NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Liu JJ, Davidson E, Bhopal RS, et al. Adapting Health Promotion Interventions to Meet the Needs of Ethnic Minority Groups: Mixed-Methods Evidence Synthesis. Southampton (UK): NIHR Journals Library; 2012 Nov. (Health Technology Assessment, No. 16.44.)

Cover of Adapting Health Promotion Interventions to Meet the Needs of Ethnic Minority Groups: Mixed-Methods Evidence Synthesis

Adapting Health Promotion Interventions to Meet the Needs of Ethnic Minority Groups: Mixed-Methods Evidence Synthesis.

Show details

7Qualitative interviews with researchers and health promoters who have adapted interventions for ethnic minority populations



Because of the space limitations of many journal publications and the tendency to report normative accounts of research studies, much of the information on the process of adapting health promotion interventions for ethnic minority populations may not be accessible through published reports. This information may, however, be more forthcoming from detailed discussions with researchers and health promoters with particular experience of working with ethnic minority populations. Qualitative research is especially useful to understand processes and experiences and to gather in-depth knowledge to get beneath public accounts. The rationale for this qualitative component was to extend our understanding of how adapted interventions are delivered and why they were successful (or not) for smoking cessation, increasing physical activity and improving healthy eating for African-, Chinese- and South Asian-origin ethnic minority populations.


To understand the processes and experiences of, as well as the rationale for, adapting health promotion interventions for ethnic minority populations and to summarise lessons learned.


We conducted semi-structured interviews with researchers and health promoters adapting health promotion interventions for smoking cessation, increasing physical activity and improving healthy eating aimed at African-, Chinese- and South Asian-origin ethnic minority groups around the world. Interviews were transcribed and two researchers independently coded the interview transcripts using NVivo 8 (QSR International, Doncaster, Australia). Thematic analysis was undertaken, this being informed by the literature reviewed in earlier phases of the work while also remaining open to more emergent issues.


Twenty-six interviews were conducted. Findings both complemented and extended the existing literature on adapting interventions for ethnic minority populations. Participants discussed how adaptations to promote healthy behaviours for ethnic minority populations could occur at every stage of the research process, from conceptualisation to dissemination, but they in particular highlighted the importance of sustainability considerations. Newer insights emerged including the need for a greater recognition of the lived dimensions of ethnicity, these including dynamic relationships with particular spaces and places, intergroup heterogeneity and previous experience with other health interventions that impacted on how new adapted interventions were received. These insights are generally unaccounted for in conventional descriptions of ethnicity. Recognition of what we have framed as the ‘contextual elements of ethnicity’ can facilitate process-oriented thinking on ethnicity and its potential impact on health behaviour. Finally, interviewees discussed the benefits of informal research networks, which constrained or enabled access to unpublished findings and ‘lessons learned’, these being readily available in the USA but less available elsewhere.


Researchers have employed a broad array of approaches in adapting health promotion interventions for ethnic minority populations, these mapping onto the Typology of Adaptation and the Programme Theory of Adapted Health Promotion Interventions we have developed. In developing and delivering these interventions, participants emphasised the need to understand and work with a range of relevant contextual considerations, and, given the complexity of this task, the need to be able to draw on the experiences of colleagues is critical. This engagement was more easily achieved in the USA, where critical mass of researchers appeared to exist, in contrast to, for example, Australia, where researchers and health promoters felt more isolated.


Within economically developed nations with large proportions of ethnic minority populations it is now well recognised that members of some ethnic groups experience elevated risk of disease in comparison to the general population and relative to other ethnic groups (see Chapter 1); this has, for example, been demonstrated for chronic diseases such as cardiovascular disease and diabetes mellitus.61,62 A preventive health approach has been adopted in the UK and increasingly elsewhere to address the constellation of underlying causes of these chronic diseases, which are, to a large extent, potentially avoidable through the elimination of a core set of risk factors such as smoking, physical inactivity and unhealthy eating.66 The preventive health approach is, as discussed at length in earlier chapters (see Chapters 1 and 46), increasingly multifaceted, with intervention approaches ranging from those that are individually centred to more population-based approaches. The bulk of the evidence on effective interventions to date, whether for the general population or for ethnic minority communities, relates predominantly to approaches aiming to change individual behaviours or risk factors, rather than changing settings or environments (see Chapters 46).

In tackling these shared risk factors, ‘ethnicity’, along with other demographic characteristics including education, income, social status and gender,31 have all been identified as important in enhancing either susceptibility or resiliency to disease.258,377 ‘Ethnicity’ was, in the framework of this project, conceptualised as encompassing the dimensions of ancestry, culture, language, physical features and religion (see Figure 1).12 These dimensions, however, offer only limited insight into how to operationalise ‘ethnicity’ in preventive health approaches, hence the need for detailed work with experienced investigators. When planning this work we anticipated that there might be little in the way of description in the published accounts of why or how interventions had been adapted, and what lessons had been learned from this process, and this was to a large extent verified from our systematic review (see Chapter 6).

Building on the systematic review (see Chapter 6), in which we identified and analysed the evidence on adapted health promotion interventions for smoking cessation, increasing physical activity and improving healthy eating for African-, Chinese- and South Asian-origin groups, we wished to better appreciate some of the experiences and lessons learned in delivering these adapted interventions. We had hypothesised that these accounts were unlikely to be documented in the very public peer-reviewed literature accounts.

Research questions

The following research questions addressed in this chapter relate to Phase 2D of the project proposal (see Chapter 2):

  • How and what factors (contextual, theoretical, practical) are considered in the adaptation of health promotion interventions for ethnic minority groups?
  • Which factors were found to increase or decrease the effectiveness of adaptations for interventions?



Ethical approval was obtained from the School of Health in Social Science Research Ethics Committee at the University of Edinburgh (see Appendix 4).


We employed purposive, maximum diversity sampling aiming to recruit 20–30 participants; based on our previous experience with undertaking a number of related qualitative studies, we anticipated that these numbers of interviews would allow us to achieve saturation on the issues at the heart of our enquiry. We primarily wished to sample researchers delivering and evaluating innovative interventions for smoking cessation, physical activity and healthy eating for African-, Chinese- and South Asian-origin groups. We were, in addition, also interested in hearing the perspectives of individuals involved in delivering the intervention, such as counsellors and health educators/promoters. To ensure that we had a maximum variation sample we devised a three-by-three recruitment approach (Figure 11), which was used to inform sampling and recruitment decisions. Participants with different roles in health promotion interventions and programmes were approached for interviews with the explicit aim of capturing a range of opinions and perspectives.

FIGURE 11. Sampling matrix for qualitative interviews.


Sampling matrix for qualitative interviews.

Participant selection and recruitment

Guided by the sampling matrix (see Figure 11) we initiated four interview recruitment waves. Each wave satisfied a different sampling goal, namely piloting, generating interest, maximising diversity and ensuring representativeness. A description of the recruitment waves and the corresponding numbers of participants approached and who participated is presented in Table 19. Recruitment wave two was particularly innovative as participants were recruited from a sampling frame derived from studies included in the systematic review (see Chapter 6). We continued to interview until we reached saturation,378 that is, when additional interview data no longer generated any major new insights. Although no formal assessment for saturation was used, saturation became apparent as we iteratively coded interview transcripts, such that no new NVivo codes were generated with the last few interviews.

TABLE 19. Numbers of participants approached and who participated in each recruitment wave.


Numbers of participants approached and who participated in each recruitment wave.

We supplied an information sheet explaining the purpose of the study (see Appendix 21) to all potential participants and allowed them sufficient time to consider participation and ask questions. We provided a link to our project website ( and our contact details should those approached wish to request more information, and also contact details for someone outside of our project should they have questions they wished to ask an independent person. Written consent (see Appendix 22) was obtained from all consenting participants (both researcher and participant signed the consent form) and all participants completed a demographics questionnaire (see Appendix 23).

Data generation

Semi-structured interviews with participants were conducted either by telephone or face-to-face. Semi-structured interviews provided an opportunity to build rapport, elicit detailed accounts and explore new topics as they arose while keeping focus on the topic of interest. The interviews were structured around an interview guide, which was first piloted and refined through expert consultation (see Appendix 24). Interviews were conducted in English by either JJL or ED or jointly by both researchers and were digitally recorded and transcribed by the project secretary (FA). All transcripts were subsequently checked for accuracy and completeness by one of the two researchers (JJL, ED).

Data analysis

Analysis was comparative and iterative to synthesise the body of included studies.379,380 A coding frame was developed drawing on both the theoretical concepts and emerging information and applied to the corpus of data for thematic extraction (and the subsequent realist evaluation, discussed in Chapter 8). The identified themes from the theoretical literature (see Box 5) were key concepts framing this analysis. As these themes were mirrored in the empirical literature, we therefore anticipated that they would also be apparent in the interviews and thus they informed the development of some of the deductive codes. Furthermore, the systematic review (see Chapter 6) gave rise to 46 adaptations derived from the empirical studies – these too further informed our deductive code generation. These findings together and along with the discussion of ethnicity in Chapter 1 constituted what we knew on this topic before conducting the interventions. We anticipated being able to locate and follow some of these common threads in the interview data; this was expected given that our sampling frame was in part populated by authors of the empirical studies included in our systematic review (see Chapter 6). We therefore anticipated uncovering some of these earlier discussed ideas. These qualitative findings were further important to contextualise and illustrate how adaptations and intervention stages work under the realist framework introduced and discussed in Chapter 8.

The coding frame (see Appendix 25) included both deductive codes generated through familiarity with the body of literature12,13,16,26,35,254,255,257259,262,263,265,268270 as described above (see Chapters 5 and 6) and inductive codes generated from an initial read-through of interview data. This hybrid approach381 has the potential to yield a rigorous thematic analysis as new insights and emerging themes are accommodated alongside themes already identified from the literature. The coding frame was independently piloted by two coders (JJL and ED) and disagreements were resolved by discussion or, when necessary, a third coder (AS). The two authors independently coded all transcripts using the NVivo 8 qualitative data analysis software. Multiple independent coding allowed for discussion of new and emerging insights and, importantly, the transparent documentation of how analysis developed.382

The two coders grouped and mapped the identified codes with attention to how and what factors were considered to be important in the adaptation of health promotion intervention for ethnic minority groups. Thematic analysis383 was used to analyse text: 65 inductive and deductive codes were generated from the five topic areas and were grouped into conceptual clusters (see Appendix 26). These clusters formed 24 basic themes, which were further grouped into 11 organising themes. In light of our overall study question, namely ‘How can interventions be adapted to be more effective for ethnic minority groups?’, these themes were further abstracted into three global themes to answer this central question. The organisation of the codes into this hierarchical thematic network is described more fully in Appendix 26. This systematic method of building up the themes was grounded in the text-linked codes and helped to identify consensus or conflict across themes, as well as the absence of expected themes,384 to explore how adapted interventions worked or were thought to work.

Reflexive considerations

Our interest in interviewing researchers and health promoters with considerable experience in delivering adapted interventions meant that we were, in some instances, ‘studying up’. Research, and particularly interviews, usually take place between people of unequal socioeconomic status (SES). Participants in health-care studies, for example, are usually patients or subjects in trials, while those conducting the research are health professionals with institutionally sanctioned degrees. In contrast, ‘studying up’ refers to conducting research with participants who have a higher social/economic status and/or are members of bureaucratic or institutional bodies. ‘Studying up’ can help uncover the logic behind how and why interventions are designed and delivered, as well as help deconstruct the ideologies that both enable and constrain the success of adapted interventions.385

The background of the team, with wide-ranging interests and expertise, aided recruitment as members initially reached out to those in their networks. This multidisciplinary and multiethnic research team provided thoughtful suggestions in grouping the deductive and inductive codes, cautioning against generalisations, and particularly in illuminating the boundaries of an analysis focused on ethnicity, to suggest considerations of SES, age and education.

The ethnic positionings of the two interviewers (JJL identifies as a Chinese Canadian who has also lived and worked in Asia, Africa and Europe, and ED identifies as Celtic European who has lived and worked in New Zealand for the past 10 years) positively facilitated identification and familiarity with the researchers interviewed. However, as the majority of our interviews were with researchers identifying as African American, although we were able to appreciate the historical situated experiences they were drawing on, we were largely unable to grasp the gravity of such a history that has been compounded by contemporary health-care and structural issues. At times these unfamiliarities widened our social distance and perhaps prevented us from ‘digging deeper’.

In analysis, our collective backgrounds in social sciences and public health framed our approach and likely directed our attention to the influence of social, structural and environmental factors on ethnicity and health over, for example, cognitive factors. Furthermore, there was a tension between two of the researchers (JJL and ED) over the use of the term ‘community’, with one opting for a more critical perspective and the other for a more practical perspective. This tension is likely apparent in our use and disuse of ‘community’ and who it can represent in our research.


We approached 37 participants and 27 agreed to be interviewed (see Table 19). The majority of declines were implicit, as the participant did not respond to our request. Those who responded to our e-mail request cited time constraints as the reason why they were unable to participate; some of these people referred us to their colleagues. We therefore conducted 26 semi-structured interviews with researchers and health promoters delivering adapted health promotion interventions to African-, Chinese- and South Asian-origin ethnic minority groups around the world. The qualitative data were first independently analysed (this chapter) and then synthesised with findings from the other components of this mixed-methods study using a framework of realist synthesis (see Chapter 8).

Our maximum diversity sampling captured a wide range of participants of varying ages and occupations and working with different populations around the world (Table 20). The largest group of participants (35%) was aged between 46 and 55 years. Participants were located in the USA (n = 14; one interview included two participants), the UK (n = 7), Australia (n = 2), New Zealand (n = 2), Norway (n = 1) and China (n = 1). The participants’ occupations included community outreach workers and consultants (n = 5) and research associates (n = 4), with the majority being university professors (n = 17). Participant 25 did not return a demographics form and thus we have no information to report except to researcher-identify him as ‘male’. It should be noted that a participant’s self-identified ethnicity was not always matched to the ethnicity of the target group with whom the researcher had previous working experiences. This diverse sample allowed us to explore the process of adapting health promotion interventions in a broad range of sociopolitical contexts.

TABLE 20. Interview participant demographics.


Interview participant demographics.

In the following summary we first present an overview of the findings from the qualitative interviews and then, in subsequent sections, we elaborate and provide more detailed descriptions with illustrative quotes as well as a discussion of how these findings add to the existing literature that we reviewed in the previous chapter (see Chapter 6).

Overview of findings

We identified three main ‘global’ themes: constructing ethnicity, adapting interventions and sustaining adapted interventions.

Constructing ethnicity emerged as a new insight in terms of broadening the thinking on ethnicity from its more conventional dimensions (i.e. religion, language, culture, physical features and ancestry) as described in Chapter 1. Furthermore, although demographic variables such as age, gender and SES interact with ethnicity, these distinctions were maintained so as to retain the utility of ‘ethnicity’ as a concept and not render it all encompassing. We found that expanding the concept of ‘ethnicity’ to also include contextual elements (e.g. social environments, physical space, past exposures to health research and services, diverse life experiences including stress) was particularly valuable for consideration on how to adapt interventions for ethnic minority populations.

The theme of adapting interventions discussed the kinds of adaptations undertaken and considerations around implementing the adapted intervention. These adaptations and considerations reinforced the adaptations identified in the systematic review (see Table 13). This was expected as many of the interview participants were, as noted above, recruited from a sampling frame derived from studies included in the systematic review. Many of the identified themes complemented what was already known about working with ethnic minority populations, for example the importance of community engagement and building trusting relationships,97 providing incentives to individuals or organisations for recruitment and retention,386 adapting messages and materials,35 matching personnel and using preferred methods to deliver messages and materials257 and, finally, building linkages with existing organisations.387

The final global theme, namely sustaining adapted interventions, was an unexpected finding as we did not anticipate the degree of in-depth discussion in relation to the practicalities of delivering adapted interventions and sustaining this field of research (e.g. funding, formalising the intervention, recognition of work, staff safety and capacity building).

More detailed findings on each of these overarching themes are presented in the following sections in which each of the three global themes is elaborated on and discussed in turn by drawing on its constitutive organising and basic themes (italicised). A schematic diagram precedes each global theme to outline the hierarchical and associated relationships between these themes. Verbatim quotes are provided to further illustrate and support the three overarching themes [the speaker is identified by their participant (P) number, e.g. P1]. In the following sections we use the term ‘researchers’ to refer to the participants of this study, and the term ‘participants’ to refer to those for whom the intervention was delivered.

Global theme 1: constructing ethnicity

The process of constructing ethnicity incorporated commonly used population demographics such as age, gender and socioeconomic factors, which were, generally speaking, acknowledged to intersect with ethnicity in both expected and sometimes surprising ways. Constructing ethnicity also drew on familiar dimensions of ethnicity such as ‘culture’, food, language, religion/spirituality and ancestry, which were discussed in Chapter 1. Finally, in considering ethnicity in the practice of delivering adapted interventions for ethnic minority groups, constructing ethnicity also stimulated discussions of participants’ exposure to health care and research, setting-based considerations in which space and place are prominently featured, participants’ social environment and researchers’ recognition and acknowledgement of differences. Consequently, population demographics, conventional ethnicity and contextual ethnicity represent the three organising themes contributing to the global theme of constructing ethnicity (Figure 12). Together, these three subthemes have the potential to shed light on ethnicity as a dynamic process rather than a limited set of immutable characteristics.

FIGURE 12. Global theme 1: constructing ethnicity.


Global theme 1: constructing ethnicity.

Population demographics

In the theme of population demographics, age, gender and SES, for example, were reported as common factors considered alongside ethnicity. These factors can intersect with and mediate how ethnicity is experienced.

Age-associated preferences were recognised to interact with ethnicity:

We have the facilitators and the experts. All of them are black women in mid-life. It was just as important to these women in, in the focus groups when we started, they said now don’t bring in here as our fitness expert a 20 year old that weighs 110 lb and is perky right. They don’t understand what we’re going through with our back problems, our knee problems.

(P23, general health, USA)

Gender, like age, appeared to intersect with ethnicity as women and men were observed to have different health problems, health concerns, levels of participation and ability to participate in interventions:

Women couldn’t get them on their saris, the saris not the best thing for a pedometer so they were trying their knickers and, and things like that. The men of course had no trouble on their belts. And the men would go out for a walk and their wives would be at home cooking dinner.

(P8, healthy eating, UK)

Different approaches to health behaviour change were also apparent:

The Chinese population prefer, the men, the men like to become smoke free and then celebrate it back with their family so they won’t lose face if they relapse.

(P7, smoking, New Zealand)

Socioeconomic status was another factor that intersected with ethnicity. In many cases, these interventions took place in lower socioeconomic settings:

Sure em the area in which we did the, em where we had where our offices were located was in a low socio economic em area [okay]. Em most em of our participants were em low em poverty level or below, em most of my clients did not ah have jobs or were between jobs, em, very transient sort of lifestyles. Eh we did have a few that were em you know middle class em that I can remember, eh most did not have any sort of em college experience, a lot of them did not em graduate from high school. Em so you know we were working with a very relatively low educated low income population.

(P26, smoking, USA)

Conventional ethnicity

The theme of conventional ethnicity encompassed the ethnicity dimensions, which were presented and discussed in Chapter 1.

‘Culture’, food, language, religion/spirituality and ancestry represent some of the ethnicity dimensions already discussed in Chapter 1. Broadly speaking, in the interviews ‘culture’ was discussed with a focus on shared qualities and understandings. ‘Culture’ was perceived to be particularly significant for dietary behaviours. Researchers felt that, despite their participants’ expressed willingness to change, behaviours surrounding food in social situations were quite durable and difficult to alter.

They view within their home that healthy food isn’t acceptable, either to themselves or their family, or visitors, they come over and say oh those were horrible cause we know we recommend for instance bake your samosas and the women are like oh wait a minute, I would be considered a horrible hostess, if I did that, if I didn’t offer them a [mithai] you know the Indian sweets whenever people were there then I would be a horrible hostess, and the pressure is so strong it would make them feel, you know, in a sense that they were ostracised from their community.

(P8, healthy eating, UK)

Language was another important dimension of ethnicity that was identified and a prime objective of these adapted interventions was to ensure the availability of language-appropriate materials, that is, that translations were accurate and/or that these reflected the literacy level of the target population. Interestingly, in one intervention, English versions of written intervention materials were perceived by the participants as providing an opportunity to learn English. In other interventions, written materials were eschewed in favour of pictorial information, as this was felt to be more accessible.

The role of religion and religious leaders in interventions varied. Interventions delivered for African American populations were more likely to embrace the role of religion and spirituality in assisting behaviour change, whereas interventions delivered to Muslim populations generated a mixed reception – for example some participants advocated behaviour change during Ramadan whereas others felt that it should be reserved for faith-based activities only.

Finally, ancestral ties were observed to affect health practices and behaviours, cutting across the many other aspects of ethnicity:

Where they come from influence what they eat, but not the location and the cities in United States . . . where they came from, from their original country, is it China but if they, they came from Hong Kong eh they like certain food, and from Taiwan they like a certain different kind of food, so it would be different.

(P21, healthy eating, USA)

We can’t do it in groups, again for confidentiality issues because if a community is quite close-knit, so you have to make sure that they don’t you know they’re not in a group where somebody knows them from Bangladesh back home so there’s a lot of village ties, so they don’t really want, somebody older wouldn’t want somebody younger than them knowing about their personal issues and why they smoke.

(P6, smoking, UK)

Interestingly, discrimination was not mentioned or raised by any of the interview participants, even though it is often cited in the literature to play an important role in determining health status.258 Historical discrimination was, however, mentioned in the context of discussions on trust when attempting to engage African American populations.

Contextual ethnicity

The theme of contextual ethnicity discussed processes related to the concept of ethnicity relevant to the participants in the interventions or programmes. These processes included participants’ health-care/research exposure, considerations of their settings and their social environment and researchers’ acknowledging differences within the target groups.

Participants’ exposure to and past experiences with health care or health research affected the way that they perceived the intervention and influenced their level of interest and participation, particularly if their previous experiences with research projects or researchers were negative. Degree of exposure to information and services was identified as potentially hampering or increasing participation. For example, information on smoke-free environments was advertised through mainstream media and thus was perceived to represent a systematic exclusion of a group of Chinese-origin people from accessing the same health information:

P: Back in 2003 New Zealand em brought in the eh smoke-free environments legislation.

I: Yes, yes. So one of the leaders in the world weren’t they?

P: That legislation came in but of course most of the promotion and everything was done in the mainstream profession and what we found is that a lot of the em communities that weren’t accessing mainstream media [mm mhm] weren’t getting those messages across about to take the smoke outside and about the fact that it was now em you know legislated against em smoking in public places, bars, restaurants and things like that, so there was a whole section of the New Zealand population that kind of missed that message.

(P7, smoking, New Zealand)

Setting based included researchers’ observations of factors related to space and place and are thought to effect intervention outcomes, for example immigration status, availability of cheap and illegal tobacco products, migration influxes and movements in and out of the city, as well as changes in practice over time, such as increased dining out:

P: Em but it, it, there was a, just a complete lack of understanding from the business owners’ perspective of why they should care less about this.

I: Mm mhm. Em why, why were they suspicious of the pro, of the programme?

P: I think it’s because there’s a lot of suspicion, again I, you know, this is, you’re in Edinburgh and the United States, I mean you have to under, it, it, there’s all kinds of issues around immigration now in the United States [mmm], there’s a lot of ah fear and you know that people are gonna be accused of hiring illegal workers.

(P18, smoking, USA)

Migration, for example, was observed as a salient stressor, particularly in the context of providing for the family and dealing with family concerns ‘back home’ (this was particularly relevant for Bangladeshi-origin populations). Immigration status was another source of stress, which appeared to be prevalent within ethnic communities (more relevant for recent migrant Chinese-and South Asian-origin populations).

We tried to talk of you know discuss issues that were more relevant to men, other race-related issues that might be relevant such as the stress of dealing with you know recent immigration or other immigration-related issues, trying to find a job, trying to be able to support one’s family and things of that nature.

(P20, smoking, USA)

These stressors were widely acknowledged by researchers as inhibitors to intervention success, and therefore stress-reduction strategies featured prominently in smoking cessation interventions, for example for Chinese and African American populations:

The first thing to do for white women, now of course this is a broad generalisation, but in general if you ensure . . . that they are also involved in weight reduction plans, because if young white women, as a group, start to gain weight, when they stop smoking, they’re going to go back to smoking. If you’re trying to stop black women from smoking and you don’t have a stress management programme in place they, are going to go back to smoking when they get stressed.

(P23, general health, USA)

Where the intervention was delivered was another important setting-based consideration. Holding the intervention in a location and physical environment that was perceived to be relevant to the target ethnic group was thought to increase participation and salience. The implicit assumption made by researchers was that there were, in general, stronger spatial ties, particularly for recent immigrants. Community centres, schools, churches, libraries, mosques, homes, cultural centres and ethnic enclaves were all enthusiastically cited as convenient locations, and this was generally accompanied by a wholesale avoidance of academic institutions as sites of intervention.

There was a mixture of settings [mmm]. Em the initial meetings quite often were actually held em at the university, em although that tended to, that was in the early days of the intervention, because the participants in terms of transport sometimes had issues with that, not all the women drove and I, I suspected they didn’t feel as comfortable in that setting. We had access to a em community centre and some of the meetings occurred there and particularly once the intervention was rolling, so the initial meetings might be at the university or community centre, I would suggest that the community centre was preferable, em the women from what I can gather appeared to be more comfortable there [mmm] eh but the subsequent meetings say over the 12 weeks of the intervention quite often would end up being at one of the participants em homes and the meeting would occur there.

(P12, physical activity and healthy eating, Australia)

A consideration of physical environment also captured the effect of population density and spread, with less dispersed populations having a greater opportunity for participation in interventions. Therefore, the type of ethnic groups captured by intervention efforts appeared to be tied to the spaces they occupied.

I think the individuals we recruited in our group tended to be ah, some of the Korean American participants ah probably were less underserved than the Chinese American participants, if that makes sense [mm mhm]. Em the, a lot more of the Korean community sites that we recruit from, they tend to ah actually sprawl out into the suburbs, suburban regions. Em whereas the Chinese American communities that we’ve been working with have been a little bit more concentrated around the [city name] area.

(P20, smoking, USA)

Related to physical space, the social environment also shaped participants’ relationships with family and friends, and determined the kinds of social support one could receive, or be expected to receive. For healthy eating and physical activity interventions for women (South Asian and African American), for example, the family was often seen as a pivotal motivator for health improvement:

But then of course as these are women, we always hook it into if you don’t change your behaviour you’re going to see in your grandchildren the statistics of women dying prematurely from preventable deaths. So we really try to do both, I mean we, we focus on expanding their sense of responsibility to themselves in terms of health but we always link it to how, if they change, it can help in terms of their community, in terms of their family and, and the rest of their community.

(P23, general health, USA)

The family was also a strong motivator in smoking cessation interventions for men (of Chinese and South Asian origin):

I think and a lot of it would have to do with coming back to the family . . . we would say you know try to bear in mind that you know your children really want you to stop smoking and you know you, you’ve moved to this country to help your children have a better life and you know they really want you to, to be here with them for a long time and you’re doing this, you’ve done all these other things for your children you can do this one more thing too.

(P20, smoking, USA)

Another aspect of the social environment was social support, which was well documented in the interviews. Overall, the idea of organic social support was thought to be more effective than purposefully formed support groups; however, the use of the term ‘organic’ differed across interviews. In some cases it referred to friends and family as the support provider:

I mean peer support is likely to be successful if it’s done well. Even though we, we didn’t find this an effect, I’m not sure that that really means that that isn’t successful. But yes for sure, I think organic social support would be even better and that’s why we’ve, we’ve created the third study the way it’s designed.

(P14, physical activity and healthy eating, USA)

At other times, organic support referred to the groups that formed in the interventions:

In the intervention group you know changed more initially but then those changes washed out faster, and even reversed in some cases, versus the control group that changed more slowly without all the handholding but you know made use of the social support of you know the other women in their, in their classes, and they were actually able to make a more sustainable change in terms of being able to stabilise their waist circumference.

(P13, physical activity, USA)

Therefore, it remains unclear which kinds of organic support were considered to be effective, as this has been defined or interpreted differently across interviews. Furthermore, social support was also contingent on opportunity; for women who may not have received support at home for physical activity, for instance, an intervention may have acted as the catalyst for these women to form activity groups, which in turn provided support during and well after the end of the formal intervention period.

Lastly, researchers acknowledged differences that exist among supposedly similar intervention group members. In many cases, intervention effectiveness was difficult to determine because of the heterogeneity of participants. This heterogeneity was often related to factors such as educational attainment, gender and age, as discussed earlier, while degree of acculturation also contributed to heterogeneity:

There are those older people as well they would say . . . oh yeah I only eat a baked potato because my granddaughter makes this so I eat that as well, you know eh so before they were never eating but now with time eh there’s been a cultural shift and there even if it is pasta or so they could be cooking eh the pasta with eh Asian touch to it so there is, yes acculturation definitely there.

(P3, healthy eating, UK)

Global theme 2: adapting interventions

We classified researchers’ descriptions of implementing adapted interventions for ethnic minority groups (regardless of the ethnic group targeted) into the following six organising themes: development, engagement, materials, methods, description and linkages (Figure 13). Not all researchers provided insights for each theme as some tended to focus on, for example, engagement, whereas others more readily discussed materials and methods. These themes contributed to the global theme of adapting interventions. The interviews interpreted together revealed the additional considerations and frequent negotiations that required attention at different stages of the intervention. These considerations and negotiations shaped and mediated what adaptations and programme components were feasible and acceptable and, therefore, ultimately possible to deliver.

FIGURE 13. Global theme 2: adapting interventions.


Global theme 2: adapting interventions.


The theme of programme development encompassed both political considerations (research environment) and personal factors (project rationale).

The initial steps of intervention development required a consideration of the research environment. This sometimes meant aligning the intervention with government policies but also ensuring that these policies, for example smoke-free policies, are understood and comprehended by ethnic communities; and sometimes delivering interventions for a group that had been largely overlooked by government health promotion policies. However, funding for these activities was often tied to government-prioritised areas, and when priorities changed, earmarked funding became scarce and determined the kinds of activities that could be undertaken.

We just didn’t have the resources to promote it the way we had hoped to promote that one, that, that event and that was partly because of the turnover at the City Health Department and the commissioner who had committed funds to this intervention left and the new commissioner came in and had his own agenda.

(P18, smoking, USA)

The researchers’ own backgrounds were usually the main driving force behind the project rationale for the intervention, sometimes aided by PhD students who had identified hitherto unexplored areas. Their past experiences usually influenced the adaptation rationale, as well as the kinds of intervention theory they were likely to adopt. For many US-based researchers, their research background and experiences allowed them to tap into a network of researchers working in this emerging field, and thus they were able simultaneously to draw on their collective knowledge regarding adaptation rationale and intervention theory in the development of the project.


The theme of engagement involved two equally important steps: engagement with the targeted ethnic group or community (community involvement) and engagement with the characteristics of the targeted community (knowing the population).

Approaching the community and assessing their needs and considerations at the outset received near unanimous support from almost all of the interview participants. Community involvement was further improved when community advisory or steering groups were convened for the purpose of the intervention to advise on numerous issues, including appropriateness of incentives as well as health messages. Other times, pre-existing and established community organisations or leaders were approached to guide the interventions, and this was particularly evident at the recruitment stage.

The second kind of engagement required the researchers to know the population through various methods, including formative work, which encompassed focus groups, interviews, surveys and literature reviews. This was important to understand the community as a shifting, changing entity and to avoid inappropriate characterisation. This appreciation of the community continued throughout the lifecycle of projects through, among other things, participant feedback. For example, a weight loss intervention was developed to be culturally sensitive for an African American population, incorporating appropriate goals and motivators:

This is a journey this is a long process this is not a diet that you go on for a short period of time. A journey, but also for your best body so it, it’s for you to look and feel better but not necessarily to be thin, and we conveyed a lot about how you really can’t be there to take care of your family if you yourself are the one who is disabled or ill because stemming at least partly because of your weight.

(P14, physical activity and healthy eating, USA)

During the actual intervention recruitment period, participants also fed back to researchers, for example one group felt it was discriminatory within the church context to have an intervention for the exclusive use of one ethnicity:

Well I think some women who were not African American wanted to participate and felt excluded. Other women just felt that that was ridiculous that their friends were being excluded purely based on the colour of their skin, I mean it really is racially discriminatory against women who are not African American or black. And so and it was by design because the higher risk group is African American or black women, but within the context of the church that, that health risk is, is really immaterial.

(P14, physical activity and healthy eating, USA)


The theme of materials represented the practical strategies employed for recruitment and retention and included consideration of the resources delivered to participants.

In the USA, incentives were commonly used as motivators to improve recruitment and retention. The types of incentives employed included both direct financial rewards (money and travel costs) and indirect benefits (ranging from free toothbrushes to gym memberships). The value of the incentive was largely dependent on project funding.

There was mixed opinion on whether or not incentives improved individual-level participation. There was more support for providing incentives to organisations to enhance their participation, which was thought, in turn, to facilitate participation by its members. Providing incentives to established organisations was seen as an effective strategy to increase participation.

The resources delivered included materials and messages, which were targeted to the participants for language and message format, for example, and this was not surprising given our topic of enquiry. What was surprising was that some participants mentioned that these resources should be ‘evidence based’ and this was perceived to be an important criterion for those delivering the intervention as well as for some participants receiving the intervention.

So there may be a higher level of sort of sophistication and em and this, people feel that if it’s delivered through the church that’s enough you know but they want the information to be high quality and not necessarily delivered by a pastor anymore, they want it to be delivered by a health professional. So there seems to be a shift taking place, or it may be more of an urban rural shift, we’re not really sure.

(P22, healthy eating, USA)


The theme of methods described how intervention components were delivered (mode of delivery) and important researcher qualities that enhanced or constrained effective delivery of intervention components.

There was great emphasis on researcher qualities, particularly their ability to develop relationships. Relationships, although perceived as time-consuming pursuits, were also described as an absolute necessity.

One of the kind of ways of you know building trust and building partnership really have to do with spending time, going to the services, being there for events, listening.

(P22, healthy eating, US)

Perseverance was felt to be necessary to develop relationships with the target community, but also a necessary quality for researchers to secure resources for the community:

So it took me about 2 years to get permission to open up the swimming pool for these women at eh on the weekly basis, for few hours, which was just women only, and I had to go through many many steps.

(P9, physical activity and healthy eating, Australia)

Researchers also needed to be flexible within their roles and strike a balance between delivering a research project and responding to the target population, such that sometimes research components needed to be modified, for example translation of documents, timing of sessions or even completely dropping some component of the work if necessary:

One of the things that we found was that kids just abhorred having their fingers stuck, these are you know African American and Latino kids and so we just dropped the the serum glucose measure because . . . we were losing too many of our research participants you know even though they had parental consent and they had originally signed [consent] after the first round of data collection.

(P13, physical activity, USA)

The mode of delivery was often described as more important than the message itself; strategies such as peer educators or peer counsellors were seen to be appropriate for ethnic minority communities, although reasons were not always provided. Often interventions were delivered in group settings, and one motivation for this was that there may have been transferable modelling skills that could be shared with children or other family members who may be present during the intervention, or that could be taken home after the intervention.

So we knew that we had their attention, that was one main reason cause they’re a major target group because of their own health, and that’s what we emphasised. But number two we also know that a woman in that age group, and certainly African American women, we know are the matriarchs of their families, so they are influencing not only their adult children but their grandchildren, their great grandchildren, their churches, their workplaces, they are very influential [mmm]. And so if we can change the lifestyle of one single woman, just one single woman in that age group, we have a major ripple effect that’s going to impact on their entire community.

(P23, general health, USA)

Beyond the family there was thought to be potential peer modelling effects within tight-knit communities:

But, but the, the, the eh the and the higher risk group persons, at least some of them, very successful in eh change their behaviour, started to walk eh in the district every, every evening eh every afternoon or evening, attend the classes eh walking classes which walked in the district and, and eh some of them lose weight so that anyone could, anybody could see that. Eh and they became very important role models for the others.

(P19, physical activity, Norway)

Across the interviews, staff and staff characteristics were reported as contributing to successful uptake and delivery of the intervention. Having someone who was of the same gender, who spoke the same language were some of the practical attributes described:

Of the same culture . . . South Asian and . . . struggled with obesity herself.

(P8, healthy eating, UK)

Having an ethnically matched staff member was preferable; however, this was not always possible. It was felt that a person who was not from the same ethnicity (or background) would still be able to deliver the intervention if they were enthusiastic; however, they could be perceived by participants as not understanding their particular situation. It was also seen to be useful if the staff were well known in their community and respected. These were generally more nebulous characterisations, but the justification given was that:

Qualifications for working in those fields you know, we could teach the skills but we couldn’t teach the community connectiveness, that’s either there or it’s not.

(P7, smoking, New Zealand)


The theme of study description encompassed descriptive accounts of the programme or intervention. Across the interviews an eclectic mix of approaches, various settings and multiple activities and actors were reported. Participants were similarly diverse, with a mix of generations, genders and ethnicities reported. Groups ranged from 10 participants in small pilot studies to thousands of participants in a citywide survey. There was no obvious model intervention for adaptation considerations.


The theme of linkages described both the limits of collaboration and project scope (research boundaries) as well as the benefits of collaboration with other professionals and institutions (professional collaborations).

Health interventions delivered for ethnic minority populations often drew clear boundaries around health and social issues, which were outside of their remit. Although some projects acted as a link between services, they did not, for example, deal with mental health or abuse cases. In these situations they made referrals to the appropriate services. On the other hand, some interventions were purposefully involved in related services for their population, for example pregnancy services. There did not appear to be a clear pattern in determining research boundaries, as responsibilities were continuously negotiated:

You know again in some ways we were sort of battling [partner organisation] sometimes as well, wanting to do something that wasn’t just standard.

(P8, healthy eating, UK)

Overall, professional collaboration with existing institutions, services and professionals was reported to positively enhance the intervention. However, linkages with some organisations required more complex manoeuvring than others, and this feature was largely dependent on the characteristics of the organisation, for example the church and its religious leader:

There’s also a lot of differences between churches and denominations in terms of the governance and the structure and some denominations have, like a Baptist denomination, has a consistent pastor that will be there for many years and so if they commit to working with you they’re gonna stay with you, whereas other denominations like the AME [African Methodist Episcopal] we found they rotate pastors every couple of years and so you start into it and once the new pastor comes in they may have a different priority eh and may decide not to, that they’re not so enthused about working with you [laughs]. So that has been a problem in terms of you know often if, if a church, quits in the middle it’s usually because there’s a change in leadership [okay] so that was one thing in terms of working with an organization.

(P22, healthy eating, USA)

Global theme 3: sustaining adapted interventions

Sustaining an adapted intervention involved interpreting intervention findings, reflecting on what was learned and what could be taken forward, and fostering the capacity of the workforce as steps towards continuity. The three organising themes of evaluation, insights and strengthening the workforce thus contributed to the global theme of sustaining adapted interventions (Figure 14). Discussions of how to disseminate project findings, sustain the programme of research and manage project personnel demonstrated the constraints and challenges faced by the researchers we interviewed as well as revealing the rewards and motivators for continuing their work.

FIGURE 14. Global theme 3: sustaining adapted interventions.


Global theme 3: sustaining adapted interventions.


The theme of evaluation discussed issues relating to intervention cost (research costs) and effectiveness (intervention assessment).

Additional research costs were felt to be necessary, particularly for community engagement, and particularly justified when carrying out work with ethnic minority populations.

The truth is that any time you’re gonna do a, whether we were in the black community, Hispanic community, we, we were gonna have to fund community organisations to do this work because you can’t do this work unless you’re in the community. So I don’t think, I think doing any kind of tailoring or creating you know tailored programmes is a little, adds an expense to the existing programme, but then if you want to be effective in your, if you’re in a city like [city name] you, you don’t really even think about it that way because it’s the cost of doing business, it’s, it’s not a question of should we do this, its, you have to do it you know.

(P18, smoking, USA)

At times, researchers felt that funders had little understanding of these additional research costs (P10), which added to the difficulty of accessing sufficient funding for research or for sustaining interventions beyond the research phase. A lack of funding, as well as the lack of personnel and overlapping project phases, meant that many interventions were not formally evaluated, even though they had been delivered. It was, however, suggested that reporting in the usual peer-reviewed journals might not always be the preferred route for communicating results:

Dissemination in the community can be much more important than producing a peer-reviewed publication as far as being effective goes.

(P17, physical activity and healthy eating, New Zealand)

When the intervention assessment was published, the outcomes and results were not always straightforward and often raised more questions than answers in relation to the components that were likely to be effective in promoting behaviour change:

The structure is, is very critical. But what we, what we don’t know is really one site different from another? Another thing we think we know, we haven’t documented it, is the facilitator the key person to the success of the outcome?

(P23, general health, USA)

When ‘successes’ were mentioned, this was usually discussed anecdotally, as the evaluation either was carried out informally or had yet to be published:

I: So would you consider this intervention that you ran to be quite successful?

P: Oh without a doubt, yeah. No, it, it, in terms of, we did some informal, I mean, more informal follow-up and that would say for example the fact that the swimming sessions continued on you know a number of the ladies who, we, we got some data on that eh who had participated in the pilot study and so some of them we collected bloods most probably 18 months down the track, after they’d finished and again yeah we most probably had you know in the pilot I think we ended up with nearly 20 women, and of those I would think most probably 50% had actually made a positive change in terms of their lifestyle. And we had, and that, that was with no active intervention since we eh we started the pilot.

(P12, physical activity and healthy eating, Australia)


The theme of insights discussed growth experienced by the researchers (researcher development), confirmation and correction of initial research assumptions and new findings generated in the research (intervention insights), and how these understandings were translated into strategies to sustain health promotion actions (intervention continuity).

Researcher development was something that occurred over the intervention period. Researchers reflected on their own experiences by discussing personal growth and reward, and how this incentivised them to continue to work delivering and implementing adapted interventions and programmes. Other researchers used their work as a springboard to study ethnicity and health at a systems level.

Intervention insights were often derived from unexpected findings and through reflection on initial assumptions about the target community. Identifying and sharing their own insights on effective strategies and then relaying their lessons learned to other researchers were cited as reasons for participating in this present qualitative study. One participant was adamantly against maintaining anonymity because she wanted everyone to know about her programme:

I don’t have any problem if we are identified in your research because the whole thing as I see it is about sharing and caring so it was quite interesting when I went down to my boss and I said oh look I really don’t think I’ll sign off on this part where it says here, the your point number 6 where it says I understand you won’t put anything in your report that could be used to identify me and, and [name] says oh well scrub that out and oh no identify away.

(P7, smoking, New Zealand)

Widely publicising and marketing a programme was a reported method for increasing sustainability and intervention continuity. Formalising the intervention through ‘manualisation’ (developing manuals or guides on how to deliver the intervention) or the development of project logos appeared to coincide with the longevity of the intervention. Furthermore, continuity was also reported to require, on some level, a continued presence or advocacy on the part of the researchers. Developing training centres with set materials and teaching methods to facilitate the consistency of the intervention for participants in a broad geographical area was a model for sustainability reported in one intervention:

We said we’re gonna look at [name of a fast food chain] and see how they do it [laughs], a [name of a fast food chain] hamburger in [city name] is the same as the one in Edinburgh. But anyway, so the quality, one of the ways to maintain our quality em because we know it works and, and there are many ways, reasons that we know it works is to really eh be in charge of the training of future facilitators and community health leaders so that’s a real important piece of what we’re doing.

(P23, general health, USA)

Strengthening the workforce

The theme of strengthening the workforce discussed the management of a positive and safe working environment for these staff members (project management) and the training of intervention staff and recognition of their work (workforce development).

Project management largely involved facilitating positive working environments for staff, particularly a group of multiethnic staff working across sectors:

I mean our common goal was to provide an intervention that was going to support the population. Em and you know so we did have some disagreements, or should I say some very, very robust debates, but em getting everybody on the same page, it did take some time but once we did we just steamrolled, we really steamrolled . . . because of the robust discussions we had we forged such strong alliances out of them em and really paved the way for other tri-party initiatives to be able to happen a lot faster.

(P7, smoking, New Zealand)

Project management also involved the responsibility of ensuring the safety of staff, particularly in the context of delivering interventions in the community and through home visits, which were common for some ethnic minority groups:

Well as a manager of a multi-site project my concerns are safety, as I said we had a homicide last week we were with, we’re experiencing a lot of shootings and, and you know violence in the neighbourhoods and to have outsiders come in, especially when we first go in its, it’s alarming.

(P16, smoking, USA)

Staff, personnel and principal investigators working on interventions for ethnic minority health tended to originate from the target population. Across interviews the importance of this feature of the workforce was emphasised and that, for example, it was more appropriate to:

[take] lay people and [upskill] them rather than taking people and trying to skill them in the community.

(P7, smoking, New Zealand)

Thus, workforce development involved training in technical health promotion skills, rather than softer social skills already presumed to be held by staff members. Along with skills development, recognition of their work was also cited to strengthen the workforce:

It was a big buzz for the team to realise you know that they were, they were this important, that they should have this priority placed upon them and resource, resources spent on them.

(P5, smoking, UK)


By naming the first global theme constructing ethnicity we have deliberately sought to invoke ‘ethnicity’ as a concept that has been built up and deconstructed over decades of discussion and debate, with boundaries around what it can encompass, for example the boundaries explicated by the five dimensions of ethnicity that we considered earlier (see Chapter 1), as well as ‘ethnicity’ as a concept that remains an unfinished process. In turn, the insights drawn from the theme of constructing ethnicity alert one to different kinds of ‘ethnicity’ at work and discussed when adapting interventions. The conventional elements of ethnicity relate to the five dimensions of ethnicity discussed in Chapter 1, which capture, more or less, the common sense usages and the commonly used definitions prevalent in health-care research. The contextual elements of ethnicity, as illustrated by the above data, extend our thinking around ethnicity, in particular by taking into account past health-care and research exposures and experiences, relationships with spaces and places, the significance of social dynamics and the inherent heterogeneity within the target group. These process-focused dimensions of ethnicity can offer a more dynamic conceptualisation of ethnicity beyond descriptive characteristics attributed to persons or groups, which at times appear entrenched and immutable (see Chapter 1).

The contextual elements of ethnicity can further offer a platform from which to consider ethnicity when adapting health promotion interventions. Ethnicity when considered in context can involve a mapping process that captures both conventional elements, such as religion, language, culture, physical features and ancestry, as well as demographic variables, such as age, gender and SES, and also the relationships that members of these ethnic groups have and share with their social groupings and their physical space as well as their diverse experiences. Mapping these processes and relationships has both practical applicability and analytical value for both intervention delivery and analysis. Ethnicity therefore interacts with a number of factors that affect behaviour change and understanding the complex nature of these interactions is useful in advancing the concept of ethnicity itself as well as the extent to which we can adapt health promotion interventions. These elements of conventional and contextual ethnicity represent a preliminary effort to reconceptualise a definition of ethnicity with greater utility for practice and theory.

The second and third global themes of adapting interventions and sustaining adapted interventions suggest that considerations for adapting interventions appear at every step of the intervention process, from initial thinking about funding and political and personal factors to community involvement, incentives for recruitment and retention, professional and institutional linkages and researchers’ qualities. The third global theme in particular indicates that fostering intervention and programme continuity requires an additional set of considerations, which include conducting formal and informal evaluation, sharing insights, formalising intervention/programme strategies and strengthening the workforce. These findings echo the Programme Theory of Adapted Health Promotion Interventions derived from empirical intervention studies (see Chapter 6, Figure 9), while the strategies for adaptation map onto those identified in the 46-item Typology of Adaptation (see Chapter 6, Table 13). These qualitative findings can help to contextualise and illustrate how adaptations and intervention stages work.

Finally, reflecting on this study as a whole, we observed that the majority of the researchers who were available for interview (i.e. identified from the sampling frame) were based in the USA and as such this revealed a geopolitical imbalance in the area of research on ethnicity and health, particularly for adapted interventions. For example, researchers working at institutions along the east coast of the USA identified an informal network that they accessed to share ideas and collaborate on projects. More importantly, this informal network was likely to facilitate the sharing of findings from projects. As we uncovered, projects may go unevaluated and unpublished, leaving those outside of the informal networks without access to this information. In essence, many of the lessons learned from these projects would be invaluable to other researchers in developing future interventions and this ‘lost’ knowledge hampers a sustained growth of evidence in this field.

Internationally, particularly in Australia and New Zealand, there was little connection to this emergent informal network in the USA and the researchers interviewed reported working within smaller groups with less government-financed and collegiate support. In Europe, connections between people and projects from the UK and continental Europe were reported and observed. Conspicuously absent were reports of an organic global network of researchers, as geography appeared to dictate the limits of collaboration and sharing, despite the wide availability of communication resources such as the internet. It was encouraging to hear that identifying effective strategies and relaying lessons learned for other researchers and the potential for generalisability were cited as reasons for participating in the qualitative interviews and this willingness to share insights could contribute to sustained work in this area.

Strengths and limitations

This qualitative component was developed alongside an extensive and rigorous systematic review (see Chapter 6). Included studies from the review published after 2005 were used to create a sampling frame to recruit participants for interviews. This integration represents an effort to link up the two separate components; it proved to be particularly fruitful in identifying a maximum diversity sample and therefore represented a particular strength of this component. However, despite the implementation of a cut-off date of 2005, some of the papers reported on studies that were conducted 10 years previously and consequently project recall was understandably poor. In the future, a later cut-off date might be used. The risk of such a strategy, however, is that forthcoming projects may not yet have analysed their data, or researchers may be unwilling to share data before publication, which was the case in one of our interviews.

A further limitation is that the majority of participants were researchers based in academic institutions. Although we made attempts to find projects developed within the community, this was not always possible as they tend not to be evaluated or written up for publication. This introduces a bias in our sampling towards university- and health professional-conceived interventions, rather than perhaps de novo community interventions. It is, however, worth bearing in mind that we did interview some community workers and that some academics are very much embedded in the communities they work with, and that the dichotomy between university led and community led may be less apparent. Widening the sampling frame and employing a longer search period for participants may mitigate this bias, but, more importantly, may enrich the accounts of adaptation.

The qualitative interviews themselves provided access to information that was not published or in some cases to information that may never be published. This was particularly important and necessary to learn about the processes associated with adapting interventions, of which there are few papers in the published literature. This access provided unique insights far beyond what a reader (ourselves included) could glean from accessing only peer-reviewed journal articles on the same study.

Another strength of this component of the research was the opportunity to speak to those delivering and implementing programmes; this was because they were ideally positioned to provide practical insights on adapting health promotion interventions for ethnic minority groups. On the other hand, speaking only to researchers and health educators about adapting health promotion interventions without supplementary interviews with the participants of those interventions was a limitation of this qualitative component. Therefore, we tried to exercise caution when interpreting interview participants’ views on ethnic groups, particularly when presenting the target community’s acceptance and satisfaction with the intervention, as these opinions were filtered through the lens of their own values and thinking.

Overall, the interview participants did not shy away from discussing aspects of their interventions that were ineffective or that simply did not work. Participants were willing to share some of their shortcomings, as well as their successes, with the explicit goal of advancing the field.


Qualitative studies undertaken alongside systematic reviews, and other linked studies of this nature, are important in teasing out the processes and thinking that underpin adapting interventions for ethnic minority populations. Our findings indicate that a process-oriented concept of ‘ethnicity’, encompassing both conventional and contextual elements of ethnicity, may be helpful to understand the factors and interactions that affect behaviour change and therefore to better adapt health promotion interventions for ethnic minority populations. Furthermore, our qualitative findings confirm many of the main findings from our systematic review of empirical intervention studies, which led to the development of our Programme Theory of Adapted Health Promotion Interventions (see Chapter 6). The programme theory illustrates that adaptations can take place throughout the entire intervention process, from conception to dissemination. Synthesis of these research components will be discussed in greater depth in the following chapter.

Image ch1f1
Image ch6f2
© 2012, Crown Copyright.

Included under terms of UK Non-commercial Government License.

Bookshelf ID: NBK115633


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (3.4M)

Other titles in this collection

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...