Tuberculosis Treatment Adherence as a Relational and Cultural Process: A Qualitative Study From Urban Indonesia

Study Design and Epistemological Positioning

A qualitative exploratory descriptive design was used to examine how TB patients and family members experienced and interpreted treatment adherence in an Indonesian urban setting. This design is appropriate when the research goal is to describe and understand phenomena as experienced by participants, rather than to test hypotheses or produce statistically generalisable findings (Sandelowski, 2000). The epistemological orientation of this study is broadly interpretivist: we assumed that adherence is not a fixed, observable behaviour but a socially constructed and contextually embedded experience that can only be understood from within the perspectives and meaning-making practices of those involved (Creswell & Poth, 2018).

Reflexive thematic analysis was used as the primary analytic method (Braun & Clarke, 2006, 2019). This approach was chosen because it offered flexibility to work inductively across multiple participant accounts, to remain closely grounded in participants’ own accounts, and to attend to both individual and relational dimensions of the adherence experience, which were central to the study’s aims.

The study was conducted at across urban primary health facilities in Kupang, the capital city of East Nusa Tenggara Province, Indonesia. This setting was selected because it serves a concentrated TB patient population and was accessible for in-depth qualitative recruitment.

Participants

Participants were adults diagnosed with pulmonary tuberculosis and currently receiving anti-tuberculosis treatment, along with family members who were actively involved in supporting their care. Eleven participants took part in the study, including six patients (three men and three women) and five family members (two men and three women). Participants were recruited through purposive sampling conducted in collaboration with TB programme officers at the study facilities. We sought variation in age, sex, educational background, and family caregiving role in order to capture a range of experiences relevant to treatment adherence.

Inclusion criteria for patients were: (a) adult (aged 18 years or older); (b) confirmed diagnosis of pulmonary tuberculosis; (c) currently enrolled in and receiving anti-tuberculosis treatment at one of the study facilities; and (d) willingness to participate and provide informed consent. Family members were included if they were identified by the patient or treating team as actively involved in treatment support. Participants were excluded if they were acutely unwell at the time of interview, or if they did not consent to audio recording.

The sample size of 11 participants reflected the depth-oriented aims of qualitative interviewing and the practical boundaries of recruitment within the study setting. In exploratory qualitative research, sample size is guided by the richness and sufficiency of data rather than statistical power (Malterud et al., 2016). We assessed information sufficiency iteratively during data collection and determined that the interviews provided adequate conceptual depth to address the study aims.

The adherence trajectories of participants varied. Among patients, three had maintained consistent medication-taking throughout their treatment course; two had experienced periods of inconsistent adherence, particularly during the first weeks of treatment; and one had initially refused to commence treatment due to cultural beliefs about the illness. Treatment duration at the time of interview ranged from six weeks to five months. Family members’ levels of engagement also varied: two were described as primary caregivers who assisted with daily medication supervision and clinic attendance; two played a more intermittent support role (accompanying patients to appointments and providing reminders); and one had limited direct involvement in medication management but was engaged with emotional and spiritual support. All six patients lived with at least one family member. No participant lived alone.

Data Collection

Data were collected in September 2025 in a private room at the health facility, chosen to offer participants privacy and comfort during discussions of sensitive and stigmatised experiences. Individual face-to-face semi-structured interviews were conducted by the lead researcher, who had prior experience working in community-based TB programmes and was familiar with the sociocultural context of care in this region.

At the study facilities, anti-tuberculosis treatment was delivered through family-supervised self-administered therapy, in which patients took their medication at home under the daily oversight of a designated family member rather than under direct observation by a health worker. This modality provided the structural and systemic context within which the relational and household dynamics described below unfolded, even though the analytic focus of this study remained on the relational and interpersonal dimensions of adherence.

Each interview lasted approximately 45–60 minutes. Conducting interviews individually enabled participants to discuss emotionally difficult and socially sensitive experiences, including shame, delayed care-seeking, family tensions, and spiritual distress, with greater confidentiality than a group setting would permit. A second researcher kept field notes during and immediately after each interview, recording non-verbal expressions, emotional tone, pauses, and contextual observations that informed interpretation.

A semi-structured interview guide was used to explore a series of topic areas: emotional responses to diagnosis; experiences of treatment over time; interactions with health services; stigma and social perception within the community; cultural understandings of what caused the illness; the practical roles family members played in supporting treatment; and the role of religious or spiritual practice. The guide was used flexibly — probing where accounts were ambiguous, encouraging elaboration, and following issues that participants themselves identified as important. Interviews were conducted in the local language (Bahasa Indonesia) and audio-recorded with all participants’ explicit consent. Audio recordings were transcribed verbatim and translated into English by a bilingual researcher on the team. Field notes were finalised immediately after each interview to capture contextual details and early analytic impressions.

No participants declined to take part after being approached. No repeat interviews were conducted; however, two participants were contacted by telephone after their interview for brief clarifications regarding specific statements they had made. These clarifications were documented in field notes and incorporated into the analytic process.

Data Analysis

Data were analysed using reflexive thematic analysis as described by Braun and Clarke (2006, 2019), working inductively from participants’ accounts rather than applying a pre-determined theoretical framework. Reflexive thematic analysis treats theme development as an active, interpretive process in which themes are constructed by the researcher through close engagement with the data.

The analytic process proceeded through six phases. First, both researchers read the full interview transcripts multiple times, making reflective notes before any formal coding began. Second, initial codes were generated line-by-line across the transcripts, attending closely to how participants described emotional experiences, interpreted their illness, described family interactions, and talked about what helped or hindered their adherence. Third, the two researchers coded the transcripts independently, then met to compare their coding and work through interpretive differences. Disagreements were treated as analytic resources: they often signalled passages where meaning was genuinely ambiguous or where our respective positionalities had led us to read the same words differently. Fourth, from the initial codes, preliminary thematic clusters were built and refined iteratively by moving between coded segments and the full dataset. Fifth, themes were reviewed against the full dataset to assess coherence and grounding. Sixth, themes were defined and named in ways that captured the interpretive claim each carried.

Coding was conducted manually, with annotations in printed transcripts and digital memos. Negative and deviant cases were actively sought during analysis by examining accounts that did not align with identified patterns and using these to refine or qualify our interpretation. We paid particular attention to convergences and contrasts between patient and family accounts, especially where family members framed adherence as collective responsibility while patients described it through emotional disruption or shame.

Themes were finalised when we were satisfied that each was analytically coherent, grounded in the data, and meaningfully distinct from the others. Six themes were constructed, each with a single subtheme. The subthemes were developed to specify the primary mechanism or analytic claim within each theme — they were not intended to be exhaustive of all content within a theme, but to highlight the most analytically significant pattern. This structure was adopted to maintain analytic focus rather than to suggest that each theme contained only one meaningful dimension.

Methodological Rigour

Rigour was addressed through strategies consistent with the interpretivist tradition and the specific approach of reflexive thematic analysis, addressing credibility, dependability, confirmability, and transferability (Lincoln & Guba, 1985). Credibility was supported by comparing accounts across patient and family interviews, and through researcher triangulation, independent coding by two researchers, followed by iterative interpretive dialogue. Dependability was addressed through the maintenance of an audit trail, documenting analytic decisions at each stage of the thematic process. Confirmability was supported through reflexive journalling by both researchers throughout analysis; this was used not as a rhetorical gesture toward transparency but as a practical tool for surfacing and interrogating the assumptions brought to the data. Transferability is supported through thick description of the study context, facility setting, participant characteristics, and the specific sociocultural conditions in which the data were generated, enabling readers in comparable settings to assess relevance.

Reflexivity

Both researchers live and work within the Indonesian cultural and healthcare contexts that participants also inhabit. This shared positionality was both a resource and a risk. It meant that basic features of everyday Indonesian social life — the meaning of family obligation, the social weight of community reputation, the role of prayer in daily routine — did not require explanation before we could understand what participants were describing. The risk was that familiarity could slide into assumption: we might read cultural patterns into data that were more varied or contested than we recognised, or our shared embeddedness with participants might lead us to overlook tensions present in the data. We used reflexive journals and regular team discussions to work against this tendency, explicitly asking at each analytic stage whether an interpretation was grounded in what participants actually said or in what we expected them to say. Where assumptions were identified, we documented them, either revising the interpretation or acknowledging the assumption explicitly in our analytic memos.

Ethical Considerations

Ethical approval for this study was obtained from the Faculty of Nursing Airlangga University Health Research Ethics Committee (Approval No: 3387-KEPK). The study adhered to the principles outlined in the Declaration of Helsinki for research involving human participants. All participants received a full written and verbal explanation of the study’s purpose, procedures, and their right to withdraw at any time without consequence for their medical care. Written informed consent was obtained before each interview. Personal identifiers were removed from transcripts and replaced with participant codes. Audio recordings and transcripts were stored in password-protected files accessible only to the research team, and findings are reported at the thematic level to prevent identification of individual participants.

Theme 1: Diagnosis as Psychological Disruption

Participants’ accounts indicated that the diagnosis of tuberculosis was experienced not only as confirmation of illness but as a disruption to identity. Emotionally heavy language predominated, suggesting that TB carried a burden extending well beyond bodily symptoms. Two patients described their initial reactions in terms that condensed shame, fear of judgement, and a sense that a TB diagnosis could alter how they were perceived by others. One woman recalled: “When I found out, I felt like something was wrong with me as a person, not just my body” (Patient 2, woman, age 32).

The brevity of some responses was itself analytically significant. One patient (Patient 1, man, age 28), asked to describe the experience of beginning treatment, offered just two words: “It felt heavy.” Here, heaviness did not simply denote physical fatigue or medication side effects; it appeared to condense shame and social fear in a single phrase, suggesting that diagnosis threatened social identity as much as bodily wellbeing.

Family interviews reinforced this interpretation by showing that the burden of diagnosis extended beyond the patient. One family member described an initial household response shaped by shame, community gossip, and spiritual fear rather than immediate recourse to formal care: “At first, she felt ashamed. People said it was a curse. So the family only prayed and did not want to bring her to the hospital. It took a very long time before she recovered” (Family Member 3, woman, mother of patient).

This account suggests that delay in seeking treatment was not driven by ignorance of where care could be found; rather, it reflected anxiety about what hospital attendance would communicate to the community and how the illness would be morally interpreted. The psychological burden of diagnosis thus appeared relational from the outset, distributed across the household rather than contained within the patient alone. This pattern is consistent with ecological models of health behaviour (Bronfenbrenner, 1979), which situate individual responses to illness within nested social systems.

Theme 2: Fragile Illness Acceptance and Early Non-Adherence

Several patient interviews described a period early in treatment when medication-taking was inconsistent. What was notable was how participants explained this: not as forgetfulness, and not as deliberate resistance, but as something closer to emotional non-adjustment or incomplete incorporation of the diagnosis into one’s sense of self. One patient explained: “In the beginning, I skipped some doses because I had not fully accepted my condition. I knew I had to take the medicine, but in my heart I was still refusing to believe it” (Patient 4, man, age 38).

The phrase “had not fully accepted” is analytically significant. It points to a gradual process of incorporating a TB diagnosis into one’s sense of self — a process that is separate from, and precedes, behavioural compliance. Early non-adherence appeared less as a knowledge deficit than as a struggle to inhabit an identity marked by illness and stigma. Medication-taking required not only knowing what to do, but accepting what treatment symbolised.

Family accounts added another dimension. Relatives described stepping in to monitor routines, remind patients, or respond to missed doses, suggesting that families temporarily carried the structural scaffolding of adherence while patients were still psychologically adjusting. One patient recalled: “If I missed taking my medicine, my mother would immediately scold me, asking why I did not take it” (Patient 5, woman, age 24). Across interviews, adherence became more stable once acceptance deepened. In this sense, acceptance did not always precede adherence chronologically; rather, family supervision often sustained the routine until emotional acceptance caught up. This dynamic parallels findings from comparable settings in which psychological burden is identified as a key early barrier to adherence (Aibana, Rybak, et al., 2020; Nasir et al., 2024).

Gender influences appeared evident among patient participants, with women more likely to describe non-adherence in terms of shame and concealment from extended family members, whereas men more often described early resistance as denial of illness severity.

Theme 3: Stigma as Social Identity Threat

Stigma ran through the interviews as a persistent undercurrent. It surfaced in embarrassment about taking medication where others could see, reluctance to allow health workers to visit the home, withdrawal from social life, and delayed care-seeking. Two patients described this experience. One man stated: “What made it difficult was the feeling of shame if others saw me taking TB medication” (Patient 3, man, age 45). Similarly, a woman stated, “I tried to hide my condition because I didn’t want people to treat me differently” (Patient 6, woman, age 62). The visibility of medication-taking mattered because it made the diagnosis socially legible. Being seen with TB medication risked public reclassification as contagious, shameful, or morally tainted. Stigma therefore operated not only through what participants felt internally, but through what treatment disclosed to others.

A family member’s account illustrated how the stigma threat extended beyond the individual patient: “He felt inferior and ashamed. He did not want to socialise because people said it was a curse, and he refused to go to the doctor. There is still a lot of stigma. Patients feel embarrassed and do not want health workers to visit their homes” (Family Member 1, man, brother of patient).

Taken together, these accounts indicate that stigma in this setting functioned as a social identity threat at the household level, placing the collective reputation of the family at risk. Managing adherence therefore required managing visibility, disclosure, and collective social standing — not only taking medication correctly. These findings align with existing literature on TB stigma (Chen et al., 2021; Sekandi et al., 2024), while adding a household-level dimension that is less commonly foregrounded.

Gender intersected with stigma in specific ways in patient accounts. Women described particular concern about community perceptions related to marriage prospects and social acceptability, while men were more likely to frame stigma in terms of occupational and breadwinner identity.

Theme 4: Cultural Explanatory Models and the Obstruction of Care

Before biomedical explanations for TB could take hold, many participants described interpreting the illness through cultural and moral frameworks in which tuberculosis was understood as a curse, a hereditary punishment, or the consequence of wrongdoing. These beliefs were not peripheral to treatment decisions; they shaped whether biomedical treatment appeared meaningful in the first place. For example, one participant recounted: “A relative refused [TB] treatment because he believed the illness was a curse” (Patient 1, man, age 28). If an illness is understood as the product of supernatural or moral forces, biomedical treatment can appear incomplete because it appears to address the wrong cause.

What enabled re-engagement with formal care in these accounts was not abstract education about pathogens, but concrete evidence that made an alternative interpretation of the illness credible. One family member explained: “To break the belief that it is a curse, health workers must show the laboratory results. The lab proves that this illness is not a curse. At first, we thought this illness came from something spiritual, not from germs” (Family Member 4, woman, wife of patient). Laboratory evidence served not simply as correction, but as a form of epistemic reframing: it introduced a biological explanation that participants and families could accept without necessarily abandoning the wider cultural worldview within which curses or moral causation still made sense. This distinction matters because it suggests that effective communication may depend less on dismissing cultural beliefs than on creating an intelligible bridge between them and biomedical care, a point consistent with existing Indonesian TB literature (Fitri & Krianto, 2023; Pratiwi et al., 2012) and with health communication frameworks that emphasise meeting patients within their explanatory frameworks rather than confronting them.

Theme 5: Family as Behavioural Infrastructure

If one finding most directly challenged conventional adherence frameworks, it was that family support was described not primarily as emotional encouragement but as practical organisation. Across interviews, family involvement resembled an operational system more than a background source of comfort. This was reinforced by patient statements, such as “The one who helped me the most was my family” (Patient 2, woman, age 32), and by the recollections of family members, as illustrated in this example: “We took turns bringing him to the hospital and collecting his medicine. The family must work together” (Family Member 5, woman, sister of patient). Family members coordinated transport, collected medication, managed household finances related to care, monitored daily medication routines, and intervened when treatment faltered. One family member articulated this obligation in terms that carried a strong moral character: “We had to make sure he took it every day, otherwise he would forget” (Family Member 2, man, father of patient).

Rather than being framed as respect for independence, allowing a patient to manage TB alone was perceived as abandonment. For example, one family member stated, “Family is the most important. If the family lets the patient walk alone, it is the same as abandoning them” (Family Member 1, man, brother of patient). In these accounts, adherence was not something the individual achieved with optional support from others; it was something the household actively produced through shared labour and relational obligation. This extends prior findings on family supervision in TB care (Charyeva et al., 2019; Chen et al., 2020) by offering a more process-oriented account of how families operationalise that support in everyday household life.

Negative dimensions of family involvement were also present, though less predominant in participants’ accounts. Two patients described family monitoring as emotionally difficult, at times experiencing it as surveillance rather than support. One patient also reported that family pressure to pursue traditional remedies before seeking formal care contributed to delayed treatment initiation. These accounts indicate that family involvement is not uniformly positive, and that programme designs engaging families must attend to the potential for family pressure to obstruct as well as facilitate adherence.

Gender also shaped how family involvement was experienced by patients. The women described family support primarily in terms of accompaniment and reminders, while the men were more likely to describe family members managing finances and logistics on their behalf. Family members who were women were consistently described as the primary day-to-day caregivers, consistent with broader patterns of gendered care work in Indonesian households.

Theme 6: Spirituality as Therapeutic Reinforcement

Spiritual practice appeared in the interviews not as an alternative to biomedical treatment but as something woven into it. Participants described prayer and faith as functional components of the treatment experience. For example, one patient stated, “Taking medication without prayer feels incomplete” (Patient 1, man, age 28). In a similar vein, a family member shared, “We believed that recovery needed both medicine and prayer” (Family Member 3, woman, mother of patient).

In one account, prayer completed the treatment act; in another, it supported acceptance and emotional endurance. One family member recalled: “The most important support was family and prayer. We kept praying so that he could accept his condition and make peace with it” (Family Member 2, man, father of patient). Spirituality did not merely comfort patients alongside what might be considered the ‘real’ work of treatment; it helped make treatment bearable and meaningful. Family narratives also suggested that shared prayer practices strengthened solidarity around the treatment goal, reinforcing the sense that illness was a burden to be carried collectively rather than individually. These findings are consistent with reports from comparable settings in which spiritual practice supports psychological coping and persistence in TB care (Nasir et al., 2024).

Across these six themes, the interviews suggested that tuberculosis treatment adherence in this setting was not experienced as an isolated act of willpower. Participants described adherence as a process that had to be worked into daily life through emotional adjustment, family coordination, management of social exposure, and interpretive frameworks that made treatment meaningful. As such, there was not a single determinant of adherence, but a set of interacting processes unfolding within household, community, and spiritual life.

Limitations

Several limitations should be acknowledged. The study was conducted in a single urban municipality (Kupang), East Nusa Tenggara, and involved eleven participants; the findings therefore cannot capture the full diversity of TB experiences across Indonesia’s many regional, ethnic, and socioeconomic contexts. The single urban setting limits what can be said about rural communities, where kinship structures, health system access, and the cultural texture of illness experience may differ substantially. The study was conducted in a predominantly Christian city in eastern Indonesia, and the role of spirituality in these accounts may reflect that specific religious context. The findings, therefore, may not transfer to predominantly Muslim or otherwise differently constituted settings. In addition, interview data rely on retrospective self-report and may underrepresent experiences participants found too painful, private, or socially risky to articulate. It is also possible that some accounts presented family support or spiritual practice in more coherent ways than they were lived day to day.

Each participant was interviewed once, with brief follow-up telephone contact for two participants to clarify specific statements; the study did not include repeat interviews or triangulation with other data sources, such as treating clinicians or community health workers, who might have offered an additional perspective on the accounts presented here. The purposive sample, recruited with the assistance of TB programme officers, may also have been weighted toward patients and family members who were already engaged with care, and this potential selection bias should be considered when interpreting the findings.

The small and purposively recruited sample also limits what can be said about gender differences, adherence trajectories, and the influence of living arrangements on treatment experience. These dimensions were addressed where possible in the analysis, but the sample size did not permit systematic comparison across subgroups. Future research with larger and more diverse samples would allow more sustained examination of how gender-related influences, household composition, and treatment trajectory interact with the relational processes described here.