposal) are expected to have greater impact. In a study of 19 tobacco cessation centers across India, tobacco users who received behavioral counseling demonstrated 12.5% cessation rates at six weeks post-intervention, only 5% less than those who received counseling and medication. If this level of impact was sustainable over the long-term, it could translate into 33 million fewer tobacco users. According to the 2016-2017 Global Adults Tobacco Survey (GATS) in India, only 48% of smokers and 31% of smokeless tobacco users were advised to quit by a provider, highlighting a potential role for brief advice interventions. The use of diverse health professionals (rather than only highly credentialed individuals) is an important opportunity for brief advice interventions. India meets only 25% of the WHO benchmark for health care staffing, and available educational facilities are insufficient to grow the workforce to needed levels. To reach needed staffing levels, the number of physicians would need to increase by 151%, but the current annual growth rate is 10-14%. Task-shifting is a strategy that emphasizes moving tasks from highly credentialed workers to less credentialed (but more readily available) workers and can promote access to preventive services and health. The task-shifting model has been effectively used to deliver mental health EBPs in LMICs, showing that providers with little or no prior mental health training, and often little formal education, can deliver EBPs effectively with positive clinical outcomes. In a recent report, the Ministry of Family Health and Welfare, Government of India, recommended a task shifting approach suggesting that cessation programs be delivered by a range of health professionals, including health workers, community outreach workers, and other professionals. An important consideration for task-shifting models is the potential changes to practitioner training. For example, an EBP for autism spectrum disorder was adapted from the United Kingdom to sites in India and Pakistan. The use of non-specialists to deliver the program necessitated a change from a 2-day training for therapists to a 4-week training for non-specialists and new referral patterns and measures for non-specialist competencies. Practitioner training is a critical driver of program success, and capacity-building programs can increase practitioners' knowledge, skills, motivation, and ability to implement EBPs. Successful training requires achieving sufficient dose and offering supports, connections, and resources, a challenge in low-resource settings. Communication technologies, such as mobile phones, offer a tremendous opportunity and have been used with frontline health workers to increase knowledge and skills; deliver learning materials and reminders; improve guideline adherence; increase patient follow-up rates; increase worker credibility in the community, and support knowledge generation. Many studies with health workers focus on one-way communication, such as text messages, or two-way communication, such as information menus and tailored responses, but fewer have relied on social media as this team proposes to do. The team draws on the communities of practice model, in which networked peers share problem-solving skills, spread new information and innovations, and model exemplar practices. Given financial and logistical barriers of in-person participation, the team will turn to social media-based peer networks, which offer asynchronous connections, reach, social capital, and social support. Leveraging connected health technologies to support practitioner networks adds an innovative dimension to this brief advice EBP. There are about 478 million smartphone users in India, a number that is expected to increase rapidly as penetration increases. The team will use social media and an app to promote networking, support, and training for practitioners receiving the intervention. Smartphones allow for low cost, flexible communication in low-resource settings and opportunities to access support regardless of time or location. Proposed Study: The proposed study will assess the effectiveness of an adapted brief advice program, with a secondary goal of collecting pilot data regarding program implementation, a hybrid of effectiveness and implementation research. The team will train practitioners from three low-resource/high-reach settings to deliver brief advice on tobacco cessation to their patients. The following will be assessed: (a) tobacco control outcomes among patients; (b) implementation outcomes among practitioners; and (c) the extent to which connected health technologies may support capacity-building and impact practice change. Mumbai is an optimal study site for two reasons. First, the population is highly diverse, in terms of socioeconomic status (SES), region/state of origin across India, language(s) spoken, religion, caste, and other cultural groupings making the findings more generalizable. Second, the team has an outstanding local partner (NSF-Mumbai) that has been leading cancer prevention and control efforts for 15 years and tobacco cessation for five years. The NSF team will provide the local support and expertise to increase the likelihood of immediate and long-term implementation. The proposed study offers four major innovations. First, it will be conducted in diverse, low-resource settings with high reach. With scalability in mind, the team will use task shifting and technology-enabled training and support to make the intervention appropriate for low-resource settings. Second, the proposal brings together inter-disciplinary expertise: behavioral science, technology-driven healthcare, communication and implementation sciences, and tobacco cessation (both from research and practice perspectives). The multi-disciplinary expertise and early emphasis on implementation science through the hybrid design to study program effectiveness and implementation increase the likelihood of developing an intervention that can be adopted, adapted, and sustained. Third, findings will build the literature for translating evidence-based programs from high-income countries to LMICs and from high- to low-resource settings. Fourth, the proposal strategically combines multiple implementation strategies (task-shifting, capacity-building, technology-based support) to increase implementation success. The framework used here draws on behavioral and implementation sciences. The brief advice components come from the US Public Health Service, using the 5A's (Ask, Advise, Assess, Assist, and Arrange) strategy to promote tobacco cessation, which is endorsed by the WHO. For implementation science, the team draws on the EPIS model and its extension, the Dynamic Adaptation Process. Building on previous work, this proposal focuses on the Preparation/Adaptation and Implementation phases, while maintaining a focus on Sustainment, to ensure that a design for scale. The team has also included the Stages of Implementation Completion measure to assess progress towards integrating the EBP. The team incorporated the UK's Medical Research Council guidance on the development of complex interventions and Patel's processes to support the translation of evidence from high-income countries to LMICs. Finally, the team will draw on the literature related to capacity-building for EBPs to emphasize the role of peer networks and engaged learning approaches. The framework drives both the content of this research as well as the methods used. Aim 1 Methods: Aim 1 involves adapting and pilot-testing a brief advice program for use in three low-resource settings with high potential for scale-up: NGO-run health centers, TB clinics, and dental clinics. Activities include formative research (Phase 1), program adaptation (Phase 2), and pilot-testing (Phase 3). For Phase 1 of Aim 1, key informant interviews and focus groups will be conducted to capture the drivers influencing EBP implementation and patient-level outcomes, guided by the theoretical framework. The results will be complementary, with the interviews providing critical context regarding the intervention and training opportunities and the focus groups offering detail about practice settings and tobacco users served. The team will conduct interviews with leaders in each of the practice settings, policymakers, and patients to gain insight into the views of diverse stakeholders. These interviews will complement focus group discussions with providers. Verbal consent will be obtained from all participants as part of the intake process. The team will recruit approximately 30 individuals from Greater Mumbai to provide a high-level view of program delivery context. The team will recruit three leaders from each of the three practice setting types (a total of 9), three leaders from the policy sector, and three leaders with expertise in delivering health programs to the underserved in Mumbai. The team will also recruit five tobacco users per practice setting (for a total of 15). Purposive sampling will be used to maximize diversity among respondents regarding populations served/represented. The theoretical framework prompts attention to four areas of interest: 1) system-level factors (e.g., the policy environment related to tobacco control); 2) organizational-level factors (e.g., leadership buy-in, culture and climate, extent to which technology is utilized to support practice); 3) practitioner-level factors (e.g., experience with / attitudes towards EBPs and preventive services such as tobacco use and the use of different types of technology to support practice); and 4) patient-level factors (e.g., social context, culture, triggers for tobacco use and cessation). The team will adjust the questions based on the type of key informant. For example, with patients, the team will focus on the context in which they approach their health and tobacco control as well as how they engage with the providers, healthcare institutions, and relevant systems. The interviews will be conducted by an experienced qualitative researcher overseen by the NSF-Mumbai team. Interviews will be conducted in English, Hindi, or Marathi, depending on participant preference and will take about one hour. The team will conduct interviews until reach theoretical saturation is reached, the point at which no new themes arise. Prior experience and the literature suggest that 30 interviews will provide sufficiently rich data. Interviews will be audio-taped and translated/transcribed by an independent, professional transcriptionist. With an orientation toward framework analysis, comprehensive analysis will focus on systematic indexing and charting, and draw comparisons both within and then across interviews, as well as between sectors to understand better the multi-level context in which tobacco control EBPs will be deployed. Each stage of coding and analysis will be discussed and verified by the research team. These methods will be enhanced by the use of NVivo 11, a state-of-the-art ethnographic data management software program. The team will conduct member-checks with a subset of interviewees and refine data interpretation based on feedback. The team will conduct focus group discussions with practitioners, a technique that brings 6-10 individuals together for a semi-structured, interactive discussion that yields rich insight around shared social meanings and norms. Verbal consent will be obtained from all participants as part of the intake process. The team will conduct two discussions per setting type, for a total of six discussions including approximately 48 individuals. The team may pivot to key informant interviews with practitioners, given the status of the COVID-19 pandemic in India and the challenges of conducting virtual focus group discussions. The team will use a semi-structured guide that draws on the theoretical framework to explore the four areas described in the key informant interviews (system-, organizational-, practitioner-, and patient-level influences). The team will extend the organizational-level focus to explore practitioners' perceptions of organizational characteristics (including, but not limited to size and structure); leadership (e.g., potential champions); and implementer characteristics (e.g., values and goals, social networks, and perceived need for change). The team will also conduct a detailed workflow assessment to ensure that the adapted EBP can be integrated into practice. Finally, the team will also ask questions about their use of technology and how they engage with computers, smartphones, etc. in their professional context. An experienced moderator (overseen by the NSF-Mumbai team) will moderate the focus group discussions, which will be held in easily accessible settings and will take approximately 60-90 minutes. The team will conduct focus groups until reach theoretical saturation is reached, the point at which no new themes arise. Prior experience and the literature suggest that six focus groups will provide sufficiently rich data. Interviews will be audio-taped and translated/transcribed by an independent, professional transcriptionist. Once more, a framework analysis approach will be used and the team will extract themes within and across focus groups, exploring in detail the areas of overlap and divergence between practice setting types. The formative work will yield critical information about the implementation context for the intervention, which will drive adaptations to increase impact and scalability. Phase 2 of Aim 1 involves EBP adaptation. The team will apply formative research results to adapt the brief advice intervention for the three target settings in Greater Mumbai. EBP adaptations for global health applications are often made for cultural appropriateness and to match delivery context and settings. While cultural adaptations will be critical here, a major component of the adaptations relates to task-shifting to allow for resource- and system-constraints. There are two applications of task shifting. First, instead of physicians and nurses, the team will train outreach workers, TB treatment providers, and private dentists. Each of these practitioners will deliver a brief intervention (about 30 seconds to 2 minutes) for their patients. Second, most brief advice EBPs prompt practitioners to offer counseling and referral to counseling services. At this time in India, referral services are mainly only accessible to high SES individuals. Thus, a referral system is being created, using counselors overseen by the NSF-Mumbai team. These counselors will administer the two scheduled follow-up sessions, delivered by phone to reduce costs. The counseling will follow an evidence-based 5A's program and emphasize motivation and encouragement to quit, followed by a focus on relapse prevention, e.g., coping with external triggers that prompt tobacco use, and support with planning another quit attempt should relapse occur. A referral from a trusted practitioner to this service is expected to make counseling with a new provider acceptable to patients. The team will recruit counselors with modest qualifications and salaries to support the goal of creating a low-cost, scalable intervention. The theoretical framework highlights attention to: 1) training and coaching, including adaptation support and 2) peer networks. The training and coaching piece is critical given the use of task-shifting strategies. The capacity-building literature makes it clear that practitioners need ongoing support and technical assistance to gain and apply new skills. The capacity-building intervention will start with a training workshop (delivered over two half-days) to introduce practitioners to tobacco control and the brief advice EBP. The training will be jointly developed by members of the study team with rich experience in developing capacity-building training (the Dana-Farber/Harvard team) and expertise in teaching tobacco cessation counseling and brief advice interventions in India (the NSF-Mumbai team). The team will limit each training to roughly 20 practitioners per session to allow for an interactive experience. This design is consistent with a recent review of capacity-building interventions, that highlights the importance of practitioners' engagement with trainers as well as each other while building skills and learning to use EBPs. Practitioners will use a smartphone-based app to access support for intervention delivery. Equally important, the app will be linked to a social media group (powered by WhatsApp) to promote networking among practitioners as well as between practitioners and support staff. In this way, practitioners can use the app to answer questions and access technical assistance. The technology partner, Dimagi, has worked with the NSF-Mumbai team previously to build a cessation counseling app for use in high-resource hospitals. The team will leverage that knowledge to create a new app for use in low-resource settings and emphasizing ongoing training and support. The team will build in the capability to adjust the app in an ongoing manner, using rapid assessment cycles to determine what is or is not working for practitioners. The assessments will leverage the Plan-Do-Study-Act methods for assessing the situation, making changes, assessing impact, and then determining the next course of action. An important component of the technical support will include support for adaptation of the brief advice program. Allowing practice sites to make adjustments is vital for implementation success and supporting a systematic process for adaptation allows for practitioners to balance fit with fidelity. Phase 3 of Aim 1 involves pilot-testing the intervention. Pilot-testing is a critical part of the process of adapting an EBP for use in a new setting. The team will conduct one pilot session for each of the three types of practice settings (n=10 practitioners per session). The training will be delivered with breaks between sections to elicit feedback from practitioners. Detailed feedback will also be solicited (through group discussion) at the end of the pilot training. To evaluate the app, two practitioners per pilot-test sessions will be recruited to participate in usability testing. Drawing on prior experience with websites and smartphones, the team will use a combination of think-aloud methods and usability testing activities to assess the functionality and acceptability of the app. Aim 2 Methods: Aim 2 involves using a cluster randomized control trial to assess whether the adapted brief advice program results in increased cessation rates among tobacco users attending three types of low-resource healthcare settings in Mumbai, compared to usual care. For Phase 1 of Aim 2, the team will conduct a two-arm RCT comparing: Arm 1 - the adapted brief advice intervention (developed in Aim 1) with Arm 2 - distribution of educational pamphlets. The unit of randomization will be the practice site to avoid potential contamination. Practitioners in the intervention arm will receive training and support for the adapted brief advice intervention, including screening, offering brief advice, and referral to screening as appropriate. Practitioners in the control arm will initially receive copies of a tobacco cessation pamphlet to distribute to patients, but using a delayed control design, they will be offered the brief advice training and intervention supports after the experimental data have been collected. Arm 2 will use a high-quality, evidence-based tobacco cessation pamphlet created by the NSF-Mumbai team for low SES audiences in Mumbai. Given that practitioners and staff at targeted practice sites are over-burdened, a research assistant will attend the practice on a randomly selected day and attempt to enroll every nth patient based on the appointment schedule. The number for n will be determined on-site based on the number of patients who present and the time it takes to enroll a patient. The process will continue for a given site until the enrollment quota has been reached. Practitioners will be aware of assigned study arms but not of study hypotheses or outcomes. Signed consent will be obtained from all patients as part of the intake process conducted by the practitioner; additional details are in the Human Subjects Forms. Stratified random assignment among practice sites by the three setting types will be used. Within each setting, the team will randomize individual eligible practice sites using random number generators in SAS. The staff member who enrolls and provides practice sites their assignment will not know the treatment allocation status. The team will use standard items from the Global Adult Tobacco Surveys (GATS) Group to assess tobacco use status (smokeless and smoked), intention to quit, quit attempts in the last 12 months, and receipt of cessation advice from a provider. The team will also collect sociodemographic information using standard items created for India by the Demographic and Health Surveys program. For Phase 2 of Aim 2, after the initial results from the implementation assessments are analyzed, the team will follow up with a subset of practitioners to understand the barriers to and facilitators of program implementation. The team will select a total of nine practitioners, using a purposive sample to maximize diversity across setting type to recruit equal numbers of successful and unsuccessful implementers. A sequential, explanatory, mixed-methods design (in which the qualitative data "explain" key quantitative results) offers the opportunity to capture the "why" and "how" of challenges or successes related to program implementation. The team will interview nine practitioners (three from each type of practice setting). The framework prompts an integrated view of implementation influences as well as processes by which programs are integrated and adapted by practitioners/organizations. The team will conduct semi-structured, in-depth interviews to characterize the policy and funding environments, the impact of practice settings on implementation, the ways in which the program did or did not fit client needs, program adaptations, and practitioners' experiences with training. The interviews will be taped and translated/transcribed. Two researchers will lead data analysis, drawing on framework analysis approaches to draw comparisons both within and then across interviews, as well as between practice setting types. Once more, the team will utilize NVivo 11 to manage data. The RCT will assess the effectiveness of the adapted brief advice intervention that relies on task shifting to account for resource constraints in Mumbai. The implementation data will provide important data regarding the scalability of the intervention as well as support for systematic adaptations. Aim 3 Methods: Aim 3 involves evaluating the use of communication technologies, such as social media and apps, to support ongoing training and networking among practitioners in the intervention arm. Aim 3 examines the opportunity to use communication technologies, including apps and social media, to overcome key challenges faced in capacity building, including sustaining engagement, increasing intervention dose, and building/expanding social capital among participants. This aim will focus on practitioners in the intervention arm and assess: 1) practitioners' views on and engagement with technology-based supports for capacity-building, 2) the relationship between technology-based engagement and program implementation. Technology-based supports for practitioner engagement: The team will use an app and a popular social media channel, WhatsApp, to facilitate ongoing engagement and learning related to the adapted brief advice intervention. The app will offer: 1) a repository of training materials and additional resources, e.g., support material for promoting cessation, 2) models of delivering brief advice, 3) periodic text-based reminders regarding app content and recent activity, and 4) a forum to share practitioners' innovations and successes. The team will assess engagement and make necessary adjustments using rapid cycle assessments, once more using the PDSA cycles. Dimagi, the technology partners, have key expertise relevant to this aim as they build mobile platforms to support frontline health workers to deliver care in India and other LMICs. The company's rich (and well-studied) platform, CommCare, will be the basis for the app, which will engage and support practitioners to deliver the intervention. The team will also create a WhatsApp group for practitioners. This free, social media platform reported over 200 million users in India as of early 2017 and has over 1 billion users in 180 countries. The platform offers end-to-end encryption to ensure privacy and supports the sharing of text, photos, videos, documents, as well as voice calls. The team will collect data regarding user engagement and perceptions of the technology-based supports, which will be integrated with implementation data from Aim 2. First, drawing on the user engagement evaluation literature, the team will capture data from the app to assess patterns of navigation; depth, duration, and frequency of engagement; the contribution of user-generated content; and attention and interest. These markers can be captured using existing routines in the app without adding burden to respondents. The team will run regular checks to ensure that data are being collected as planned and that users are not experiencing software-related difficulties. User engagement data from the WhatsApp group will be collected to assess social-media based interactions. The team will capture the individual conversations as well as social media engagement markers, including content creation, replies, and shares for the social media components. Second, the team will capture pilot implementation outcomes related to the technology-based engagement supports. These items will be included in the practitioner assessments described in Aim 2 (delivered at 1-, 3- and 6-month follow-up). The team will examine implementation outcomes measures focused on technology-based supports for practitioners (the app and WhatsApp groups), including acceptability, appropriateness, feasibility, and potential for sustainability, using recently developed short-form scales. The team will conduct preliminary assessments to ensure these measures are appropriate as part of the pilot-test in Aim 1. Third, the content of the WhatsApp conversations will be captured. This will be collected from the WhatsApp group in an unobtrusive manner, without adding burden to respondents. Data Analysis Plan: In terms of analysis, the team will examine: 1) practitioner engagement with technology-based supports for capacity-building, 2) the acceptability, appropriateness, and feasibility of technology-based supports for capacity-building, and 3) the relationship between technology-based engagement and program implementation. First, the team will examine practitioner engagement with the app and social media platform and create a typology of users. This analysis will include a quantitative assessment of usage patterns as well as an exploratory content analysis of the WhatsApp data, drawing on methods from the study team's previous studies of social media data use among community organizations. Second, the team will characterize the acceptability, appropriateness, feasibility, and potential for sustainability of the technology-based supports for capacity-building. By separating out the technology component, the utility of this theoretically low-cost, scalable implementation strategy can be assessed. Third, the team will conduct an exploratory assessment of the relationship between app engagement and program delivery at the 1-, 3-, and 6- month post-training periods. The team will use Spearman (rho) correlations to examine the association between the user engagement metrics described above with the data describing practice change (Stage of Implementation Completion). With 66 practitioners in the intervention arm, the data will be preliminary but will offer useful insight regarding the link between technology-based engagement and program implementation. This aim will offer insight into the ways in which technology-enabled engagement can support training and capacity-building goals for practitioners working in low-resource settings.