What will it take to put millions more South Africans onto antiretroviral therapy?

This article originally appeared in HIV & AIDS treatment in practice, an email newsletter for healthcare workers and community-based organisations in resource-limited settings published by NAM between 2003 and 2014.
This article is more than 12 years old. Click here for more recent articles on this topic

For their support in accomplishing our coverage of the 5th South African AIDS Conference HATIP would like to thank:

  • Lance Sherriff for reporting and editorial assistance.

  • Dr Francois Venter and the South African Clinician’s Society for their help with travel and lodging.

  • Fellow rapporteurs for the clinical track: Dr Nesri Padayatchi of CAPRISA; Dr Kedar Mate of the Institute for Healthcare Improvement; Dr Koleka Mlisana University of KwaZulu-Natal; Dr Sengeziwe Sibeko CAPRISA; Dr Doug Wilson, of Edendale Hospital, KZN.

  • HATIP would also like to thank: Professor Yunus Moosa, University of KZN; Dr Lucie Cluver, Oxford Univerity; Dr Ahmad Haeri Mazanderani of the Good Start Foundation.

“The rapid expansion of antiretroviral therapy (ART) programmes within a limited budget will require innovative and cost-effective approaches to service delivery,” Professor Yunus Moosa of the University of KwaZulu Natal said during his plenary presentation on the last day of the 5th South African AIDS Conference. Some of the innovations that could help South Africa meet its massive need for treatment that were discussed over the course of the conference included making better use of the private sector, task shifting, decentralisation of services to primary health clinics, improving service quality (and efficiency), and making health services more user-friendly for the patient on life-long chronic care.

After a slow start, South Africa has within a fairly short time launched the largest ART programme in the world, putting over 1.4 million people onto treatment, 400,000 of whom initiated within the last year. This is remarkable but the programme must continue to grow. It has not yet reached its targets for getting all of those who currently qualify for ART onto treatment — and will need to almost double that number by the end of 2012. If the country follows WHO’s 2010 ART guidelines and increased the CD4 cell count threshold for initiating ART to 350, the number qualifying for treatment would immediately jump significantly. One can expect most of those people to thrive and continue on care, while most of the remaining South Africans currently infected will also eventually be in line for treatment.

How can the system manage, when there are already some indications there may be a decrease in quality of care and retention in care as patient volume on ART in some facilities has increased. How will South Africa deliver ART to so many in need?

To come up with potential solutions, it may help to identify and better understand the obstacles that limit the successful delivery and scale-up of ART services. Prof. Moosa described a number of patient-related, programmatic and provider-related challenges to expanding care.

Patient-related barriers to the delivery of care



Social attitudes that suggest that having a particular illness or being in a particular situation is something to be ashamed of. Stigma can be questioned and challenged.

retention in care

A patient’s regular and ongoing engagement with medical care at a health care facility. 


A healthcare professional’s recommendation that a person sees another medical specialist or service.

linkage to care

Refers to an individual’s entry into specialist HIV care after being diagnosed with HIV. 


When blood pressure (the force of blood pushing against the arteries) is consistently too high. Raises the risk of heart disease, stroke, kidney failure, cognitive impairment, sight problems and erectile dysfunction.

A number of patient-related factors can serve as obstacles to HIV testing and counselling, such as making referrals, ensuring strict adherence to medication and retention in care. For instance, delivery of care for HIV may be complicated simply by the age at which the disease affects most people — during the prime of their lives. Good health seeking behaviour is less likely during at a stage of life when people may still be emotionally immature, with a “reckless, indomitable attitude to life leading to an immortality complex. Death is not somethign they want to discuss or think about,” said Prof. Moosa.

Likewise, he noted that poor social and financial support complicates the lives of many HIV-positive people. When dealing with basic survival issues, such as the need to keep a roof over one’s head or to feed one's children, the need to deal with one’s HIV may seem quite remote. When managing so many competing priorities and other crises, making clinic appointments and staying on treatment can get put on the back burner.

It doesn’t help matters when clinical services are not easily accessible or user-friendly for the people who need them. Patients often risk losing their jobs because they time wasted sitting in queues at clinics waiting for medication.

“The stuff we want as healthcare workers is not what the patient wants,” said Professor Francois Venter. “They want appointments and they want clinic hours when they don’t have to be at work. It’s good to say we have this patient-centered approach but we’ll only do it from 8 am to 4 pm during the weekday - it’s not patient-centred! That’s healthcare worker centred.”

The topic of how patients want their healthcare delivered was addressed by a number of speakers as well as a dedicated session at the conference. Prof. Moosa stressed that the programme has to design services “for the lifelong delivery of care [which means it has to] seamlessly integrate into the routine of life. It’s critical to ensure each visit is a positive experience,” he said.

Much was made of evidence several years ago showing treatment adherence was better in Africa than in industrialised countries. Indeed this was important because it soundly refuted the racist constructs of those who said that ART would never work in Africa. Nevertheless, there is evidence showing that adherence declines over time in a resource-limited setting, so programmes need to consider ways to make adherence simpler, such as the use of aids like pill-boxes.

Without adequate support, adherence may be complicated by the not-uncommon changes in the packaging, strength or appearance of tablets as the programme switches to drugs supplied by different generic pharmaceutical companies. Packaging can also enhance adherence (co-packaging) and delivery. The switch to using fixed dose combinations (especially once a day) should also make adherence somewhat easier.

But stigma topped Prof. Moosa’s list of patient-level barriers to care. Social stigma against HIV has a number of sources, and affects access at all levels. People living with HIV are stigmatised due to often irrational fears about how they might spread the infection. They are also stigmatised because HIV is associated with already stigmatised at-risk populations, including men who have sex with men, people who use drugs and sex workers, and linked with sexual promiscuity. Prejudice and moral judgements about how HIV is acquired dehumanise people as less deserving of compassion or care, and it is important to note that stigma is also usually internalised, leading to, or worsening, already existing feelings of depression and low self-worth.

Regardless of the source, stigma seriously affects health-seeking behaviour and particularly obstructs the main entry point into care — HIV counselling and testing (HCT). It may also be part of why many people do not achieve an effective referral after testing positive.

Social stigma often seriously impacts on the support structure available to someone who tests positive. It also causes the person with HIV to fear loss of their existing social support structures. These fears are not entirely unfounded, as there are people who act out on social stigma with violence. Women are particularly threatened by violence. Given such fear, it is easy to see how stigma could lead to poor adherence to ART and to medication for prevention of parent to child transmission, when people fear that pill taking will arouse suspicion.

To help reduce HIV-related stigma, Prof. Moosa said that HIV needs to be demystified “to be as matter of fact as possible. Health care workers need to be part of the solution and not part of the problem when it comes to stigma. Education is the only way to deal with stigma.”

That may be part of the solution to addressing stigma due to ignorance about the risk of transmission. But it won’t address the moral judgments, stigma, prejudice and violence directed against the vulnerable or marginalised populations in Africa.

“Between 2001 and 2011, the political environment in South Africa has vastly improved leading to the availability of ART, PMTCT and children on treatment. But the social environment, with stigma and violence among young people, has remained the same ... or has it worsened?” said Mark Heywood of Section 27, giving the second part of the Nkosi Johnson Memorial Lecture. He cited the case of Noxolo Nogwaza, a young lesbian who was murdered on April 24, 2011, “one of many,” he said.

Heywood believes the cycle of violence and vulnerability can only be broken by demanding that everyone be given his or her basic human right to a decent education, including adult education. But it will take time to implement and years to produce those outcomes.

In the meantime, the effects of stigma and discrimination can still be tackled — for instance through laws protecting people living with HIV, together with practical means for reporting and human rights violations.

In addition, alternatives to the family and community support that may be lost for people who test positive are needed. Consequently, a number of NGOs and programmes are recognising the need to foster the development of peer supporters and caregivers to reduce the vulnerability of people living with HIV, help them make effective linkages to care and navigate the health system, bolster adherence, and eventually contribute more to their own management for life-long ART.

Programmatic factors affecting ART delivery

The dramatic increase in uptake in HCT is a major step forward, but it matters little if those people never make it onto ART. Programmatic barriers appear to contribute to substantial losses to follow up in the linkage between HCT and ART programmes — resulting in a failure to initiate timely ART. Prof. Moosa cited a paper last year in AIDS that examined factors related to commencement of ART within the first year of a new diagnosis of HIV in an outpatient setting among ART-eligible patients.1 Out of 1474 people who tested positive, 462 (31%) never made the linkage to the ART programme to get their CD4 cell counts measured. At least 583 (53%) of those who did complete the referral had CD4 cell counts below 200. Of  these, 62% did not initiate ART within twelve months although they qualified for it. Among ART-eligible subjects, there were 108 known deaths; 82% occurred before ART initiation or with unknown ART initiation status.

Males or anyone without an HIV-positive friend or family member were significantly less likely to initiate ART within the year (knowing someone living with HIV facilitates disclosure and increases social support).2

The same team looked at factors associated with pre-treatment losses to care and loss to follow-up, which they defined as failure to get a CD4 cell test within 8 weeks of diagnosis, which occurred with nearly half the subjects testing HIV-positive and referred to care at the sites in Durban. The study found that people were significantly more likely to be lost to care if they lived ten or more kilometres from the testing centre, had a history of tuberculosis treatment or were referred for testing by a health care provider rather than self-referred. Of course, given the high cost of transportation to the clinic, financial constraints could be part of the reason the distance from the clinic mattered. Likewise, clinic hours may lead to greater losses among working men, who can’t afford the time off work.

The need for peer-based tracking systems to ensure that people complete their follow-up to the ART site, get timely CD4 cell counts and commence treatment, was emphasised by several speakers. Clinic services also need to become more user-friendly.

Another programmatic challenge that will clearly have an impact on patient load and other factors is the question of when to start treatment. Virtually everyone agrees that it would be better for a person living with HIV to start before their CD4 cell count falls to 200. Increasing the CD4 threshold for initiating treatment to 350 would reduce the risk of opportunistic diseases, death, serious non-AIDS events, immune reconstitution inflammatory syndrome (IRIS), the chances of achieving a normal CD4 count, drug related toxicities.

Furthermore the recent results of the HPTN 052 study have confirmed that antiretroviral therapy clearly reduces the risk of HIV transmission in couples where the HIV-positive partner is receiving treatment.

But regardless of the guidelines, people in South Africa have only been coming in for care when their CD4 cells are below 100, according to Moosa. However, this is likely to change with more people getting tested for HIV before they start falling ill.

“Current data is compelling enough to revisit the debate on when to start treatment in South Africa,” said Prof. Moosa,” though the sickest must be prioritised.” At the same time, it is logical that the one way to reduce the pool of the ill (who consume far more health system resources) is to prevent the well from becoming sick. It may be in the national interest to target particular groups intensively, such as nurses and teachers, for early commencement of ART.

But even without increasing the CD4 cell threshold, the number of patients in need of treatment will increase — and the current doctor-based model will not be able to meet the increasing demand for care.

An alternative in resource-limited settings, Prof. Moosa noted, is the public health approach. This strategy to scaling up access was promoted by WHO’s 3x5 initiative, and has even allowed the poorest of countries, such as Malawi, to rapidly scale up ART. The approach relies on the use of simple standardised regimens, decentralised service delivery and the delivery of care by non-specialist healthcare workers, including medical officers, nurses (the most practical and widely implemented in resource-constrained countries) and community workers.

Nurse-led models of care have been shown to be effective in the management of other chronic diseases, such as hypertension.3 Prof. Moosa also cited the CIPRA-SA study, concued by Ian Sanne and colleagues. CIPRA-SA was a prospective, non-blinded, non-inferiority randomised controlled study comparing doctor-managed versus nurse-managed ART.4 Care was either provided by two experienced primary health-care nurses vs. two doctors. Both groups had little or no experience with ART at the beginning of the study, but each group received similar structured didactic and clinical training.

“The study found that nurses are as effective and efficient at providing ART, with regards to early virologic failure; late virologic failure, toxicity management, and loss to follow up,” said Prof. Moosa. (Similar findings were presented from the STRETCH study this year at the conference and are described in a separate aryticle in this edition). “Note however, that both doctors and nurses received protocol-specific training, lectures on clinical management and ongoing telephonic clinical support. To duplicate these results you need to duplicate [the level of ] training and clinical support.”

With larger facilities at capacity, there is a need to decentralise care, and shift more of the workload to primary healthcare clinics, and step-down facilities.

A recent retrospective cohort analysis study from the team at Kheth’Impilo, compared routine adult ART in three tiers of health care sites, primary, district and regional hospitals in four provinces from 2004-2007, with seven per cent of the subjects on ART. Analysis of baseline characteristics showed that primary health care facilities were seeing patients that are sicker than at higher levels of care. Despite seeing a sicker population  the primary healthcare sites consistently out-performed higher levels of care, with regards to better retention in care, and fewer losses to follow-up. Mortality was similar to that seen in regional hospitals but better than district hospitals, with comparable rates of virological suppression.5

Since it was not a prospective randomised study, differences in outcome could have been specific to these facilities. Nevertheless, the data seem reassuring that “ART can be adequately provided at PHC level,” according to Prof. Moosa.

Improving diagnostics and monitoring

Also required for the continued expansion of ART programmes is the maintenance of an uninterrupted chain of drug supply as well as good laboratory service delivery (between the clinic and the reference laboratory and visa versa). Notably there are a number of point of care (POC) tests becoming available to some facilities that might reduce turn around time, including the CD4 cell tests, the GeneXpert TB test, and tests for haemoglobin, ALT, creatinine, and lactate.

Point of care testing could have the potential to speed up the interval between presentation at the health facility and diagnosis, reducing the risk that patients will be lost to follow-up, suffer delayed treatment or treated inappropriately.

Dr Lesley Scott, of the National Laboratory Service (NLS) described some potential challenges introducing on site, nurse-operated venepuncture for POC testing.6 The NLS developed standard operating procedures, quick reference charts, provided training and monitored outcomes. On the positive side, the nurses did a good job, with lab results that were consistent with the NLS on the same specimens. The big problem is that there’s no easy way to link all these devices with the central database (the Laboratory Information System). They all use non-standard protocol communications; each device requires different proprietary software and data management systems to interface with the LIS systems, and some of the devices don’t support bi-directional communications. To list but a few of the problems. The net effect is that this creates much more work than it should.

So NLS has initiated a collaborative effort to investigate options, ideally trying to find something that could operate as a universal interface program, and data management system for all the POCs. Dr Scott said they still need to perform a randomized controlled study to make certain this will work in more remote settings, whether it improves patient outcomes, is safe and cost-effective.

It isn’t absolutely clear that it will be. For instance, the introduction of a BD FACScount, which produces a CD4 cell count result in about 40 minutes, did not turn out to have quite the impact desired according to when it was first piloted at the Themba Lethu Clinic in Johannesburg, according to a retrospective analysis presented at the conference by Buyiswa Ndibongo, a health economics researcher at the Wits Health Consortium.7 She evaluated whether people newly diagnosed with HIV who were known to have been offered a POC CD4 cell test that provided results while they waited, started ART any sooner than either an historic control and a control group of other patients who were not offered the test.

They did, but not profoundly. The group offered POC initiated ART 32.2 days after HCT, 9.8 dayssooner on average than the historical control (p = 0.003 (95% CI 3.49, 16.11) (and no differences between historical and concurrent controls). 54% of those offered POC who were initiated onto ART within 16 weeks, 8-16% higher compared to comparison groups, p < 0.05. “This reached statistical significance,” said Ndibongo, “but the effect was only modest.”

Notably, “a large proportion of walk-in HCT patients have low CD4 counts, below 250, on the day they completed HIV testing,” she added, 63-71% of those who had their CD4 cell counts measured (depending on group).

One theoretical benefit of POC testing is that it could reduce the number of clinic visits required. But the expectation had been that offering POC CD4 would lead to a lot of participants starting treatment almost right away, and that people wouldn’t be lost between the time they had blood drawn for CD4 cell counts and the next clinic visit. What no one had counted on was that a lot of people simply didn’t seem to be ready for this step — 39% rejected the offer of POC CD4 — which is perhaps not so dissimilar to the number of people who wouldn’t come back for their CD4 cell results. These people clearly need a different type of intervention.

Another potential challenge to optimising the programme is “the flow of information from the coal face to the centre,” said Prof. Moosa. “Data is a powerful tool to rapidly improve care. It allows staff to measure their performance against national benchmarks and nearby clinics.” Data analysis can reveal the reasons for poor practice, and lead to the detection and dissemination of best practices.

Some of the quality improvement efforts in the country have initially been stymied by the lack of good quality data. According to Dr Wendy Dhlomo-Mphatswe of the 20,000plus Partnership, a quality improvement process was targeted at each step of the PPTCT cascade, in order to reduce PPTCT in KwaZulu Natal to below 5%.8 “This would require data for each step to map the processes of PMTCT care and track progress with data,” she said. The data feedback loop would be complete when the local staff began to put their local data to use in monitoring their own progress. But an initial data audit found problems. “We needed targeted data improvement initiatives first!” she said.

Care provider factors

A number of care provider factors such as knowledge and its application are obstacles to the successful delivery of ART. Prof. Moosa said that HIV should be seen as the silent killer. “Every patient is infected with HIV until proven otherwise, just like every patient has hypertension until he has his blood pressure checked,” he said.

People are going to be living with HIV into old age, added Prof. Moosa, with risk factors of non-AIDS conditions in multiple organ systems. He said that it will be critical not to create a health system within a health system to manage these conditions but to rather integrate the HIV programme with mainstream medical care where these conditions are routinely managed.

However, all the specialists and other healthcare workers will need to have adequate knowledge about HIV. “No practicing healthcare workers (phlebotomist to psychiatrist) should allowed to practice without some basic knowledge of HIV,” said Prof. Moosa. It is important to put processes into place to ensure that health care workers have adequate knowledge, and that they know how to apply that knowledge. Prof. Moosa cited a study which reported that although 95% of doctors were aware of the recommendations for managing their patients with cardiac disease, only 50% of their patients were at treatment goals for those risks. Quality of care must be monitored. Prof. Moosa added that care providers should have current maps and contact details of all up and down-referral centres.

Ultimately, as suggested earlier, ART will need to be provided for life using a chronic care delivery model. Ideally, the patient should be an active participant in their own care, adequately trained to manage their illness, with the assistance of family/peer/community support systems, while the healthcare providers need to be prepared and adequately supported by protocols, data, and the appropriately designed clinic system.


Prof. Moosa reiterated that the key obstacles to moving millions more South Africans living with HIV onto ART in the coming years will be the burgeoning patient load that will become unmanageable for doctor-based services, as long as primary health clinics are underutilised. Rapid scale-up can only be achieved if treatment is initiated at the primary health care level using less specialised but well-trained staff.

“The ongoing challenges faced in the struggle to scale up high-quality ART are formidable,” said Prof. Moosa. “This is a fight we have to fight. Failure is not an option. If we fail, it should not be from lack of trying. It will be a delicately balanced burden but if everyone pulls their weight we will overcome.”


[1] Bassett V et al. Who starts antiretroviral therapy in Durban?... not everyone who should. AIDS 24(suppl 1):S37-S44, 2010.

[2] Losina E et al. The "ART" of linkage: pre-treatment loss to care after HIV diagnosis at two PEPFAR sites in Durban, South Africa. PLoS One. 5(3):e9538, 2010.

[3] Logan AG, Milne BJ, Achber C, Campbell WP, Haynes RB Work-site treatment of hypertension by specially trained nurses. A controlled trial. Lancet 314: 1175-1178, 1979.

[4] Sanne I et al. Nurse versus doctor management of HIV-infected patients receiving antiretroviral therapy (CIPRA-SA): a randomised non-inferiority trial. The Lancet, Volume 376, Issue 9734, Pages 33-40, 3 July 2010.

[5] Fatti G, Grimwood A, Bock P Better Antiretroviral Therapy Outcomes at Primary Healthcare Facilities: An Evaluation of Three Tiers of ART Services in Four South African Provinces. PLoS ONE 5(9): e12888. doi:10.1371/journal.pone.0012888, 2010.

[6] Scott L et al. Anticipating hurdles for the implementation of HIV/TB diagnosis and monitoring using multiple point of care (POC) platforms.

[7] Ndibongo B et al. Point-of-care CD4 testing after HIV diagnosis to reduce losses to initiation of antiretroviral therapy:  an evaluation of a pilot program at the Themba Lethu Clinic, Johannesburg, South Africa. 5th South African AIDS Conference, Durban, 2011.

[8] Dhlomo-Mphatswe W Using quality improvement to reach public health targets. 5th South African AIDS Conference, Durban, 2011.