Phase 1: Defining study scope and research questions

We sought to understand the design and impact of interventions aimed to reduce TB related stigma. We obtained input from stakeholders with lived experiences of TB as well as clinician researchers to formulate this question and to assist with identifying relevant research and operational examples.

Phase 2: Identifying relevant studies

An initial search of PubMed was undertaken followed by an analysis of the text words contained in the title and abstract of relevant articles, and of the index MESH terms used to describe the articles. Based on the initial search, we refined our search terms for TB and stigma using identified keywords and index terms (S1 Appendix) and undertook a revised search across the following databases (PubMed, EMBASE, Web of Science). All searches were conducted by a medical librarian and results imported into Covidence software (Fig 1) [18].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. PRISMA diagram illustrating study selection process.

https://doi.org/10.1371/journal.pgph.0000989.g001

Phase 3: Study selection

We considered any studies that implemented interventions to reduce the stigmatization of people with any form of TB (including drug resistant TB). We did not restrict by study design and included studies reporting on quantitative, qualitative or mixed data. We included published studies that assessed interventions that included reducing stigma alongside other clinical outcomes, and studies that assessed impact on stigma alone. We restricted to studies published in English without restriction by publication year.

Articles were screened for relevance based on the title and abstract. This was followed by a full-text review of articles identified from initial screening. At both stages of the screening process, articles were screened by a review group, with decisions regarding article inclusion made by two independent reviewers (IF, MG, WH, DW, HM). Disagreements were resolved through discussion with a third reviewer (RRN).

Phase 4: Data charting, quality appraisal and stakeholder consultation

Data was extracted from the included papers using a pre-piloted charting table. The charting table incorporated data on study authors and year of publication, country, health system context, study design, target population, sample size, details about how and when the intervention was delivered, analysis methods, primary study findings, and the impact of the intervention. Two reviewers (IF and MG) extracted data independently using NVivo software to organize the data. Any disagreements that arose between the reviewers were resolved through discussion with a third reviewer (RRN). The quality rating of all studies included was graded by two reviewers (IF and WH) using a modified version of the 2018 mixed-method quality appraisal tool (MMAT [19]). Upon answering screening questions for each study type (qualitative, quantitative, and non-randomized control), a total percentage was calculated. Studies which scored less than 50% were of low quality, between 51% and 70% average quality and above 70% high quality. An assessment of implementation strategy and outcomes was conducted using guidance from Proctor et al. [20].

To build on and contextualise our findings for discussion, we undertook a stakeholder consultation, as recommended in the Arksey and O’Malley framework [15], with TB survivors, advocates, policy makers and clinician researchers. Using a focus group format to facilitate knowledge exchange and group feedback, we shared preliminary findings of the review and asked them to discuss issues related to interventions to reduce stigma drawing on their expertise and experiences and to identify emerging priorities for research for such interventions.

Phase 5: Collation and summary of findings

Three reviewers (IF, MG and RRN) classified interventions aimed to reduce stigma by the socio-ecological level at which the intervention was targeted (individual, interpersonal, organizational, community, or policy level) [14]. Individual interventions targeted individual behaviour change among those with diagnosed with TB to modify individual characteristics such as knowledge, attitudes, and self-esteem. Interpersonal interventions aimed to modify the person living with TB’s environment such as their household, work environment and/or friendship networks by establishing or strengthening interpersonal relationships to restore or promote the person with TB’s health. Organizational interventions targeted organizational change to modify health and stigma-related aspects of an organization, such as through training programs or organizational policy. Community interventions aimed to increase knowledge related to TB and regarding stigma in specific community groups, and to increase community development skills, such as support network to facilitate access to services for people with TB. Public policy level interventions use national and local laws and policies to effect change [14] (Fig 2). We used recently developed guidance on TB stigma measurement designed with expert consensus input [21] to standardize the definitions used by the reviewers to classify stigma manifestations, rather than definitions used in individual studies to enable comparison across studies. For example, in one study the phrase ‘I stay away from people… to avoid being rejected’ had been classified as internal stigma [22]. However, as per the stigma manifestation classification used in this review [21] this indicator was reclassified as a measure of anticipated stigma.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Potential stigma reduction interventions targeting stigma drivers and/or manifestations that could be implemented at different socio-ecological levels.

https://doi.org/10.1371/journal.pgph.0000989.g002

Characteristics of included studies

Studies focused on the following populations: people with drug-sensitive TB (n = 4), people with multidrug-resistant (MDR)-TB (n = 1), health care workers (n = 3) and community members (n = 1). All studies were conducted in the pilot stage of implementation. Studies focusing specifically on TB patients and their families were implemented at either the diagnosis (n = 1) or the treatment (n = 4) stage of the TB care cascade. Interventions are classified according to the socio-ecological level at which they were targeted as follows: (1) individual or interpersonal (n = 3), (2) interpersonal (n = 2), organizational (n = 3), and community (n = 1). All studies were conducted in lower- and middle-income countries. The characteristics of included studies and findings related to stigma are in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of included studies, organized by socio-ecological level.

https://doi.org/10.1371/journal.pgph.0000989.t001

Quality assessment

All included articles underwent methodological quality appraisal using the MMAT. One study that was a randomized controlled trial (although stigma was not a primary outcome) scored 100%. However, the majority of studies (range 28–100%) were judged to be low quality, often due to lack of coherence between qualitative data sources, analysis and interpretation and high risk of non-response bias in quantitative studies. No studies were excluded based on quality.

Stigma measurement

There was considerable heterogeneity in the approaches to TB stigma measurement (Table 1). Four studies used quantitative indicators to compare changes in stigma pre and post intervention [22–25]. Of these studies, two differentiated indicators based on stigma manifestations [22, 25], however these studies classified these indicators into stigma manifestations that were not aligned with the definitions of the stigma manifestations used in this review [13] so these were reclassified. Three studies gave no definition of stigma nor the drivers or domain(s) that were being measured [23–25].

The remaining five studies used qualitative data [26–30], four of which did not incorporate any specific analysis approach or framework and focused on the use of quotes. One qualitative study conducted thematic analysis [27].

Implementation details evaluated

Detailed descriptions of interventions and indicators of implementation effectiveness (distinct from clinical effectiveness) should be included to enable comprehensive evaluation of interventional studies [20, 31, 32]. We assessed the included studies using Proctor et al.’s recommendations for naming, defining and operationalizing implementation strategies across seven categories: actor, action, action targets, temporality, dose, implementation outcomes addressed and theoretical justification [31] (S3 Appendix). Assessment of implementation outcomes included the following measures: acceptability, adoption, appropriateness, cost, feasibility, fidelity, penetration and sustainability [20]. All studies were implemented in the pilot study phase. No studies adopted any formal theoretical frameworks that are used for implementation research [33]. Appropriateness and feasibility were the most frequently reported implementation outcomes (S3 Appendix). Studies varied in the degree of description of the intervention. Some included intervention descriptions that were sparse and lacked key details that are needed for systematic evaluation and replication [23, 25, 29, 34]. Few interventions documented how they were tailored to different contexts. Those that did, conducted an operational analysis and tailored interventions based on findings and the judgement of local health teams [22]. One intervention described modifications during study implementation for example, when nurses were used to mitigate a shortage of psychiatrists to conduct support groups in Peru [26]. Although only one intervention was at the community-level and used community volunteers as part of the implementation team [23], several others used community actors such as household counselors [25], or CHWs to deliver the intervention and local leaders to provide support [27].

Combined individual and interpersonal level interventions

No included studies focused solely on targeting stigma experienced at the individual level. Three studies involved interventional components that targeted both individual behavior and behavior of familial and/or social networks of participants. This was done with the aim of reducing anticipated or internal stigma. All three interventions used support groups (often termed TB clubs consisting of people with TB and a peer or clinician facilitator) [22, 26, 27] at the individual level to equip TB patients with mechanisms to counteract negative beliefs. Similarly, all three interventions also incorporated interpersonal components to the intervention including home visits [22], family workshops [26], and additional support from community leaders [27]. Stigma reduction was the primary intervention aim in only one identified study [22]. The other three studies primarily aimed to assess the impact of an intervention on clinical outcomes: adherence [26] and treatment completion [26, 27] with impact on stigma measured as a secondary outcome.

Interpersonal level interventions

Two studies intervened at the interpersonal levels of the affected person’s household or social networks with the aim of reducing stigma experienced by those living with TB by addressing drivers of stigma as well as internal and enacted stigma manifestations. This included reducing blame associated with the disease [25, 28] and misunderstanding about the biomedical causes of the disease [25].

Organizational level interventions

Three interventions were implemented at the organizational level, and were focused on health workers. These studies employed a range of information-based tools, including visual messaging, participatory theatre, and training, to address the drivers of stigma within health care settings.

Community level interventions

One intervention at the community level aimed to reduce stigma by increasing knowledge through an educational campaign delivered to community members regarding TB disease or reducing stigmatizing behaviors [23].

Barriers to implementation

TB related stigma was found to be highly contextual, as demonstrated after the prevalence of internal stigma after the same household counselling intervention applied to two study sites in South Africa and Zambia was reduced in one setting but increased in the other, although the confidence intervals were wide [25]. This was suggested to be due to heterogeneity in community sites between Zambia and South Africa in terms of local contextual factors. Recruiting and appropriately training intervention facilitators and addressing mental health in people with TB were identified as important barriers across several studies. When psychosocial support groups were used to reduce stigma in Peru, it was challenging to finding trained facilitators that did not show prejudice or fear of MDR-TB patients. However, despite having little or no training, nurses who served as co-therapists with the trained psychiatrists were able to develop therapeutic strategies to facilitate interventions in the groups and to lead a group when one psychiatrist had to withdraw [26]. In the community based study in which enacted stigma increased after an awareness campaign, the authors hypothesized that the community volunteers who administered both the campaign and survey could have passed on discriminating views of their own to participants [23]. One of the studies implementing support groups targeted towards reducing individual level stigma found that at baseline, some participants had suicidal ideations and/or depression (Box 1) [26]. Analysis of these barriers highlights the need to ensure high-quality training for staff delivering interventions in studies and moreover to programmatic staff delivering TB care, and for integrated mental health services to be available for people with TB, in addition to interventions targeting stigma.

Stakeholder consultation

This review was conducted as part of a community-engaged research study to evaluate TB stigma using community-based stigma assessments in South Africa and use these findings to develop stigma reduction interventions. To complement and contextualize the results, a stakeholder consultation was conducted in the form of a group discussion with policy makers, TB programme specialists, former TB survivors and clinical researchers, all with relevant expertise and experience related to TB management in South Africa. Their opinions closely reflected the findings of the review: more research must be done to highlight at which point(s) of the cascade of care, with whom, and through what implementation mechanisms interventions should be targeted. The lack of interventions at the policy level was identified as a significant research gap. While national documents detailing strategies to end TB stigma exist in countries such as India, stigma remains all too pervasive in practice settings [40]. A TB programme specialist mentioned that the law should be utilized to a greater degree to ensure that the rights of those living with TB are protected from stigma at the organizational and policy level.

Stakeholders noted that despite growing attention of the need to address stigma, misunderstanding of TB and risk of infection among community members remains widespread. They suggested a greater emphasis on highlighting the universality of TB risk to the general public rather than only focusing on high-risk groups such as people living with HIV, to minimize the exacerbation or transfer of additional stigma. Additional actors both within the healthcare system and in the community should also be engaged to reduce stigma, who may include traditional healers and church leaders who often overlook key entry points to the TB cascade of care. Interventions that equip these groups with tools to mitigate stigma drivers may have great potential to reduce stigma manifestations. Stakeholders also reiterated the importance of ensuring access to both counselling and quality mental health services for people affected by TB.

Strengths of our scoping review included conducting a systematic search of TB stigma reduction interventions literature, that was not limited by country setting or date, in consultation with a research librarian, followed by screening, extraction and evidence grading by at least two independent reviewers. By including only peer-reviewed articles, ensuring that our eligibility criteria were rigorously applied, and performing methodological quality appraisal at the individual study using the MMAT tool, we have endeavored for this scoping review to uphold similar quality standards to a systematic review and maximised our ability to evaluate the impact of stigma reduction interventions. Through the stakeholder consultation that we undertook as part of this review, we were able to incorporate key perspectives from people directly affected by TB. Our study was limited by the inconsistent classification of stigma manifestations and the interpretation of these findings. A further limitation of this review is holding the stakeholder consultation only within the South African context. While the TB burden in South Africa is one of the highest globally, findings may not be applicable to other high incidence settings. Further, since we restricted our search to published literature, grey literature of relevance may have been missed in this review. However, through the stakeholder consultation, relevant international policy documents were identified.