Background and approach

This study was conceptualized jointly by the US and Botswana-based team. It emerged from an ongoing TB genomic epidemiology study in Botswana led by CM and SS that began in 2020 stemming from a collaboration started in 2013 [30–34]. CM has also led programmatic TB and HIV work in Botswana since 2008. In 2021, building on the existing study infrastructure, we received supplemental funding to conduct research about bioethical issues involved in uses of NG-WGS for TB research and disease control.

In 2022, we conducted a qualitative study using semi-structured interviews and deliberative dialogues with key TB stakeholders in Greater Gaborone health district, Botswana–which includes the country’s capital and largest city. The recruitment period began on May 2^nd, 2022, and data collection ended on November 17^th, 2022. We aimed to understand how participants viewed the benefits, risks, policy challenges, ethical dilemmas, and practical considerations that ought to accompany a scale-up of NG-WGS for TB in Botswana, with relevance for global stakeholders [35–38]. In addition to assessing the acceptability of implementing NG-WGS for TB among stakeholders in Botswana, we sought to “socialize” the technology among participants, educating them about NG-WGS to learn about their understanding of the tool to inform a hypothetical future scale-up [39]. Our design drew on frameworks from Science and Technology Studies (STS), anticipatory governance, and empirical bioethics. We situated our inquiry within the constructivist paradigm of STS, which treats technologies and health interventions not as pre-constituted things, but as complex socio-technical assemblages enacted as much by the functions of a technology as by the socio-political dynamics, people, infrastructures, policy frameworks, and processes that shape implementation in a specific setting [39–41]. From empirical bioethics, we took a two-pronged approach. On the one hand, we foregrounded ethical issues using etic, deductive, or “top-down” frameworks that emphasized bioethical principles (e.g., autonomy, justice, nonmaleficence, beneficence) and ethical concerns related to pathogen genomics regardless of implementation context (e.g., publication, data sharing, consent, revealing transmission dynamics). On the other hand, we also utilized an emic, inductive, or “bottom-up” approach that emphasized participants’ responses and local conditions to identify ethical issues particular to the context in Botswana or which arose from our context but carry more generalizable implications [27, 37, 38, 42]. We drew on anticipatory governance by building our study around what we perceived to be a likely policy future, in which genomic sequencing becomes part of routine TB control in LMICs [6, 26, 35, 36, 43, 44]. Indeed, this anticipated future is somewhat materializing with WHO’s 2023 recommendations for tNGS to become part of routine TB control [3]. While our study focused on stakeholders’ views on NG-WGS, our findings are relevant to tNGS, owing to the technologies’ similar functions (e.g., diagnosing drug resistance, use for hotspot mapping).

Protection of human subjects and ethics statement

Human subjects research approvals were obtained from the University of California, Irvine Institutional Review Board (IRB), the University of Texas Medical Branch IRB via reliance, and the Botswana Health Research and Development Committee. Participants provided signed consent and were given 50 pulas (~4 USD) per research engagement. Participants could provide consent on forms written in English or Setswana, the official languages spoken in Botswana.

Participant characteristics, sampling, recruitment, and data collection

Participants were Greater Gaborone-based TB stakeholders purposively sampled and recruited from three key groups: (1) TB policy stakeholders (e.g., from non-governmental organizations, district health management team (DHMTs), and public health officials); (2) TB community stakeholders (e.g., TB survivor-“champions” and other community-based advocates); and (3) TB clinical and research stakeholders (e.g., physicians, nurses, and research staff). Participants were recruited by Victus Global Botswana Organisation (VGBO) staff (CC, SB, and OM) through direct outreach to individuals, organizations, and facilities based on the team’s deep contextual knowledge. Individuals who were enrolled in the parent TB genomic transmission study or who worked for the parent TB study were excluded from recruitment. VGBO is made up of individuals with longstanding ties to the TB care, prevention, and advocacy community in Botswana, and is the host organization for a TB advocacy network. Recruitment outreach included seeking permissions from DHMT offices and related entities, following procedures that are standard in the country, appropriate within Botswana’s regulatory frameworks and cultural norms, and part of our approved protocol [37, 38]. While all participants lived or worked in the Greater Gaborone health district, many had experience working across the country or in roles with a national scope, and our questions were about a hypothetical full-country implementation of TB NG-WGS. Our findings thus provide insight into the views of TB stakeholders in Botswana as a whole, not only those who live in or near the capital.

The study had two branches, with the first involving 30 semi-structured interviews and the second involving four deliberative dialogues with a range of 10 and 22 participants each; the study had 48 total participants (see Table 1). Individuals could participate in either or both branches, including in multiple dialogues. Almost all interviews were conducted by SB at the VGBO offices, in healthcare facilities, and locations suggested by participants (e.g., their offices). Several interviews were conducted virtually. Deliberative dialogues are a focus group-like method designed to allow stakeholders to air out thoughts and feelings about a topic, with the facilitator guiding the group toward articulating areas of consensus and disagreement [45–47]. Each dialogue was structured to bring together different groups within the study population, culminating in the final dialogue with participants from all stakeholder groups (see Table 2). SB facilitated dialogues with assistance from OM, CC, BK, and CM.

Download:

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

Table 1. Participant characteristics.

https://doi.org/10.1371/journal.pgph.0002479.t001

Download:

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

Table 2. Deliberative dialogue branch structure.

https://doi.org/10.1371/journal.pgph.0002479.t002

Interview guides, deliberative dialogue facilitation guides, and all other study materials were developed collaboratively by US- and Botswana-based team members. Participants could speak in either English or Setswana, with any additional translation assistance provided by the bilingual interviewer (SB) or dialogue facilitation team (SB with assistance from BK, OG, and CM). Interview and deliberative dialogue guides were organized around first assessing participants’ knowledge, and then moving through a discussion of key topics around benefits and risks of the technology such as hotspot and cluster mapping, uses of NG-WGS for TB contact tracing investigations, consent for data uses, data security, the identification of drug resistant types of TB, sharing clinical NG-WGS data hypothetically collected by the national TB program with researchers, as well as transparency and engagement during implementation. The guides are included as (see S1 and S2 Files).

One key feature of our study was an animated video series developed by the team that was shown to participants prior to interviews and dialogues, to educate them about NG-WGS for TB. RP led the conceptualization and creation of the videos, drawing on the fields of graphic medicine and graphic public health [48]. The eight videos told the story of a fictional family in Gaborone that experienced TB diagnosis, treatment, and contact tracing by community health workers (CHWs) that was informed by NG-WGS. The videos also showed scenes of ministry epidemiologists debating whether NG-WGS data ought to be published to show TB hotspot or cluster maps in Gaborone. The full series was shown to participants before interviews. Before dialogues, a condensed version was shown to assembled participants. The videos were designed to help communicate complex details about the information that NG-WGS can reveal, with the aim of opening conversations about ethical issues and implementation challenges.

Rather than providing possible answers or posing questions in a yes/no or either/or style, the videos explained various uses of NG-WGS along with potential ethical issues, and presented situations in which the technology might be used. The aim was to demonstrate potential socio-medical situations and ethical dilemmas that could result from a hypothetical future scale up. For example, in the videos, NG-WGS shows that the husband-and-wife characters are infected with different kinds of TB–drug-susceptible TB and multidrug-resistant TB (MDR-TB)–thus requiring different treatments. Fig 1 below provides brief summaries of the content, purpose, and plot of each video. During development, videos were tested with VGBO personnel for comprehension. Versions in English and Setswana were developed using local voiceover talent. In interviews and dialogues, SB would refer to the videos to guide and enhance discussions with participants; who would also often independently refer to the videos. A methodological paper reporting details about the video production process and participants’ reactions will be published elsewhere, with a focus on implications of our approach for the fields of medical humanities, empirical bioethics, and graphic public health. The video series is provided as (see S1 Text).

Download:

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

Fig 1. Video summaries.

https://doi.org/10.1371/journal.pgph.0002479.g001

The interview phase took place before the deliberative dialogue phase. At the end of interviews, participants were asked if they would be interested in taking part in a dialogue; this recruitment method was part of our approved protocol and a question included on the consent form. Most interviewees were enthusiastic about participation in a dialogue and agreed to be re-contacted. Additional dialogue participants were recruited who had not been interviewed, to ensure that each had an adequate number and range of participants. The dialogues thus involved a mixture of people familiar with the study and individuals who had not been interviewed. This was intentional and led to a variety of perspectives by people with differing levels of familiarity with TB NG-WGS, but who had a common baseline of information provided by the videos. In both interviews and deliberative dialogues, the interviewer and main facilitator (SB) worked to take a neutral position on the question of whether NG-WGS for TB should be scaled up, noting that the purpose of the study was not to advocate for this, but to understand stakeholders’ views.

The process of interviewing participants and talking about interviews and new transcripts in weekly team meetings led to refinements of the interview guide and interviewing approach. Data from the interviews informed the structure of the dialogues. At the start of the fourth and final dialogue, SB showed the videos and presented a slideshow on high-level points of consensus and disagreement about NG-WGS based on data from the interviews and the first three dialogues.

Data analysis

English transcripts were generated by VGBO staff as data were collected and were analyzed using reflexive thematic analysis in a combined inductive and deductive framework [49, 50]. SM conducted coding using Atlas.Ti. He first generated an initial list of deductive codes driven by known areas of ethical debate or interest in TB genomic epidemiology and topics that came up in the interviews and dialogues. This initial codebook was circulated to the team and further developed by the group over email and during several team meetings. Inductive codes were added iteratively by SM as coding proceeded and were workshopped in team meetings. Interview transcripts were coded first, followed by deliberative dialogue transcripts. Additional deductive codes from Felton et al.–which pertained to rhetorical strategies used in deliberation–were added before coding deliberative dialogues, to better capture consensus, disagreement, and strategies of rhetorical persuasion used by participants [51]. While intercoder reliability is not supported within the qualitative paradigm of reflexive thematic analysis, coded transcripts in Atlas.Ti were shared and discussed with study team members (SB and VS) as coding proceeded, and the process was a recurring topic in weekly team meetings and one-on-one conversations. Discussions about the coding process, the content of interviews and dialogues, and nascent themes were a substantial focus of team meetings during the data collection, data analysis, and writing phases. This process of coding, sharing, and open discussion within the team was the foundation of the reflexive thematic analysis.

The study team also held a dissemination session with TB stakeholders in Gaborone on March 29^th, 2023, after data collection had ended, coding had concluded, and initial findings had been workshopped within the team. The half-day event involved sharing findings that were precursors to the themes reported in this paper in a slideshow and handout. The brief presentation was followed by full-group discussion about key issues in the study and participants’ thoughts on next steps for TB NG-WGS in Botswana. The conversation involved both English and Setswana discussion, led by SM, SB, OG, BK, and CM. Attendees included participants and other TB and health stakeholders in Botswana. The session was followed by a lunch where conversation continued. While the dissemination session did not involve data collection, feedback from attendees informed ongoing data analysis and future plans.

Botswana TB stakeholders broadly supported scaling up NG-WGS, provided that the country engages in robust planning, capacity-building, and education

Participants voiced their overall support for scaling up NG-WGS for TB in Botswana, owing to many perceived benefits. Benefits frequently discussed included the identification and treatment of drug-resistant TB, revealing transmission hotspots, and assisting in contact tracing investigations by showing clusters. Many participants were very enthusiastic, with one in the second dialogue saying emphatically “Yes, we want it–like, yesterday!”

Similarly, a TB champion (survivor-advocate) spoke to the promise of NG-WGS diagnostics for care and prevention:

It would make a massive difference because these very people that we help…will be able to access their results much quicker and it would also help them with regard to contact tracing and identifying the people they have been in contact with much faster. (P01)

There was near-universal support for scaling up TB NG-WGS among participants. However, this support was frequently qualified with assertions that doing so should only take place alongside proper planning, capacity-building, community engagement, and other initiatives to ensure that the technology could be used robustly and sustainably. One TB clinician noted that:

I think it’s a good test to do nowadays. My little concern was with the cost implications. But other than that, with capacity, I think it could actually help for proper diagnosis as mentioned…If there is something that can actually get us on the right track of medications to treat, the better, rather than doing guesswork. (P28)

Table 3 provides more statements that reflect the range of reasons and conditions that participants articulated for supporting a scale-up of NG-WGS for TB in Botswana.

Download:

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

Table 3. Participants expressed broad-but-conditional support for a scale-up of NG-WGS for TB.

https://doi.org/10.1371/journal.pgph.0002479.t003

In sum, TB stakeholders in Botswana expressed broad support for a scale-up of NG-WGS. However, participants also recognized the complexity of this proposal and offered support that was contingent on the country taking a robust and well-considered approach to implementation. Participants’ comments in this regard spanned discussions of costs, proper resourcing, ongoing training, updating guidelines, undertaking community and stakeholder engagement, informing patients about NG-WGS, and a wide range of other issues. In the following themes, we explore how participants discussed these and other topics against the backdrop of their broad support for scaling up NG-WGS for TB in Botswana.

Efforts to scale up TB NG-WGS must learn from past experiences, such as the introduction of other molecular diagnostics into Botswana’s TB program

When discussing the potential scale-up of NG-WGS, many participants invoked past attempts to improve Botswana’s national TB system with novel diagnostic technologies. They argued that lessons learned from previous initiatives should guide any scale-up of NG-WGS for TB. One reference point that came up repeatedly was the introduction of the GeneXpert molecular diagnostic system in the 2010s, which is used for the diagnosis of TB and rifampin resistance (first-line TB treatment) [34, 52]. While participants spoke about GeneXpert as a general success and as being much faster than culture-based TB drug sensitivity testing, discussions of GeneXpert sometimes referred to lacking funding and resources to ensure its sustainability. Stories about GeneXpert often functioned as lessons to inform a potential NG-WGS scale-up.

One civil society advocate who had previously worked for the national TB program spoke about their ongoing advocacy to resource the GeneXpert technology, which reflects other participants’ comments about the tool:

I had to lobby with the bosses of the Ministry of Health that ‘[GeneXpert] is going to work but we need money for maintenance.’ That is where we have a problem because even now, people still call me and say, ‘the reagents are finished,’ you see…It was helping people. When we went to install it in [a village clinic], people were very happy. (P05)

While reflecting on a complex clinical case that involved a use of GeneXpert that apparently did not identify drug resistance in a patient, P06 –a nurse and TB focal person–noted that NG-WGS could have been helpful because NG-WGS can identify types of drug resistance that GeneXpert cannot [13, 14]. They said “Before we would use GeneXpert. If it was [after an NG-WGS scale-up] and she came back positive, we could have sequenced and if she came back resistant…we would know that [she had] MDR [multidrug-resistant TB]” (P06). Quotations in Table 4 show how participants invoked the Botswana experience with GeneXpert when discussing an NG-WGS scale-up.

Download:

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

Table 4. Participants noted that Botswana’s experience with GeneXpert could inform a potential TB NG-WGS scale-up.

https://doi.org/10.1371/journal.pgph.0002479.t004

Partly because of the experience with GeneXpert, participants were acutely aware that if a costly and complex new TB diagnostic technology was introduced into the public health system without adequate resourcing, it might not contribute to a successful response. Scaling up NG-WGS for TB thus appeared contingent upon making these sorts of monetary and logistical investments to ensure long-term sustainability and benefits.

Public health and research re-uses of TB NG-WGS data were generally supported, even without specific consent, and informing patients about how their data are used was seen as important

Participants were generally supportive of the national TB program using patients’ clinical TB NG-WGS data for routine public health purposes and of sharing de-identified data with researchers for secondary analyses. For most participants, both forms of data use were still supported after it was explained that people receiving TB treatment are not afforded a chance to consent to re-uses of their data for either public health surveillance or retrospective analyses by researchers. Per P06 (a nurse), who held this view, “I don’t have any power [at the ministry]. But I don’t personally see any problem with [the ministry] taking information. This is done in an attempt to improve the situation‬. So, there is no problem…Everyone [who works with TB program data] will have their own code of conduct until the information eventually helps us improve the situation. I don’t believe that there are any major challenges.” Consent was one focus of our study because its relationship to public health practice is an enduring topic in bioethics and has been an area of controversy in molecular HIV surveillance [25, 28, 42, 53]. While lacking consent affordances for routine uses of patient data in TB public health programs were not a concern among Botswana TB stakeholders, some participants did express skepticism about sharing TB program data with researchers without consent. For example, P20 (a civil society advocate) discussed the possibility of introducing “broad consent” into Botswana’s healthcare system to govern future uses of patients’ TB data for both routine public health uses and secondary research purposes, but also noted that broad consent “is quite a new thing” and would pose feasibility challenges.

In dialogue with the issue of consent, numerous participants said that TB patients should be informed about how their clinical data are used by the public health system. Per one TB clinician, “we should put all over the walls of the clinic that ‘in the interest of public health, all data that is collected here is anonymously saved and analyzed to help you‬.’ I think they can be informed by that” (P21). Similarly, P22 –a TB survivor–stated that patients should be informed about how their data are used for public health: “as someone is getting data from me, they should be telling me that my data might be used for this and this and that.” When the interviewer made it clear that public health uses of patients’ TB data and research re-uses are done without specific consent, this participant said that “when things get tough and there is a national crisis, they should use the data; when it’s things that really have to do with a national concern and national security, they should use the data. The Ministry of Health will see how to go about it.” These and other similar statements reflect a general attitude of trust toward state institutions held by most Botswana citizens stemming from effective statecraft and development policies following independence from colonization in 1966 [37, 54–56]. Even so, participants held mixed views about whether people in Botswana trust the country’s healthcare system, with P13, a policy stakeholder in government, saying that “it’s 50–50.”

Support for re-uses of TB patients’ clinical data by public health authorities without specific consent, along with informing patients about how their data could be used, were topically linked areas of consensus. However, as with any consensus, there was not universal agreement. For example, P14 (a public health policy stakeholder in government) took the position that asking patients for their consent for secondary research re-uses of their reportable TB data should be part of the process of enrolling a patient into care, but that routine public health uses of patients’ clinical TB data should not require consent. P14 noted their belief that retrospective studies of public health data by researchers were “not the same as we do [at the ministry]” in TB disease control. When asked about potential data quality ramifications of requiring TB patients to give specific consent for secondary research re-uses of their public health data, P14 argued that ‬‬“if you see a large number of people refusing, that means we need to improve our education system–that means people are not educated and they are not aware of the benefits of submitting that data‬.”‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬

Implicit in P14’s position is that patients would give consent to secondary research re-uses of their reportable public health data if benefits to society were explained to them. While that position would need to be verified through further studies, P14’s support for patient education is aligned with our overall finding that informing patients about how their TB data could be used for public health programs and research purposes was supported across our sample. Table 5 shows the varied views participants held about how, where, and when Botswana’s health system should seek consent or educate patients about uses of their data.

Download:

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

Table 5. Participants supported public health uses of TB data without specific consent, with some reservations.

https://doi.org/10.1371/journal.pgph.0002479.t005

In sum, statements by participants reflected a consensus that public health uses of individuals’ clinical data often do not allow for specific consent. This view generally extended to secondary uses of de-identified TB program data for retrospective research, but not for all participants. One remedy offered for lacking consent affordances was that information be given to TB patients about how their clinical data could be used to support Botswana’s TB response.

Ongoing community and stakeholder engagement along with education for patients, healthcare staff, lab workers, and the public should be part of any TB NG-WGS scale-up

Participants stressed that community and stakeholder engagement–along with educational programs aimed at the TB workforce and the public–should be central to any scale-up of TB NG-WGS in Botswana. These efforts were seen as initiatives that should begin prior to scale-up, continue during implementation, and then become ongoing. There was consensus that education and engagement ought not be limited to consultations prior to implementation or merely aimed at securing stakeholders’ support. Rather, these efforts should be central activities and seen as crucial to successful implementation.

One participant who worked in a lab that handles TB samples emphasized the need for health workforce training. They said “Obviously, it is a new thing…so people will have to be thoroughly trained. Not just the testing staff but everyone involved, because there is a new kid on the block. So, introduce it” (P19). P02, a research staff person, also noted that clinical staff “have to be taught; we shouldn’t take it for granted just because they are health care workers‬‬–‘so, sequencing is here!’–then we just stop there…Let’s train them so that they know what sequencing is, how is it going to benefit patients, and how should they be working around it as health care workers so that once it comes, they are prepared.‬” P26, a TB program coordinator and nurse, connected the issue of health workforce education for NG-WGS to patient rights and education:‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬

Health care workers are also ignorant about the rights of the patients. We need to teach them faster. There is community TB care in the guidelines; it has the patient charter in there–the rights, right to privacy, right to information, right to confidentiality, etc.…Those should also be emphasized for each and every patient who starts on treatment: being taught about TB, the basics of TB, counselling should be given to them about expectations from the day of diagnosis until the end of treatment.‬‬

Participants also emphasized that educational efforts ought to include sustained and coordinated public communications, community engagement, and stakeholder engagement. A statement by one TB survivor-advocate captures these overall sentiments:

I think [the government] should contact the relevant stakeholders…the affected communities, the community leaders, the chiefs [kgosis], and parliamentarians. They have to spread the message and inform us about the technology they will be introducing. So, we have to engage the parliamentarians, call kgotla meetings and inform the public of the government’s intention to introduce such a technology–to explain to them that they have been used in detecting COVID-19 but want to now start using it for tuberculosis…‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬

They should consult stakeholders…spread the message through radio or social media so that this message can reach a larger population…engage journalists, the print media. They should be taught about this thing and be sensitized about it and then publish articles in their newspapers. I think that could also help. (P01)

Other comments about public education and communication included anti-TB stigma messaging and working with existing civil society organizations to enhance government-led efforts. Two participants in the first dialogue suggested forming a new civil society organization that would play a role in education related to a scale-up of NG-WGS. One noted that “as a civil society person, I would advise the TB survivors or the professionals to form a very‬, very strong advocacy group so that we can speak with one voice‬.‬” When discussing the logistics of a scale-up, P07, a TB nurse and program coordinator, noted that “I think even the community could be informed; they can be informed that TB nowadays is tested this way. Those are the resources I would need to equip for training, for facilitation, and to teach people.”‬ Table 6 shows other statements by participants emphasizing the need for ongoing forms of public engagement and patient education, including engaging kgotla meetings (traditional community gatherings designed to support local governance and decision-making) and kgosis (hereditary chiefs).‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬

Download:

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

Table 6. Participants advocated for ongoing engagement and education about TB NG-WGS in the event of a scale-up.

https://doi.org/10.1371/journal.pgph.0002479.t006

Broad and ongoing stakeholder engagement, patient education, and healthcare workforce education were roundly supported. This reflects emerging best practices in global and community-based health practice. However, participants’ recommendations also reflect the distinctive makeup of Botswana’s political system, which combines elements of traditional culture such as kgosis, kgotla meetings, and generally-trusted state institutions, including a comparatively robust healthcare system [37, 54, 55]. These entities, along with media organizations and advocacy groups, were all seen as playing different-but-vital roles in an NG-WGS scale-up.

Scaling up TB NG-WGS in Botswana should involve deliberations about ethical norms for publishing data, particularly regarding transmission hotspots, clusters, and stigmatized groups disproportionately impacted by TB

There was general agreement that introducing NG-WGS into Botswana’s TB program should be accompanied by assessing some of the ethical norms that govern how the country uses TB public health data. Participants spoke about this regarding data uses for disease control, such as for diagnosis and contact tracing. However, issues about ethical data use were most pronounced around how TB NG-WGS data ought to be published. P22 (a TB survivor) noted that widely disseminating TB NG-WGS data might not be beneficial and could “instill fear in people.” They then said that “maybe [the Ministry of Health] can just tell the part of the city that is affected. Maybe they should just go to the kgotla and inform them, ask them to do tests and sequencing and what not‬.” Others supported widely publishing data, with P16 (a community health worker) noting that “I think that if the community is aware, they can also help fight the disease, because if we try to keep certain things secret, things will only get worse‬.” In sum, participants acknowledged the need for new ethical norms regarding the publication and dissemination of TB NG-WGS data; however, no clear position emerged about what those norms ought to be.‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬

To this point, there was a lack of consensus regarding the ethics of publication practices regarding TB NG-WGS hotspot maps and cluster maps, two common ways of displaying genomic TB data that were shown to participants in the animated videos. Hotspot maps generally display geographic areas where transmission is concentrated (e.g., a city district), whereas cluster maps show networks of individual people represented by dots connected to each other by lines showing molecular or non-molecular (i.e., behavioral) epidemiological connections.

Some participants did not see clear ethical distinctions between these different ways of displaying TB genomic data, whereas others saw cluster maps as being higher risk for potentially identifying individuals. There were also concerns that publishing either kind of map could stigmatize neighborhoods or subgroups affected by TB that are already disenfranchised by further associating them with TB (e.g., poor districts, migrant workers, people living with HIV). P17, who worked in civil society and community healthcare, said that sensitive TB NG-WGS data about subgroups should “only be shared to platforms at such levels as the ministry…but not necessarily to the community, [kept] in the space of healthcare workers‬.” P17 took this position “so that we can have specific interventions that target certain groups of people, whether they are sex workers, whether its [men who have sex with men], whether its drug users,” and that “it would be ok to share this information within the health space, for those that are going to be assisting in programming. But in terms of [sharing information with] the community, it can bring about some discrimination.‬”‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬

Others saw benefits to broadly publishing TB NG-WGS data, arguing that doing so could contribute to TB prevention education. Per one TB policy stakeholder who supported the publication of hotspot and cluster maps:

If that area has been labelled as a TB hotspot, and those people really know the modes of transmission of TB, and they are taught how to prevent from contracting TB, then they are going to practice whatever it is that is needed to prevent TB infection. So, it will not be stigmatization…of course the risks are not outweighed by the advantages. (P14)

On the other hand, P08 (a TB clinician) had a mixed view, noting that “Information is empowering, right? Information is important, decisions should be made with all possible information.” P08 then added that “information can also breed biases and negativity and unwanted, unforeseen consequences…You don’t mean for people to judge you or your area for being a hotspot.” Further, in the words of one participant in the first dialogue who advocated sharing data with “community leaders” and “champions” before deciding whether to share specific maps publicly: “I always say that when it comes to sharing, do so with caution because for some, you are chasing away their people.‬”

Table 7 shows other participants’ reflections and statements about potential benefits or pitfalls of publishing TB NG-WGS data that could highlight disproportionately affected areas or populations. The quotations show the ambivalence of participants on the question of when or how hotspot or cluster data ought to be shared, and their rationalizations. The exchange presented at the bottom from the second dialogue dramatizes the extent of participants’ mixed and uncertain positions about the benefits of publishing cluster maps.

Download:

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

Table 7. Participants debated the ethics of publishing TB NG-WGS data.

https://doi.org/10.1371/journal.pgph.0002479.t007

In sum, there was no clear agreement on the exact kinds of TB NG-WGS information that should be published in surveillance reports, prevention messaging, or research articles. Views varied widely on issues related to sharing hotspot maps, cluster maps, and data about marginalized subgroups. Participants often recognized complexity involved in making these decisions and indicated the need to develop context-specific best practices, ethical guidelines, and policies about how, when, and in what manner to display and disseminate TB NG-WGS data.