Background
Voluntary counselling and testing (VCT) for HIV first evolved in Western settings, with a strong focus on coping and support mechanisms for those diagnosed HIV-positive, as well as on the promotion of behaviours to lower the risk of onward transmission or acquisition of HIV [
1]. While there is a paucity of evidence to guide the theoretical constructs upon which HIV prevention counselling advice is expected to act, many individual-level models of behaviour change focus on aspects of individual agency such as intentions, attitudes, beliefs and feelings of self-efficacy regarding sexual risk reduction [
2]. However, social and cultural norms, as well as structural factors such as poverty, are likely to be important determinants of behaviour in African and other settings where identity is strongly influenced by family, community and other social factors [
3].
As Western HIV testing norms have been imported into African settings, this has been accompanied by assumptions that, as in Western settings, counselling advice can help to encourage reductions in high risk sexual behaviours such as multiple concurrent partnerships or increases in the frequency and consistency of condom use [
4,
5]. The sharing of personal information by the client with the counsellor is seen as essential to the counselling process, as is the joint identification of risk behaviours and possible strategies to reduce them [
1,
6]. A number of studies have highlighted some of the challenges associated with the provision of HIV risk reduction counselling in African settings, including difficulties experienced by counsellors when trying to balance the contrasting aims of VCT as a public health intervention with specific goals, but also an individually targeted counselling intervention which aims to respect client autonomy and choice [
7]. Other challenges include difficulties adapting standard HIV testing guidelines to local settings [
8,
9], and difficulties and discomfort experienced by counsellors when talking with clients about sexuality and sexual behaviour [
10,
11]. Quantitative evidence to support the role of HIV testing and counselling (HTC) in sexual behaviour change in sub-Saharan Africa is weak, particularly among HIV negative individuals [
12‐
14]. While a few qualitative studies have explored clients’ perspectives of and responses to HIV prevention counselling messages in African settings, many of these have focussed on HIV-positive individuals [
15‐
17] and few have included individuals testing HIV-negative [
18,
19].
In order to better understand the processes through which HTC may contribute to sexual behaviour change and HIV prevention in a rural area of northwest Tanzania, we conducted a qualitative study to explore healthcare worker, community and both HIV-negative and HIV-positive clients’ perceptions of HIV prevention counselling advice provided at VCT or at an antenatal provider-initiated testing and counselling (PITC) service (antenatal PITC has a greater focus on diagnosis and referral of HIV-positive patients compared to VCT, but still includes counselling on HIV risk reduction regardless of HIV-status [
20]).
Methods
Study setting and design
This study was carried out within the context of a community HIV cohort study in Kisesa in rural northwest Tanzania, where regular rounds of demographic and HIV serological surveillance (sero-surveys) have been ongoing since 1994 [
21]. HIV prevalence in the study area was estimated at 6.5 % in 2010, and most residents are subsistence farmers. VCT services were first provided as part of the 2003–2004 sero-survey, with further provision of VCT during three additional sero-survey rounds between 2006 and 2013. A permanent VCT clinic opened within the study area’s main health centre in 2005, while PITC has been routinely offered to pregnant women attending the health centre’s antenatal clinic (ANC) since the end of 2008. Temporary outreach VCT services are occasionally provided by governmental or non-governmental organisations. Although they have improved over time, rates of HIV testing uptake in Kisesa remain lower than the national average, with coverage of testing being approximately 40 % among men and 50 % among women in 2010 [
22], compared to 47 % among men and 62 % among women in the country as a whole in 2011/2012 [
23].
Qualitative data collection took place between February and April 2012. Nine group activities incorporating participatory learning and action (PLA) activities were conducted in order to gain general community perspectives of HTC services, as well as to serve as a source of recruitment for in-depth interviews (IDIs). Thirty IDIs with HIV-negative and HIV-positive service users explored individual experiences of completing HTC, while five IDIs were carried out with nurses or counsellors offering HTC in order to explore provider perspectives. The IDIs with service users and healthcare workers provided the richest source of information in terms of understanding responses to HIV prevention counselling messages, and so the majority of findings reported here are drawn from these.
Sampling, recruitment and data collection
PLA activities were carried out with four female and five male groups of six to twelve participants, stratified by rural or urban residence and age-group (18–34 or 35–60 years). A combination of purposive sampling and snowball techniques were used. Firstly, four to six individuals were purposively selected from a sampling frame constructed from the demographic and sero-surveillance datasets, which included some ‘seeded’ individuals who were known or reported to have used HTC services, as well as some who had not used HTC. The ‘seeded’ focus group approach has previously been used in this setting [
24], and allowed us to include HTC clients in the group activities without inadvertently revealing their HTC-use status during recruitment or data collection activities. Next, each purposively sampled individual was asked to invite a neighbour or friend from their village from the same age group to participate. In total 48 purposively sampled men and women were invited to participate, of whom 40 attended and 32 brought a neighbour or friend, resulting in a total of 72 participants (41 men and 31 women). The PLA activities included brainstorming activities using flip-chart paper to express ideas and opinions relating to the perceived usefulness of HTC services, positive and negative aspects of service provision, and opinions relating to counsellors.
IDIs were carried out with 15 men and 15 women who had used HTC services, 10 of whom were HIV-positive (five men and five women). The majority of interviewees (n = 25) were recruited via the group activities. At the end of the activity, each participant was asked privately whether s/he was willing to discuss previous use of HTC. If the participant agreed and was an HTC user, s/he was invited to participate in an IDI on experiences of going for HTC. The discussion guide included topics on motivations for going for HTC, impressions of counsellors, information and advice shared with and received from counsellors, attitudes and feelings regarding the counselling advice received and regarding sexual risk reduction, and experiences of disclosing HTC use to partners or other family members. Five HIV-positive individuals were recruited for IDIs via the HIV care and treatment clinic (CTC) at Kisesa health centre in order to recruit individuals who may and may not have accessed antiretroviral therapy.
Five IDIs were carried out with healthcare workers, purposively sampled to include individuals providing VCT (n = 2), PITC at the ANC (n = 2), and one member of the District Health Management Team who provided training, supervision and support for counsellors. Topic guides for healthcare workers providing HTC services included length of time and experience providing HTC, the content and information shared during pre and post-test counselling sessions, community perceptions of HTC, and levels of training and support received. The IDI with the HTC trainer covered issues relating to the content of counsellor training programmes, as well as ongoing professional development, support and supervision for counsellors.
All group activities and IDIs were conducted in Swahili by same gender researchers, with the exception of the IDI with the HTC trainer (a woman who was interviewed by a man). Study activities were audio-recorded following informed consent from participants, and then transcribed and translated into English.
Data analysis
Coding and analysis of transcripts was carried out in Nvivo10, guided by a framework approach [
25]. After familiarisation with the data, an initial coding frame was developed based on the research questions, but also drawing on concepts derived from theories of behaviour change. Codes were also generated inductively, grounded in the accounts of study participants. The initial coding frame developed by the lead author (CC) was reviewed by co-authors (AW, SL and JW) and revised according to emerging concepts, which were grouped under overarching themes. The refined coding-frame was then applied to all transcripts. Data were then synthesized, distilling the coding to explore the emerging patterns and meanings in terms of sexual behaviour change and HIV-prevention, and to provide explanations for findings.
Discussion
Our study highlighted several barriers, influenced by social and contextual factors, which seemed to limit plans or intentions regarding sexual behaviour change following HTC, including limited communication between clients and counsellors during counselling sessions, and inequitable gender-power relations.
Previous studies in sub-Saharan Africa have highlighted counsellors’ difficulties talking about sexuality or intimate matters during HTC, for example because the topic was embarrassing or difficult to talk about [
10,
11], and it is likely that clients share similar concerns. In some African cultures there are prescribed social norms about who should be consulted for advice regarding sexual matters, and involving an outsider in family problems may go against traditional practice [
26]. Such findings may help to explain why candid discussions between clients and counsellors appeared difficult for many participants in our study, and indicate that counselling approaches which include techniques to put clients and counsellors at ease when talking about sexuality would be useful.
Previous qualitative research in Tanzania and elsewhere in sub-Saharan Africa has noted the dominant role of men within relationships [
18,
27], and that men’s sexual respectability is often linked to greater sexual activity and the conquest of multiple partners, in contrast to women’s sexual respectability which is characterised by obedience, faithfulness and subservience to male partners [
28,
29]. One study in Tanzania reported that men may exaggerate their numbers of sexual partners, while women were more self-effacing and tended to under-report non-marital partnerships [
30]. These findings may help to explain why men in our study appeared more at ease discussing their sexual behaviour during HTC compared to women, and why men felt more empowered to act upon counselling advice.
A hierarchical relationship between healthcare provider and ‘patient’ was seen to inhibit communication between both male and female clients and counsellors during HTC, and other studies in sub-Saharan Africa have noted similar hierarchical relationships [
9,
31,
32]. Lie et al. [
33] note that HIV counselling guidelines tend to be prescriptive and directive by endorsing preferred behaviours (e.g. using condoms, remaining faithful to one partner), in contrast to more traditional person-centred counselling approaches which have a greater focus on client autonomy. As also observed in this study, counsellors seem to face challenges when trying to balance the contrasting aims of HTC as both a public health intervention with specific goals (i.e. to encourage sexual risk reduction), but one that is simultaneously meant to respect individual clients’ choices [
7]. Training courses and HIV counselling guidelines which place greater emphasis on maintaining a neutral stance, and on techniques for engendering more equitable client-counsellor interactions, may help to improve the outcome of HTC in terms of sexual risk reduction. For example, one study in South Africa noted that counsellors who attempted to engage clients by ‘making appeals’ rather than ‘prescribing rules for living’ seemed to be more successful in gaining clients’ engagement and commitment to sexual risk reduction [
7].
Research has suggested that in sub-Saharan Africa, family and community are more central to the construction of health and well-being than the individual [
3,
9]. In support of this, one study reported that HIV-positive Ugandan men and women’s main motivations in preventing HIV transmission were altruistic wishes to protect sexual partners and to avoid orphaning children [
34]. In African settings, HIV training modules and counselling messages which focus on broader culturally relevant altruistic messages relating to partners, children and the wider community, for example emphasising the benefits of protecting unborn children from HIV, or of maintaining the integrity and health of the family unit, may be more successful than ones which focus solely on individual interest [
34].
Although studies have reported that counsellors sometimes struggle to balance confidentiality requirements with what they may perceive as being in the best interest of the patient or the community at large (for example, by disclosing an HIV-positive individual’s status to a partner or other family members without the patient’s consent, in order that the partner may protect himself or herself from HIV, or that the family may provide care and support for the patient) [
8,
35], other studies have shown that confidentiality is of paramount importance for clients [
9,
33]. Although most participants in our study reported that they felt counsellors were trustworthy, even small fears relating to breaches of confidentiality may have a disproportionate impact upon clients’ willingness to share sensitive information. Thus counsellor training modules should cover the importance of maintaining client confidentiality in sufficient detail, while clinic buildings and infrastructure should be sufficiently private so as to put clients (and counsellors) at ease.
As highlighted by counsellors in this study, additional aspects of counsellor training which require attention include the depth in which topics are covered, and the availability of refresher training courses and supervision and support for counsellors. Furthermore, it was also apparent that the risk reduction strategies proposed by counsellors were often limited to only include condom use or abstention, and tailored discussions that covered a fuller range of available, relevant HIV prevention tools (e.g. HIV testing for partners, treatment for partners who test positive, testing and treatment for sexually transmitted infections, and eventually, pre-exposure prophylaxis) which might represent more realistic prevention methods for some women, should also be covered in counsellors’ training courses.
A number of clients reported finding the process of learning their HIV status reassuring in itself, regardless of counselling messages received. However, this knowledge alone seemed unlikely to lead to reductions in sexual risk behaviour. Several participants reported the desire to know their HIV status as their main motivation for going for HTC, and although counselling sessions were not observed as part of this study, it is possible that a strong focus of the counselling advice was on the different possible HIV test results and how they should be interpreted, rather than on sexual risk reduction. Previous studies have suggested that ‘de-linking’ testing and counselling service provision might help clients to focus on counselling messages and thus more effectively support behaviour change [
19,
36], but this hypothesis remains untested.
A relatively small number of clients reported having gone for testing together with their partner, however some expressed a desire to do so. The few individuals who reported having gone for couples-HTC reported positive experiences, including that couples might get ‘motivated’ together. Although previous studies have shown that couples-counselling is challenging for counsellors, particularly in the case of sero-discordant results [
37,
38], others have reported the approach to be beneficial because discussions between husband and wife can be mediated by a third party, and because the process can help to renew commitment to the relationship by both members of the couple [
39]. Much of the focus of VCT in sub-Saharan Africa to date has been on the provision of an individual service, however couples-counselling likely represents a promising option in terms of increasing the uptake of HTC services as well as helping to encourage behavioural change [
40].
A limitation of this study was that HTC users may have been reluctant to share details of what was discussed with counsellors with study researchers, due to the sensitive nature of the subject matter, and due a similar hierarchical relationship between participants and researchers to that observed between clients and counsellors. In order to minimise this bias, interviewing techniques to build rapport and trust with participants were addressed during fieldworker training. In order to maximise confidentiality, all IDIs were conducted either in private rooms near Kisesa Health Centre, or in the case of rural residents, in private locations near rural dispensaries. For confidentiality reasons, individual counselling sessions were not observed as part of this study. Had we been able to do so, we might have gained more a more in-depth understanding of the nature of the interactions between clients and counsellors during counselling sessions, and of the topics covered.
Acknowledgements
We would like to thank the study participants, as well as the clinicians and healthcare workers at Kisesa Health Centre, without whom this research would not have been possible. Thanks are also due to the research assistants involved in data collection (Ray Nsigaye, Joyce Chuwa, Grace Bulugu, Mpyanjo Chagu) and the wider staff at the National Institute of Medical Research in Mwanza who are involved in running the Kisesa cohort study.