Background
Despite long-standing tobacco control efforts, a persistent socioeconomic gradient in smoking prevalence exists in a number of western countries [
1‐
3]. In Australia, markedly higher smoking rates are found among highly disadvantaged groups including Aboriginal and Torres Strait Islander people (38%), individuals who are unemployed or with a low income (25-30%) and people with a mental illness (36%) compared with the general population (15.1%) [
4‐
6]. There is also evidence of a prevalence gradient within disadvantaged groups. For example, Aboriginal and Torres Strait Islanders typically experience several types of disadvantage, with research showing increasing smoking prevalence as a function of decreasing household income [
7]. Effective tobacco control initiatives which address smoking in all socioeconomic groups are critical to addressing this social disparity.
There is growing debate about how to best address disparities in tobacco use and subsequent health outcomes [
8‐
10]. Behavioural interventions have shown promise amongst some disadvantaged groups although the evidence is mixed [
11]. There remains a lack of robust evidence about the effectiveness of a number of tobacco control strategies for disadvantaged groups [
11,
12]. Evidence on the relative effectiveness of major strategies such as mass media and telephone support does not address the full range of disadvantaged groups, so is insufficient for decision making and policy development. For example, studies of the effectiveness of anti-tobacco mass media campaigns across socioeconomic groups commonly omit the highly disadvantaged [
12]. Inter-group comparisons of the relative effectiveness of population-based strategies have focussed on gender, age and some racial groups rather than Indigenous status, poverty or mental health [
13].
The lack of methodologically rigorous evidence about the most effective tobacco control strategies for redressing the socioeconomic gradient in smoking rates can impede decision making, funding and agenda setting. Consequently, decisions can be easily influenced by individual opinion. In the absence of robust trial data, representative data on the views of relevant stakeholders can be helpful in guiding decision making, setting a research agenda or simply understanding where resources are likely to be channelled in the absence of evidence. While little is known about the views of experts, views about effective tobacco control strategies have been sought from disadvantaged smokers. For example, remote indigenous community members indicated brief advice and pharmaceutical quitting aids were perceived as important and effective, as was introduction of smoke free areas [
14]. In contrast, there were conflicting views on Quit programs and tobacco taxation increases. Qualitative and quantitative studies of disadvantaged groups attending social and community service organisations have identified financial or material incentives and subsidised pharmacotherapies as popular choices for cessation support [
15,
16]. Understanding the views of the tobacco control community is important not only to assist with decision making, but to understand the perceptions guiding current decision making, and identifying targets for strategic research funding.
This study aimed to explore the views of a sample of Australian and New Zealand tobacco control advocates, researchers and workers regarding:
1.
Perceptions of the effectiveness of i) Population-level strategies such as mass media campaigns, taxation increases, limits on expenditure of government payments; and ii) Cessation support such as telephone or SMS cessation support, subsidised pharmacotherapies and web-based cessation programs; for the general population and for each of three socially disadvantaged groups: Aboriginal and Torres Strait Islander people, those with a low income and people with a mental illness.
2.
Additional strategies perceived to be effective for these disadvantaged groups.
Discussion
This survey of key tobacco control stakeholders found high levels of perceived effectiveness of population-based strategies including mass media campaigns and increased taxes for the general population. Fewer respondents were convinced that these strategies would be effective for disadvantaged sub populations with high smoking rates. In general, the number of years working in tobacco control did not affect perceptions of strategy effectiveness. The one exception to this was that those with more experience were more likely to rate increased taxes on tobacco as an effective strategy for Aboriginal and Torres Strait Islander peoples and people with a mental illness, compared to those who had not been working in tobacco control for as long.
Approximately half of the sample considered mass media to be effective for Aboriginal and Torres Strait Islander peoples and people with a mental illness, significantly lower than the 84% perceiving an effect for the general population. It is not possible to judge whether respondents believed this was due to the medium (e.g. television) or the message (e.g. lack of culturally relevant actors or issues). However, the data suggest that one of Australia’s largest tobacco investments is not therefore, considered to be an avenue for reducing the social disparity in smoking rates. While there are data to suggest that mass media can be effective for lower income groups [
17], the issue of differential effectiveness of mass media for very disadvantaged groups has not been adequately addressed in the literature [
12]. Recent Australian campaigns aimed at delivering culturally-appropriate mass media tobacco control messages for Aboriginal people require sound evaluation and dissemination of data to ensure the tobacco control community is aware of developments in this area [
18,
19].
Similarly, widely available cessation strategies such as telephone, SMS, and web-based programs were perceived by a much lower proportion of respondents to be effective for disadvantaged groups (16%-52%) than for the general population (35%-61%), most notably for Aboriginal and Torres Strait Islander people. A survey of low-income clients accessing social and community support organisations, also found low levels of support for receiving smoking cessation advice via Quitlines or SMS services [
16]. Given that telephone, SMS and web-based support are the most promoted and available forms of cessation support in Australia, there is a need for robust research demonstrating whether very low levels of relative effectiveness are the case for such groups compared to the general population. If this is the case, it is necessary to develop effective strategies for improving the reach, efficacy and effectiveness of cessation support strategies for groups with particularly high smoking rates. One Australian trial of the telephone support service (Quitline) with a disadvantaged sample found no evidence of effectiveness at 12 months follow-up [
20]. More recent evidence also suggests lower levels of engagement and less effectiveness for Aboriginal and Torres Strait Islander people accessing the Quitline [
21], a finding which is echoed by a survey of Indigenous community members and health workers, suggesting unmodified Quit programs may lack appropriateness in such settings [
14].
In contrast, ensuring government payments are not spent on tobacco was perceived by more respondents to be effective for each disadvantaged group than for the general population. However, the levels of endorsement were low (20%-35%), suggesting this was not an approach that would be broadly supported by the tobacco control community. Subsidised medication was the sole instance where the proportion who perceived this strategy to be effective was high (72-84% of respondents) for all population groups, with higher proportions of endorsement for some disadvantaged groups than for the general population. This suggests the respondents perceive addiction or finances (or both) to be of major importance in reducing smoking in disadvantaged groups. Given the high cost of this approach and concern about the real-world effectiveness of pharmacotherapies [
22,
23], careful testing of the relative cost-effectiveness of medications for disadvantaged versus general population groups is needed.
Also of interest are the additional strategies nominated by respondents as potentially effective as an indication of what key stakeholders may endorse or advocate. Increased restrictions on smoking in public areas were nominated as likely to be effective for all groups other than for Aboriginal and Torres Strait Islander people, while tailored quit programs were suggested for each disadvantaged group. Tailoring of cessation programs to disadvantaged populations has intuitive value, particularly for Indigenous groups given the unique social and cultural factors which influence smoking behaviour in this population [
24]. Evidence regarding tailored cessation programs in disadvantaged populations is limited, and has been the subject of some debate particularly in relation to tailoring for Indigenous populations. One review advocated for the development of tailored and targeted approaches to smoking cessation [
25], while another review argued that not all smoking cessation programs need to be culturally adapted to be effective [
26]. However, both reviews only included a small number of studies, suggesting further research is necessary to fully inform the tobacco control community on the value of tailored programs. In relation to people with a low income or mental illness, programs delivered by health care providers and service providers were nominated. While there is a growing evidence base in this field [
27‐
31], this work is in its infancy. The dissemination of effective approaches should also be studied, given the identified challenges in achieving high rates of cessation advice in primary care [
32]. Strategies nominated as likely to be effective for Aboriginal and Torres Strait Islander people suggest community-based, culturally relevant approaches that involve elders and other community leaders are preferred. Given the weakness of the evidence-base around culturally-relevant programs [
33], it is important to establish robust science around the implementation of cultural relevance to ensure tobacco control programs for Aboriginal and Torres Strait Islander people provide the greatest possible benefit.
When considering the evidence for the effectiveness of strategies for Aboriginal and Torres Strait Islander people, it is important to consider the implications of this for the Indigenous population of New Zealand. The Maori population makes up a larger proportion of the New Zealand population (approximately 15%) [
34] compared to Aboriginal and Torres Strait Islanders, who make up approximately 2.5% of the Australian population [
35]. However, both Indigenous populations experience many health disparities compared to the non-Indigenous population including a significantly lower life expectancy, and a higher smoking prevalence [
35,
36]. Therefore it remains important to continue research into effective strategies for promoting smoking abstinence in these populations, and where possible, identify approaches that are most effective and acceptable for each group.
The interpretation of these findings should take into account some limitations. Firstly, although the response rate is acceptable, it does not preclude response bias. It is possible that those who did not respond may have expressed different views to those of the respondents. Those working with disadvantaged groups may have been more likely to respond than those without such involvement, and so may have stronger views about the needs of such groups in relation to tobacco control strategies. Furthermore, it is acknowledged that this survey was conducted at a time in which the Indigenous tobacco control workforce was undergoing substantial change and growth. As such, some of these workers may not have been on the contact list, and therefore would not have been invited to complete the survey.
It is also likely that the sample was heterogeneous in terms of familiarity with data about the effectiveness of the various tobacco control strategies. While the heterogeneity of the sample may be considered a limitation, it may also be considered a study strength. It should also be noted that no explanation was provided regarding definitions of the sub populations. This may be a concern in relation to interpretation of the term ‘low income’ which can be interpreted as less than average income or below the poverty line which are quite different groups, the latter of which is not well-reflected in studies of the effectiveness of increased taxes on smoking cessation [
37]. These groups also have significantly different smoking prevalence rates- 24.6% in the most disadvantaged quintile [
4], compared to 61% among individuals with multiple forms of disadvantage accessing social and community service organisations [
16]. As the survey did not provide respondents with a definition of low income, it may be that their responses reflect wide variations in definitions of low income groups.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
All authors conceptualised the study and contributed to the design of the project. CP and HT were involved in data collection, HT and PM were involved in data analysis and all authors contributed to data interpretation and drafting and revising the manuscript. All authors read and approved the final manuscript.