Introduction

Multidrug-resistant tuberculosis (MDR-TB) is a major public health concern that is threatening global TB control. MDR-TB is defined as strains of TB resistant to the two most effective first-line anti-TB drugs, isoniazid and rifampicin [1]. The disease is often a consequence of inappropriate or interrupted treatment of drug-susceptible TB. In 2018, in Eswatini an estimated 509 patients were notified to have confirmed drug-resistant (DR-TB) compared to 110 DR-TB patients in 2008; that is more than a quadruple increase over a decade [2,3]. The rural areas in Eswatini carry a disproportionately high burden of MDR-TB. This can likely be ascribed to the primary health clinics (PHC) often being far away and geographically inaccessible from patients’ home. Rural health services are also typically characterised by a lack of frontline TB human resources for health (HRH) [4,5].

The control of TB has been highlighted as a priority in the post-2015 global TB strategy (the End TB Strategy) and the Sustainable Development Goals agendas [1,6]. In 2018, Eswatini reported a 66% treatment success rate for MDR-TB patients (2016 cohort), that is, well below the World Health Organisation (WHO) target of 75% or higher [3,7]. As a coping mechanism to address MDR-TB treatment access and HRH challenges in the predominantly rural Shiselweni region, Médecins Sans Frontières (MSF) established a community-based MDR-TB treatment model in 2008. The main feature of this model is the task-shifting of directly observed treatment (DOT) supervision and MDR-TB injection administration responsibilities traditionally restricted to professional nurses, to incentivised trained lay community members within the existing National TB Control Programme (NTCP) [4].

In this model of care, MDR-TB patients discharged from the MDR-TB inpatient hospital are linked to a local lay community member known as a community treatment supporter (CTS), of their choice. The selected neighbour must have sufficient literacy skills to be able to comprehend the English training manual and document administered injections. Instead of making trips to the, MDR-TB patients receive their daily injections and DOT from CTSs in their (the patients’) homes. Core clinical decisions such as monitoring of MDR-TB treatment and progress remain the purview of formal facility-based professional healthcare workers.

CTSs are typically recruited and trained to focus solely on MDR-TB. Their training, coordinated by community MDR-TB nurses, comprise on-the-job practical learning followed by a 3–5 days theoretical workshop at a later stage. During the first component of the training, the CTSs first practice safe injection techniques on orange fruits before administering injections to their patients. Before starting to administer injections on their own, CTSs should complete at least three observed injections assessed by a community MDR-TB nurse. Considerable emphasis is placed on drawing the right dose of the injectable drug, TB infection prevention and control (IPC), DOT and waste disposal. Newly recruited CTSs wait for up to a month to attend the theoretical component of the training in groups of ten.

The second component of the training focuses primarily on theoretical themes relating to MDR-TB epidemiology, transmission, diagnosis, treatment/strategies (DOT), safe injection handling, adverse drug reactions, and CTSs’ responsibilities including patient confidentiality and procedures for patient referral to community MDR-TB nurses. The training is conducted by members of the formal health services through classroom instruction, group work and open discussions. Although the workshop uses pre- and post-training assessments, no certificates or any formal form of accreditation are given to CTSs on completion of the training. Each CTS receives a monthly stipend and irregular on-going supervisory visits from and at the discretion of the community MDR-TB nurse.

Although task-shifting in TB control is not a new concept in Eswatini, a key tension in the delegation of MDR-TB injection administration responsibilities to CTSs has been the absence of a national policy framework to regulate their recruitment, training and accreditation. Professional bodies have raised ethical concerns about standards of care and safety risks for both patients and CTSs ranging from potential errors in dosing to transmission of infections through unsafe injection handling and inappropriate community MDR-TB IPC practices [8]. As a result, there are fears that task-shifting injection administration to CTSs may create a two-tiered system of MDR-TB management, with “superior” and “inferior” tracks.

Previous extensive systematic literature reviews on task-shifting have established the safety, effectiveness and acceptability of using appropriately trained and supervised lay community health workers (CHWs) in the provision of injectable contraceptives [9] and delivery of vaccines and medicines to mothers and children through compact pre‐filled injections [10]. The experience in this case study setting is important given the paucity of programmatic experience globally in the provision of MDR-TB injections by lay CHWs.

This study was conducted to assess the KAP levels of CTSs that may facilitate or impede the supervision of DOT and administration of injection treatment in the community-based MDR-TB programme. Results will help in formulating recommendations for optimising community-based MDR-TB management training for CTSs in the Shiselweni region.

Methods

Instrument development and measures

An interviewer-administered structured questionnaire consisting of 64 items (S1 File) was developed based on literature review and CTS training materials and job descriptions [12–16]. The content validity of the questionnaire was assessed by expert opinion. The internal reliability of the KAP scale was found to be acceptable and had a satisfactory discriminating power with subscale Cronbach’s alpha co-efficient for knowledge, attitude and practice of 0.72, 0.68 and 0.62 respectively [17–21].

The questionnaire collected socio-demographic details as shown in Table 1. There were 25 knowledge-related questions that examined the definition of MDR-TB, its aetiology, transmission, main symptoms, diagnosis, treatment, safe injection handling, and IPC as shown in Table 2. Respondents were asked to choose the correct response on a given statement on a scale that ranged from 1 (“yes”), to 2 (“unsure”), and 3 (“no”). Responses to knowledge questions were assigned a score of 1 for correct and 0 for inappropriate or uncertain responses. CTSs’ attitudes towards community-based MDR-TB management were measured using 21 statements. Participants chose either “strongly agree” (5 points), “agree” (4 points), “unsure” (3 points), “disagree” (2 points) and “strongly disagree” (1 point) on a 5-point Likert-type scale as shown in Table 3. CTS’s self-reported practices regarding MDR-TB were assessed based on responses to 12 questions as shown in Table 4. The self-reported practices were verified by carrying out observations in a sample of 20 CTSs using a structured observation checklist, as shown in Table 5.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Socio-demographic characteristics.

https://doi.org/10.1371/journal.pone.0271362.t001

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. CTSs’ knowledge about MDR-TB.

https://doi.org/10.1371/journal.pone.0271362.t002

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. CTSs’ attitude towards MDR-TB.

https://doi.org/10.1371/journal.pone.0271362.t003

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Self-reported CTSs’ community based MDR-TB practices.

https://doi.org/10.1371/journal.pone.0271362.t004

Download:

• PNG
  larger image
• TIFF
  original image

Table 5. Community based MDR-TB practices observed at patients’ homes.

https://doi.org/10.1371/journal.pone.0271362.t005

The questionnaire was availed in both English and the local language siSwati. The instrument was pretested for practicality among 10 CTSs who were no longer administering MDR-TB injections and were excluded from the study. A structured observation checklist (S2 File) for verifying CTSs’ self-reported community-based MDR-TB management practices was developed based on a schedule used by community MDR-TB nurses in supervising CTSs and literature review [22–25]. The checklist comprised of 26 items recording CTSs’ TB IPC, DOT, and injection administration practices.

Results

Table 1 presents a summary of the demographic characteristics of the study sample. A large majority of participants (n = 78; 95.1%) were female and 33 (40.2%) were older than 50 years. Half (n = 41; 50.0%) of the respondents had low literacy and numeracy skills (primary education or lower) compared to the other half with secondary or higher education.

The length of service of respondents administering MDR-TB injections ranged from 1 to 8 months, with a mean duration of 5.8 months. More than one-tenth (n = 12; 14.6%) of newly recruited CTSs administering injections had not yet attended the 3–5 days theoretical training on community-based MDR-TB management.

Predictors of good community-based MDR-TB management practice

Socio-demographic factors, i.e. duration administering injections, having attended MDR-TB training, sound MDR-TB knowledge and positive attitudes, were considered as potential explanatory variables for good community-based MDR-TB management practice. A bivariate logistic regression analysis (Table 6) found no significant association between individual socio-demographic variables such age, level of education, duration administering injections, MDR-TB training, knowledge and attitudes and good community-based MDR-TB management practice. A binomial logistic regression analysis revealed that after controlling for other variables in the model, none of the predictor variables were statistically significant.

Download:

• PNG
  larger image
• TIFF
  original image

Table 6. Bivariate and binomial logistic regression predicting community based MDR-TB management practice among CTSs.

https://doi.org/10.1371/journal.pone.0271362.t006

Discussion

This study contributes to extant knowledge about task-shifting by assessing the KAP of CTSs in community-based MDR-TB management in the Shiselweni region. While the overall level of CTSs’ KAP relating to community-based MDR-TB management was satisfactory, the study identified some important gaps that deserve attention, in particular, the timing, adequacy, design and content of the training provided to CTSs.

The overall satisfactory KAP scores in this study masked substantial weaknesses in some aspects of CTSs’ community-based MDR-TB management. Regardless of socio-demographic characteristics, some respondents were not aware of the definition of MDR-TB, its transmission, the rationale behind swabbing before administering an intramuscular injection and demonstrated unfamiliarity with drugs and regimens used in the treatment of MDR-TB. Similar results were found among healthcare workers in other countries [15,16,27,28]. Failure to identify medication names and fundamentals behind MDR-TB treatment care plan may adversely affect the ability of CTSs to recognise the doses, schedules and possible drug side effects during their interaction with patients.

Given CTSs’ limited formal education and lack of familiarity with special medical terminology and technical concepts, the present study provides impetus for developing appropriately tailored training material on community-based MDR-TB management. The use of lay conceptualisations and illustrations, and where possible, translated into the local language rather than using training packages developed for formal healthcare workers is recommended. The training course should integrate the theoretical and practical components to enhance a deeper understanding of community-based MDR-TB treatment concepts before delegating injection administration tasks to CTSs. Refresher training should be considered to be as important as the initial training.

The study findings highlighted salient erroneous and potentially stigmatising attitudes among CTSs. For instance, a substantial proportion of respondents inappropriately perceived that MDR-TB patients are to blame for their own condition. Similar to findings reported in previous studies among professional healthcare workers, most CTSs in the current study (86.6%) perceived themselves to have a high occupational risk of acquiring MDR-TB infection [26,29]. Studies from Ghana [30,31] and South Africa [32] have suggested that fear of infection is a major cause of TB stigma and can adversely influence patients’ behaviours in accessing MDR-TB services [33–35]. These stigmatising attitudes should be addressed in the CTS training design to ensure the development of appropriate relationships between CTSs and their patients [36]. These findings also call for further qualitative research to assess the grounding for MDR-TB-associated stigma among CTSs.

The use of cross ventilation by opening windows and doors in this study was higher compared to observations in studies in Ethiopian (89.2%) and South African (69.0%) hospitals [37,38]. Nevertheless, of concern was the finding of low compliance with basic hand hygiene practice (after injection administration) among CTSs–similar to other studies [26,39]. Key strategies in this setting include the reinforcement of educational initiatives with written reminders for CTSs to recognise hand hygiene opportunities as well as the availability and use of low cost alcohol based hand rub to interrupt the cross-contamination chain [26,39–41].

The observations conducted, albeit limited in number, found the overall intramuscular injection practices of CTSs to be generally satisfactory. These findings were comparable to results reported among professional healthcare workers in a large Indian hospital [42] and CHWs administering injectable contraceptives [9,43]. In the current study, CTSs invariably administered injections and oral medications daily thereby optimising MDR-TB patients’ adherence to treatment plans. Nevertheless, and similar to results from studies in Kenya [44] and Nigeria [45], of concern was the underreporting of needlestick injuries by CTSs. Needlestick injuries among healthcare workers can be considered as a cardinal indication of unsafe injection handling practices. Future research should identify strategies to reduce needlestick injuries and improve reporting among reporting-averse CTSs.

Without an enabling national policy on task-shifting, the results highlight potential patient safety and liability risks for the lay CHWs who undertake the delegated tasks [8]. With many health facilities stretched thin by the COVID-19 pandemic, task-shifting to appropriately trained, equipped and adequately supervised lay community members will optimise delivery essential TB treatment to communities. Apart from that, lay community members may play a critical role in facilitating, administering and expanding the reach of the COVID-19 vaccine. As such, there is need to develop conducive task-shifting guidelines to regulate the careful selection, appropriate training, accreditation and continuous supervision of CTSs.

In the present study, a majority of CTSs were female, indicating patients’ gender preference in selecting their DOT and MDR-TB injection treatment provider. Research elsewhere has demonstrated that TB patients ought to pick the supporter of their inclination as selection of treatment supporter outside the health system does not adversely affect TB treatment outcomes [46]. Nevertheless, future studies could explore strategies that can strengthen gender equity in the recruitment of these frontline HRH and support male involvement in community-based MDR-TB care.

Converse to findings from previous research, there was no statistically significant correlation between individual socio-demographic variables, knowledge and attitudes and good community-based MDR-TB management practice [16,47]. Future research with a sufficiently larger sample is required to better establish the predictors of good community-based MDR-TB management practice.

With this study mostly founded on self-reported practice and compliance, social desirability bias may have occurred. However, this was partially countered by carrying out structured observations of CTS administering MDR-TB injections to complement survey responses. Respondents were also assured of anonymity of the questionnaires, reporting of findings using aggregate analysis and that the outcomes of the study would not affect their job security or incentives in any way. Although verification of self-reported practice was limited to 20 CTSs due to logistical feasibility, the observations provided the researchers with reasonable insights into CTSs’ self-reported KAP and what they actually practice in community-based MDR-TB management.

Conclusion

The study results indicate that in this setting there is the need for considerable strengthening of initial and on-going theoretical and in-service training programmes to reinforce awareness, address gaps in current knowledge and dispel misperceptions and potentially stigmatising attitudes regarding community-based MDR-TB management among CTSs. Taken together, these findings raise questions about the timing, adequacy, design and content of community-based MDR-TB management training provided to CTSs.

Although some problems remain, including the absence of a formal regulation framework for task-shifting and the limited literacy and numeracy skills of most CTSs in this study setting, trained lay CHWs can be an option to complement an overstretched health workforce in providing DOT supervision and administration of intramuscular MDR-TB injections at community level.

Supporting information

Acknowledgments

The authors thank the NTCP and MSF for authorising the research. We also thank the community MDR-TB nurses and CTSs in the Shiselweni region for their support and participation in this study.

Availability of data and materials

The data analysed during this study are not publicly available as individual privacy would otherwise be compromised. Data is available on request from the corresponding author Ernest Peresu.