I thought the struggle for TB/HIV integration was over. We won.

‘TB and HIV and AIDS will now be treated under one roof.’ These were the exact words of President Jacob Zuma on 1 December 2009, World AIDS Day. ‘This policy change will address early reported deaths arising from undetected TB infection among those who are infected with HIV.’

I was wrong: this was no victory. We won one battle, but the war with TB is far from over and the level of integration remains low. Five years after the president’s landmark speech, fully TB/HIV integrated clinics are hard to find, TB remains the first cause of death of people living with HIV, and in 2012 only 54 percent of people with both HIV and TB were started on ART in South Africa. As with revolution, implementing real change is hard. When it comes to TB/HIV integration, it seems particularly hard. Why?

A less well-known fact about the South African HIV denialism era is that Manto Tshabalala-Msimang, the infamous Minister of Health under Thabo Mbeki, in addition to denying the benefits of antiretroviral medicines, also fiercely opposed the integration of TB and HIV care. She encouraged HIV and TB services to continue to function in their own silos, with devastating consequences for patients. The political tide has turned, but Manto’s legacy lives on: TB and HIV services continue to be divided at all levels and TB/HIV integration is poorly implemented – if at all.

What is TB/HIV integration?

Part of the problem is that here are many different interpretations of what TB/HIV integration means. Different models have been described, and evolved over time, from minimalist TB/HIV collaboration – where TB and HIV services remain separate but in communication – to full TB/HIV integration. However, from a patient perspective it is quite simple. A fully integrated model is centered on the patient, not the disease. One patient is seen by one and the same nurse and/or doctor, in the same room and during the same visit, for both HIV and TB. Each patient has one folder, containing all their medical information, and collects all their medication from the same pharmacy. It is a one-stop shop.

But clinic-level integration needs to be supported by integration at the higher levels. While this integration has happened at district level to a certain extent, with the creation of HIV/AIDS/TB/STI (HAST) departments, at provincial and national level HIV and TB directorates remain very separate. At the international level it is even worse with the World Health Organisation’s Global TB Programme and HIV/AIDS and Global Hepatitis Programme completely separate. As long as these programmes continue to operate in a vertical as opposed to an integrated manner, it is likely that TB/HIV integration will remain a paper ideal rather than a reality on the ground.

Why is it important to integrate TB and HIV care?

TB is the leading cause of death in South Africa, killing 89,000 people in 2014, of which 64,000 were living with HIV. The majority of people with TB also have HIV; out of 295,000 new TB patients tested for HIV, 62 percent were HIV-positive. Yet, out of 450,000 new TB cases in 2013, only 312,000 (69 percent) were diagnosed. A recent autopsy study of adults who died at home in a high HIV prevalence area in South Africa, found a quarter of deaths had evidence of undiagnosed, likely infectious TB1. About a third of people with active TB are not started on treatment or notified of their disease2. This is probably the biggest challenge with TB at the moment, as it is with HIV: ensuring that people are diagnosed in time to get treatment.

Multidrug-resistant TB (MDR-TB) is spreading, even if numbers of patients with drug-sensitive TB (DR-TB) are slowly dropping. The number of reported new cases with DR-TB increased from 3,200 in 2004 to 14,000 in 2012. MDR-TB represents 2.2 percent of SA’s TB cases but accounts for 32 percent of the national TB budget. The per patient cost of treatment for extensively drug-resistant TB (XDR-TB) in 2013 was US$26,392 – four times greater than MDR-TB ($6,772), and 103 times greater than drug-sensitive TB ($257)3. For comparison purposes, antiretroviral therapy (ART) cost $113 for one patient for one year. Similarly to drug sensitive TB, DR-TB is mostly transmitted from person to person, and disproportionately affects people living with HIV4.

While the treatment success rate (i.e. the proportion of patients who successfully complete treatment) isn’t great for TB in South Africa (77 percent), it is worse for people living with HIV (74 percent), and abysmal for MDR-TB (45 percent) and XDR-TB (15 percent). For patients in whom active TB has been diagnosed, mortality is particularly high during the first months on treatment5 and effective management of both diseases – especially early initiation of ART in patients with TB – is critical to improve survival6,7. A study in a non-integrated South African setting showed that ART initiation in TB clinics can be delayed up to 116 days, mostly due to the long referral time between TB and ART services8. Full integration of TB and HIV care in a clinic in Khayelitsha increased the chance of co-infected patients to start ART by 60 percent, and reduced the time to ART initiation by an average of 72 days9.

To manage a patient co-infected with TB and HIV, clinicians have to know both diseases and the medicines used to treat them well. Medicines for TB and HIV interact and have common and sometimes additive side-effects. TB has different symptoms and deteriorates faster in people with HIV. In a fully-integrated system, nurses and doctors learn about both diseases and thus provide a better quality of medical care.
A recent mathematical modelling study estimated that adding annual community-based TB/HIV case findings to current recommended TB/HIV control strategies would avert a further 17 percent of TB cases, 8 percent of MDR cases and 8 percent of TB/HIV deaths10, suggesting that there is a need to increase community models of care for TB and HIV testing.

What is the situation of TB/HIV integration in South Africa?

There has been tremendous progress at rolling out antiretroviral therapy to all clinics in South Africa, and TB and HIV are now mostly treated under the same roof – but not in the same room. While the National Department of Health has issued a Practical Guide for TB and HIV Service Integration at Primary Healthcare Facilities11, implementation is lagging behind12 and in most clinics we remain far from ‘One patient. One Nurse. One Folder.’. Patients often continue to see different nurses for TB and HIV, to have one folder for TB and another for HIV, and to get their medication in different places. There are three different electronic monitoring systems for HIV, TB and drug-resistant TB, and management of TB and HIV services remains vertical, continuing the separation of services and missing the benefits of full integration.

Several recent South African studies describe many missed opportunities that could be addressed by good implementation of TB/HIV integration13,14 : low numbers on isoniazid preventive therapy, low rates of TB screening and HIV testing, insufficient referral of TB patients for ART, and ongoing TB deaths among people with HIV.
In summary, one-stop service TB/HIV integration has the potential to improve access to prevention, diagnosis and treatment, as well as adherence and treatment outcomes for patients with HIV and/or TB. Yet much remains to be done to achieve this potential and this will necessitate increased political will towards the ideal of one patient, one nurse, one folder, treated in one room, not just under the same roof. Integration needs to happen at higher levels as well as in the clinics. In the end it is about shifting towards an approached that is truly centred on patients.