Ten years of nurse-initiated antiretroviral treatment in South Africa: A narrative review of enablers and barriers

Background The roll out of nurse-initiated and managed antiretroviral treatment (NIMART) was implemented in 2010 by the National Department of Health (NDoH) in South Africa in response to the large numbers of persons living with HIV who needed treatment. To enable access to treatment requires shifting the task from doctors to nurses, which had its own challenges, barriers and enablers. Objectives The aim of this narrative is to review content on the implementation of NIMART in South Africa over the period 2010–2020, with a focus on enablers and barriers to the implementation. Method A comprehensive search of databases, namely, PubMed, Google Scholar and Cumulative Index to Nursing and Allied Health Literature (CINAHL), yielded qualitative, quantitative and mixed-method studies that addressed various topics on NIMART. Inclusion and exclusion criteria were set and 38 publications met the inclusion criteria for the review. Results Training, mentorship, tailored tuberculosis (TB) and HIV guidelines, integration of services and monitoring and support have enabled the implementation of NIMART. This resulted in increased knowledge and confidence of nurses to initiate patients on antiretroviral treatment (ART) and decreased time to initiation and loads on referral facilities. Barriers such as non-standardised training, inadequate mentoring, human resource constraints, health system challenges, lack of support and empowerment, and challenges with legislation, policy and guidelines still hinder NIMART implementation. Conclusion Identifying barriers and enablers will assist policymakers in implementing a structured programme for NIMART in South Africa and improve access, as well as the training and mentoring of professional nurses, which will enhance their competence and confidence.


Introduction and background
The implementation of nurse-initiated and managed antiretroviral treatment (NIMART) was a direct response to the high rate of persons living with HIV and requiring treatment. Initially, antiretroviral treatment (ART) was provided in hospitals and initiation was performed by doctors. With more patients requiring treatment, as HIV infections soared and doctors' capacity exceeded, a task shifting model was implemented, with nurses in the public sector having to initiate ART to scale up HIV treatment and increase access for more South Africans living with HIV. 1 Poor socioeconomic conditions and distances that patients had to travel to access care brought the need to decentralise HIV management services to primary healthcare (PHC) facilities. This, in turn, increased pressure to have more nurses trained to initiate ART and manage stable patients following national guidelines. 2 The evidence that task-shifting may improve health outcomes, quality of care and patient satisfaction, 3,4 together with the additional benefits of decentralisation of treatment 5 and the growing numbers of persons living with HIV in South Africa, necessitated the wide-scale implementation of NIMART training. The World Health Organization recommendations and guidelines for taskshifting 6 advocate that task-shifting should be implemented alongside efforts to increase the skilled workforce, health systems reorganisation and an enabling regulatory framework. Continued quality of care can only be maintained with standardised competency-based training, supportive mentoring and effective referral systems. 6 Whilst there has been evidence that NIMART-trained nurses can initiate and manage patients successfully, researchers cautioned that we may not know enough about key patient-, provider-and organisational-level enablers and barriers of wide-scale implementation. 7 It has been 10 years since the initial NIMART implementation in South Africa in 2010. Human immunodeficiency virus treatment and management guidelines have been revised several times since the ART implementation, and HIV care has been integrated into various other services such as general PHC, 8 tuberculosis (TB) management 9 and antenatal care. 10 Nurse-initiated and managed ART has also expanded to include the management of children and patients with virological failure. 11 With such wide-scale implementation and the evolving role of nurses in the context of NIMART and HIV management, it is inevitable that there may be challenges. Recent reviews that have summarised the enablers and barriers of the implementation of NIMART in the context of South Africa could not be found in the literature. It therefore became critical to search the literature in order to identify enablers and barriers and to make recommendations that will improve NIMART implementation.

Aim
The aim of the article is to review published literature on NIMART in South Africa, with particular focus on the enablers and barriers to implementation.

Method
In order to provide a comprehensive synthesis of the evidence and a broad perspective on NIMART, articles on the topic, its history and development were searched and presented in a narrative format. 12 A narrative overview or review is a nonsystematic narrative synthesis of previously published literature. 12

Search strategy and study selection
PubMed, CINAHL and Google Scholar databases were searched for relevant South African articles published between January 2010 and June 2020. Different search strategies were applied, using the MESH term combinations. In PubMed and CINAHL, we used Boolean operations such as (['nurseinitiated'] AND NIMART) AND 'South Africa') and in Google Scholar we used a string (NIMART South Africa). We identified additional records by reviewing master's or PhD e-theses, conference abstracts and published studies known to the authors. Relevant grey literature, such as Department of Health documents, was also included.
One of the authors and a research assistant screened the abstracts for relevancy. Articles were included if they met the following criteria: published in English, between January 2010 and June 2020 and reported studies conducted in South Africa. Records were excluded if the results did not relate directly to NIMART or if the study was not conducted in South Africa.

Data extraction and evidence appraisal
Data on the study aim, methods, sample and key enablers and barriers were extracted in tabular format. The quality of evidence was appraised using the John's Hopkins Evidence Level and Quality Guide. 13

Ethical consideration
This article followed all ethical standards for research without direct contact with human or animal subjects.

Results
A total of 479 records were identified: 7 from PubMed and CINAHL, 475 from Google Scholar and 15 through the authors of this article. After removing duplicates and excluding studies not relevant to the topic, 38 publications were included in our narrative literature review. The review includes qualitative, quantitative and mixedmethods studies and literature reviews that reported on evidence related to the NIMART implementation in South Africa.
Almost all the studies were classified as level III, B (nonexperimental studies of good quality). 13 Most of the quantitative and qualitative studies were descriptive in nature. However, the studies had sufficient sample sizes and provided reasonably consistent results and recommendations.
The studies' findings are presented narratively under the headings 'enablers' and 'barriers' to NIMART implementation. A summary of the included studies is provided in Annexure 1. A summary of the enablers and barriers is depicted in Table 1.

Enablers to nurse-initiated and managed antiretroviral treatment implementation
The Streamlining Tasks and Roles to Expand Treatment and Care for HIV (STRETCH) trial, conducted before the official implementation of NIMART in 2010, reported on key enablers in the South African setting. 14 Key components of effective implementation included tailored guidelines, training and support to build the clinical confidence of nurses and health services reorganisation. 14 19 and a course developed by the University of KwaZulu-Natal. 2 Across the courses, various challenges were reported such as difficulty to complete assignments, dispensing certificates and POEs for ART initiations. 18 Success stories included improvement in pre-and post-course knowledge and confidence and increased rates of POE completion because of a team of roving mentors. 2,18,19 Adequate training results in improved knowledge of HIV management, greater confidence and clinical competence, particularly if accompanied by mentoring. 14,20,21,22 A study conducted in an urban and rural district in the Western Cape found that the majority of NIMART-trained nurses had adequate HIV management knowledge and were very confident to manage adult patients. 22 The survey was specific to adult HIV management as nurses are only involved to a limited extent in the management of children in the Western Cape.
Nurse-initiated and managed antiretroviral treatment training leads to feelings of empowerment because of expanded roles. 20 Appropriate training and support can lead to increased quality of patient care, confidence and professional development. 17 Nurses may experience work satisfaction because of the difference they are making in patients' lives. 23 In the North West province, the ability to work independently boosted nurses' self-esteem and selfworth. 24 In Venda, some nurses reported that they felt proud that they were contributing to the NIMART programme. 25 In Johannesburg, the training of nurses in NIMART increased access to ART as shown by an increase in the number of monthly initiations. It also resulted in reduced workloads at referral facilities. 26 A qualitative study in Johannesburg reported that referrals to tertiary hospitals were significantly reduced after the introduction of NIMART. Nurses observed an improvement in the quality of life of their patients and the retention of patients in care, which they felt reflected the success of NIMART. 27 Similarly, in KwaZulu-Natal, a crosssectional study revealed that in 98% of the primary care clinics in one district, nurses initiated patients on ART. The majority of nurses also indicated that children were initiated in their clinics, with only some still being initiated by a doctor. 28 Accelerating NIMART in paediatric patients, coupled with mentoring and support were considered enablers in the provision of ART in infants and children living with HIV in the Eastern Cape. 29 Clinical exposure further contributes to knowledge and confidence as was found in a study conducted in the Western Cape where a higher caseload of HIV patients was associated with higher knowledge and confidence. 22 In the North West province, effective placement, specifically at Community Health Centres, contributed to POE completion. 24 The mentoring of nurses had a positive effect on patient care by improving the clinical skills of professional nurses and by raising the standard of HIV care in the Eastern Cape. 30 In Khayelitsha, Cape Town, a mentoring programme led to improvements in the quality of care nurses provided and their knowledge and confidence. 17 Nurses initiated 77% of ART-eligible patients after completing a mentorship programme and improvements in their ART management were observed. 17 If the majority of persons living with HIV are managed by nurses, doctors can attend to the more complicated cases, which, in turn, may result in better overall patient outcomes. Another study in the Western Cape found that a 2-week mentoring period was associated with greater HIV confidence compared with other periods. A period of 2 weeks of 'dedicated mentoring-time' with a mentor may be sufficient to complete competencies. 22 Jobson et al. 31 conducted a qualitative study, evaluating the role of mentoring in the support of the scale-up of ART across three provinces of South Africa. The study explored the role of targeted mentoring using a needs-based approach to support NIMART, pharmacy management and data management. 31 The study adopted a two-stage approach, which they classified as proactive and reactive mentoring. For NIMART-trained nurses, the proactive approach entailed having nurse mentors accompanying them during patient consultations, which provided support in transferring skills learnt during training to actual implementation when providing care. Participants stated that they had gained knowledge after training and mentoring, which empowered them to up-scale services through NIMART. The reactive role required the mentor (nurse) to be a problem solver and a source of support, as required at various stages during HIV care and management. This type of support is important for NIMART-trained nurses who were starting to initiate and manage patients on ART. Mentorship strengthened skills and knowledge transfer, allowed mentees to develop their own skills and provided a source for psychosocial support as mentees became confident and felt empowered to take on new responsibilities in providing HIV care. 31

HIV and TB management guidelines
The introduction of tailored guidelines such as the Practical Approach to Care Kit (PACK) was critical to assist nurses to implement NIMART. 14 A mixed-methods study conducted in the Limpopo province found that nurses practiced rational prescription and followed ART guidelines, with lower patient mortality (below 1%) and loss-to-follow-up rates compared with that of surrounding hospitals. In addition, 91.1% of NIMART-managed patients had undetectable viral loads after 1 year on treatment. Patients also reported high levels of satisfaction. 32 A study conducted in the North West province found that nurses preferred guidelines to be user-friendly, easy to follow and keep on their person and available in all consulting rooms. 33 Nurses also wanted to be supported and supervised until they are familiar with applying the guidelines in practice. They preferred continuous training and education about guidelines. Organisational and structural changes such as manageable workloads and improved communication were thought to enable adherence to ART and TB treatment guidelines. 33

Integration of services
Initially, NIMART was implemented as a vertical programme. Now it is increasingly being integrated into a range of primary care services. 8 The integration of HIV care into PHC services is reported as a structural facilitative factor for the implementation of NIMART. 27 The integration of HIV care into antenatal care and the introduction of option B+ meant that all pregnant women living with HIV needed to be initiated on ART. A retrospective record analysis conducted in the Gauteng province comparing time to initiation of ART amongst antenatal clients before and after NIMART implementation showed no significant reduction in the time to initiation on ART. 10 However, the study was conducted in the early stages of NIMART training and not all midwives may have been trained; therefore, they could only initiate clients on certain days.
Contrary to this, a study conducted in KwaZulu-Natal, soon after NIMART implementation in 2011, found that 97% of women were initiated on ART in antenatal care settings, illustrating a shift in care from ART clinics to nurse-managed antenatal clinics. This also resulted in reduced time to ART initiation from 38 to 4 days. 34 One study evaluated the effectiveness of the NIMART programme in the Waterberg district of the Limpopo province and found the number of patients initiated in the hospital dropped as did the number lost-to-follow-up, after the introduction of NIMART in primary care. 35

Monitoring and support
A study conducted in the Western Cape found that regular feedback about clinic and personal performance was associated with higher HIV management knowledge, 22 with a study in Limpopo highlighting that support from visiting doctors and management were viewed as very helpful, even if management could not resolve all their problems. 23 Support amongst nurses is a further enabler to the implementation of NIMART. 25 Supportive teamwork (ongoing support from facility managers and colleagues), motivation and support from mentors were found to be key contributors to POE completion in the North West province. 24

Barriers to nurse-initiated and managed antiretroviral treatment implementation
The implementation of NIMART may have been too hasty, not providing enough time for crucial capacity-building interventions such as mentoring and systems reorganisation. 20 Barriers identified in this review include: (1) non-standardised training and inadequate mentoring, (2) human resources constraints, (3) health system challenges, (4) lack of support and empowerment, (5) challenges with legislation, policy and guidelines and (6) patient-related factors.

Non-standardised training and inadequate mentoring
Inadequate or non-standardised training and the lack of continuous clinical mentoring and supervision by clinic managers were mentioned in most studies as barriers to the implementation of NIMART. 19,20,23,24,25,27,28,35,36,37 As mentioned before, NIMART training was, and is being conducted by various NGOs, the NDoH and private training providers. One of the challenges is that training is not standardised, ranging from a few days to a few weeks or even months, which makes it difficult to evaluate its effectiveness or quality. Each provider determined their own programme's content, duration, instruction methods and assessment strategies. This limits the ability to set guidelines for structured training and mentoring. In some cases, nurses initiated patients even before attending or successfully completing NIMART training. 19 The process of submitting POE's and doing an additional course in dispensing was found to be dysfunctional as many nurses could not complete it. 15,18,38 In the North West, prerequisites such as PACK training and IMCI were barriers to the completion of POEs. 24 In rural North West province, challenges relating to NIMART training included the lack of a standardised curriculum, lack of involvement of quality assurance bodies, nursing colleges and universities and inadequate continuous professional development (CPD). 39 Another barrier identified in the North West province was disorganisation at the level of the Regional Training Centre (not receiving POE's directly after the training) and at the level of the trainee (lack of planning). 24 This partly defeated the aims of NIMART, which was meant to be an intervention intended to improve healthcare access and equity, ideally without compromising the quality of care and a key strategy for expanding access to HIV treatment at PHC level.
There is a lack of evidence in South Africa regarding the effectiveness of different NIMART training programmes and the impact on patient outcomes. 2 One should also question the sufficiency of a once-off NIMART training and further explore the role of continuing education and CPD. A study conducted in the Western Cape found that NIMART nurses with recent training on guidelines (less than 3 years) had better knowledge compared with nurses who reported to be trained more than 3 years ago. 22 In general, NIMARTtrained nurses identified the need for continuing education. 2,19,20,21,35,39 Whilst NIMART training enabled access, a lack of mentoring following training was a barrier in terms of the gap between the number of nurses who received training and those who could initiate treatment. Despite the large-scale training to upskill nurses, not all nurses initiated ART after receiving training. A study conducted in rural North West province found that although NIMART increased access to care, there was no steady increase in the initiation of adults, children and pregnant women on ART. Initiation was especially low amongst children. This was despite the fact that 75% of nurses in 99% of healthcare facilities were NIMART trained. 40 Facilities were also performing below the targets for retention in care and viral load completion and suppression rates, indicating poor quality of care and non-compliance to guidelines. Another study conducted amongst NIMARTtrained nurses across 7 provinces found that of the 126 nurses sampled, only 79 initiated treatment and only 9 initiated treatment in children. 19 Factors contributing to low initiation rates include inadequate training, lack of confidence and lack of mentoring and supervision. A study conducted in the Western Cape amongst 77 NIMART nurses working across 29 healthcare facilities on factors influencing the knowledge and confidence of professional nurses prescribing HIV treatment concluded that training, mentorship and clinical practice experience were associated with confidence and knowledge. 22 A qualitative study conducted in Johannesburg found that a lack of mentoring was likely to have contributed to a lack of nurses' confidence to initiate ART. They recommended a nurse-mentor model where experienced nurses can supervise and support colleagues; as well as access to telephonic support. 20 Nurse-initiated and managed antiretroviral treatment-trained nurses in Venda reported that because of the limited 1-week didactic training, they ended up learning most of the competencies at work, highlighting the need for mentoring support. 25 The need for supportive mentoring was underscored by the fact that only 63% of the nurses who completed a training course in KwaZulu-Natal felt confident enough to initiate ART following the training. 2 With the escalating number of persons living with HIV requiring treatment, it became crucial to nationally upskill the nurses who were not initiating as a result of a lack of mentorship and standardised training.
Whilst some of the NGOs had a mentorship programme as part of their training, this was not a norm across all trainings.
The NDoH subsequently developed a national Clinical Mentorship Programme, which was aimed at providing practical on-site support to NIMART-trained nurses and ensure a competent and confident workforce. A manual was developed, with didactic content and practical tools that could be used to design, implement and evaluate a clinical mentorship programme at facility level and following clinical guidelines on HIV care. Of importance was that the manual could also be used by pharmacists, medical officers and other members of the multidisciplinary team providing HIV care services. 16 However, even with this Clinical Mentorship Programme in place, challenges with regard to mentoring exist. One of these challenges include that during proactive mentoring, it often happened that the mentors ended up delivering the actual service themselves and were sometimes not available when needed because of being responsible for several facilities. 31 There is also no standardised time or method for clinical mentorship and it varies from a minimum of 40 h in the Western Cape 17 to targeted mentoring in other settings. 15,31 Several studies identified key knowledge and confidence gaps amongst NIMART-trained nurses, further indicating a need for improved training and mentoring. A lack of confidence to manage children was identified in a study in the North West province. 39 Naude 21 similarly found low levels of competency in certain areas of HIV management, particularly for ART initiation and follow-up in children, in the North West province. In Venda and Limpopo, nurses reported that they faced problems in performing certain tasks such as obtaining blood from children. 25,41 Mashudu 35 also found that few children were initiated on ART in Limpopo. Low initiation rates amongst children should be investigated further. It may be that the policy in many provinces, such as the Western Cape, is that children must be managed by doctors or that paediatric NIMART training and mentoring is not sufficient. In a study to explore the experiences of healthcare professionals regarding the provision of ART for children in PHC settings in Nelson Mandela Bay Health District in the Eastern Cape, the need for training, mentoring and debriefing was expressed as one of the challenges related to providing decentralised ART to children. Nurses were apprehensive to work with children and noticed incongruence in the interpretation of ART side effects in children. 29 An analysis of nurses' queries to the National HIV and TB Health Worker Hotline showed that 66% of the queries received were from NIMART-trained nurses and most were related to ART initiation and adverse drug reactions. The most common knowledge gap identified was on the interpretation of blood results before initiation of ART, 42 which was also confirmed by Rasalanavho. 25 In the study conducted in the Western Cape, low confidence was reported in prescribing for concurrent illnesses and in identifying the signs and symptoms of immune reconstitution inflammatory syndrome. Nurses had less than optimal knowledge of virological failure and drug-drug interactions. 22 In KwaZulu-Natal, most NIMART nurses knew the correct ART regimens, ART eligibility criteria and when blood for CD4 count and viral load should be taken, although worrying gaps were identified. 28

Human resources constraints
Human resources constraints, increased workloads and administrative duties such as paperwork, were cited in many studies as negatively influencing the implementation of NIMART, 14,19,20,21,23,25,27,32,35,39 with doctors reportedly not fully supporting the NIMART programme. 25 Excessive workloads also interfere with the completion of mentorship programmes. 24 In one study, 55% of nurses reported seeing more than 30 patients a day and 30% saw more than 40 patients a day. 21 As a result of integration of HIV management in primary care, NIMART-trained nurses also have other tasks to perform. 22,25 Nurses reported the inability to delegate tasks to lower cadres and also reported having to do non-nursing tasks such as collection of drugs from depots 27 or using their own transport to collect drugs. 25 Despite human resource challenges, NIMART nurses display resilience. In the Western Cape, even though 44.4% of nurses felt that their workload was unacceptable, and 48.1% were dissatisfied with their work environment, salary and work hours; 88.3% were nonetheless still motivated to work. 22 The challenge of high workloads was mitigated by nurses in Venda through problem-solving and innovative strategies such as allocating different times for collecting tablets and reviews as well as group counselling. 25 The growing number of patients on ART necessitates looking at decentralising ART care further and utilising chronic care models for stable patients on ART to decongest clinics. 20 Several studies mentioned the utilisation of lower cadres of healthcare workers such as lay workers to trace patients, 27 the education of the community, increased community management of HIV and addressing poverty and stigma, as crucial to the continued success of the HIV programme in South Africa. 23,43 Health systems challenges Although integration of services can be an enabling factor, it is not without challenges. High HIV-TB co-infection rates and high antenatal HIV prevalence rates in South Africa necessitate integration of HIV into TB and antenatal care. This means that primary care nurses and midwives have to expand their competencies to include NIMART. These changes also have an effect on the scope of practice and the training of nurses. It may also impact the quality of care that can be provided to other patient populations in primary care and women's healthcare settings. A study conducted in the Free State to evaluate patients' perceptions of the effect of the integration of NIMART into primary care on quality of care showed that there was no decrease in patient satisfaction with staff. This was despite increases in patient numbers. However, satisfaction scores were lower for child health and chronic care patients (except TB), suggesting that there may be a knock-on effect on other services. 44 A study in the Limpopo province on the views of registered nurses of the prevention of mother-to-child transmission (PMTCT) programme revealed considerable challenges to integrate NIMART into antenatal and postnatal care. 43 These challenges related particularly to having to spend more time with patients when initiating ART during the antenatal period; the additional workload was not accompanied by additional support.
http://www.sajhivmed.org.za Open Access Integration of services should be accompanied by the provision of adequate resources and reorganisation of services. 14 However, several studies reported inadequate infrastructure and resources such as consultation rooms, inadequate workspace compromising patient privacy and confidentiality, as well as the lack of resources such as stationery, drugs, equipment needed for blood collection, telephones, electronic drug ordering systems, information systems and slow turnaround times of blood investigations. 15,19,21,23,25,27,32,33,35,41,43 Poor data management, including lack of compliance with standard operating procedures, incomplete records and inadequate audits hampered performance. 23 Nurses also experience poor referral feedback systems. 27 There is a necessity for quality improvement teams to ensure quality of care. 20

Lack of support and empowerment
Health system issues and lack of managerial support was identified in several studies. 20,23,25,33 In Venda, nurses reported a lack of support from doctors and refusal to see complicated cases requiring expert management. 25 Naude, 21 through a mixed-methods study, explored empowerment amongst NIMART-trained nurses. Quantitative findings revealed that nurses were only moderately structurally empowered.
With regard to psychological empowerment, professional nurses felt that they only minimally influenced their work environment. Management and organisational processes were identified as being central to the empowerment of NIMART nurses. 21 In the Eastern Cape, barriers to paediatric NIMART implementation included ineffective management, disharmony and non-conducive work environments. 29 A negative attitude towards the management of HIV and TB patients was identified by Mboweni and Makhado. 38 Negative attitudes and discrimination from non-NIMART-trained nurses and refusal to manage HIV positive patients was also reported in a study conducted in Venda.
Nurses reported that patients questioned their ethical values to keep information confidential and had negative attitudes towards the NIMART programme. 25 A study conducted in the Limpopo province found that nurses also had fears of infecting themselves with HIV, 41 indicating that there may be a lack of workplace-based support.

Challenges with legislation, policy and guidelines
The lack of enabling legislation and regulations, 45,46 the lack of clear guidelines and standard operating procedures, 37 unclear roles or scope of practice -nurses sometimes perform the roles of medical practitioners and pharmacists, 21 and salary or remuneration challenges were also reported. 32 Provincial governments and the NDoH were mandated to maintain records of all nurses authorised to prescribe ART and to communicate this to the South African Nursing Council. Anecdotal reports suggest that this has not happened. There is no formal scope of practice for NIMARTtrained nurses and they do not receive additional remuneration for expanded roles.
Some studies reported the lack of updated guidelines, particularly in rural areas. 28 A study in KwaZulu-Natal and North West provinces revealed that nurses may not be following HIV and TB guidelines. Barriers to adhering to guidelines included a lack of agreement with the guidelines or guidelines not being clear or understandable; insufficient knowledge as a result of not being updated or involved in guideline development; frequent guideline changes; poor motivation; and lack of supportive supervision. 9 Continuous revision of guidelines requires frequent updated training, but training programmes are not always communicated timeously to ensure that staff can attend. 43

Patient-related factors
Many studies also highlighted patient-related factors hindering the implementation of NIMART such as poverty, stigma, lack of transport and non-adherence to treatment. 25,27 However, the discussion of patient factors is beyond the scope of this review and could perhaps be a focus for future studies.

Discussion
It has been 10 years since the implementation of NIMART in South Africa. Evidence from observational studies indicates that there are several enablers and barriers to implementation. Based on the evidence presented here, we discuss implications for practice, education and research.

Implications for practice
The narrative review identified two frameworks developed from rigorous research that could be implemented in practice to strengthen NIMART training and implementation and empower NIMART-trained nurses. 21,37 The conceptual framework for strengthening NIMART training and implementation developed by Mboweni and Makhado 39 outlines structural attributes that advocates for adequate resources, infrastructure, supportive legislation and policy, effective training strategies and a supportive healthcare system culture. 37 This is supported by the framework of Naude 21 who outlined the following elements of structural empowerment: training and mentoring (knowledge, technical skills and professional growth); resources (human, equipment and supply chain); support (clinical supervision, guidance from managers and feedback); communication (information, protocols and guidelines) and power (formal in the form of role clarification and decision making and informal in the form of team acceptance and remuneration). 21 Many of these infrastructural and resource inadequacies may be addressed by investing in and improving the leadership and management of PHC services.

Implications for education
Mboweni and Makhado's framework identified that effective NIMART education and implementation requires (1) an integrated curriculum and effective training strategies for inand pre-service education and (2) a healthcare system culture that includes support through mentoring or coaching, discipline, communication and referral systems. 37 Although NIMART training is still being provided by NGO's and various private organisations, there is now a greater effort to incorporate comprehensive HIV training, including NIMART, in undergraduate pre-service programmes. 2,46,47,48 Evidence suggests that pre-service nurses still have inadequate HIV management knowledge. 47 Some researchers advocate for specialisation programmes for the management of chronic conditions, including HIV and TB. 21 One study recommended future research regarding the role of the postgraduate diploma in Primary Care Nursing in the implementation of NIMART as nurses felt that those who had the qualification were better equipped to manage patients on ART. 27

Implications for research
Further research is needed on the standardisation and effectiveness of NIMART training and mentoring programmes and the long-term impact on patient outcomes towards attaining national and international targets for HIV care and management. Evidence-based guidelines for the integration of NIMART into primary care services and continuous monitoring and support are urgently needed.

Strengths and limitations
The search strategy was not comprehensive and some relevant studies may have been excluded.
However, the authors are confident that we have provided an exhaustive description of barriers and facilitators to NIMART implementation in South Africa.

Conclusion
Although training, mentorship, guidelines, integration of services and monitoring and support have enabled the implementation of NIMART, several barriers such as non-standardised training, inadequate mentoring, human resource constraints, health system challenges, lack of support and empowerment, and challenges with legislation, policy and guidelines still hinder its effectiveness.
Various key role players such as the NDoH, the South African Nursing Council, training providers and researchers need to work together to standardise training and provide evidencebased guidelines for mentoring, integration of services and continuous monitoring and support.

Authors/year Study aim Methods Sample Key enablers and barriers Level and quality of evidence
Cameron et al. 19 To determine the percentage of nurses initiating new HIV positive patients on therapy within 2 months of attending the NIMART course and to identify possible barriers to nurse initiation.