Decolonising Tuberculosis Treatment Pathways: A Participatory Action Research (PAR) Approach Within the Western Cape,South Africa

Abstract

Tuberculosis (TB) continues to pose a major public health challenge in South Africa, particularly in the Western Cape, which records some of the country’s highest TB incidence rates. Despite well-established TB control programmes, treatment access and adherence remain uneven due to structural barriers such as stigma, non-medical patient costs, and rigid clinical routines. Understanding how patients navigate these constraints is essential to improving TB treatment delivery and health outcomes. This paper outlines the qualitative methods used to create a taxonomy of pathways that patients use to access first-line TB treatment in the Western Cape, South Africa. The study was conducted in one substructure of the Cape Metropolitan region and demonstrates a methodological approach that co-creates research with frontline TB staff and key TB stakeholders to ensure that findings are contextually grounded and relevant to real-world service delivery. Using a participatory action research (PAR) framework, combined with a single-round Delphi approach and the SynNovation technique, the study systematically validates consensus among frontline TB staff on process mapping, prioritisation, feasibility, and the impact of TB treatment pathways currently used in the Western Cape. A consensus-building workshop incorporating the SynNovation technique enabled participants to refine and rename the identified pathways, promoting both cultural relevance and linguistic neutrality. The resulting taxonomy reflects staff-derived insights into TB treatment delivery and offers a foundation for future validation from patients. This study represents an initial but meaningful step toward practical decolonisation, as it repositions implementers of TB care as co-creators of knowledge and language.

Introduction

Tuberculosis (TB) remains one of the most persistent public health challenges in South Africa, despite decades of targeted interventions and significant global investment (Department of Health, 2025). The Western Cape province continues to record some of the highest TB incidence rates globally, with many patients experiencing delays, treatment interruptions, or defaulting before completing therapy (Marx et al., 2012; Western Cape Government, 2025). Despite well-established national TB control programmes, treatment access continues to be shaped by social, economic, and structural barriers. In many areas, patients navigate a fragmented landscape of healthcare services ranging from directly observed treatment short-course (DOTS) models to more flexible, community- or home-based approaches. While the national DOTS strategy has achieved progress in diagnosis and treatment adherence, it has also revealed structural weaknesses such as rigid clinical routines, fragmented communication between primary and community levels, and limited integration of patient social realities into care delivery (Adisa & Durojaye, 2025). These limitations have contributed to treatment fatigue, stigma, and inconsistent continuity of care - factors that perpetuate infection cycles and increase the risk of drug resistance (Sazali et al., 2023).

In response to these ongoing challenges, researchers and practitioners have called for new frameworks that move beyond purely biomedical or compliance-based perspectives (Mhimbira et al., 2016). Decolonising health research offers one such approach (Kumar et al., 2024). Rooted in post-colonial and indigenous methodologies (Denscombe, 2025; Keikelame & Swartz, 2019), decolonisation in health systems seeks to challenge the epistemic dominance of Western biomedical models and elevate local knowledge, practices, and voices. It covers the experiences of healthcare workers and communities, recognising that effective interventions must emerge from within the contexts they aim to transform (Mehjabeen et al., 2025; Rosario et al., 2024). In the South African TB response, this means repositioning frontline TB staff, ¹ community health workers, and programme managers not merely as implementers but as co-creators of knowledge and innovation.

Within this context, there is a pressing need to understand how patients in the Western Cape actually access and sustain TB treatment, and to map the diverse pathways that exist beyond policy intentions. Few studies have systematically captured the range of patient trajectories or considered the perspectives of frontline TB staff who mediate these pathways (Aibana. et al., 2020; Pradipta. et al., 2021; Kielmann. et al., 2021). This taxonomy was created to fill the knowledge gap on the different pathways through which patients with drug-susceptible TB access their TB treatment, particularly the lack of literature on how policy, facility practices, and patient realities intersect in the Western Cape. By describing these pathways through a participatory process, this taxonomy provides a practical framework for understanding treatment access that reflects both health-system and community perspectives.

Participatory action research (PAR) offers a robust methodological bridge between decolonising theory and practical system-redesign. As described by Fogg et al. (2022), PAR is characterised by its deliberate and iterative cycles of planning, action, and reflection, enabling those directly involved in care delivery to collaboratively identify problems, test solutions, and co-produce contextually grounded innovations. This approach aligns closely with the principles of decolonisation, as it shifts power from external academic researchers and institutional authorities to frontline practitioners and community stakeholders, thereby validating practitioner expertise and redistributing decision-making within the research process. This method was chosen based on the understanding that a bottom-up, collaborative approach would yield study results that strongly resonate with the primary beneficiaries of the findings, in this case, the patients. There has been a steady rise in recognition that PAR should be used more frequently in research (World Health Organization, 2022; Guy et al., 2020), particularly in TB research.

This study applies PAR alongside a structured Delphi approach process to co-develop a taxonomy of TB treatment pathways in the Western Cape. The research brings together frontline TB staff, particularly nurses, community health workers, and health-system stakeholders to examine how patients access first-line TB treatment, identify barriers to continuity, and propose locally grounded alternatives. The study contributes to the growing body of qualitative and methodological literature that demonstrates how participatory, decolonial approaches can yield more responsive and sustainable health-system knowledge and innovations. Through combining practitioner insight, structured consensus methods, and creative co-design techniques such as the SynNovation technique ² (SynNovation, 2024), this paper aims to illustrate a replicable model for decolonising health systems research and for improving the accessibility and inclusiveness of TB care.

This paper is part of a broader research project focusing on improving access to TB treatment within the Western Cape. While the larger project includes qualitative and quantitative components, this submission details the qualitative research process, emphasising its independence and contribution as a standalone study. The qualitative findings provide in-depth insights into treatment pathways, complementing the broader mixed-methods framework, but are critical on their own for informing policy and practice. The rest of this paper will set out the conceptual framework that forms the foundation of the methodology, followed by a description of the methods used, results, a succinct discussion, and a conclusion.

Conceptual Framework

This study is grounded in a participatory and decolonial conceptual framework that positions knowledge as co-constructed between researchers and practitioners, rather than hierarchically produced. By linking decolonial theory to practice, this framework acknowledges that knowledge generation within health systems is a collective, iterative process shaped by dialogue, reflection, and shared agency. PAR combines participatory and action research, emphasising dynamic cooperation between researchers and participants to enhance empowerment, motivation, self-esteem, and solidarity. It prioritises equitable participation to address issues affecting specific groups and is well-suited for primary care settings due to its collaborative nature (Magwenya & Ross, 2023; Mash, 2014; Nelson et al., 1998; Piercy et al., 2014).

PAR follows a cyclical process: the “look” phase, which identifies problems, the “think” phase, which analyses data, and the “act” phase, which develops plans, implements actions, and evaluates outcomes (Chauke & Ngoepe, 2022). It aims to correct power imbalances affecting marginalised populations, generate knowledge, promote cooperation and empowerment, and improve practice by involving participants in research activities (Cockburn & Trentham, 2002; Robertson et al., 2017).

In the context of this study, PAR provides both an ethical and methodological foundation for decolonising TB research. It helps researchers understand and mitigate health disparities and seeks to reorganise power dynamics, acknowledge community knowledge, and confront inequities to advance social justice (Gill et al., 2012; Olshansky et al., 2005). PAR also benefits community transformation, making it suitable for developing a taxonomy that addresses community-specific issues and objectives (Franquesa-Soler et al., 2023; Magwenya & Ross, 2023).

Partnership and Co-Creation in PAR

Co-creation in PAR entails involving stakeholders in collaborative processes to produce knowledge and solutions. This strategy is apparent in diverse domains such as public health, community development, and business. Previous research emphasises the use of co-creation in research approaches such as PAR to overcome obstacles, promote introspection, and establish frameworks for practical implementation (Longworth et al., 2024; Pearce & Magee, 2024).

Partnerships between academics and various stakeholders with indigenous knowledge, in jointly generating knowledge, promote mutual value, reciprocity, and collective decision-making (Bessette et al., 2023; Fisher & Doering, 2023). The incorporation of indigenous knowledge is essential as it significantly improves the quality and pertinence of research outputs, guarantees a more thorough comprehension of matters, and fosters considerate and principled research methodologies (Louis, 2007; Peltier, 2018).

This methodological paper highlights these partnerships through the involvement of frontline TB staff and key TB stakeholders ³ in the TB management programme. The taxonomy was derived from key informant interviews, and the findings were verified using a Delphi approach that incorporated the SynNovation technique.

Practical Decolonisation

Practical decolonisation in this study represents an initial but meaningful step toward re-orienting research practices to centre local knowledge and expertise (Keikelame & Swartz, 2019). Engaging local experts in the consensus-building process challenges traditional Western-centric paradigms that often dominate global health research and the traditional top-down approaches to knowledge generation, thereby promoting a more inclusive and representative decision-making process (Foroughi et al., 2023; Lewins, 2001). In practice, this study engaged frontline TB staff as participants and co-researchers, ensuring that their lived experiences and insights directly informed the study’s outcomes. The participatory methodology served to equalise power dynamics between researchers and participants, fostering a collaborative and inclusive process. This can be seen in the Delphi questionnaire that was used below.

Decolonisation also extended to the taxonomy of TB treatment pathways. The preliminary pathways were named and described in ways that resonated with local linguistic and cultural contexts, avoiding stigmatising or overly-clinical terminology. This ensured that the framework is academically rigorous, practically relevant, and accessible to those directly involved in TB care. The decolonisation of theory and practice within the co-creation and PAR framework is a crucial undertaking that seeks to question and modify existing power hierarchies, advance impartiality, and prioritise the inclusion of marginalised perspectives in the research process. By integrating decolonial methodologies, researchers can confront past injustices, empower communities, and cultivate research procedures that are more inclusive and fairer (Gill et al., 2012; Greenhalgh et al., 2016; Kuttner et al., 2024). In this way, decolonisation was not an abstract value in this study, but rather a lived practice embedded in each research decision. Practical decolonisation is elaborated further in Figure 1 below.

Figure 1.

Unpacking practical decolonisation in research.

Figure 1 illustrates the study’s overall design, integrating PAR, the key informant interviews, the Delphi approach, the SynNovation technique, and final reporting into a single iterative framework. The pyramid represents the cyclical nature of PAR, with each stage building on the previous one through reflection and collaboration. Icons indicate the main phases: preparation, interviews, Delphi approach validation, consensus-building, and reporting. Together, these steps operationalised practical decolonisation by ensuring equitable participation across all research phases.

Methodology

This study used a PAR design underpinned by principles of co-creation, reflexivity, and mutual learning. PAR was selected to align with the study’s decolonial orientation, enabling those closest to the TB care system, frontline TB staff, community health workers (CHWs), and programme managers, to actively shape both the research process and its outputs. Rather than viewing participants as subjects of inquiry, the study positioned them as co-researchers who collaboratively identified problems, interpreted data, and generated actionable solutions. This participatory design was complemented by a Delphi approach and a SynNovation technique, which together enabled systematic validation and creative refinement of the emerging TB treatment taxonomy.

Ethical Considerations

Ethical clearance for this study was obtained from the University of the Western Cape Biomedical Research Ethics Committee (BM23/3/1). Permission to conduct the study within the Western Cape Department of Health services was also obtained before data collection. All participants provided written informed consent prior to taking part in key informant interviews, the Delphi approach questionnaire, and the SynNovation renaming technique. Confidentiality and anonymity were ensured through the use of pseudonyms, and participation was voluntary, with the option to withdraw at any stage. Ethical rigour was maintained through ongoing reflexive engagement between the researchers and participants to ensure respect, reciprocity, and transparency throughout the process.

Participants and Sampling

Participants were purposively recruited based on their active involvement in TB care, programme management, and service delivery in one substructure of the Cape Metropolitan region of the Western Cape to ensure a comprehensive understanding of both clinical and operational perspectives. The research team initially aimed to conduct twelve (n = 12) key informant interviews, but due to scheduling constraints and clinical workloads, nine (n = 9) key informant interviews were successfully completed.

District TB coordinators facilitated recruitment by identifying suitable participants based on their roles in tuberculosis management. Potential participants received an invitation letter and participant information sheet outlining the study’s purpose, voluntary nature, and confidentiality measures. Those interested in participating contacted the research team directly by email or phone to confirm their involvement.

This taxonomy represents a staff-derived framework, reflecting the professional perspectives of those directly engaged in TB service delivery. While patients were not included at this stage, frontline staff occupy a dual role as implementers and intermediaries between patients and the health system, enabling them to reliably report on the practical realities of treatment access and adherence. Their embeddedness within daily TB care delivery provides a credible basis for mapping how patients navigate existing treatment pathways. Future work will include validation from patient participants to complement and refine this staff-informed taxonomy.

Data Collection

Data collection unfolded across three interrelated phases designed to progressively build understanding and consensus around the taxonomy of preliminary TB treatment pathways. Each phase built on the insights and validation achieved in the previous stage, allowing for iterative refinement and participant engagement throughout the process, as illustrated in Figure 2 below.

Figure 2.

Overview of participant engagement and outputs across the three study phases.

Figure 2 illustrates the number of participants invited and retained at each stage, key informant interviews (n = 9 of 12 invited), Delphi approach round (n = 7 of 9 invited), and the consensus-building workshop (n = 7 of 9 invited), culminating in a validated taxonomy of TB treatment pathways, discussed in more detail in the results section.

Data Analysis

Data from all phases of the study were analysed using an iterative thematic analysis approach informed by Braun and Clarke’s 2006 framework (Ahmed et al., 2025; Braun & Clarke, 2012). The principal investigator and three researchers independently reviewed transcripts from the key informant interviews, Delphi questionnaire feedback, and SynNovation notes to identify recurring themes and validate emerging insights. Analytical rigour was ensured through regular debriefing and peer review among the researchers, enabling reflection and consensus on coding decisions.

Themes developed during the interviews informed the structure of the Delphi questionnaire, while feedback from the consensus-building workshop refined and validated the preliminary TB treatment pathways. This iterative process reflects the cyclical nature of PAR, where reflection and action occur simultaneously. Triangulation across data sources - key informant interviews, Delphi questionnaire responses, and SynNovation renaming technique discussions ensured coherence and strengthened the credibility of the findings. Each phase of the study is detailed below in three parts.

Part 1: Key Informant Interviews

To gather in-depth insights into the different treatment pathways for accessing TB treatment, key informant interviews were conducted with frontline TB staff and key TB stakeholders (n = 9) involved in TB management in one sub-district in the Western Cape, South Africa. The interview process followed several steps, as shown in Figure 3 below.

Figure 3.

Steps followed in conducting key informant interviews.

The preparation phase began with the purposive selection of participants based on their expertise and direct involvement in TB care. Recruitment sought to reflect the diversity of roles involved in TB care delivery, ensuring representation across clinical (nursing), community (CHW), and administrative functions. As noted above, district TB coordinators supported the recruitment by identifying potential participants, who were then invited via email and follow-up calls.

A semi-structured interview guide was developed, encompassing key questions related to each step along treatment pathways for accessing TB treatment, as well as the challenges and opportunities associated with each pathway. Interviews were scheduled at the convenience of the participants, and written informed consent was obtained before participation. The consent process included a detailed explanation of the study’s objectives, methods, and participants’ rights, including confidentiality and the option to withdraw at any point in time.

Emphasising confidentiality before conducting the interviews was crucial to encourage frontline TB staff to share the most granular and honest information. The semi-structured interview guide facilitated open-ended questions to explore key themes, allowing participants to provide detailed responses. Additional probing questions were used to gather deeper insights and clarifications. Interviews were recorded using more than one auditory recording and transcription software to ensure accurate transcription and analysis, these included Microsoft (MS) Teams transcription and recording, as well as Fathom transcription.

During the data analysis phase, all interviews were transcribed manually. The transcripts were then analysed using thematic analysis to identify patterns, themes, and key insights that led to the emergence of mutually exclusive categories for TB treatment pathways. The findings were summarised, highlighting critical challenges, opportunities, and recommendations identified by the participants. The preliminary findings served as the foundation for the next phase of the study, which used the Delphi questionnaire (Barrett & Heale, 2020). Figure 4 below encapsulates the process followed during this stage.

Figure 4.

Tuberculosis treatment pathway analysis.

Part 2: Consensus Building Workshop

The second step involved conducting a consensus-building workshop. ⁴ An initial invitation to attend a consensus building workshop was sent to the same key informant interview participants (n = 9), of whom seven were accepted (thus n = 7). The first step of the workshop was to present a taxonomy of the preliminary treatment pathways for accessing TB treatment based on analysis of information from the key informant interviews. This step ensured that all participants understood the different pathways before issuing out the questionnaire, which relied on the accuracy of these pathways. The consensus-building workshop was held as a single-day collaborative session that incorporated two complementary components: a structured Delphi questionnaire and a SynNovation-based co-design activity. This combined approach allowed participants to validate the emerging taxonomy of TB treatment pathways while also contextualising and renaming the pathways in ways that reflected local realities and language, thereby removing overly clinical and stigmatising labels.

The Delphi Questionnaire

A taxonomy infographic was developed to illustrate the four preliminary treatment pathways that were identified in the key informant interviews. Thereafter, schematic representations and detailed process maps were used to make each pathway’s steps clear and understandable. This visual and descriptive approach helped participants grasp the complexity of the pathways, ensuring that they could effectively complete the subsequent questionnaire, which tested the rigour of the findings. The four distinct treatment pathways for accessing and adhering to TB treatment - each differing in location, level of support, and frequency of healthcare facility contact, are discussed more in the results section below.

The Delphi questionnaire ⁵ was used to build consensus on the taxonomy developed from the key informant interview analysis, ensuring robust findings that would contribute meaningfully to understanding the different treatment pathways used to access TB treatment. Participants were asked to assess and refine the four preliminary TB treatment pathways identified from the key informant interview analysis. During this phase, participants responded to a structured questionnaire presenting the initial findings from the key informant interviews. They were asked to rate the importance and feasibility of each identified pathway. The steps followed are summarised in Figure 5 and explained in further detail below.

Figure 5.

Steps used in the Delphi approach.

As seen in Figure 5 above, the first step involved the development of a questionnaire instrument, which consisted of quantitative and qualitative questions to elicit detailed responses on various aspects of TB treatment pathways. The instrument that was developed can be found in Appendix 1. When developing the questionnaire instrument, five key themes were covered, namely process mapping, prioritisation, feasibility, impact assessment, and consensus building, discussed in more detail below.

Process Mapping

Participants were asked to rate the accuracy of the process map for each of the four identified treatment pathways. A 10-point Likert scale was used, where 1 indicated the least accurate description and 10 indicated the most accurate description of the pathways. The rating provided by the participants indicated a measure of the perceived accuracy of the described pathways to actual patient experiences in the real world, ensuring the reliability and validity of the process mapping in representing first-line TB treatment.

Prioritisation

Participants were encouraged to provide detailed feedback on the main challenges they perceive in each pathway and suggest opportunities for improvement. The participants then ranked pathways based on their impact and frequency of use. High-priority pathways often represented those serving the most vulnerable populations or those with significant bottlenecks hindering timely access to treatment.

Feasibility

Participants assessed the feasibility of implementing each pathway in different contexts, considering resource availability, infrastructure, and potential impact on patient outcomes. This assessment aimed to identify pathways that could realistically be implemented, considering the practical challenges that patients might face.

Impact Assessment

Participants discussed the potential short- and long-term impacts of each pathway on patient outcomes and healthcare delivery. Participants were asked to rate various aspects of the different TB treatment pathways, focusing on their effectiveness, accessibility, and overall patient experience. This section of the questionnaire aimed to forecast the outcomes of implementing each pathway, highlighting the benefits and drawbacks for patients.

Consensus Building

The second step in Figure 5 was the workshop itself, which constituted the consensus building phase of the study. Achieving agreement among diverse stakeholders was critical to the study’s success. Through iterative discussions and the Delphi questionnaire round, participants reconciled differing perspectives to create a shared understanding of TB treatment pathways. This section aimed to highlight areas of contention and diverse opinions regarding the different TB treatment pathways. Participants were asked which pathways they believed would generate the greatest consensus and which would generate the most disagreement.

The survey instrument, provided in Appendix 1, included both quantitative and qualitative questions. Participants rated each pathway using a 10-point Likert scale for accuracy, feasibility, and impact, and provided open-ended feedback on challenges, opportunities, and implementation barriers. Consensus was defined as ≥70% agreement or an interquartile range (IQR) ≤1, consistent with accepted Delphi standards (Hsu & Sandford, 2007; Sinead, et al., 2010). The instrument was administered online during the workshop using a secure platform. Participants provided informed consent before completing the survey, and all data were collected in real time. A 100% response rate was achieved.

SynNovation Technique – Renaming the Pathways

While the discussion of each pathway in the Delphi round provided a clear picture of strengths as well as limitations, it was equally important that these models be contextualised in a way that resonated with local realities. To this end, the SynNovation technique was used to encourage indigenous knowledge creation and help participants develop names for the TB treatment pathways in the taxonomy. This method, adapted from synectics, encourages innovative and actionable solutions with strong stakeholder buy-in (SynNovation, 2024). Various physical items were provided to inspire creative thinking and encourage participants to think innovatively about how these objects could inspire new names for the pathways. Each stage introduced new objects to maintain a steady flow of ideas for each different pathway. The participants were split into 3 different groups (ranging from 2-3 participants in each group), where 3 researchers acted as the note takers, taking notes of each emerging idea that the participants came up with using the items they were provided. The items in this process included a teabag, sunglasses, a greeting card, batteries, a toy car, a postcard, two plugs that connected to each other, a flower, a butterfly, a balloon, nail polish, and a plastic bag, among a few others. These objects were used to stimulate metaphorical thinking and encourage lateral connections between daily experiences and TB care.

Throughout the exercise, participants answered specific questions related to the items and the pathways, facilitating a dynamic and inventive brainstorming session. This unique approach was chosen to ensure that the participants were empowered to come up with the TB treatment pathway names that were ‘homegrown’ and not imposed on them by an external body. The renaming of these pathways using the SynNovation renaming technique ensured that the language used was neutral and non-stigmatising. This technique facilitated a more respectful and inclusive dialogue and encouraged participants to think more broadly about the negative connotations often associated with TB-related jargon. Throughout the session, participants discussed how each object could symbolise various aspects of treatment adherence, accessibility, or patient experience. This participatory activity fostered indigenous knowledge creation, encouraged open reflection, and ensured the final pathway names were neutral, non-stigmatising, and culturally meaningful. The exercise provided an inclusive space for dialogue and co-ownership, reinforcing the collaborative ethos of the study.

The third step involved presenting the notes and ideas generated by the three groups back to the full participant group for collective validation. Each group’s output, recorded by the researchers during the SynNovation activity, was summarised and shared for open discussion. Participants were encouraged to reflect on, agree with, or challenge the ideas proposed by other groups, as well as to elaborate further where clarification or nuance was needed. This step fostered rich peer dialogue and critical reflection, with several participants expressing appreciation for the creativity of others’ contributions, noting that the innovative ideas “sparked new ways of thinking” about patient engagement and communication. This reflective exchange not only strengthened agreement on the final pathway names but also deepened participants’ sense of ownership and collective authorship of the taxonomy.

The fourth and final step was the analysis of the consensus questionnaire, which happened after the consensus building workshop. The study initially identified four distinct preliminary TB treatment pathways in the key informant interviews, these were then validated through high consensus during the consensus building workshop, discussed in more detail in the results section below. Each pathway represents a unique set of patient experiences, reflecting the diverse challenges and opportunities within the healthcare system. The taxonomy includes descriptive elements and schematic process maps for each pathway, providing a robust framework for implementation and further research. Figure 6 below summarises the process of achieving consensus.

Figure 6.

Achieving consensus on tuberculosis treatment pathways.

Together, these activities produced a validated, contextually grounded taxonomy of TB treatment pathways that served as the foundation for the subsequent analysis. The outcomes of the workshop are discussed in further detail in the results section.

Results

The PAR process produced a co-created and contextually grounded taxonomy of preliminary TB treatment pathways in the Western Cape. This unfolded in two interlinked phases: key informant interviews that identified initial pathways and a one-day consensus-building workshop (with Delphi and SynNovation components) that refined and validated them. District TB coordinators facilitated recruitment by identifying suitable participants based on their roles in tuberculosis management. Potential participants received an invitation letter and participant information sheet outlining the study’s purpose, voluntary nature, and confidentiality measures. Those interested in participating contacted the research team directly by email or phone to confirm their involvement.

The final sample of this study comprised of nine female healthcare professionals, including an HIV, AIDS, Sexually Transmitted Infections (STIs), and Tuberculosis (HAST) medical officer, a HAST manager, a pharmacy supervisor, and six professional nurses (commonly referred to as “TB sisters”) working in primary healthcare and TB treatment settings. This composition ensured representation across clinical, managerial, and pharmaceutical perspectives in TB service delivery. Their professional experience ranged from approximately 1 to 15 years in TB management and patient care, ensuring that the study captured grounded and contextually informed insights. To preserve anonymity, detailed demographic information such as age was not systematically collected, as the participants’ professional roles could make them identifiable, given the small sample size. This diversity ensured a comprehensive understanding of both clinical and operational aspects of TB treatment.

Part 1: Key Informant Interviews

In the first phase, key informant interviews with frontline TB staff revealed that TB patients typically access treatment through a variety of informal and formal routes, some adhering strictly to directly observed therapy short-course (DOTS) models, while others were more flexible, relying on self-administration or community health worker (CHW) supervision. The interviews also highlighted a more innovative approach to TB treatment access through home delivery methods. Four main treatment pathways were identified, reflecting varying levels of patient autonomy, clinical supervision, and community involvement:

Daily Clinic-Based Supervised Treatment

In this treatment model, the TB patient travels to a healthcare facility every day for a specified period to receive and swallow TB medication/pills under the direct supervision of a nurse. This routine involves bypassing general patient queues, waiting to be seen in a TB-specific room, reporting on health status, and occasionally producing sputum samples, which are tested. This pathway, although highly structured, demands more time and non-medical patient costs from the patient, with round-trip travel and in-facility wait times extending the daily burden of care.

Community-Supervised Home-Based Treatment With Periodic Clinic Visits

In this model, a CHW visits the TB patient at home daily to supervise medication intake. The CHW verifies the medication, checks on the patient’s condition, and monitors side effects. The patient also attends a clinic every two weeks for check-ups. This approach offers convenience and accessibility, especially for patients facing transport challenges, but is resource-intensive from a staffing/nursing perspective, and ultimately, the public sector.

Self-Administered Medication With Biweekly or Monthly Check Ups

This less intensive approach involves patients taking their medication on their own each day, often tracking doses independently. They visit the clinic every two weeks or monthly to receive their medication supply and for a period ranging from 2 weeks to a month. This model provides autonomy and reduces clinic visits but relies heavily on patient adherence.

Monthly Delivery of Medication With Minimal Supervision

In this model, medication is delivered to the patient’s home once a month, and the patient self-administers their treatment daily. Clinic visits occur monthly or less frequently for clinical monitoring. This pathway minimises patient travel and burden but also assumes high patient responsibility.

Part 2: Consensus Building Workshop

The insights from part 1 informed the second phase, which involved the Delphi-style questionnaire and the SynNovation renaming technique used to refine and validate the illustrative pathways. Each phase built on the previous one: the key informant interviews generated the preliminary taxonomy of TB treatment pathways, while the Delphi questionnaire quantified consensus on their accuracy and feasibility, and the SynNovation renaming technique contextualised and renamed the pathways to reflect local terminology and cultural relevance.

Through process mapping, participants traced the patient journeys from diagnosis to treatment completion, indicating the points where patients commonly disengage or “fall out” of care. These included delays between diagnosis and treatment initiation, breakdowns in communication between clinics and CHWs, and logistical barriers such as high transport costs and rigid clinic schedules. Mapping these flows visually helped frontline TB staff compare official protocols with what happens in practice, allowing them to identify the inefficiencies, as well as adapt the sequencing of care to be more patient-centred.

Prioritisation involved identifying which care trajectories required urgent reform. Participants discussed the pathways in terms of both patient vulnerability and systemic strain. Pathways where patients faced frequent clinic visits and where socio-economic constraints (for example, job loss, daily transport costs) were most severe were often ranked as high priority for reform. Feasibility considerations revealed the delicate balance between ideal care models and the realities of implementation. Participants weighed staffing capacity, administrative processes, the strength of CHW networks, and the trust between patients and providers. Some pathways were viewed as highly effective in theory but were rated lower on feasibility due to current resource constraints, spotlighting the importance of aligning innovation with operational capacity.

Impact assessment focused on both clinical outcomes and system efficiency. Small procedural adjustments, such as aligning TB follow-ups with other existing clinic appointments, were recognised as having the potential to improve adherence, detect side effects much earlier, and reduce patient travel costs. Participants emphasised that pathways offering patients greater autonomy often reduced stigma but risked lower adherence when support structures were inadequate.

Finally, the results from the SynNovation renaming technique was a critical step in refining the pathway taxonomy. This process went beyond numerical agreement to ensure shared understanding of pathway designs and the language used to describe them. Participants brought varied professional and lived experiences to the table, creating an environment where competing perspectives could be integrated into a shared framework. The pathways were further renamed using this technique, and the TB staff, along with the researchers, agreed on the names to be used.

The participants assessed the four pathways, each representing a distinct balance between patient autonomy and clinical oversight. The percentage of consensus for each pathway can be seen in Figure 7 below. Participants rated each pathway according to its feasibility and potential impact on TB treatment outcomes, using a Likert scale from 1 (very low) to 10 (very high). The percentages indicate the level of consensus (proportion of participants rating 8 or above) for each pathway.

Figure 7.

Participant consensus on the perceived feasibility and impact of four TB treatment pathways.

The now final validated treatment pathways include pathway A, which involves daily clinic visits at the start of treatment, ensuring high clinical oversight, early detection of side effects, and opportunities for peer support. It was regarded as offering strong clinical benefits but was also the most burdensome for patients, with financial strain, stigma, and inconvenience cited as key barriers. It achieved a 96% agreement rate, indicating strong consensus on its clinical merit despite operational challenges.

Pathway B reduced clinic visits by relying on CHWs for more frequent community-based oversight. Participants valued its ability to maintain a degree of clinical monitoring while supporting patient autonomy and enabling continued employment. However, issues of confidentiality, inconsistent CHW reporting, and the need for reliable supervision were raised. This pathway recorded an 89% agreement rate, reflecting its perceived promise if paired with adequate CHW support structures.

Pathway C offered patients significant independence, with medication self-administered and clinic visits kept to a minimum. This model reduced stigma and was convenient for patients who struggled with clinic attendance, but concerns were raised about poor adherence, missed detection of side effects, and limited patient support. Agreement reached 87% - the lowest among the four, but still above the 70% consensus threshold.

Pathway D represented complete self-management without clinic visits, maximising convenience, eliminating travel costs, and freeing patient time. However, participants highlighted serious risks, including disengagement from treatment, medication errors, and limited traceability of patients. Despite these concerns, it secured an 89% agreement rate, largely due to its fit for a small subset of highly motivated and stable patients.

Overall, consensus was achieved on all four pathways, confirming strong agreement among participants on their relevance and viability, with a mean agreement rate of 90.3%. The process not only validated the pathways quantitatively but also enriched them qualitatively through collaborative interpretation of findings during the SynNovation renaming technique. As summarised in Appendix 2, the Delphi ratings demonstrated high convergence (median ≥ 9 across treatment pathways), reflecting broad alignment in both numerical assessments and qualitative discussions. This combination of structured consensus methods and participatory dialogue ensured that the taxonomy was grounded in real-world experience and adapted to the Western Cape’s health system realities. The high level of consensus underscores the robustness of the key informant interview findings, suggesting that the proposed pathways are well aligned with participant expectations and experiences, providing a strong foundation for future implementation. Owing to this high degree of agreement in the initial round, no additional Delphi rounds were required. ⁶

Together, these sequential stages: key informant interviews, Delphi-style questionnaire validation, and SynNovation renaming technique renaming produced both quantitative and qualitative outputs that reinforced one another, demonstrating how practitioner insights guided each phase of the research process and informed the final taxonomy of TB treatment pathways.

Discussion

This study contributes to the growing movement in health research that calls for deeper involvement of patients, frontline TB staff, and community stakeholders in the generation of health research knowledge. This research thus highlights the potential of PAR and the Delphi approach to address systemic challenges in TB treatment. The participatory approach ensured that findings were deeply rooted in local knowledge and practical realities. The use of the SynNovation technique exemplifies how decolonising methodologies can create inclusive communication and innovative solutions. Through the use of PAR, frontline TB staff and key TB stakeholders were explicitly included in co-creating indigenous knowledge, thus contributing to this evolving research paradigm. By prioritising co-creation, this study contributes to a growing body of literature that advocates for equitable and impactful health research. Figure 8 below further highlights the advantages of using PAR, specifically in TB research.

Figure 8.

Unveiling participatory action research in tuberculosis treatment.

Importantly, the findings demonstrate that improving TB treatment outcomes requires addressing non-clinical barriers, including stigma, non-medical patient costs, rigid treatment schedules, and the social conditions that shape adherence. These issues emerged directly in the key informant interviews and were reinforced during the consensus-building phase, where participants identified stigma and financial strain as major challenges in pathways requiring daily clinic attendance (Pathway A), and limited confidentiality in community-based models (Pathway B). By integrating these contextual dimensions into the taxonomy, the study provides a practical framework for patient-centred reform, aligning with global efforts to strengthen people-centred health systems. The SynNovation renaming exercise further demonstrated how linguistic and cultural relevance can reduce stigma, as participants replaced technical labels with locally meaningful terms that reflected respect and patient dignity. ⁷ The emphasis on linguistic, cultural, and operational relevance responds to the World Health Organization’s call for inclusive approaches that reflect patients’ lived experiences rather than universalised models of care.

Consensus methods such as the Delphi approach and SynNovation technique played a key role in bridging experiential knowledge with systematic validation. They moved beyond consultation toward true co-design, creating a structured yet flexible environment for collaborative reflection. For example, the Delphi questionnaire allowed participants to collectively assess feasibility, prioritising community-supervised pathways as both high-impact and realistic under current health system constraints, while the SynNovation session empowered participants to reframe pathway names in ways that promoted ownership and cultural resonance. In doing so, the study modelled how consensus-building tools can be embedded within participatory research to enhance ownership and the sustainability of outcomes. These processes validated experiential knowledge by translating the lived realities of frontline practitioners into actionable policy insights, an essential feature of decolonised knowledge production. This is particularly relevant in South Africa, where health systems remain shaped by historical inequities and sociopolitical legacies that often marginalise community knowledge (Masenya, 2022; Ramani et al., 2023).

Each of the four pathways also carries distinct implications for practice and feasibility. Pathways A and B were prioritised for maintaining clinical oversight and supporting adherence in high-burden communities, whereas Pathways C and D emphasised autonomy and flexibility but required additional adherence support and monitoring mechanisms. These findings provide a foundation for tailored interventions, balancing patient empowerment with sustained clinical engagement. This alignment between structured supervision and patient independence underscores the need for adaptable, context-specific TB care models rather than one-size-fits-all approaches.

While this taxonomy reflects the insights of frontline staff and programme implementers, it does not yet incorporate direct patient perspectives. As such, it should be regarded as a staff-derived framework that captures provider-level experiences of TB treatment access and delivery. Future validation through patient participation will be essential to ensure that the taxonomy reflects lived experiences and user-centred realities of care.

From a decolonising perspective, the study operationalises the principles of equity and epistemic justice by elevating the voices of those traditionally excluded from research design. The use of participatory consensus methods redistributed power between researchers and practitioners, demonstrating how decolonisation can move from rhetoric to practice. By engaging TB nurses, CHWs, and programme staff as co-creators, the study challenged hierarchical knowledge production and generated a taxonomy grounded in indigenous insight and professional experience.

Conceptually, this research offers a replicable model for integrating participatory and consensus methodologies in complex, low-resource settings. It demonstrates how bottom-up system redesign can yield both theoretically rich and operationally feasible outcomes. Practically, the taxonomy of TB treatment pathways developed here provides policymakers and practitioners with a structured yet adaptable tool for improving service delivery in the Western Cape and similar contexts. The study’s limitations include its focus on a single geographic region and a relatively small sample size. Future research should explore the applicability of the taxonomy in other contexts and examine its impact on patient outcomes over time. Longitudinal studies could provide further insights into the sustainability and effectiveness of the proposed interventions. In addition, future work should extend this participatory framework to include patient perspectives, ensuring that the taxonomy captures not only provider and system-level insights but also the lived experiences of those receiving TB treatment. Incorporating patient voices would further strengthen the decolonial orientation of the study by grounding the taxonomy in community realities and promoting equitable knowledge co-creation.

This study offers a preliminary yet tangible step toward practical decolonial research by repositioning frontline staff as co-creators of knowledge and decision-making and by integrating culturally resonant naming and framing of treatment pathways. Finally, it is worth reiterating that this submission does not represent the “qualitative component” of a mixed-methods study but rather a complete qualitative inquiry designed to stand alone. The methodology and findings are sufficient to address the research aims independently while complementing the broader project’s objectives.

Conclusion

This study presents a rigorous and participatory process for developing a taxonomy of TB treatment pathways that are both contextually grounded and ethically engaged. Through PAR, the Delphi approach, and the SynNovation technique, frontline TB staff were able to name, prioritise, and assess the feasibility and impact of real-world TB care trajectories. This study also contributes to the field of global qualitative health research by demonstrating the value of PAR and decolonising methodologies in developing actionable frameworks for tuberculosis care. Furthermore, this study highlights the effectiveness of participatory methodologies in developing a taxonomy of TB treatment pathways that centre local expertise and prioritises inclusivity. By advancing understanding of how PAR can address health disparities, particularly within the South African context, this study provides a replicable model for global application. Future research should build on these findings to refine and implement participatory approaches in TB care, ensuring that marginalised voices remain at the forefront of health system innovation. As global health continues to reckon with colonial legacies and systemic inequities, this research offers a methodological blueprint for building more inclusive, equitable, and effective health systems.

Supplemental Material

Supplemental Material - Decolonising Tuberculosis Treatment Pathways: A Participatory Action Research (PAR) Approach Within the Western Cape, South Africa

Supplemental Material for Decolonising Tuberculosis Treatment Pathways: A Participatory Action Research (PAR) Approach Within the Western Cape, South Africa by Yuvir Naidoo, Jill Cupido-Masters and Carmen S. Christian in International Journal of Qualitative Methods.

Supplemental Material

Supplemental Material - Decolonising Tuberculosis Treatment Pathways: A Participatory Action Research (PAR) Approach Within the Western Cape, South Africa

Supplemental Material for Decolonising Tuberculosis Treatment Pathways: A Participatory Action Research (PAR) Approach Within the Western Cape, South Africa by Yuvir Naidoo, Jill Cupido-Masters and Carmen S. Christian in International Journal of Qualitative Methods.