Stigma relating to tuberculosis infection prevention and control implementation in rural health facilities in South Africa — a qualitative study outlining opportunities for mitigation

Abstract

Background

Tuberculosis (TB) is a stigmatised disease with intersectional associations with poverty, HIV, transmission risk and mortality. The use of visible TB infection prevention and control (IPC) measures, such as masks or isolation, can contribute to stigma.

Methods

To explore stigma in this condition, we conducted in-depth individual interviews with 18 health workers and 15 patients in the rural Eastern Cape of South Africa using a semi-structured interview guide and narrative approach. We used reflexive thematic analysis guided by line-by-line coding. We then interpreted these key findings using Link and Phelan’s theoretical model of stigma, related this to stigma mitigation recommendations from participants and identified levels of intervention with the Health Stigma and Discrimination Framework.

Results

Participants shared narratives of how TB IPC measures can contribute to stigma, with some describing feeling ‘less than human’. We found TB IPC measures sometimes exacerbated stigma, for example through introducing physical isolation that became prolonged or through a mask marking the person out as being ill with TB. In this context, stigma emerged from the narrow definition of what mask-wearing symbolises, in contrast with broader uses of masks as a preventative measure. Patient and health workers had contrasting perspectives on the implications of TB IPC-related stigma, with patients focussing on communal benefit, while health workers focussed on the negative impact on the health worker-patient relationship. Participant recommendations to mitigate TB IPC-related stigma included comprehensive information on TB IPC measures, respectful communication between health workers and patients, shifting the focus of TB IPC messages to communal safety (which could draw on ubuntu, a humanist framework) and using universal IPC precautions instead of measures targeted at someone with infectious TB.

Conclusions

Health facilities may unwittingly perpetuate stigma through TB IPC implementation, but they also have the potential to reduce it. Evoking ‘ubuntu’ as an African humanist conceptual framework could provide a novel perspective to guide future TB IPC stigma mitigation interventions, including policy changes to universal IPC precautions.

Background

An estimated global total of 10.6 million people developed tuberculosis (TB) disease in 2021 [1]. Yet despite being common, TB remains a stigmatised disease with intersectional associations with poverty, social marginalisation, HIV, transmission risk and the risk of dying due to the disease [2]. Understanding TB stigma and developing mitigation strategies are important components of improving TB prevention and care.

Infection prevention and control (IPC) policies may contribute to actions that could be experienced as stigmatising, for example by isolating those who are believed to be infectious. Both the 1999 and 2019 WHO guidelines on TB IPC caution that stigma is an important implementation consideration, specially also mentioning its relevance to mask wearing [3, 4]. Viewing this through the Health Stigma and Discrimination Framework, this places a focus on healthcare facilities as important spaces where patients may have TB stigma experiences and where there may be observable stigma practices [5].

In a study based in Ghana, a number of examples of discrimination relating to TB IPC in healthcare facilities were identified [6]. Junior members of staff were delegated the care of TB patients, health workers avoided or shortened interactions with TB patients and health workers treated TB patients with contempt by shouting at them. Fear of becoming infected with TB was identified as one of the drivers of the stigma [6]. Pervasive TB stigma can lead to outcomes including being a barrier to seeking a TB diagnosis, hinder commencement of treatment and long-term poorer health outcomes [7].

We use Goffman’s seminal conceptualisation of stigma as a physical marking that, within the context of social relations and interaction, reveal the tainted moral status associated with the person [8]. Stigma can be further explored through considering the following: enacted stigma which is overt discrimination (for example someone with TB being dismissed from their job) and felt stigma, the sense of shame and a fear of encountering enacted stigma (for example someone with TB avoiding contact with other people as they fear they may react with disgust) [9]. This can be further differentiated as anticipated stigma (which may be perceived but may or may not become enacted stigma) and internalised stigma (feeling shame or blame for having TB disease) [9,10,11]. Despite receiving prominence in the policy and academic literature, there has been little progress in finding ways in which TB IPC-related stigma can be mitigated [6, 10, 11]. Instead, avoiding the stigma related to mask wearing by people with TB is sometimes cited as a reason for abandoning the use of masks in health facilities [12]. This contributes to the ongoing spread of TB to health workers, patients and visitors — affecting those with underlying health conditions more severely.

When looking at the literature beyond application to TB on the links between infection control measures and stigma, some possible mitigation strategies are identified. In a literature review on the stigmatisation of source isolation (people who have a transmittable infection), Gammon and colleagues found that isolation can be a barrier to contact time with health workers, and that protective equipment can impair communication [13]. They recommended additional training for health workers to mitigate the social and psychological impacts of isolation on patients through increasing social interaction between health workers and patients.

During the COVID-19 pandemic, in some settings, people with Asian heritage wearing masks were stigmatised and treated as carriers of disease [14, 15]. A multi-country meta-ethnography by Tsang and colleagues explored how positive social meanings attached to mask wearing contributed to higher uptake of masks during respiratory epidemics and sustained use outside of epidemic contexts, predominantly in countries in Asia [16]. They observed that positive social identities support the use of mask wearing including associations with solidarity, civic responsibility and scientific modernity. This shifted the meaning of mask wearing from being exclusively for the prevention of a specific disease (such as SARS) to that of offering broader protection against a range of respiratory illnesses, pollution and radiation [17, 18]. This contrasts with the requirement in health facilities that patients with TB wear masks to protect others but no requirement that people who do not have TB wear masks as protective to the wearer. It is likely that in some settings, due to widespread use of masks for COVID-19, mask wearing may have become less stigmatised for TB as well [19, 20].

Nyblade and colleagues suggested an array of intervention strategies to address stigma in health facilities. This included the following: providing information about the stigmatised condition, building skills through interactive learning activities and promoting empowerment, for example by offering counselling for the stigmatised group [7]. Nyblade and colleagues also recommended increasing contact with people who form part of the stigmatised group (for example through rotations in clinical roles for health workers or exposing staff to individuals with the stigmatised condition in non-clinic interactions) and using structural approaches to address stigma through reviewing policies and service integration [7].

Stangl and colleagues, through their Health, Stigma and Discrimination Framework, call on researchers who develop stigma interventions to ‘move away from psychological models that see stigma as a thing which individuals impose on others, and instead emphasize, the broader social, cultural, political and economic forces that structure stigma’ [5]. They present multiple levels that stigma interventions could focus on, highlighting that a combination of levels should be preferred.

In this study, we aimed to explore (a) the underlying contributors to why TB IPC measures in health facilities in the study context were perceived as stigmatising and (b) how the perspectives of health workers and patients differ regarding the implications of TB IPC associated stigma. (c) We then present participant recommendations for how the stigma associated with TB IPC could be mitigated or overcome.

Methods

Study setting

This qualitative research study was conducted in a rural district of the Eastern Cape of South Africa, the province with the highest TB incidence rate in the country at 692 per 100,000 people [21]. The district has an HIV prevalence of 15.3%, with people with HIV disproportionately affected by transmission of TB in health facilities [19, 22]. The district’s high tuberculosis burden is rooted in the contributions of migrant mining labour, racially discriminatory tuberculosis policies during the apartheid government, and a focus on TB treatment as the main intervention, to the neglect of improving social and economic conditions [20].

Data collection

Data sources include in-depth individual interviews with 18 health workers and 15 patients conducted between August 2019 and January 2020 which lasted between 30 and 90 min each. (see Table 1) We recruited participants at one district hospital and four primary care facilities. Sampling was purposive and was from waiting rooms in the health facilities and at staff meetings. We specifically looked to include health workers and patients with key perspectives such as health workers who previously developed occupational TB, patients who have TB currently or had TB before and patients who had had experiences of using a mask while visiting a health facility. It is important to note that some patients with experiences of TB were not asked to wear a mask (n = 3), and some patient participants who have not had TB also had experiences of being asked to wear a mask when visiting a health facility (n = 2); this was while being investigated for symptoms of TB. During this study period, the most commonly used type of mask by patients was surgical masks, while health workers used surgical masks or particulate filter respirators. Our sample size was guided by data saturation, which relates to the degree to which new data repeat what was expressed in previous data [23].

Table 1 Participant characteristics

We used a semi-structured interview guide with a narrative approach, which was refined based on pilot testing (see additional file 1 for our interview guide.) Interviews were conducted in isiXhosa or English (based on the participant’s preference), audio-recorded, transcribed, translated into English where applicable and checked for accuracy by the research team. Interviews were conducted at the preferred site of the participants, which included health facilities, the research centre and homes.

Data analysis

Initial analysis was inductive and grouped according to patient and health worker interviews. We were guided by Braun and Clarke’s reflexive thematic analysis, using line-by-line coding led by HvdW with NVivo software [24] (see additional file 2 for our coding tree). We then interpreted the thematic areas relating to TB IPC and stigma in health facilities using Link and Phelan’s theoretical model of stigma to identify the contributions of different stigma components and how this relates to mitigation recommendations and identify levels of intervention with the Health Stigma and Discrimination Framework [5, 25]. This study formed part of a broader programme of doctoral research which also includes organisation practices influencing TB IPC implementation and TB IPC in community settings.

Link and Phelan’s theoretical model of stigma draws on disability and sociological studies and has five co-occurring components [25]. Two stigma sub-concepts relate to the material properties of a mask and physical implications of TB IPC measures (as visible label of having TB and emphasising separation between ‘us and them’). Two stigma subcomponents relate to TB IPC measures reflecting the stigma associated with having TB more broadly (loss of status and discrimination and stereotyping of people with TB), and the final component considers the influence of power which can relate to stigma practices [26]. Link and Phelan argue for including a focus on how power is asserted through existing socioeconomic inequalities and through differences in class, race, religion or gender. In relation to TB, power can also be asserted through institutions, policies, TB IPC measures, diagnostics and treatment delivery [27]. Another example where power could be expressed is in the doctor-patient relationship, where Shei highlighted that this relational power depends on trust — which the patient trusts that the health worker will use the power asymmetry in an ethical way and make constructive changes in the lives of their patients while treating patients as rational, intelligent, fellow human beings [28].

Research team

The research team was led by HvdW, a clinician-researcher who had previously worked in this setting and who led the health worker interviews. A research assistant (N. S.) who lives in the area and has experiences of receiving care at the health facilities supported recruitment. She received training in qualitative research methods and led the isiXhosa interviews (with HvdW taking notes for most of these interviews). We kept detailed journals and had regular debriefing conversations about how our presence including gender (both being female) might influence the responses from participants. We piloted different approaches to conducting the interviews (for example where HvdW would not be present) and found that participants did not focus attention on her as note-taker. Co-investigators provided expertise in occupational health (RE), primary care (CB and TG), behavioural science (STC) and social science (TG) and supported critical discussions of the data collection processes and interpretation.

The reporting of this research has been done according to the COREQ checklist [29] (see additional file 3).

Results

We start by describing the stigma experiences and practices that our participants linked to TB IPC measures in this context. We then contrast health worker and patient perspectives on what implications these stigma experiences have for TB IPC implementation. Finally, we present patient and health worker recommendations on how TB IPC stigma could be mitigated.

Stigma experiences and practices relating to TB IPC in health facilities

Participants felt that mask wearing for source control (protecting others from becoming infected) was a practice that could often lead to someone with TB experiencing stigma.

When you look at a person [wearing a mask] you think to yourself that maybe this one has a huge illness which is bigger than mine. Even though a person doesn't eventually say anything, the eye speaks, to see how people see you. – participant 32, male MDR-TB patient, previously asked to wear a mask.

This description ‘the eye speaks’ shows how much of this stigma is based on the perceptions of others and how it could be communicated through a glance in a facility’s waiting room. It is an intersection between felt and enacted stigma — the person wearing a mask may anticipate and interpret certain body language from people in the waiting room as stigmatising without this being the intention, but the small gestures could also communicate a message that becomes enacted stigma. One participant described how she unobtrusively tried to cover her baby’s head with a blanket to protect them from TB when they saw someone who was wearing a mask in a waiting room. Two other participants described trying to move away from a masked person, one saying she will do it in a way not to ‘offend the person’ which we consider examples of enacted stigma.

This shows how in our study setting a mask was seen as a label of TB and by some as a label of MDR-TB, which through its shared symbolic meaning led to fear of infection being a driver of stigma:

[Should I be asked to wear a mask] means that I am infected—I have TB then.—Participant 8, female patient, no previous TB, not asked to wear a mask.

Yet whether a mask was a valid label for TB was questioned by some health workers, as some of the health facilities included in the study would provide masks for patients with respiratory symptoms while their TB investigations were pending. In our sample, there were also some patients who had TB who were not asked to wear a mask (see Table 1) or patients with MDR-TB who were expected to wear masks irrespective of whether they still posed an infection risk to others.

Isolation was also described as a practice that could lead to experiences of stigma. Patients felt this was both a physical separation but also communicated a reduction in their value, when they were asked to join a different group:

What made it hurtful is that I was taken away from other people and be placed alone there. Maybe I was placed with people who had this thing on [wearing a mask]. … I thought I should persevere. – participant 15, male previous TB patient, asked to wear a mask.

I did not get offended. Because I know the TB disease is contagious. It is a must that you agree when you are put aside because you will infect a lot of people. – participant 5. female patient previous TB, not asked to wear mask.

The emphasis that this participant places on communal safety is further explored in the “Results” section of this manuscript.

While improving natural ventilation was considered the least stigmatising TB IPC measure, some participants still felt that it could communicate a message of carelessness on the part of health workers:

[There are people] hurt by that. “They open [windows], we are cold, and they do not care about us”, they would say. I would say “It is so that the disease passes and gets out through that window and door because it will infect you if these doors and windows are closed, you will not be safe”.—participant 25, female patient, twice had previous TB, not asked to wear mask.

Some participants felt the stigma related to infection control measures should be viewed more widely, beyond the experience linked with infection control practices, to the broader experience of being a person who is ill. TB IPC is a painful moment, often at the beginning of a new illness experience, but also symbolises a broader experience of now having a different identity.

I was afraid of people in that moment they were giving the mask to me. I thought to myself that it means my illness is wrong. [But] my illness shouldn't affect other people, I thought I should endure the mask. What was making me feel embarrassed was continually coming to the hospital. Everyone seeing me going from the hospital carrying the pills. That causes so much embarrassment for me. But then I would think that there is nothing they [the people looking strangely at me] are going to do for me, my health is here [at the hospital].- participant 15, male patient, previous TB, asked to wear a mask.

This participant links his experience of receiving a mask to the shame of being seen to visit the health facility or carrying tablets — which is broader than being ashamed to have TB. He felt ashamed to be an ill person, because in this context for his age group and gender it has the intersectional stigma of being connotated to having HIV.

Patients and health workers think differently about the implications of TB IPC-associated stigma for health facilities

While in our dataset health workers and patients broadly agreed that TB IPC measures in health facilities can be stigmatising to patients, they had contrasting perspectives on the implications. In facilities where masks and respirators were commonly used, health workers perceived patients to oppose TB IPC measures, but that they ‘endured’ it as part of receiving care. It was interesting that none of the health workers we interviewed considered patients to value TB IPC for communal safety. In facilities where masks and respirators were not commonly used, health workers said making a patient feel ‘dehumanised’ through TB IPC measures contradicted the underlying values of empathy, respect and establishing a therapeutic relationship. One health worker described disregarding the use of respirators to show they care for patients above their own needs:

Patient care and outcome at the end of the day is more important than PPE. – participant 29. Mamele hospital, female doctor.

Health workers described being concerned that TB stigma would deter patients from coming for care at the facility in the future and would also cause patients who wear masks in facilities to be stigmatised in the community. In a different clinic, a health worker wearing a respirator as personal protective equipment was seen to be stigmatising to patients:

The [patient] asked me, "Sister, why are you all wearing a mask and we're not wearing them? Sister, are we disgusting you?” They felt like they were dirty. So [we] stopped wearing them. – participant 6, Aphiwe clinic, female nurse.

Yet in contrast with these health worker perceptions and despite their own strongly felt stigma, patients did not perceive stigma as a reason to discontinue mask wearing, instead of focussing on the protection offered by masks, emphasising that having TB is transient, and that the health of health workers is important. This was similar for patients who were asked to move to an isolation area — they mentioned trusting the health facility to make a recommendation in the communal best interest.

The overwhelming majority of patient participants described the use of masks as a way to protect other people but also the wearer. One participant who had undergone TB diagnostic tests multiple times saw wearing a mask as a process that she could initiate when arriving at a facility and did not view it to be stigmatising:

I would wear it without any problem. If I have a fever perhaps, when I walk in, let's say I am coughing, I ask for it as I get out of the nursing area, because I don't know what I have. I will know only when my results are back. I have no problem [wearing a mask]. – participant 14, female patient, no previous TB, asked to wear a mask.

These findings contrasted with health worker views that patients did not understand why masks were being used and who simply used masks when they were told to do so. This contrast is interesting, because many of the health workers and patients form part of the same community and share a similar cultural background. This suggests that there may be additional underlying differences in the way that health workers and patients conceptualise mask wearing. For example, one participant articulated both the benefit and discomfort:

[When I was asked to wear a mask] there was no problem because I know that I have a certain sickness so I should not infect the next person. When I am to speak to another person I have to wear it. I noticed that they were looking at me. It was their first time to see what I was wearing. I asked myself whether these people looked at me for what reason? I was given a mask and they were not given one, so I did feel like my heart was hurting, but a little. Not much. – Participant 21, male patient, currently has TB, asked to wear a mask.

While some health workers focussed on their duty to care for an individual patient, the patient participants emphasised collectivism and the importance of communal safety. Health workers viewed TB IPC as interfering with their work, instead of being part of their work.

On a descriptive level, these findings indicate a mismatch between what health workers perceive patients to feel and what patients described in the interviews. Yet to explore and understand stigma, it is also important to consider how power is exerted and how this may intersect with trust. Examples of power that we specifically looked for in our data set was the asymmetrical power relationship between health worker and patient (which relies on the health worker’s expertise to provide the care and services patient needs), through the power that medicine holds as institution (with medical jargon and training) or by looking at the contextual power of an organisation (a clinic in a rural area being the only facility that provides care). The most pertinent example was during our interviews, when asked for recommendations that could be presented to health workers, some patients described being unwilling to engage with health workers on TB IPC. They described this as being willing to ‘submit’ to the treatment they receive and suggested that it would be a sign of not trusting health workers’ expertise to provide their opinion.

Some of the health workers also mentioned that they felt the influence of race and educational privilege in this rural context on how willing patients are to communicate about TB IPC:

I'm the white doctor. I never get opposition. I never get resistance. I never get people that say "I refused. I'm not going to wear it [the mask]". I don't get people that challenge me. I don't get people that question me. Perhaps they want say something but they’re just too scared to say it. – participant 4, Mamele Hospital, female doctor.

However, many of the health workers come from the same community as the patients. One nurse recommended that through having more discussions about TB and TB IPC, it would be possible to reach a shared understanding and increase the motivation of health workers to make preventative efforts:

We can move and we can educate—we are from this communities—we live with these people. And we have information about TB, but [what is lacking is] a will. And to talk with them and make sense with them so that they can understand. Mostly people understand but they need motivation to start doing something to prevent TB. – participant 7, Aphiwe Clinic, male nurse, manager.

Encouraging communication between health workers and patients on TB IPC preferences may be one way to engage with these power dynamics and ensure both perspectives are heard.

Four recommendations from patients and health workers on mitigating the stigma associated with TB IPC in health facilities

We asked participants how the stigma associated with TB IPC in health facilities could be mitigated or overcome. We first present an overview of how participant recommendations link with the multiple levels of possible stigma interventions, based on the Health Stigma and Discrimination Framework in Fig. 1. We then elaborate on these recommendations in Table 2 and the text by providing illustrative quotes and identifying the stigma subcomponent that the recommendation focusses on, through drawing on Link and Phelan.

Fig. 1

Summary of participant recommendations for multi-level, TB IPC stigma intervention targets

Table 2 Patient and health worker recommendations for implementing TB IPC while mitigating TB stigma

An underlying principle of the recommendations presented in Table 2 is that destigmatising TB IPC does not involve omitting the preventative measures but rather paying closer attention to the way TB IPC measures are being communicated and used.

Several participants conceptualised this by asking to be spoken to ‘in a human way’. This contrasts with earlier descriptions of having TB, experiencing stigma and feeling like they were less than human.

What I would advise [health workers] is that you must be polite to people, because a person is a person through other people [a Nguni proverb, often used to describe the meaning of ‘ubuntu’]. So, maybe they are a nurse, but they work with people, if the people aren't there, there is nothing they will work with, so they must treat them well. – participant 20, male TB patient, previously asked to wear a mask.

This participant suggests looking at ‘ubuntu’ as a way to understand how TB IPC measures should be adapted to emphasise interconnectedness and a shared sense of humanity, which contrasts with the separation, discrimination and disempowerment that Link and Phelan described as subcomponents of stigma (discussed in Table 2). In addition to this, it also introduces the community as point of focus, where people’s wellbeing is interdependent and challenges the idea that the health facility is separate and should have interventions only targeted towards it.

Several participants pointed out that visible infection control measures, due to its symbolic meaning, disclose to others that a person has TB. One approach already outlined above would be to change the symbolic meaning of these measures. A different approach would be to create an environment where people are comfortable to disclose that they have TB, which would in turn reduce the stigma associated with TB, and visible TB IPC measures.

TB doesn’t only attack the poor [people]. [When I tell this to patients] they say, "No, sister. No." I say, "Yes, I'm the one who has [had] TB before, but now my TB is cured. I took treatment regularly and I'm well." That's why I counsel them because I just took that sickness from her to me. Patients must know that this is everyone's disease. – participant 13, Mamele Hospital, female nurse, previous TB.

This health worker voluntarily disclosed that she had TB to patients to challenge the stereotype that only poor people are affected by TB. This recommendation uses her position of power relating to the status that a health worker has in this context, based on their employment and education, to reframe conceptions of who gets TB.

Several participants pointed out that encouraging people with TB to voluntarily disclose their illness (instead of involuntary disclosure through cues like masks) could be uncomfortable and could lead to isolation or psychological harm. However, they felt that on a community level, it could contribute towards decreasing the stigma for other people affected by TB.

I’ve thought about it in terms of normalising. I sometimes feel very uncomfortable [disclosing sensitive information about myself and risk stigma], but I do it anyway. Because I think that’s what has to be done. So that’s why I say with TB [if I developed it], I would be open about it. With HIV I hope I would, for that exact same reason. – participant 29, hospital doctor.

There is an implied responsibility for the ‘receiver’ to respond tactfully and supportively, to meet the vulnerability of the person disclosing their illness. The benefit may not necessarily be directly for the individual disclosing their illness, although it might enable more support from family and friends:

What kills us sick people is that a person overthinks it, they think, “O God – maybe this person is disgusted by what I am.”– participant 18, female previous TB patient, not previously asked to wear mask.

The largest benefit would be for people from the same community affected by TB who may then face less stigma, as the conceptualisation of who falls ill is broadened and challenged. What is relevant to an infectious disease like TB is that disclosure may also make colleagues or family members aware of the potential risk of developing TB after exposure. ‘Ubuntu’ brings an interconnected, community-focused perspective (in contrast with an individual perspective) that could contribute to addressing some of the underlying causes of TB stigma.

Discussion

We had three main findings: TB IPC measures can both generate stigma experiences of patients but also reflect the stigma associated with the disease. Patient and health workers had contrasting perspectives on the implications of TB IPC-related stigma, with patients focussing on communal benefit, while health workers focussed on the negative impact on the health worker-patient relationship. Finally, recommendations to mitigate TB IPC-related stigma included comprehensive information, respectful communication, shifting the focus to communal safety and universal IPC precautions.

An important question when considering IPC-related stigma is as follows: where is the boundary between IPC in the interest of public health and an example of direct discrimination against someone who is ill? From a consequentialist philosophical perspective applied to TB, it would be when TB IPC stops serving a purpose, when it is futile and unnecessary. There were examples in our data where TB IPC measures caused harm when it evolved into broader social norms of engaging with TB patients in all settings, irrespective of whether they pose a risk of infection. TB IPC can in certain contexts be used to rationalise not only direct discrimination (through individual interactions) but also structural discrimination (which may be through social structures of institutional policies). An example of structural discrimination from our data was some of the health facilities having norms that expected MDR TB patients to wear masks irrespective of how long they have been on treatment and whether they are still infectious to other people. Guidelines that provide specific advice on the likely duration of infectiousness for people who have started TB treatment and whether and how TB IPC should be implemented in different settings (household, community and health facility) would make an important contribution and should include considerations for harm caused by prolonged isolation. See for example the National Tuberculosis Coalition of America’s recent guidelines on isolation in community settings that makes significant progress on this [26].

The differences in health worker and patient perspectives on TB IPC-related stigma suggest that stigma mitigation interventions will need to account for these power dynamics to ensure that both perspectives are engaged with. It would also be interesting to present the emphasis that patients place on communal safety to health workers and see whether this may be a different way to train health workers on IPC, instead of emphasising individual benefit through improving the safety of the health worker. Using participatory research methods for TB IPC stigma interventions may be supporting co-creating interventions that engage with power dynamics between health workers and patients [30].

Future work could focus on piloting stigma interventions based on the recommendations in Table 2 and should ideally focus on different levels of the Health, Stigma and Discrimination Framework as we outlined in Fig. 1. Two recommendations for mitigation (namely requests for detailed information and respectful communication) align with the work of Gammon and colleagues who advised that health workers should compensate for potentially stigmatising isolation practices with increased social contact with patients [13]. One recommendation focussing on policy (universal instead of targeted use of IPC measures) was made by our participants pre-pandemic and implemented as part of the broader COVID-19 pandemic response. This widespread use of masks has been viewed as having a positive impact on the stigma associated with mask wearing for TB [31]. Drawing on evidence from previous epidemics, the broader conceptualisation of the utility of mask wearing (for COVID-19, TB and other respiratory illnesses) may support ongoing use of masks and could also be applicable to other airborne IPC measures [16]. However, this needs to be researched further, as the connotations with COVID-19 may bring new symbolic meanings to mask wearing.

Existing research in South Africa has highlighted the need for TB stigma interventions to be adapted to local contexts — as there can be differences between cultural groups [32]. In this rural Eastern Cape context, ‘ubuntu’ as a conceptual framework place emphasis on shared humanity and collective wellbeing. The participant recommendation to focus the framing of TB IPC messages on community-level benefits and the use of principles of ‘Ubuntu’ warrants further exploration. There is a rich history both in the academic literature and in community contexts that could be drawn on.

‘Ubuntu’ could refer to a moral quality in a person but also an intellectual tradition that falls within African humanism [33, 34]. It can be defined as follows: ‘You are what you are because of other people. We don’t live in isolation; we live in a community. That sense of community is what makes you who you are, and if that community becomes broken, then you yourselves also become broken. And the restoration of that community, the healing of that community, cannot happen unless you contribute to the healing of it in a broader sense. I am because of others, in relationships with others. I am not an island of myself, I am part of the community, I am part of the greater group’ [33].

‘Ubuntu’ is a way to foreground African constructions of the world and has been used as a conceptual framework to inform diverse fields such as social work practice and African philosophy [35, 36]. It could engage with some of the deep-rooted contributors to TB stigma — such as feeling less than human. This shift towards collectivism is also supported by research during the COVID-19 pandemic that suggested public communication framing mask use as important for communal safety was more effective than messages describing it to be important for individual safety [37, 38].

Voluntary disclosure of TB, particularly of people who are seen to have higher status in a certain context, aligns with this community-focussed approach and could reduce stigma more broadly. There may also be insights from other illnesses that used this approach, for example the use of HIV-positive T-shirts (with former President Nelson Mandela wearing this) as part of advocacy for the Treatment Action Campaign [39]. As Stangl and colleagues point out, in the example of HIV, stigma experiences fuelled patient advocacy efforts, resilience and improved access to healthcare [5]. Further research should evaluate our recommendations for a stigma intervention and translating it to other contexts.

Our research had key strengths, in that the team collecting data combined insider perspectives (having lived and worked in the setting) with outsider perspectives (approaching the topic through a research lens) and a multidisciplinary team. Limitations were that we relied on self-reporting of TB IPC and stigma practices, since planned observational fieldwork was disrupted by the COVID-19 pandemic. We also acknowledge that interventions that only focus on TB IPC-related stigma in health facilities may not be sufficient to address the broader stigma associated with the disease in different contexts and recommend that IPC stigma interventions should be sensitive to this. Due to our recruitment strategy that was based at health facilities, our study did not include participants for whom TB IPC-related stigma may have entirely prevented health care engagement. The sense of community responsibility reported by our participants may therefore not represent the perspectives of people with TB who did not engage in care.

Conclusions

In the rural health facilities in South Africa we studied, TB IPC measures both reflected and contributed to the suffering caused by TB through stigma. Health facilities can play an important role in mitigating or perpetuating this stigma, and context-sensitive interventions should explore this further. We recommend focussing on providing comprehensive TB IPC information to patients, respectful communication that also pays attention to non-verbal cues, shifting the focus from individual benefit of IPC implementation to communal safety and using universal airborne IPC precautions instead of measures targeting potentially infectious patients. Evoking ‘ubuntu’ as an African humanist conceptual framework could provide a novel perspective to guide these stigma mitigation interventions.