Article Text

Download PDFPDF

Cost-effectiveness of a smokeless tobacco control mass media campaign in India
  1. Nandita Murukutla1,
  2. Hongjin Yan1,
  3. Shuo Wang1,
  4. Nalin Singh Negi1,
  5. Alexey Kotov1,
  6. Sandra Mullin1,
  7. Mark Goodchild2
  1. 1 Vital Strategies, New York, USA
  2. 2 World Health Organization, Geneva, Switzerland
  1. Correspondence to Dr Nandita Murukutla, Vital strategies, 61 Broadway, Suite 2800, New York, NY 10006, USA; nmurukutla{at}vitalstrategies.org

Abstract

Background Tobacco control mass media campaigns are cost-effective in reducing tobacco consumption in high-income countries, but similar evidence from low-income countries is limited. An evaluation of a 2009 smokeless tobacco control mass media campaign in India provided an opportunity to test its cost-effectiveness.

Methods Campaign evaluation data from a nationally representative household survey of 2898 smokeless tobacco users were compared with campaign costs in a standard cost-effectiveness methodology. Costs and effects of the Surgeon campaign were compared with the status quo to calculate the cost per campaign-attributable benefit, including quit attempts, permanent quits and tobacco-related deaths averted. Sensitivity analyses at varied CIs and tobacco-related mortality risk were conducted.

Results The Surgeon campaign was found to be highly cost-effective. It successfully generated 17 259 148 additional quit attempts, 431 479 permanent quits and 120 814 deaths averted. The cost per benefit was US$0.06 per quit attempt, US$2.6 per permanent quit and US$9.2 per death averted. The campaign continued to be cost-effective in sensitivity analyses.

Conclusion This study suggests that tobacco control mass media campaigns can be cost-effective and economically justified in low-income and middle-income countries. It holds significant policy implications, calling for sustained investment in evidence-based mass media campaigns as part of a comprehensive tobacco control strategy.

  • Media
  • Economics
  • Prevention
  • Low/Middle income country
  • Advocacy

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

Tobacco use, a serious global public health priority,1 is of particular concern in populous low-income and middle-income countries (LMICs) like India, where it leads to high morbidity, mortality and health costs.1–4 Efforts to reduce tobacco consumption have thus been identified as a key ‘best buy’ intervention, particularly to avert the rising tide and the economic costs of tobacco-related non-communicable diseases in LMICs.5

A large body of evidence, from high-income countries (HICs) and LMICs, has demonstrated that tobacco control mass media campaigns can be effective in reducing tobacco prevalence and in preventing tobacco uptake.6–10 Hence, the WHO’s treaty, the Framework Convention for Tobacco Control,11 and its MPOWER package of effective interventions,12 obligates all 180 countries that are parties to the treaty to implement ‘public education’ campaigns, including via mass media, as part of a comprehensive strategy to reduce tobacco consumption.

The cost-effectiveness of interventions has increasingly become a prime consideration for public health programming, particularly in resource-constrained LMICs. Studies in HICs have found that — despite their significant upfront costs — tobacco control mass media campaigns are cost-effective and offer good value for money by increasing rates of quitting and reducing health burden.6 However, to date, there has been no published evidence of their cost-effectiveness in LMICs, where resources are more limited. Hence, this paper seeks to address this gap in knowledge by assessing the cost-effectiveness of a recent tobacco control mass media campaign in India.

‘Surgeon’ smokeless tobacco mass media campaign

Smokeless tobacco is more frequently consumed in India than smoked tobacco, with 33% of men and 18% of women chewing tobacco.13 It has been associated with a higher risk of mouth cancer, pharyngeal cancer, oesophageal and ischaemic heart disease, and is the likely reason for India’s high rates of oral cancer among young adults.14

To address the smokeless tobacco burden, the Government of India with the technical assistance of the public health non-government organisation World Lung Foundation (WLF)i embarked in 2009 on what was, to our knowledge, the first national mass media campaign on the harmful effects of smokeless tobacco use with the intent to improve knowledge of smokeless tobacco’s health harms, denormalise its use and prompt cessation among current smokeless tobacco users. The campaign consisted of a pretested 30 s public service announcement with hard-hitting and emotionally moving personal testimonials of patients suffering from smokeless tobacco-related cancers. It aired on national television and radio in November and December 2009. The TV component of the campaign aired predominantly during ‘primetime’ hours (81% of the time); a third of the campaign aired on state-owned national and regional TV channels that were expected to reach the target audience, including in rural areas, and the rest was allocated to privately owned, national cable and satellite channels. The campaign was supplemented by 11 460 planned radio spot exposures for the duration of the campaign (further details available in the previous publication on this campaign).9

The results of the Surgeon campaign evaluation were published in Tobacco Control in 2012.9 The evaluation compared those who were ‘campaign aware’ (the intervention group) against those who were ‘campaign unaware’ (the status quo comparison group that represented the scenario that may exist in the absence of or with limited receipt of the intervention) on key outcome indicators. Campaign awareness was measured through a series of questions that assessed accurate campaign recognition and confirmed comprehension of its key messages. Given the diffusive nature of mass communication,15 it must be noted that the ‘campaign unaware’ group may in fact have been exposed to the campaign — although unrecalled — and therefore provided a tougher benchmark against which to detect statistically significant improvements in campaign outcomes. Outcomes were measured through a battery of true/false knowledge questions and agreement with attitudinal statements. Cessation attempts were measured as self-reported quit attempts in the 2 months prior to the evaluation (the campaign period).

The Surgeon campaign was found to be effective. Campaign-aware smokeless tobacco users were significantly more likely than campaign-unaware users to show increased knowledge and changed attitudes consistent with the campaign’s messages. Most pertinently, there was an increase in quit attempts during the campaign period.9 These findings of the campaign’s effectiveness were consistent with a significant body of literature primarily from HICs that has found that hard-hitting, graphic campaigns that arouse negative emotions like fear and concern tend to be highly impactful.7

However, the question of the campaign’s cost-effectiveness remained. Hence, this paper retrospectively analyses the investment in the Surgeon campaign against its outcomes. The economic justification for the campaign is determined through the calculation and analysis of the campaign-attributable: (1) number of quit attempts and permanent quitters 1 year later, (2) costs per quit attempt and per permanent quitter, (3) the expected number of tobacco-attributable deaths averted in the future and (4) comparisons to other such interventions where available.

Study data and method

Drawing data from the Surgeon campaign evaluation, a standard cost-effectiveness methodology was used in which the costs and effects of the Surgeon campaign were compared with the status quo (‘campaign unaware’ group or the scenario in the absence of the campaign). The analysis was designed to be rigorous: wherever a range of input parameter values were available, the more conservative value was always chosen to provide a harder test of the Surgeon campaign’s cost-effectiveness. Details of the various input parameters and accompanying calculations are provided in table 1 and below.

Table 1

Parameter of campaign and status quo, and outcome and cost-effectiveness of the campaign

Campaign cost

The total campaign cost was approximately US$1 111 500 (details in table 2). The costs included research in the form of message testing to support the development of the campaign and the impact evaluation; media production costs; media air time for the campaign’s delivery, including on TV and radio; press launches and public relations; and WLF staff and administrative time. Staff and administrative time spent by Government of India (GOI) was unavailable and could not be input into the model.

Table 2

Cost of the campaign

Campaign-attributable net additional quits

A significant increase in the percentage of quit attempts due to the campaign had been found previously9: 47.5% of campaign-aware versus 34.2% in the status quo were found to have attempted to quit, representing a 13.3% net additional quit attempt rate due to the campaign. This 13.3% net additional quit attempt rate was then applied to the population of adult smokeless tobacco users in India with access to TV or radio (the campaign’s mass media channels and therefore the population potentially reached by the campaign) to calculate the number who had attempted to quit and the number who stayed quit a year later. The population of adult smokeless tobacco users in India with access to TV or radio was derived from the 2009 Global Adult Tobacco Survey13 and the Indian Readership Survey.16 The total number of adult smokeless tobacco users in India was 205 981 000, but with a 63% mass media penetration of TV or radio, the total smokeless population with access to TV or radio was calculated as 129 768 030.

To calculate the number of net additional quit attempts due to the campaign, the 13.3% net additional quit attempt rate was multiplied with the potential campaign target audience of smokeless tobacco users (n=129 768 030). To calculate the number of permanent quits 1 year later, a methodology established in a number of published studies was followed17 18: namely, it was assumed that 2.5% of quit attempts result in permanent success, based on evidence that about 4% of smokers making an attempt abstain for at least 12 months, and that about 65% of those who achieve 12-month cessation will remain permanently abstinent. In the absence of data for smokeless tobacco users, these estimates for smokers were applied.

Disease burden averted

A ‘single cohort’ approach was used to measure the impact of the campaign on the expected number of tobacco-attributable deaths averted among the cohort of permanent quitters in this study.19 20 Under this approach, the health benefit (eg, avoidance of a tobacco-related death) was calculated over the remaining life course of those who quit.

Epidemiological studies over the past 50 years have shown that tobacco ultimately kills a third to half of all people who use it.21 22 While there is some uncertainty about the mortality risk from different tobacco products, including traditional or ‘indigenous’ products such as bidis and smokeless tobacco in India, one recent study adopted an expected mortality rate of 40% for bidi users in India.23 Given the relatively similar socioeconomic profile of bidi and smokeless tobacco users in India,13 the same 40% mortality rate was applied in our study. Additionally, it was recognised that the number of deaths averted from the campaign also depends on the benefits of cessation. It is well known that quitters can avoid many of the excess hazards of tobacco use. However, some adjustment is required to account for the reality that not all users — particularly the elderly — who quit can avoid early death from a tobacco-related disease. National-level studies including for India typically use a survival rate of around 70% for permanent quitters.20 23 This rate is applied only to quitters who would otherwise have died from a tobacco-related disease. The product of the two rates (40%×70%) was applied to the net permanent quitters to derive the number of tobacco-attributable deaths averted due to the campaign. That is, for every 100 permanent quitters, there would be 28 premature deaths averted (100 quits multiplied by the 40% mortality risk and the 70% survival rate). Additionally, reflecting the epidemiological studies described above, a range of 30%–50% mortality rates were applied for sensitivity purposes.

Cost-effectiveness analysis

Incremental cost-effectiveness ratio of the campaign was calculated as the ratio of the cost of the campaign divided by the campaign-attributable outcomes gained, including quit attempts, permanent quits 1 year later and deaths averted.

Sensitivity analysis

Sensitivity analyses were conducted and are described in table 3A and table 3B. The sensitivity analyses are based on varying two key parameters. First, upper and lower 95% CIs, which were calculated using the SD from each analysis, were used to calculate the range of additional quit attempts. Second, the mortality risk was varied by 10%, giving a range of 30%–50% around our central assumption of 40%.

Table 3A

Sensitivity analysis of outcomes of India smokeless campaign

Table 3B

Sensitivity analysis of cost-effectiveness of India smokeless campaign

Results

Quit attempts, permanent quits and disease burden averted

The quit attempt rate was 13.3% higher in the campaign scenario than in the status quo (table 1). With 129 768 030 smokeless tobacco users with TV or radio access in India, the campaign was estimated to have generated an additional 17 259 148 quit attempts and 431 479 permanent quitters after 1 year. With 431 479 more permanent quitters due to the campaign, future deaths averted were projected to be 120 814 at 40% premature mortality risk.

Cost-effectiveness

With a campaign budget of US$1 111 500, the cost per quit attempt was US$0.06 and the cost per permanent quitter was US$2.6. The cost per death averted by the campaign, at 40% premature death risk, was US$9.2.

Sensitivity analysis

Sensitivity analysis of the outcomes of the campaign using 95% CIs (table 3A) revealed that the additional campaign-attributable quit attempts ranged from 15 654 758 to 18 863 538; additional campaign-attributable permanent quits ranged from 391 369 to 471 588. Campaign-attributable future deaths averted, at varying estimates of premature mortality risk (30% and 50%), ranged from 90 611 (95% CI 82 187 to 99 034) to 151 018 (95% CI 136 979 to 165 056).

Sensitivity analysis of the cost-effectiveness of the campaign using 95% CIs (table 3B) revealed that cost per campaign-attributable quit attempt ranged from US$0.06 to US$0.07; cost per permanent quit ranged from US$2.4 to US$2.8. Cost per death averted, after including our varying estimates of mortality risk (30% and 50%), ranged from US$7.4 (95% CI 6.7 to 8.1) to US$12.3 (95% CI 11.2 to 13.5).

International dollars

Cost-effectiveness studies are often converted into international dollars in order to enable the comparison of findings between countries at different underlying levels of income and development. The International Monetary Fund Purchasing Power Parity exchange rate (PPP$) for India in 2009 was used to convert our headline findings into international dollar terms.24–26 The cost per successful quit came to PPP$9.0 international dollars, while the cost per death averted totalled PPP$32.1 international dollars.

Discussion

The Surgeon campaign was found to be highly cost-effective. With a campaign cost of US$1 111 500, the campaign successfully generated 17 259 148 additional quit attempts, with an extrapolated 431 479 permanent quits and, at 40% premature death risk, 120 814 deaths averted. The cost per benefit was US$0.06 per quit attempt, US$2.6 per permanent quit, and, at 40% premature death risk, US$9.2 per death averted. Even with sensitivity analyses that accounted for uncertainty in estimates used in the cost-effectiveness analysis, the costs per benefit accrued as a result of the campaign were significantly lower than the benchmarks traditionally used in economic evaluations. For example, the cost of US$9.2 per death averted in this study is well below India’s 2010 Gross Domestic Product (GDP)/capita of US$1400, the benchmark used by the World Health Report.27 Although disability-adjusted life year (DALYs) were not computed, the result of US$9.2 per death averted compares well with other thresholds such as <US$100 per DALY.24 28

The cost-effectiveness of the Surgeon campaign compared favourably with economic evaluations of other successful and cost-effective tobacco control mass media campaigns.6 For example, the American EX campaign cost US$37 355–US$81 301 per quality-adjusted life year (QALY) saved.29 A campaign in London targeting Turkish speakers cost US$175 (£105) per life year gained and US$1375 (£825) per permanent quit.30 A Scottish smoking cessation campaign that included mass media, quit line and an information booklet costs US$507–US$1093 (£304–£656) per life year saved.31 All of these campaigns were deemed economically beneficial.6 Thus, the cost-effectiveness of the Surgeon campaign is consistent with the wider body of evidence from HICs. In light of the relatively low investment required to accrue benefits (<$10 for a permanent quit and death averted), the Surgeon campaign provides a strong economic justification for tobacco control mass media campaigns in LMICs as an effective tobacco control strategy.

There were some limitations to this study. The mortality risks from smokeless tobacco use were not available. However, parameters available in the peer-reviewed literature for smokers — particularly for bidi users in India who are of similar socioeconomic profile as smokeless tobacco users — were used. This uncertainty was also addressed in the sensitivity analyses. Further, we acknowledge that the cost of the media campaign benefited from bulk purchasing and discounted bulk media rates obtained by the government, which means that the campaign costs were likely somewhat lower than what would have been spent at that time by a private entity without similar negotiating ability. This is, however, an unavoidable constraint with media purchasing in most countries. It is hoped that the sensitivity analysis presented in this paper may serve as a guide of the possible range of investment required to accrue similar benefits.

There were also several strengths to this study. The cost-effectiveness analysis was built on highly rigorous data: the campaign evaluation data used a thorough national household survey methodology that included representative samples of hard-to-reach smokeless tobacco users, including those in rural India and in the lowest socioeconomic groups.9 Additionally, the cost-effectiveness model used the most conservative parameters wherever required: the presumed size of the impacted smokeless tobacco user population was restricted to only those with TV or radio access, even though it is known that communication can affect the wider population, which includes those without media access, through the social diffusion of messages.32 33 Next, the literature indicated that of those (smokers) who make a quit attempt, between 2.5% and 8% manage to stay permanently quit a year later. In this study, the more conservative 2.5% rate was used to calculate permanent cessation. Finally, to estimate deaths averted, a range of tobacco-attributable premature death risks were considered.

The findings of this study hold significant policy implications for LMICs. Increased quit attempts in a population are important to achieving reductions in tobacco prevalence.34 This study demonstrates how hard-hitting mass media campaigns, like the Surgeon campaign, can be cost-effective in increasing quit attempts, thereby increasing permanent quits and averting tobacco-related deaths. These returns on the campaign investment may be particularly meaningful among tobacco users who are predominantly of lower socioeconomic status by sparing the tobacco quitters — and their families — from suffering and loss. The findings are also significant for LMICs that are typically resource-constrained. They thus provide economic justification for sustained investment in hard-hitting tobacco control mass media campaign as a ‘best buy’ approach to protecting public health.

In conclusion, this study fills an important gap in tobacco control science and practice by demonstrating the cost-effectiveness of a tobacco control mass media campaign in an LMIC setting. It suggests that just as in HICs, tobacco control mass media campaigns in LMICs offer an impactful and cost-effective strategy for achieving population-level reductions in tobacco prevalence and the resultant health savings and gains.

What this paper adds

  • Results indicate that the campaign was highly cost-effective in a resource-constrained LMIC: it successfully generated 17 259 148 additional quit attempts, with an extrapolated 431 479 permanent quits and 120 814 deaths averted (assuming a 40% mortality risk). The cost per benefit was US$0.06 per quit attempt, US$2.6 per permanent quit and US$9.2 per death averted. The campaign continued to be cost-effective in sensitivity analyses. Our study provides the economic justification for sustained investment in evidence-based mass media campaigns by demonstrating the potential for achieving population-level reductions in tobacco prevalence and the resultant health savings and gains.

  • Studies in high-income countries have found tobacco control mass media campaigns to be cost-effective, offering good value for money by increasing rates of quitting and in reducing health burden. However, to date, there has been no published evidence of their cost-effectiveness in low-income and middle-income countries, like India, where resources are limited and concerns for cost-effectiveness are significant.

Acknowledgments

The authors gratefully acknowledge the senior staff at the Ministry of Health and Family Welfare, Government of India, Dr Pankaj Chaturvedi (Head and Neck Surgeon), management of TATA Memorial Hospital, India, and Dr Vinayak Prasad (formerly at GOI and currently with the WHO) for their vision in supporting the Surgeon campaign; Bloomberg Philanthropies and the Bill and Melinda Gates Foundation, as part of the Bloomberg Initiative to Reduce Tobacco Use, for their generous support that enabled several aspects of the Surgeon campaign, including pretesting research, ad production and the impact evaluation study; the dedicated researchers and field staff at ORG Centre for Social Research who conducted the evaluation study; and Vital Strategies communications staff, particularly Irina Morzova, Steve Hamill and Rebecca Perl, for their input and review through the campaign and publication process.

References

Footnotes

  • i World Lung Foundation merged with the Union North America in 2016 and is now known as Vital Strategies.

  • Contributors NM: Study design, oversight of impact evaluation study, design and direction of data analysis, data interpretation, literature review;, writing of this paper. HY: Design of analysis, conduct of the analysis, help in writing the paper. SW: Design of analysis, conduct of the analysis, help in writing the paper. NSN: Review of the analysis, help in writing the paper. AK: Review of the analysis, help in writing the paper. SM: Review of the analysis, help in writing the paper. MG: Guidance on the development of the model, review and contributions towards the analysis, writing of this paper.

  • Funding The impact evaluation and analysis was supported by a grant from the Bloomberg Philanthropies to Vital Strategies (formerly known as World Lung Foundation). However, Bloomberg Philanthropies was not involved in any aspect of the evaluation study or the writing of this manuscript. The authors have not been paid to write this article.

  • Disclaimer The authors alone are responsible for the views expressed in this article, and they do not necessarily represent the views, decisions or polices of the institutions with which they are affiliated.

  • Competing interests None declared.

  • Ethics approval Prior to participation, the study was described to all participants and their formal consent to participate was sought. The questionnaire was administered only to respondents who agreed to participate in this research. As an ESOMAR member, ORG-Nielsen complies with the professional and ethical standards of ESOMAR International Code of Marketing and Social Research Practice.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Correction notice This article has been corrected since it was published Online First. Some of the confidence intervals in the tables were initially misreported.