Sharing experiences and dilemmas of conducting focus group discussions on HIV and tuberculosis in resource-poor settings

International Health 3 (2011) 7–14

Contents lists available at ScienceDirect

International Health journal homepage: http://www.elsevier.com/locate/inhe

Review

Sharing experiences and dilemmas of conducting focus group discussions on HIV and tuberculosis in resource-poor settings Sally Theobald a,∗ , Lot Nyirenda b , Olivia Tulloch a , Ireen Makwiza b , Amara Soonthorndhada c , Rachel Tolhurst a , Grace Bongololo b , Armande Sanou d , Marjorie Katjire e , Nduku Kilonzo f , Fei Yan g , Nasher Al-Aghbari h , Najla Al-Sonboli i , Rachel Anderson de Cuevas a , Pamela Fergusson a a

Liverpool School of Tropical Medicine, Liverpool L3 5QA, UK REACH Trust, P.O. Box 1597, Lilongwe, Malawi c Institute for Population and Social Research, Mahidol University, Phutthamonthon, Salaya, Nakhon Pathom 73170, Thailand d Groupe de Recherche Action en Santé, 06 BP, 10248 Ouagadougou 06, Burkina Faso e Faculty of Medical and Health Science, University of Namibia, Windhoek, Namibia f Liverpool VCT and Care and Treatment, P.O. Box 19835-00202, KNH, Nairobi, Kenya g Department of Epidemiology, School of Public Health, Fudan University, Shanghai, China h Al Thawra General Teaching Hospital, P.O. Box 12685, Sana’a, Yemen i Paediatric Department, Faculty of Medicine and Health Sciences, Sana’a University, P.O. Box 11341, Sana’a, Yemen b

a r t i c l e

i n f o

Article history: Received 1 April 2010 Received in revised form 17 August 2010 Accepted 15 September 2010 Available online 2 November 2010 Keywords: Focus group discussions HIV Tuberculosis Research methods Resource-poor settings

a b s t r a c t Focus group discussions (FGD) are gaining in popularity in research on HIV and tuberculosis (TB) internationally as researchers seek to understand the experiences, needs and perspectives of people living with TB and/or HIV as well as their carers within the community and health sector. Conducting FGDs in resource-poor settings with vulnerable participants who are living with diseases that are frequently stigmatised poses multiple challenges. Our approach in this discussion paper is to follow the research cycle to present the practical experience of research teams using FGDs in TB and HIV in resource-poor contexts in Africa and Asia in order to contribute to effective practice. The approach highlights dilemmas and shares effective practice for negotiating initial discussions with different communities, constructing sampling frames and samples, choosing a facilitator, encouraging discussion, ethics, translation, pitfalls and dissemination. We demonstrate the techniques and adaptations needed to ensure that FGDs provide rich, high-quality and policy-relevant data on the voices and perspectives of people living with HIV and TB, community groups and health workers within the challenges of resource-poor settings. In applying theory to develop good practice in FGDs across the research cycle, a critical and reflexive approach is needed. © 2010 Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd. All rights reserved.

1. Introduction Focus group discussions (FGD) are an important tool in qualitative research on HIV and tuberculosis (TB) in

∗ Corresponding author. E-mail address: [email protected] (S. Theobald).

resource-poor settings. This paper captures the experiences of a multidisciplinary and international group of researchers working in regions with the highest prevalences of HIV and TB. HIV-infected populations are more vulnerable to contracting TB, both of which are stigmatised as infectious diseases. HIV and TB co-infection rates can be very high, and poorly resourced health services have struggled to manage the dual epidemics, although since the

1876-3413/$ – see front matter © 2010 Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.inhe.2010.09.006

8

S. Theobald et al. / International Health 3 (2011) 7–14

Table 1 Overview of the HIV and tuberculosis (TB) epidemic in the countries discussed in this paper. Country

Estimated HIV prevalence [% HIV + adults aged 15–49 years] (low–high estimate)3

Total estimated TB incidence (all forms) (2007)4

Total reported TB cases/100 000 (2007)4

Papers discussing experience of HIV- and/or TB-related stigma

Malawi Kenya Namibia Burkina Faso China Thailand

11.9 (11–12.9) NA (7.1–8.3) 15.3 (12.4–18.1) 1.6 (1.4–1.9) 0.1 (0.1–0.1) 1.4 (0.9–2.1)

48 144 132 357 15 905 33 437 1 305 770 90 878

346 284 733 27 74 86

Ethiopia Yemen

2 (1.8–2.2) 0.1 (0.0–0.2) a

314 267 17 121

155 38

Rimal and Creel,5 Somma et al.6 Turan et al.,7 Waterman et al.8 Smith and Nguyen,9 Thomas10 Sanou et al.,11 Stephenson12 Li et al.,13 Rao et al.14 Apinundecha et al.,15 Kittikorn et al.16 Sagbakken et al.,17 Admassu18 Date and Okita19

NA: not available. a The magnitude of sexually transmitted diseases/HIV/AIDS in Yemen cannot be accurately defined in the absence of reliable surveillance.2

global antiretroviral therapy (ART) roll-out, management of both diseases has improved.1,2 Co-infection is further complicated in resource-poor settings where participants are more vulnerable due to poverty and limited infrastructure. The context is, however, changing quickly owing to new and increasingly effective therapies. Table 1 highlights estimated HIV and TB prevalences, TB case finding rates and papers discussing the stigma attached to either disease in the countries included in this paper. Some key debates in the FGD literature are important to consider in qualitative research on HIV and TB, such as those surrounding selection criteria. Farquhar20 and Halcomb et al.21 raise concerns about participant recruitment and importing established hierarchies into discussions. Other debates refer to group dynamics and whether this semipublic interaction with arguably compromised confidentiality influences the depth of information provided.20,22,23 The choice of FGD facilitator is relevant as well as the extent to which they should be an ‘insider’ or an ‘outsider’, or a combination of the two, and whether the status of an ‘insider’ should be disclosed to the group (see Farquhar20 for example). Key dilemmas such as these will be discussed here, drawing on experiences of holding FGDs on HIV and/or TB in a variety of contexts. As a team of researchers with experience using focus groups as a tool in HIV and TB research in Africa, Asia and the Middle East, we aim to document and reflect on the practical experience of multidisciplinary and multicultural research teams using FGDs in resource-poor contexts where HIV and TB infection is prevalent. We highlight dilemmas around working on stigmatised health issues as well as responses and gaps across the research cycle identified by the researcher-authors. This was an opportunity to move from theory to practice and to share practical approaches that are not usually discussed in depth in academic papers owing to space limitations. Our approach was constructed around a set of open questions (see Box 1 ) and dialogue between researcher-authors on recurring themes and paper writing. We refer to ourselves in the paper as researcher-authors. The research projects referred to are summarised in Table 2 . By sharing lessons learnt in the FGD process and linking them to experiences, we guide the reader through the dilemmas encountered and the opportunities to

Box 1. Open questions used to guide discussion • How did you use focus group discussions (FGDs)? (Please give brief overview of research objective, context and funder) • How is the topic area—HIV or tuberculosis (TB)—perceived in your country context? Do these perceptions affect the planning and implementation of FGDs? For example, how was stigma addressed, how did you select your participants and decide who should facilitate? • Are there any risks in conducting FGDs on issues relating to HIV and TB and if so what strategies did you use to address these? • Did FGDs contribute insights that you did not get from other methods? • What was your experience of working with multiagency, multidisciplinary and multicultural teams in planning, implementing, analysing and writing up the findings of FGDs (where applicable)? • Any other thoughts or reflections to add?

optimise FGD data quality and usability in a range of settings throughout the research cycle. By drawing both on commonalities and diversity from the research sites, we highlight practical strategies to improve the quality and utility of FGDs. 2. Discussion: dilemmas and shared practice in conducting focus group discussions on HIV, AIDS and tuberculosis throughout the research cycle 2.1. Planning for focus group discussions: what was the methodological rationale? FGDs were deployed in a range of ways, often alongside quantitative research. For example, in the studies on health workers’ access to counselling and testing and ART (Study 1) and on paediatric access to ART (Study 11), FGDs were used alongside interviews prior to the development of questionnaires enabling concepts, language and experiences emerging from FGDs with health workers to guide questionnaire design and strengthen internal validity.30 In Study 10, FGDs were used alongside quantitative research approaches and the findings from the different methods were triangulated concurrently to develop conclusions and recommendations. Other studies were purely qualitative.

S. Theobald et al. / International Health 3 (2011) 7–14

9

Table 2 Research projects using focus group discussions (FGDs) that form the basis of the reflections in this paper. Study no.

Research focus

Key issues explored

Context

Sample

Research team

Funder

1

Health workers access to CT and ART24 Community perceptions of ART and challenges PLWH face in accessing and adhering to ART25

Barriers and opportunities to access both CT and ART Community perceptions of HIV and impact on community, PLWH experiences and challenges faced in accessing and adhering to ART Challenges and opportunities to access and adhere to ART Health-seeking pathway, TB diagnostic process, challenges and opportunities to adhere to TB drugs Experiences of care; the how to of service provision, challenges and opportunities, support received Poverty, illness and social exclusion impact on nutrition and food security Experience and perceptions of GBV and health-seeking behaviour; links between GBV and HIV TB patients’ experience of TB care, community perceptions of TB, TB services and how they can be improved Experience of TB diagnostics, perceptions of quality, support and barriers and enablers to adherence How do the general public perceive TB, how would they seek health care for symptoms suggestive of TB and why? Health workers’ perceptions and treatment strategies for TB Accessibility and acceptability of service provision, psychological, sociological and economic factors affecting children’s treatment compliance Perceptions about TB, particularly languages, aetiology and associated stigma amongst different groups

Three districts in Malawi (urban) Thyolo District, rural Malawi

Health workers

REACH Trust, Malawi, with LSTM

WHO

PLWH on ART, community groups

REACH Trust, Malawi, with LSTM

DFID through the HIV/AIDS Research Programme Consortium

Lighthouse, an urban clinic in Malawi Urban Malawi

PLWH on ART, carers of children living with HIV People who have been diagnosed with TB and started on treatment

REACH Trust, Malawi, with LSTM

Rural Malawi

REACH Trust, Malawi

Rural and urban Namibia

Providers of home-based care (separate groups of women and men); PLWH Different groups of PLWH and HCWs

Communicable Disease Control (CDC), USA DFID through the TB Knowledge Programme and National TB Control Programme OXFAM

Three districts in Kenya (rural and urban)

Groups of older and younger women and men

Three districts in Burkina Faso (rural and urban)

TB patients (male and female); community groups by age and gender

National TB Control Programme with LSTM

Two sites in Awassa, southern Ethiopia (one rural, one urban) Four provinces in China (rural and urban)

Female and male TB patients

Awassa Health Centre, with LSTM

DFID and the Economic and Social Research Council, UK

Community groups (separated by age and gender); and health workers

Fudan University China, with LATH, LSTM, UK

DFID

Urban and rural Thailand

Providers of HIV services and PLWH advocates

LSTM PhD student

Global Fund through Ministry of Health, Thailand

A rural district on the Thai–Burma border

Community members in groups by gender and age

Mahidol University, Thailand, with LSTM

British Council and DFID

2

3

4

PLWH experiences of accessing and adhering to ART26 TB patients, reflections on the TB diagnostic process

5

CHBC services for PLWH

6

Exploring PLWH perceptions of nutritional needs27

7

Community perceptions of GBV, links to HIV11,28

8

Perceptions and experiences of TB and patient-centred approaches to care Exploring patient experiences of different TB diagnostic approaches

9

10

Social assessment of TB control in China

11

Paediatric ART in Thailand—experiences and perceptions of health workers and care-givers29

12

Community perceptions and experiences of TB in Kanchanaburi, Thailand: a gender equity analysis

REACH Trust, Malawi, with LSTM

MoHSS, Namibia, University of Namibia, FANTA, with LSTM Liverpool VCT and Care, Kenya, with LSTM

USAID and FANTA

TROCAIRE, DFID through the HIV and AIDS Knowledge Programme European Union (INCO-DEV)

10

S. Theobald et al. / International Health 3 (2011) 7–14

Table 2 (Continued ) Study no.

Research focus

Key issues explored

Context

Sample

Research team

Funder

13

Voices and choices: PLWH reproductive health care

Urban and rural Thailand

Women living with HIV and their carers; men living with HIV

Mahidol University, Thailand

Ford Foundation

14

Exploring patient experiences of different TB diagnostic approaches and anti-TB treatment services

Experiences, dilemmas and choices relating to contraception, pregnancy and reproductive health Experience of TB diagnostics, perceptions of quality, support and barriers and enablers to adherence

One site in Yemen (urban)

Female and male TB patients

National Tuberculosis Institute, with LSTM

DFID and the Economic and Social Research Council, UK

CT: counselling and testing; ART: antiretroviral therapy; LSTM: Liverpool School of Tropical Medicine; PLWH: people living with HIV; DFID: Department for International Development; TB: tuberculosis; CHBC: community home-based care; HCW: healthcare workers; MoHSS: Ministry of Health and Social Services; FANTA: Food and Nutrition Technical Assistance; USAID: US Agency for International Development; GBV: gender-based violence; LATH: Liverpool Associates in Tropical Health.

For example, the research on gender, ethnicity and TB in Thailand (Study 12) used FGDs, in-depth interviews and group workshops as part of an action research cycle to develop interventions to support TB patients’ access to treatment. 2.2. Negotiating initial discussion with different communities and identifying strategic allies All researcher-authors stressed the importance of detailed planning and investing time in negotiating entry and discussion with different communities, developing allies and ensuring a joint understanding of the sampling frame and inclusion criteria. Many FGDs took place at health centres, where co-operation and buy-in from health staff was essential. The Malawian experiences (Studies 1–4) illustrate the importance of working in collaboration with Health Surveillance Assistants and other cadres at the hospital and clinic levels. Researcher-authors noted that in this context most participants had trust in these groups; health workers who knew patients’ TB and/or HIV status initially discussed their potential interest in participating in the studies, although informed consent processes were led by the researchers. Developing shared understanding of inclusion criteria is important: one researcher-author shared an experience when these criteria had not been understood and a woman was recruited by health staff when she was clearly not well enough to participate. Discussions with community leaders were also important, especially when recruitment for FGDs was negotiated through key players in the community and FGDs were held at community sites. In Burkina Faso (Study 8) this involved working closely with community leaders who have a certain moral authority within the community, such as priests, imams, chiefs and traditional healers. 2.3. Constructing your sample: is homogeneity in disease status possible and ethically acceptable? Researcher-authors’ experiences and insights on whether it was possible and appropriate to do FGDs with people living with HIV and/or TB was varied. In Thailand (Study 12), researchers explored the possibility

of recruiting TB patients for a FGD both at community and health facility levels. Neither proved possible: patients identified at community level were happy to participate in an in-depth interview but did not want to join a FGD, largely because of stigma and disclosure issues and also because they saw the logistical arrangements as being potentially problematic for their family members and carers. At the health facility level it was not feasible because of patients’ concerns and also those of health personnel who thought it would pose a health risk to the researcher and insisted that if the FGD went ahead the researcher should wear a mask, something she was not comfortable doing. In Studies 3, 4, 6, 9 and 14 it was possible to recruit TB patients or people living with HIV (PLWH) to participate in FGDs with health worker engagement and careful consideration of logistical issues and the informed consent process. Recruitment of participants infected by stigmatised diseases is sensitive and requires careful discussion with participants about disclosure and risks to confidentiality. In Thailand (Study 11), HIV-positive peer-support volunteers from the ART clinic took part in the same discussion as service providers, in a group that explicitly addressed HIV stigma. The FGD dynamic enabled constructive dialogue between two types of participant with different backgrounds and experience but a shared goal (to address stigma). The choice of an aerated room was stressed by researchers working on TB to minimise risks of crossinfection. In Malawi (Study 3), Namibia (Study 6), Ethiopia (Study 9) and Yemen (Study 14), FGDs were held with people living with HIV or TB at health facility level, where the nature of service provision meant that confidentiality was already compromised through joint waiting areas or health promotion activities. A major concern for PLWH participants of Study 3, which took place at the Lighthouse health facility in Lilongwe, Malawi, was whether they would lose their place in the queue to see the health worker if they participated in the FGD. The research team negotiated with the clinicians to ensure that FGD participants would be first in the queue once the FGD had finished; this in itself acted as an incentive to participation, which needs discussion in different contexts. In Namibia (Study 6), PLWH were recruited in the waiting area. It was agreed that a neighbour in the

S. Theobald et al. / International Health 3 (2011) 7–14

queue or a treatment supporter could knock on the door of the room where the FGD was taking place to call the participant if it was their turn to see the health worker. A potential pitfall of FGDs being held at health centres where participants receive treatment is that patients may feel reluctant to criticise services. In all cases researchers made it very clear in the informed consent procedure that the decision to participate and the airing of their views and experiences would have no affect on their ongoing health care. On the whole, researcher-authors felt that the group experience itself was empowering and gave individuals confidence to share experience and to criticise. This worked well in Malawi where direct criticism is rare—participants were more comfortable with being critical in a group than in individual interviews, despite the FGD being held in the health centre. FGDs were considered as a comfortable and familiar way of discussing issues, both in southern Africa and Asia. Researcher-authors also had experience of conducting FGDs with PLWH at the community level. In Study 5, participants were recruited through home-based care supporters and tended to include those who had disclosed their status and were active advocates of the rights of PLWH. There are dangers of this group being ‘over-researched’ and not representative of the many vulnerable PLWH who are not ready to disclose their status. 2.4. Constructing your sample: how to address power, hierarchy, gender and age? The norm in FGDs exploring perceptions and understanding of HIV or TB was to hold separate FGDs with younger women, older women, younger men and older men, with the expectation that homogeneity in age and gender would support participation and experience sharing.31 In Yemen, prevailing gender norms mean that women are expected to be accompanied by a male companion (husband, father or brother) in public spaces and when visiting a health centre.19 For example, when the research team in Study 14 invited women with TB to come to a FGD at the TB centre in Sana’a, they arrived with their male companions and children and at this point the research team negotiated with the women and their male companions to allow the women to come together into a private room so that they could speak freely with a female facilitator about their experiences. The discussion was successful and women spoke openly and critically about the levels of support they received from male and female members of their family. In all cases researcher-authors stressed the need for some flexibility in recruitment approaches and again the need to think on one’s feet and respond pragmatically to the unexpected. 2.5. Who should facilitate? Considering skills and context The importance of a skilled and appropriate facilitator was stressed by all researcher-authors. This is true of all FGDs, but was considered particularly so given the sensitivity of the subject matter. The moderator needs to have

11

excellent interpersonal skills: the ability to inspire confidence in others, be a good judge and interpreter of group dynamics, be empathetic but not judgemental, and to keep the discussion focused without driving it. Moderators also need to have a broad understanding of the issues at stake and the cultural context. Given varied and sometimes challenging group dynamics and the unexpected turns that FGDs can take, moderators need to be flexible. Researcherauthors shared experiences where FGD participants had inadvertently disclosed their status and this required much sensitivity in follow-up discussions. The role of cultural context was a recurring theme in researcher-author responses and meant that FGDs were a more challenging endeavour for moderators in some contexts than others. For example, in China researcher-authors explained that it is challenging for moderators to encourage the ‘free’ expression of ideas amongst participants who are cautious about voicing opinions in a society with constrained political freedom. In contrast, in Malawi people already use discussion to solve challenges facing their communities in rural and urban areas alike. Therefore, unlike one-to-one interviews that may raise suspicion, especially when members of the opposite sex sit together and talk ‘in the name of research’, FGDs are akin to normal life in these communities. The norm was for researchers from the host country to facilitate the FGD: this was appropriate in terms of language skills and cultural understanding. Even where a foreign researcher spoke the national language fluently (Study 11) it was seen as more appropriate for a Thai person to facilitate. In some contexts and research projects it was necessary to train facilitators who had no previous experience in conducting FGDs. When conducting training for health workers to do FGDs, researcher-authors noted the importance of stressing that they are not ‘knowledge tests’ or patient information sessions but an opportunity to listen and to understand perceptions, experiences and perspectives. The research team in Study 7 that explored community perceptions of gender-based violence, HIV and service delivery in three districts in Kenya made the strategic decision to train voluntary counselling and testing (VCT) counsellors to facilitate FGDs owing to concerns that participants might feel distressed during or post FGD participation. VCT counsellors with skills and experience in handling sensitive issues and situations were trained in the theory and practice of qualitative research and, where necessary, could deploy their counselling skills. 2.6. Thinking through ethics, informed consent and incentives Researcher-authors’ concerns relating to ethics were four-fold. First, some researchers faced challenges in the ways in which ethical concerns were constructed and prioritised in certain cultural contexts. For example, in China (Study 10), concerns related to ethics were rarely considered in-depth by large medical research teams, which led to constraints in thinking through how to ensure consent and confidentiality in a social and political context that does not highly value either of these. Others experienced challenges from ethics committees whose concerns on ethics

12

S. Theobald et al. / International Health 3 (2011) 7–14

of FGDs seemed to relate to the numbers of FGDs (‘is the sample sufficient?’) rather than the processes. Second were ethical considerations around the right to information. For example, what should researchers do if participants’ views were ‘wrong’ and could enhance their risk of contracting disease or restrict their ability to access treatment? This was addressed by taking leaflets about TB along to FGDs and distributing them afterwards, or by offering to answer questions after the FGD. A third challenge related to being realistic about maintaining confidentiality. Researcher-authors all stressed the importance of making it clear in the consent procedure the extent to which confidentiality could be maintained and reiterated the need for information to be kept confidential at the beginning and end of the FGDs. The latter was seen as particularly appropriate where participants has disclosed their status during the course of the FGD. A fourth and critical challenge related to the blurred boundary between incentives and inducements when working with vulnerable and impoverished groups. The norm was to provide a soft drink and/or a token of appreciation. But many researchers would have liked to have provided more than this and noted that expectations are high amongst all who participate in the research process, including respondents and those who assist with organising the FGDs. There is need for further debate about what is appropriate. 2.7. The focus group discussion itself: exploring pitfalls and opportunities Researchers were positive about the ways in which FGDs, with skilled facilitation, could be used to trigger indepth discussion about sensitive and challenging health issues. For example, in China (Study 10), FGDs enabled discussion of an issue where the first response of most participants is that they ‘know nothing’ (about the symptoms or causes of TB). An individual interview might have got no further, but in a FGD people were able to respond to any initial ideas offered by other participants (by agreeing or disagreeing), so that a range of ideas and perceptions were revealed. Malawian researcher-authors highlighted that FGDs have the ability to ignite debate that can bring out ‘bare reality’ rather than participants saying what they think researchers want to hear. Researchers also noted that some sensitive issues were raised in FGDs rather than in in-depth interviews and attributed this to the power of the group collective in triggering and enabling discussion of sensitive issues. For example, in Study 6 women’s concern over their financial dependence on men and the implications of their HIV status disclosure was discussed only in FGDs and did not emerge in in-depth interviews. Most researchers perceived that the group interaction engenders a confidence in individuals to discuss critically and openly issues, which is invaluable amongst participants who may well feel disempowered and in cultures where it is seen as inappropriate to complain. Other strategies to support discussion within FGDs included the use of third-person vignettes or tools adapted from Participatory Rural Appraisal or Participatory Learning and Action. Vignettes included the construction of a story about a poor rural woman who wanted to go to the

health facility to get advice on her prolonged cough. Participants were asked what barriers she might face; discussing in the third person was considered less threatening and particularly appropriate to triggering discussion in the ‘warm up’ stage of a FGD. Study 12 in Thailand used body mapping32 and asked participants to draw and share their ideas on TB risk factors and the impact of TB on the body. Researcher-authors’ experiences of pitfalls in FGDs usually related to challenging group dynamics or certain individuals. For example, in Study 11 in Thailand a senior doctor had strong pre-conceived ideas that appeared to limit participation from others. One researcher-author gave an example of the inclusion of a nurse participant in a FGD with PLWH; the other participants deferred to her presumed superior knowledge. In reality the nurse’s supposed ‘truths’ were largely untrue and if acted upon could have been detrimental to the health of PLWH participants. In this scenario the facilitator organised an immediate debrief post FGD to talk through some of the issues and counter the nurse’s inputs. This also illustrates the need for facilitators to be able to respond to the unexpected. 2.8. Translation, analysis and approaches to dissemination All FGD checklists were translated into the vernacular and, where complex concepts were discussed (such as stigma), the research teams explored the use of different wording and phrasing with people from relevant communities prior to commencing the FGDs. During FGDs, facilitators probed participants’ understanding of these concepts. Multicultural research teams had a variety of different approaches to translation, often contingent on money and resources. In some cases all transcripts were translated into English to enable those who did not speak the local language to participate in the analytical process; in others a subsection (10%) was translated, whilst in some the transcripts remained in the original language and members of the research team who were not able to read these played a role in commenting on and questioning the emerging analytical framework. Attention to the translation of complex concepts was seen as critical and in many cases the need to retain the original flavour and meaning of a particular concept necessitated much debate and consultation. For example, Chinese members of a research team (Study 10) were reluctant to label negative views about TB and TB patients as ‘stigma’ because the Chinese translation often used for this word has much stronger connotations of complete ostracism. Researcher-authors reported that when the researcher is involved in the actual facilitation of FGDs, it makes the notion of ‘analysis as an ongoing process’ easier because the researcher is not only immersed in the data but also the process, the atmosphere and the context. However, the researcher-authors also noted the importance of bringing those not involved in FGDs into the analysis as they may bring a different angle. Bringing together researchers with different perspectives (disciplines and cultures) was seen as positive in enhancing the trustworthiness of the analytical process. Involving clinicians and health workers in qualitative analysis was seen by most researcher-authors

S. Theobald et al. / International Health 3 (2011) 7–14

as a way to diversify the analytical perspectives brought to bear on the data and also had a positive effect in supporting ownership, and in many cases practical translation, of the findings. Most researchers tried to find ways to disseminate early analytical frameworks and findings back to participants but were often constrained by distance and budgets. In Study 8, male and female TB patients, members of the health centre management committee and community leaders came together in one meeting for feedback, heated discussion and consensus on the way forward. Similarly, in Study 5 stakeholders from the districts came together for a feedback meeting and had the opportunity to question and/or challenge each other based on the study findings. Thus, participant checking could also be used as an advocacy tool that gives participants the opportunity to air their experiences, taking advantage of the facilitation of the research team. Researcher-authors also noted that qualitative testimonies—the real voices and experiences on having HIV and TB and interacting with health systems—that emerged from FGDs were very powerful in policy dialogue. 3. Lessons learnt Our approach has shown that FGDs can provide a rich source of in-depth data on issues that are culturally constructed as sensitive. In the literature, some argue that FGDs facilitate discussion allowing less inhibited group members to take the lead in discussion.20 Whilst some participants can find the FGD experience empowering,20,31 others will fear the lack of confidentiality in the FGD and this may prohibit disclosure.23 Our analysis reveals that disclosure poses more difficulties than inhibited participation in most of the cultural contexts we work in, and this reflects other experiences of conducting FGDs on topics seen as sensitive or taboo (see Morgan22 and Kitzinger23 ). Where setting, recruitment strategy and facilitation of the FGD are appropriate, participants are willing to share their experiences even on difficult and stigmatised issues. A key lesson is that bringing together people living with HIV or TB takes careful planning, negotiation and buy-in. The need to think through in each context how to recruit, where to hold the FGD and how to encourage participants to be openly critical about service provision is important to enhance ethical standing, confidentiality and trustworthiness. Being aware of gender, age and hierarchical issues and how these interplay in different cultural contexts is critical in constructing approaches to sampling. Choice of appropriate facilitator is crucial—there are no blueprints for who the right facilitator is and this is dependent on context, the research question and the skill set needed. Our experience sharing reiterates the importance of being aware of and embedded within a culture and working to build ‘culturally competent practice’.23,33 Our reflections in African settings concur with findings from the indigenous North American context that some cultural groups embrace focus groups because of a strong oral tradition,34,35 whereas other participants such as those from China may find FGDs unfamiliar and uncomfortable.36 These group norms are particularly critical to consider when discussing sensitive subject areas. Facilitators must

13

be familiar with cultural norms for group interaction and how to probe for more information from participants appropriately.21 In some cases this may involve avoiding direct questioning by constructing stories or scenarios about a ‘friend or sister’. Some communities may find visual methods culturally familiar,37 and techniques such as body mapping may be successful in generating discussion,38 especially with sensitive topics. Our exercise also highlighted that one of the cornerstones of culturally competent practice is language. Facilitators must use accurate and culturally appropriate terms, especially when describing sensitive concepts. FGDs should be held in participants’ native language,39 and translation should concentrate on capturing meaning rather than a literal translation.40 The importance of working in collaboration with key stakeholders and seizing opportunities for feedback were stressed by all researcher-authors. Although challenging in practice owing to pressures of time and logistical issues, giving participants the opportunity to comment on emerging themes greatly enhances trustworthiness and builds the profile of qualitative research in the community. International health researchers from the north and south should seek to provide training and to build local research capacity for high-quality qualitative research, including FGD facilitation, in low-resource settings.41,42 Working across multinational and multiprofessional43 teams is challenging and is essential to engender knowledge exchange and mutual respect. Practically, in research on HIV and TB where clinical research can dominate, qualitative researchers need to seek opportunities to build the profile of qualitative research and stress the importance of patient and community voices and experiences in understanding health facility data and epidemiological patterns. 4. Conclusion Our collective experience has highlighted that when used appropriately FGDs can provide rich data on the voices and perspectives of people living with HIV and TB, community groups and health workers in resource-poor settings. This requires critical consideration of the different steps in the research cycle. Sharing lessons across different contexts and disciplines and reflexive practice on conducting such FGDs is important, especially given the limited space in many academic papers to discuss the methodological and ethical dilemmas of researching with marginalised and vulnerable groups in resource-poor settings. As international health researchers, we need to strive to ensure ethical and trustworthy approaches, build capacity so that there is a critical mass of researchers able to facilitate challenging FGDs in this area and develop innovative approaches for feedback and dialogue with PLWH, TB patients, the community, the health sector and beyond. Authors’ contributions: ST and PF conceptualised the paper and designed the approach; LN, IM, ASo, RT, GB, OT, ASa, MK, NK, YF, RA, NAl-A and NAl-S responded to the questions, contributed to the framework and commented on the paper. All authors read and approved the final manuscript. ST is guarantor of the paper.

14

S. Theobald et al. / International Health 3 (2011) 7–14

Acknowledgements: The authors would like to thank all of the collaborators and FGD participants from Thailand, China, Yemen, Malawi, Kenya, Burkina Faso, Ethiopia and Namibia. Funding: None Conflicts of interest: None declared. Ethical approval: Not required. References 1. Lazarus M. Tuberculosis–HIV co-infection: policy and epidemiology in 25 countries in the WHO European region. HIV Med 2008;9:406–14. 2. Corbett E, Marston B, Churchyard G, De Cock K. Tuberculosis in subSaharan Africa: opportunities, challenges, and change in the era of antiretroviral treatment. Lancet 2006;367:926–37. 3. UNAIDS. Country and regional responses to AIDS. Geneva: UNAIDS. http://www.unaids.org/en/CountryResponses/ [accessed 3 September 2010]. 4. WHO. Global health atlas. Geneva: World Health Organization; ©2003–2007. http://apps.who.int/globalatlas/dataQuery/ [accessed 3 September 2010]. 5. Rimal RN, Creel AH. Applying social marketing principles to understand the effects of the ‘Radio Diaries’ program in reducing HIV/AIDS stigma in Malawi. Health Mark Q 2008;25:119–46. 6. Somma D, Thomas BE, Karim F, Kemp J, Airas N, Auer C, et al. Gender and socio-cultural determinants of TB-related stigma in Bangladesh, India, Malawi and Colombia. Int J Tuberc Lung Dis 2008;12:856–66. 7. Turan JM, Miller S, Bukusi EA, Sande J, Cohen CR. HIV/AIDS and maternity care in Kenya: how fears of stigma and discrimination affect uptake and provision of labor and delivery services. AIDS Care 2008;20:938–45. 8. Waterman H, Griffiths J, Gellard L, O’Keefe C, Olang G, Ayuyo J, et al. Power brokering, empowering, and educating: the role of homebased care professionals in the reduction of HIV-related stigma in Kenya. Qual Health Res 2007;17:1028–39. 9. Smith RA, Nguyen LK. Searching for a ‘generalized social agent’ to predict Namibians’ intentions to prevent sexual transmission of HIV. AIDS Care 2008;20:235–43. 10. Thomas F. Stigma, fatigue and social breakdown: exploring the impacts of HIV/AIDS on patient and carer well-being in the Caprivi Region, Namibia. Soc Sci Med 2006;63:3174–87. 11. Sanou A, Dembele M, Theobald S, Macq J. Access and adhering to tuberculosis treatment: barriers faced by patients and communities in Burkina Faso. Int J Tuberc Lung Dis 2004;8:1–5. 12. Stephenson R. Community factors shaping HIV-related stigma among young people in three African countries. AIDS Care 2009;21:403–10. 13. Li X, Wang H, Williams A, He G. Stigma reported by people living with HIV in South Central China. J Assoc Nurses AIDS Care 2009;20:22–30. 14. Rao D, Angell B, Lam C, Corrigan P. Stigma in the workplace: employer attitudes about people with HIV in Beijing, Hong Kong, and Chicago. Soc Sci Med 2008;67:1541–9. 15. Apinundecha C, Laohasiriwong W, Cameron MP, Lim S. A community participation intervention to reduce HIV/AIDS stigma, Nakhon Ratchasima province, northeast Thailand. AIDS Care 2007;19:1157–65. 16. Kittikorn N, Street AF, Blackford J. Managing shame and stigma: case studies of female carers of people with AIDS in southern Thailand. Qual Health Res 2006;16:1286–301. 17. Sagbakken M, Frich JC, Bjune GA. Perception and management of tuberculosis symptoms in Addis Ababa, Ethiopia. Qual Health Res 2008;18:1356–66.

18. Admassu A. Coping with the challenges of AIDS: the experience of persons living with HIV/AIDS in Addis Ababa. Northeast Afr Stud 2000;7:81–100. 19. Date J, Okita K. Gender and literacy: factors related to diagnostic delay and unsuccessful treatment of tuberculosis in the mountainous area of Yemen. Int J Tuberc Lung Dis 2005;9:680–5. 20. Farquhar C. The analytical potential of ‘sensitive moments’ in focus group discussions. In: Barbour RS, Kitzinger J, editors. Developing focus group research: politics, theory and practice. London: Sage; 1999. p. 156–72. 21. Halcomb E, Gholizadeh L, DiGiacomo M, Phillips J, Davidson PM. Literature review: considerations in undertaking focus group research with culturally and linguistically diverse groups. J Clin Nurs 2007;16:1000–11. 22. Morgan DL. Successful focus groups: advancing the state of the art. London: Sage Publications Inc; 1993. 23. Kitzinger J. Qualitative research. Introducing focus groups. BMJ 1995;311:299–302. 24. Namakhoma I, Bongololo G, Bello G, Nyirenda L, Phoya A, Phiri S, et al. Negotiating multiple barriers: health workers’ access to counselling, testing and treatment in Malawi. AIDS Care 2010;22(Suppl 1):68–76. 25. Nyirenda L, Makwiza I, Bongololo G, Theobald S. A gender perspective on HIV treatment in Malawi: a multi-method approach. Gend Dev 2006;14:69–80. 26. Weigel R, Makwiza I, Nyirenda J, Chiunguzeni D, Phiri S, Theobald S. Supporting children to adhere to anti-retroviral therapy in urban Malawi: multi method insights. BMC Pediatr 2009;9:45. 27. Fergusson P, Chikaphupha K, Bongololo G, Makwiza I, Nyirenda L, Chinkhumba J, et al. Quality of care in nutritional rehabilitation in HIV-endemic Malawi: caregiver perspectives. Matern Child Nutr 2010;6:89–100. 28. Kilonzo N, Taegtmeyer M, Molyneux S, Kibaru J, Kamonji V, Theobald S. Engendering health sector responses to sexual violence and HIV in Kenya: results of a qualitative study. AIDS Care 2008;20:188–90. 29. Tulloch O. The treatment and care continuum in the Thai government paediatric antiretroviral programme [PhD thesis]. Liverpool: Liverpool School of Tropical Medicine, University of Liverpool; 2010. 30. Patton MQ. Qualitative evaluation and research methods. 2nd ed. London: Sage; 1990. 31. Krueger RA, Casey MA. Focus groups: a practical guide for applied research. Thousand Oaks, CA: Sage; 2000. 32. Cornwall A. Body mapping in health RRA/PRA. RRA Notes 1992;16:69–76. 33. Hughes D, DuMont K. Using focus groups to facilitate culturally anchored research. Am J Community Psychol 1993;21:775–806. 34. Strickland CJ. Conducting focus groups cross-culturally: experiences with Pacific northwest Indian people. Public Health Nurs 1999;16:190–7. 35. Berthelette G, Raftis Y, Henderson G. The sharing circle. A culturally appropriate format for a focus group? The Aboriginal Nurse 2001;16:17–8. 36. Willgerodt MA. Using focus groups to develop culturally relevant instruments. West J Nurs Res 2003;25:798–814. 37. Keller C, Fleury J, Perez A, Ainsworth B, Vaughan L. Using visual methods to uncover context. Qual Health Res 2008;18:428–36. 38. Ulin PR, Robinson ET, Tolley EE. Qualitative methods in public health: a field guide for applied research. San Francisco, CA: Jossey-Bass; 2005. 39. Barbour RS. Doing focus groups. London: Sage; 2007. 40. Esposito N. From meaning to meaning: the influence of translation techniques on non-English focus group research. Qual Health Res 2001;11:568–79. 41. Robinson N. The use of focus group methodology—with selected examples from sexual health research. J Adv Nurs 1999;29:905–13. 42. Lansang MA, Dennis R. Building capacity in health research in the developing world. Bull World Health Organ 2004;82:764–70. 43. O’Cathain A, Murphy E, Nicholl J. Multidisciplinary, interdisciplinary, or dysfunctional? Team working in mixed-methods research. Qual Health Res 2008;18:1574–85.