HIV and TB in Practice: Progress on IPT roll-out in South Africa

The HIV and TB in Practice feature is kindly supported by the Stop TB department of the World Health Organization.

Implementing isoniazid prophylaxis treatment (IPT) in South Africa

A majority of randomly surveyed South Africa health facilities have started implementing the recent National TB screening and IPT Guidelines according to a cross-sectional study by the US Centers for Disease Control (CDC) and the SA Department of Health presented as a poster at the 5^th South African AIDS Conference in Durban 7-10 June, 2010.¹

If the study is indeed representative of the country as a whole, it suggests there’s been a dramatic increase in the number of people living with HIV receiving IPT compared to the past. However, policy implementation was markedly uneven — with little to no action at all in some provinces, and concerns abou the quality of the service at some sites.

Two other reports at the conference, however focusing on IPT implementation in specfic districts or subdistricts, may provide models for the Health Department in how to do IPT right.

Background

The current guidelines, updated in June last year, recommended TB screening in all people living with HIV. All those with TB symptoms should be referred for TB diagnosis, while everyone else should be given IPT. The guidelines no longer require mandatory chest X-rays or a tuberculin skin test (TST) before starting HIV-positive patients on IPT, and its coadministration with antiretrovirals is longer discouraged.

“There is no excuse for health workers to withhold this inexpensive and effective intervention to prevent TB,” said Minister of Health Aaron Motsoaledi — and the department of health set ambitious targets for the first year of its roll-out (around 450,000 people).

But there has been a longstanding reluctance to scale up IPT among many in the healthcare establishment and some concerns that the health department was rushing to roll out the policy, without adequate preparation — with local authorities left to translate the policy into practice, develop training and operating procedures for staff based upon the guidance. Given these issues, what progress has been made in South Africa since the guidelines change? 

To assess how the policy was being put into effect in the country, researchers from the US Centers for Disease Control (CDC) and the SA Department of Health performed a cross-sectional study in 49 randomly selected clinics in South Africa’s nine provinces. These included one clinic drawn from the district with the highest antenatal HIV prevalence in each province (except for KZN where two were chosen), public health clinics, community health centre and district hospitals. Teams were trained to interview supervisors of TB and HIV services using standardised questionnaires to assess clinic policy, service delivery, supply, recording and reporting and barriers to implementation.

As of February 2011, 71.4% of the clinics were implementing IPT — with about 56% of the eligible patients in the sample receiving IPT. Among the clinics not providing IPT, the most commonly offered excuses were that they hadn’t had clear guidance or commitment from local authorities on the policy (29%), while about a fifth said they were concerned it would lead to resistance (though a number of studies have shown this is unlikely to occur). There was a high rate of implementation at PHC and district hospitals, but community health centres didn’t do as well.

Among the 35 clinics providing IPT, coverage varied markedly by province. Sampled clinics in some provinces reported that they had put more clients on IPT than the total number eligible (all those who had screened negative for active TB). So something was not quite right with these figures. There was little implementation at the sampled clinics in Limpopo or the Eastern Cape, and oddly, no implementation at all in the Western Cape. Likewise, although most clinics had at least some staff who had been trained in HIV/TB, this also varied significantly across the provinces.

Some form of IPT counselling was also being provided by all of the clinics providing IPT in each of the provinces except Limpopo, where only half of them were. However, the quality of the counselling was uneven. The researchers had identified four essential components for IPT counselling including adherence, reporting symptoms, reporting side effects and reducing alcohol intake — but only 17.3% of the IPT dispensing facilities got all these messages into their counselling.

All the facilities providing IPT were recording and reporting IPT data — all but one facility that counselled on IPT also recorded IPT data in their registers.

The researchers recommended continuing efforts to make IPT a priority intervention, especially in the Western Cape, the Eastern Cape and Limpopo, and called on PEPFAR to support the Department of Health’s efforts to increase monitoring of IPT. They added that the quality of IPT counselling should also be improved by mentorship at the facility level.

Regardless, these findings suggest that IPT is being scaled up at a much greater pace than before the recent guidelines were released — and although uneven — it represents a good start.

However, as the study’s author’s noted “while counselling of IPT is common, quality and comprehensive counselling is rare.” This raises the spectre that nurses and patients might not understand the intervention well enough for it to be given safely.

Another issue is that the study said nothing about intensified case finding itself — how widely and routinely TB screening was being performed, and how many active cases were being identified.

Finding and treating active TB cases in people living with HIV is a crucial part of the ICF/IPT strategy to prevent TB spread and acquisition, especially in health facilities. A focus just on the performance of IPT is not only incomplete, but potentially dangerous — continued screening of patients put on IPT is recommended to make certain that cases weren’t missed the previous time, and to quickly detect breakthough cases.

The study also left some other very important questions unasked —most notably, how many of the people being given IPT are adhering to treatment, being retained in care, and completing their course of treatment? Does the programme know? As the example of the IPT programme in Botswana has shown, it's easy enough to start handing out IPT to patients — it is another matter entirely to keep track of what happens to them. Given this well-known experience in a neighbouring country, why was this assessment mum on the topic?

It would also be useful to have more information on  the ways in which performances at health facilities differed — and what factors were associated with success or failure to implement.

While the lack of clear guidance from local authorities might be business as usual for the clinics that aren’t scaling up IPT — the CDC study gave little insight on how IPT was being operationalised at the other clinics.

Why are some clinics and districts scaling up so much better than others, and how are they doing it? Is it due to the kinds of support that they are receiving?

As already noted, quality improvement, mentoring, community-based outreach teams and peer-based adherence support were recurring themes at the 5^th South African AIDS Conference. Along with training and technical support provided by NGOs  these measures have improved the quality and reach of health services in a number of now model districts, contributing to the improvement of PPTCT services and helping support nurses to initiate ART.

These methods are now also being employed in the scale up of IPT (as well as the other Three Is for HIV/TB, intensified case finding and TB infection control) — while some districts have benefited considerably from technical expertise of some of the smartest NGOs.

The staged implementation of IPT in Ugu District Health services

Dr Pria Pundit of Broadreach Healthcare reported on the rapid implementation of IPT and TB infection control across an entire Ugu health district in KwaZulu Natal in a poster presentation.²

The Ugu District, on the coast at the southern edge of KZN comprises 78 health facilities, including five hospitals (one of which is a specialised TB hospital), one community health centre, five mobile clinics, and 57 primary health clinics (PHCs) It serves a population of around 757,000 — with an antenatal HIV prevalence of 40.6%. Broadreach Healthcare has been working with district health teams there for the past four years.

When changes to the IPT policy were announced, the NGO and its district health team considered how best to roll out IPT, along with infection control, quickly across the entire district.

After defining the key team responsible for project planning and coordination, implementation, monitoring and reporting, they set a target to put over 15,000 people on IPT in the district by June 2011.

The first step was to develop the IPT implementation tools, resources and training materials, all of which would need to conform to the DOH policies, including the updated IPT policy. These would include:

• Training Package: a training curriculum on IPT implementation, and infection control (IC) measures;
• IEC Materials: Infection Control poster focusing on 5 IC measures (Is there an infection control practitioner at the facility? Good cross ventilation? Is there a cough monitor? Are coughing patients triaged? Is cough etiquette taught?);
• Clinical Management Tools: IPT clinical guidelines, IPT initiation algorithm, IPT screening and decision support tools for nurses and counsellors;
• Data Management – IPT Register;
• Supportive supervision tools: tools for use by healthcare workers providing supportive supervision to health facilities

They chose to roll the programme out in phases — piloting in twelve better performing clinics first, in April 2010. Clinics were chosen that had low patient volumes, low TB defaulter rates and high TB treatment/cure rates. The clinic manager from each of these facilities was trained on IPT implementation, guidelines, tools and resources were distributed to the facilities. They would report and give feedback to the District TB Coordinator on a monthly basis.

Once that experience was under their belt, phase II was launched in October 10, with the roll-out of IPT to 66 health facilities. Two professional nurses from each facility were trained on IPT, materials and tools distributed to the facility, with monthly reporting and feedback.

Phase III, which began in January of this year, consisted of monitoring and supportive supervision, again with monthly reporting and feedback.

Results

Overall 162 professional nurses were trained on IPT and IC, and three sets of guidelines were distributed to each facility, together with IEC materials.

Between launch in January 2010 and the end of the first quarter of 2011, there had been a steady and marked increase in the numbers of new patients put onto IPT each quarter — with over 2300 enrolled in the last quarter and a total of 5151 patients by the end of March 2011. The district target is expected to be reached in the fourth quarter of this year.

Most of the health facilities (49 as of January this year) were also successfully implementing all five basic IC measures. (That’s one approach. Note, during an oral presentation Dr Adams Tongman described how Aurum Institutes is supporting implementation of TBIC at 11 health facilities in South Africa.³ Aurum developed guidelines, training manuals, a training course on TBIC, and a TBIC assessment tool (TBICAT) that was used for spontaneous audits of the facility. After conducting these audits, which would highlight gaps that needed to be addressed, Aurum provided  further training and mentorship leading to marked improvements in the implementation of simple administrative control measures, which are the most important and can be improved without going to great expense.)

Having a multidisciplinary team was helpful in project implementation, concluded the authors, and the District TB Champion (district TB co-ordination) played a core role in co-ordinating activities of the project on the ground. (It is interesting that their project works relies so heavily on the district TB coordinator — in many settings TB officials are not always eager to see IPT scaled up. Also, it is important for HIV programmes to begin to take more responsibility for reducing the burden of TB in people living with HIV.)

Additionally, Pundit and colleagues noted that the training of facility level healthcare workers to implement and monitor the project helped to develop ownership of the programme.

As the project goes forward, there will need to be continuing supportive supervision to make sure that the programme continues to be implemented, that patient outcomes are monitored, and IPT data recording and reporting is strengthened. But Pundit and colleagues suggest that their approach could serve as a model to other districts — and could be adapted across different departments of health programmes.

Mentors, community linkages and use of data for management to increase implementation of IPT

But their model may have some friendly competition from the TB/HIV Care Association, which reported their remarkable progress in mobilising healthcare providers in Sisonke District, KwaZulu Natal to scale up IPT to people who screen negative for TB, through mentorship and other technical support.⁴ Sisonke District is also in the extreme southwest corner of KwaZulu Natal— and another hot spot for HIV and TB.

As described by a poster at the conference, the TB/HIV Care Association’s Project Integrate is engaged in supporting the health system there, as well as in the Cape Metro and West Coast districts of the Western Cape.  This includes a number of interventions to increase access to HIV, sexually transmitted infections (STI) and TB diagnosis, treatment and care.⁵

These interventions involve community-based HIV counselling and testing (HCT) teams (each with one professional nurse and three lay people) that also offer STI and TB screening and refer people testing positive for HIV or STIs, and TB suspects, to the nearest health facility and track them to make sure they are linked to care. In addition, the project included nurse mentors who strive to improve the quality of care provided by DOH clinic staff, and paid DOTS supporters and community care givers to provide adherence support.

In the years that TB/HIV Care has been there, the districts it supports have reported a substantial increase in the numbers being tested for HIV (in both provinces) and entering into care in KZN, and over the last year, with the Western Cape as well. For instance, between October 2009 and September 2010, the community-based HCT team tested over 40,000 people for HIV and screened them for TB, identifying nearly 2,000 cases of HIV, and 161 TB suspects.

One of the consequences of such testing and screening campaigns is that many of the people who test HIV-positive, screen negative for TB. According to the national guidelines, they should be given IPT to reduce the risk that the might develop active and potentially fatal TB. 

But Jennifer McLoughlin, HIV Coordinator at Grey’s Hospital, who reported on the project in Sisonke District, said that there have been problems getting people to prescribe IPT because of misconceptions about the risk of it leading to drug resistance or the likelihood of serious toxicity. This is despite the fact that the new policy comes with targets, including a provincial target for KZN to implement IPT and to initiate 120,000 HIV-infected people on IPT. The target in Sisonke district, KwaZulu-Natal — from April 2010 to March 2011 was 12,000 people, which means 3,000 clients need to be initiated each quarter.

Something had to be done in order for the district to get anywhere near reaching these targets.

 “The Mentorship Programme was introduced to improve the integration of TB/HIV services — including increasing the implementation of IPT, by identifying and removing the barriers to implementation,” said McLoughlin. Each PHC supervisor/mentor was given four to six facilities to supervise.

The programme included formal didactic training on IPT and screening for active TB to ensure that IPT was part of the package of care, for HIV-positive clients. Lay counsellors and professional nurses were mentored to ensure high quality counseling and testing, and staff were trained on how to screen for TB.

“We had to change the mindset that screening of TB wasn’t taking a sputum, it was asking a symptomatic screen,” said McLoughlin. And of course, at this point, if they screened negative for active TB, they were to be given IPT.

They used an IPT register, that the TB/HIV Care Association had developed, for all clients who were initiated on IPT, who were entered into that register to ensure that clients were followed up to monitor weight, treatment adherence, and address side-effects with TB screening to be conducted monthly. The register was designed to run for 24 months, to see how many of the clients that were initiated onto IPT developed TB after completing 6 months of treatment.

They engaged national, provincial and district programme managers to audit progress. Each clinic was told about its targets and the need to use data for management. They would hold quarterly district meetings where the data was analysed and progress assessed.

“Our focus was to ensure that all HIV-positive clients that were tested in the HCT Campaign, were entered into care,” said McLoughlin. “Really, it’s an ethical right of a client who is eligible for IPT to be offered IPT at the service.” 

Only a year after the mentorship programme had begun, there had already been a turnaround in the provision of IPT.

The initial target was to get at least 25% of those testing HIV-positive onto IPT (~3000 per quarter for Sisonke District). Only 19% of the target had been met before the intervention (April to June 2010) — this increased steadily to 81% of the target by Jan-March 2011.

The data on treatment completion are among the best ever reported in a programmatic setting: ~98% of people on IPT completed six months of treatment, with only two per cent interrupting treatment. About 1 per cent developed TB despite IPT.  However, McLoughlin said stock-outs had delayed the completion of treatment for some individuals.

With the exception of supply issues, these outcomes compare quite favourably to other settings where IPT programmes have been plagued by drop outs and losses to follow-up.

McCloughlin attributes the good outcomes to several lessons the programme had learned

• Each nurse mentor became the champion for the implementation of IPT in the sub-districts of Sisonke.
• We established the Wellness Clinics to initiate pre-ART clients on IPT, and then monthly monitoring for weight, encouraging adherence, addressing side-effects and screening for active TB
• Adherence was monitored using a diary system and clients were recalled if they did not show up to appointments.
• The community health facilitators are the link between the facility and the community, and they played a vital role in ensuring that when a client was put onto IPT they were linked with a community caregiver.  This played a huge role in promoting adherence.
• A strong collaboration between HIV and TB programmes is necessary.
• Use the IPT register to summarise 1) clients started, 2) completion of six months of IPT and 3) clients developed TB on or after IPT.
• Use of data to forecast drug supplies —we really did use data to try and ensure we had a sustainable drug supply and
• Use of the data for management, with data analysed every quarter.

There are a number of ongoing challenges to the successful scale-up of IPT, including making sure they can keep an adequate and sustainable drug supply, according to McLoughlin. One issue is confusion around the strength of the tablets, which she said seems to change from 100 mg to 300 mg tablets. Keeping on top of changes in the policy, addressing the persistent fears about risk of resistance to isoniazid and alleviating fears about toxicities all continue to be challenges.

Finally, it will probably always be “difficult to motivate clients to come monthly when they feel well,” she said.

Planning ahead, she said that the Department of Health would need to standardise an IPT register for all districts and to ensure sustainable drug supply of isoniazid 300 mg tablets.

“We have to hold clinicians and facility managers accountable for reaching targets.  We have to be accountable — is it ethical when you’ve got an intervention that works, not to be offering it to patients?

Preparation, training and support are essential for successful IPT implementation

That being said, healthcare staff also clearly need to be given adequate support to provide these interventions well. It isn’t enough to say ‘just do it’ and then leave local programmes to their own devices. Some just won’t do it or won’t do it well, while the others will invest valuable time and resources setting up a programme, but in doing so, be reinventing the wheel and duplicating efforts to develop training, tools, registers and educational materials.

The Department of Health needs to do more than simply standardise IPT registers — it needs to determine out who is doing the best job of implementing the intervention — and then take the best model or put together one based on the best bits of the best models and promote it country-wide, provided that model can be scaled up affordably. One issue that was not addressed in these studies was how much was spent on developing and implementing these model programmes, which may be another reason some of the local authorities have yet to provide clinics with clear guidance about what to do — they simply don’t have adequate resources at hand to develop and support the programme.