Introduction

Rifampicin-resistant tuberculosis (RR-TB) until recently, required two-years of treatment, often resulting in debilitating side effects [1–3]. Additionally, patients are required to attend a clinic on a daily basis to receive treatment under directly observed therapy (DOT) [4]. These factors can negatively impact treatment adherence [5], resulting in loss to follow-up (LTFU) [6]. Other factors associated with LTFU are gender, age, previous TB episodes, substance abuse [7], and distance from the clinic [8]. Additionally, costs incurred by the patient during RR-TB treatment and care, including transportation costs [9], impact LTFU as there is often no social compensation for those expenses [10].

The reasons for LTFU are multi-factorial and attributable to individual and health system challenges [2,6]. In South Africa, a setting with a high burden of HIV, approximately 30% of RR-TB patients experience a treatment outcome of LTFU [11–13]. Two thirds of patients LTFU in Khayelitsha, South Africa, successfully endured and completed the painful six month injectable phase of treatment and only interrupted treatment once they reached the continuation phase [14]. The clinical status of most patients generally improves after a few months on effective treatment thus they are able to carry out normal activities of daily living by the time they reach the continuation phase. Anecdotally, most patients feel ready to resume their daily routines and responsibilities (i.e. working, studying, and caring for and supporting their families), long before the completion of their treatment course. However, these patients are still required to receive their oral medications in a clinic every single day under DOT, which becomes increasingly difficult to do over an extended period of 24 months [15]. DOT is the treatment strategy recommended by the World Health Organization (WHO) for RR-TB globally [16] and in the Strategic Framework for the management of RR-TB in South Africa[15]; however there is conflicting information about the utility of DOT in various programmatic settings [2,17,18]. Implementing differentiated models of care by intensifying support for RR-TB patients and offering patient centered care should be considered in RR-TB treatment programmes, as is common in HIV programmes[19– 22].

A pilot programme to integrate adherence support for RR-TB patient into the existing TB/HIV integrated adherence framework was undertaken in 2012 by the international medical humanitarian aid organisation Médecins Sans Frontières (MSF). The overall goal of this self-administered treatment (SAT) pilot programme, a differentiated model of care for RR-TB, was to encourage RR-TB patients to take treatment without having to attend the clinic daily. RR-TB counselors provide adherence counseling at the end of the completion of the intensive phase, at which time a patient is assessed for placement out of DOT. Patients were considered eligible for SAT if they were clinically stable (confirmed culture negativity and clinically improving), had few adverse events associated with treatment and if they had a favourable adherence record for RR-TB and concomitant diseases, as subjectively determined by the treating physician. If accepted for the SAT pilot programme, MSF RR-TB counselors conduct adherence counseling with the patients, stressing the importance of treatment and providing detailed information regarding the programme. Patient pill-boxes were allocated to ease the management of the RR-TB treatment supply at home. Patients in the programme were given the option of being supported by community care workers (CCWs). If a patient decides that they want CCW support, the CCW will visit them in their home once accepted into the programme to review the patient’s adherence. This process has been described in detail in a previous publication[17]. It was anticipated that this model of care would relieve some of the burden that receiving daily DOT places on the clinic staff, as well as on the patients. Previous publications and reports emerging from the programme showed that patients who received SAT did not experience increased rates of LTFU when compared to patients that received DOT [17,23].

To date there have been very few studies which have explored the experiences of SAT from the perspective of beneficiaries and those implementing the pilot programme. The aim of this study was to understand the experiences of those involved in the SAT pilot, including challenges and benefits, from the perspectives of patients, health care workers (HCWs) and CCWs.

Materials and methods

Results

We present quantitative and qualitative results below: quantitative data are taken from the semi-structured questionnaires, and qualitative data from the in-depth interviews and open-ended questions from the questionnaires.

Clinical and demographic characteristics

A total of 27 patients, 12 HCWs, and 44 CCWs were included in the study (Tables 1 & 2). Of note, 22 patients were HIV-infected. The median time from RR-TB treatment initiation to enrolment into the SAT pilot programme was 7.4 (5.7–10.8) months and 17 patients in the SAT pilot programme had received a monthly supply of RR-TB medication. Of the HCWs participating, 10 were female and the median length of time in their current position was two years. The CCWs had a median of six years working in the community, and 41 CCWs were female.

Download:

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

Table 1. Clinical and demographic characteristics of enrolled patients.

https://doi.org/10.1371/journal.pone.0203888.t001

Download:

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

Table 2. Characteristics of the enrolled patients who completed the in-depth interviews.

https://doi.org/10.1371/journal.pone.0203888.t002

In addition, a total of four women and five men from the quantitative sample took part in in-depth interviews (Table 2). The median age of these interviewees was 42.

Discussion

Our mixed-methods study describes the experiences of patients, HCWs and CCWs involved in self-administered RR-TB treatment in a peri-urban township with high rates of HIV co-infection in South Africa. The data suggest that the programme was acceptable and beneficial to patients because it gave them more autonomy over their treatment, allowed them to return to their daily activities and enabled them to better manage their side-effects. The programme was also believed, by HCWs and CCWs, to reduce congestion in the clinic and reduce the burden of clinic visits on patients.

The burden of RR-TB treatment on patients has been well-documented in South Africa and elsewhere. Living with dual diseases further adds to the load carried by patients [29]; 82% of those included in this study were HIV-infected. In this study, patients described DOT as restrictive and limiting daily activities of life. These challenges were reduced and in some cases relieved by enrolment into the SAT pilot programme. A previous study which investigated whether patients preferred home or hospital based care showed the acceptability of and preference towards home-based care [8]. Literature reflects on the complexities DOTs brings to patients’ lives [2,5] but it lacks discussion regarding the impact of SAT on the lives of those patients’. To our knowledge this is the first study to investigate health care providers and patients perspectives of a SAT pilot programme.

The majority of participants expressed that SAT motivated patients to have autonomy over their treatment journeys. The CCWs also felt motivated and empowered through their provision of care to these patients in need. The majority of HCWs and CCWs shared a similar, unfounded perception that LTFU rates were declining and cure rates were increasing among the patients in the SAT pilot programme; however this is not backed up by previously published quantitative data [17]. This presumption alone, given the high rates of LTFU and mortality with concurrent high rates of HIV co-infection, may have motivated these care providers in the provision of support to patients.

Implementing pilot programmes such as SAT require thorough training of staff, including refresher trainings, as well as clear explanations to patients. This is to ensure that implementers and beneficiaries alike understand the purpose of the pilot programme and the eligibility criteria [30]. Whilst our data show that almost all patients were appreciative of SAT, they were not always sure why they had been enrolled. This could point to a gap in the quality of the counseling provided with the SAT pilot programme or a lack of clarity during the recruitment phase. The lack of a clear understanding regarding the eligibility criteria among HCWs and CCWs reflects the need for enhanced training before implementation and enhanced follow-up afterwards. Studies conducted in South Africa show that task shifting patient support activities to community workers is an essential component of differentiated HIV/TB care in primary health care settings [21]. However, task shifting requires training packages that include supervision, mentoring and support [31–34]. This need is clearly reflected by several patients who reported that they did not see their CCW regularly, suggesting that more clarity about the CCWs role and relationship with patients is needed to ensure sufficient follow-up. It is important that CCWs receive frequent training on the provision of patient support, and that they themselves stay motivated and function as a team in order to stay engaged in their work. Finally, a study conducted in the Western Cape showed that combining quality improvement programmes alongside pilot programmes ensures improved outcomes and could be reflective of the limitation in the Khayelitsha setting [30].

Differential service delivery is becoming increasingly common in many settings with large HIV and TB burdens, but has not been widely implemented in RR-TB programmes. Differentiated models of care, many of which have been piloted in Khayelitsha and in other settings in Southern Africa, include clinic- and community-based clubs and Community ART Groups (CAGs) [19,20,35,36]. This study provides policy makers and programme implementers with information on a model of care which could be included in a differentiated model of care package for RR-TB. Such models should be considered in settings in which there are high burdens of RR-TB and limited public health resources. Whilst we recognise that our results are from an urban setting where it is easier to access patients at home than it may be in remote, rural areas, we believe this level of task-shifting can happen in other settings too. Our study showed that most but not all patients believed that SAT was beneficial, thus it is important to ensure that patients are involved in discussions about which model of care is best suited to them. Interestingly, whilst patients appreciated the programme, they did not think it should be expanded to everyone, again highlighting the importance of having clear recruitment and eligibility criteria and processes. This could also suggest a sense of pride and ownership over the programme, with current SAT beneficiaries not wanting the programme to be rolled out to others.

The data highlighted the significance of having someone, who was usually a relative, to support the patient with taking RR-TB treatment at home. This support person typically offered psycho-social support as well as daily reminders to take treatment, or provided help with preparing meals after taking treatment to help ease side-effects. The key role of support people or ‘treatment partners’ as they are often referred to in HIV programmes, has also been explored elsewhere [37]. The importance placed on these support networks suggests that they are essential and need to be considered when implementing similar programmes in other contexts.

The majority of the participants expressed that they were satisfied with this pilot programme and that it should be continued with little modification made to the design. Some patients reported that they sometimes forgot or were unable to take their treatment, and still struggled at times to manage their side-effects, showing that adherence remained challenging for some. This phenomena is also reflected in the data for other long term therapies [38].

These findings highlight the need for continued adherence support for patients once they are enrolled onto SAT, and the need to address the reasons for continued treatment interruption [39]. The CCWs highlighted some well-documented challenges faced by patients in the programme [1,29]. These challenges did not speak to the implementation of the programme, however, but to challenges faced by RR-TB patients in general. Whilst this study has shown that it is both feasible and acceptable for RR-TB patients to self-administer their RR-TB treatment at home instead of attending a clinic, there is still the need for improved, shorter and injectable-free RR-TB regimens with fewer side effects, fewer drugs and improved efficacy [40]. This differentiated model of care could also be adapted and offered as an option to patients receiving the short course regimen which has now been rolled out in the Western Cape and globally. As our data showed, many of the ongoing challenges faced by patients, such as side-effects and pill burden could not be resolved through the provision of SAT alone. Another point which was mentioned in this evaluation by CCWs, but not discussed at length in this paper, was the effect of substance use upon patients’ treatment adherence. There is a need for programmes which specifically focus on the provision of support to patients with substance use disorder (SUD) in order to address the multi-faceted adherence challenges SUD poses, as has also been documented in HIV cohorts [41].

There are several strengths and weaknesses to this study. As we only included patients who were still enrolled in the SAT pilot programme, and currently employed HCWs and CCWs, one of the strengths is that we did not have challenges with recall bias. Although we had a small sample size, we feel that the patients enrolled were representative of the RR-TB patients enrolled in SAT, as there were few eligible patients due to the small numbers of patients enrolled in the programme. Additionally, we were able to triangulate the data from the different groups. One limitation was a potential bias in the selection and recruitment of participants. The overall SAT pilot programme excluded patients with adherence challenges, who would be considered the most vulnerable patients, therefore the findings from this study and pilot programme cannot be generalised to all patients with RR-TB. Future studies should investigate programmes which include patients that might benefit more from this pilot programme, such as those with poor adherence. Additionally, those who participated in this study might have been more likely to discuss the pilot programme favorably and might have responded with answers they think the interviewer wanted to hear, even though efforts were taken during the consent process to ensure that this did not happen.

Conclusions

Findings from this mixed-methods study showed that the SAT pilot programme, a differentiated model of care for the treatment of RR-TB post completion of the injectable phase of treatment, was considered to be beneficial from the perspectives of patients, HCWs and CCWs. These findings should be interpreted with caution as the SAT pilot programme only enrolled adherent patients and thus the views represented here are not from patients who had adherence challenges, although such patients may benefit more from such a programme in future. Overall, patients expressed that they felt motivated and that it was easier to take their treatment and manage their side effects at home. Additionally, HCWs indicated that SAT alleviated pressure placed on them and the clinic allowing them more time to spend with sick patients. Finally CCWs felt empowered and motivated when providing support to patients enrolled in the programme, and highlighted that their involvement led to substantial knowledge gain. This differentiated model should be considered in other settings with high burdens of RR-TB and HIV and implementation should be coupled with thorough sensitisation, training, and mentoring to ensure that patients and implementers understand the purpose of the programme and who is eligible for participation. The ongoing challenges associated with RR-TB treatment, including adherence and side-effects, highlight the need for new and improved regimens with a shorter treatment duration. Future studies will be needed to determine the feasibility and acceptability of this model in other settings with RR-TB.

Supporting information

Acknowledgments

The authors would like to acknowledge the Provincial Government of the Western Cape and Cape Town City Health for collaboration in the implementation of the SAT pilot programme and for permitting us to conduct this research study. Importantly, authors would like to thank Khayelitsha HCWs, CCWs and RR-TB counselors who were integral to the implementation of the SAT pilot programme as well as the RR-TB patients residing within Khayelitsha. Finally, we would like to acknowledge the study team who were integral in recruiting and interviewing patients, HCWs and CCWs.