Our study assessed the availability of integrated TB and HIV clinical services in a sample of public medical facilities in South Africa in March 2011. Our data show that integrated HIV and TB services are being provided in the public medical facilities in South Africa. Among the 49 sites we sampled, about three quarters of newly diagnosed HIV-infected patients were screened for TB symptoms. According to South African national data reported to the WHO, 58% of newly diagnosed HIV-infected people were screened for TB symptoms in South Africa in 2010 
. Most newly diagnosed HIV-infected patients who screened positive for any TB symptom in February 2011 among our 49 selected facilities were started on TB treatment. No data on the proportion of those patients who had laboratory-confirmed TB was collected, and our findings on initiation of TB treatment may suggest over-diagnosis of active TB on the basis of clinical findings alone. Less than half of the newly registered TB patients in February 2011 were diagnosed as being HIV infected, compared to 60% of tested TB cases being reported as HIV-infected in South Africa in 2010 
, and of those, most had CD4 T-cell counts done that month.
Several interventions and programs implemented recently might account for those improvements. Firstly, the South African National Department of Health launched several national campaigns, including the Kick TB campaign in partnership with Desmond Tutu TB center and the University of Stellenbosch in December 2009, aimed at increasing TB and HIV awareness among school-aged children, reaching almost 39,000 learners to date 
, and the HIV Counseling and Testing campaign in April 2010, which has tested 10.2 million South Africans for HIV infection 
. In 2010, South Africa renewed its commitment to uphold WHO’s policy on collaborative TB/HIV activities including the three I’s, and increased the CD4 T-cell count threshold for initiation of TB patients on ART from 200 cells/mm3
to 350 cells/mm3
.Those campaigns greatly increased advocacy for integrated TB/HIV services and necessitated increased collaboration between national, provincial and district health authorities which together could have positively affected TB and HIV service delivery in surveyed facilities. Secondly, the National IPT guidelines were revised and disseminated in 2010 
, and the availability of guidelines has been shown to increase IPT uptake 
Not all TB and HIV services, however, are improving. In our study, the proportion of newly diagnosed TB patients living with HIV initiated on co-trimoxazole in February 2011 was lower than national estimates in South Africa of 74% in 2010 
. Additionally, less than a quarter of eligible newly registered TB patients newly diagnosed with HIV in February 2011 were initiated on ART, well below the reported 30% of TB patients living with HIV on ART in 2009 
. Further, less than half of eligible patients of sampled sites were initiated on IPT in February 2011, below the target of 60% set by the South African government for 2011–2012 
South Africa is shifting from a vertical programmatic approach with separate staff and service model in the early 2000s 
to a decentralized integrated approach with a strengthening of primary healthcare services. The primary health clinics are empowered through official policies to become the main mode of health care delivery, including essential TB and HIV services such as TB diagnosis and treatment, IPT and NIMART 
. Our study does show signs that decentralization is taking place with NIMART being used as a method of initiating and managing patients on ART in more than half of the ART-providing facilities. Among the 49 public medical facilities we assessed, however, IPT and ART uptake remain low. In addition, we found that primary health clinics were less likely to offer ART compared to other facility types, suggesting that increased ART coverage as a result of decentralization remains an ongoing process. With regards to low IPT uptake, the persistence of known barriers to IPT implementation, including fear of the selection of INH-resistant TB 
, could explain the slow IPT uptake among the selected facilities.
Recommendations on ART initiation among TB patients have been a moving target: until recently, South African ART guidelines followed the 2003 WHO guidelines recommending the delay of ART initiation among co-infected patients with CD4 T-cell counts of 200 cells/mm3
or above until completion of TB therapy 
. The newest ART initiation guidelines in TB patients now recommend starting ART after 2 to a maximum of 8 weeks of TB treatment in all TB patients living with HIV with CD4 T-cell counts of 350/mm3
or less. Despite those recent and significant changes in policy, barriers in ART initiation in TB patients living with HIV remain including concerns of immune reconstitution syndrome and drug-drug interactions 
, and could in part account for the low proportion (19%) of eligible co-infected patients in surveyed sites initiated on ART. The delay in ART initiation is worrisome, particularly in view of the findings of a recent trial in South Africa, which found that mortality was reduced by 56% among patients started on ART during TB treatment as compared to those initiated after the end of TB treatment, with no significant risk of increased adverse events 
Another barrier to the implementation of integrated services is under-staffing 
, which remains an issue in South Africa for several reasons. Firstly, there is a national shortage of both nurses and doctors. Secondly, there is uneven distribution of current staffing resources. A recent study revealed that staffing shortages particularly affect rural areas and public medical facilities, and that one of the biggest staffing issues was the inequities in distribution of essential practitioners between provinces, rural and urban areas and between public and private sectors 
. Overall, our study shows that both TB and HIV services are being provided at all levels of the public health care system, including the clinic level. These findings illustrate the progress the South African government is making towards a decentralized health care system 
We conducted the rapid assessment only in districts with the highest antenatal HIV prevalence, and our results might therefore not be representative of areas with lower HIV prevalence. Our assessment only included a few facilities in each province, thus limiting the generalizability of our results. Our small sample size reduced the power of the survey, and may have concealed statistically significant differences in our study population. A larger sample size could have shown an association between factors such as facility type and location of ART provision, and facility type and method of initiation and management of ART. We sought to maximize external validity through random sampling and by selecting sites based on the approximated national distribution of facility types. We randomly selected double the number of sites in Kwazulu-Natal to account for the fact that the province had double the number of districts compared to most other provinces and the highest overall HIV prevalence in 2008 
. Northern Cape only had four study sites as two of the facilities initially selected merged into a single entity shortly before data collection. Further, all selected facilities were public medical facilities as we were assessing the implementation of a national public medical program, and our results might not be representative of TB/HIV integration in private sector health facilities. We were unable to collect data on the number of patients screened positive for TB symptoms for whom TB was laboratory-confirmed. Finally, we did not collect data on HIV testing and prior knowledge of HIV status or the number of patients eligible for IPT completion in February 2011 among surveyed sites.
Our study demonstrated important progress is being made towards integration of TB and HIV services in South Africa, where nearly all facilities offered routine TB screening to people living with HIV infection, and routine HIV Counseling and Testing to TB patients. However, uptake of other essential services, such as ART and IPT, needs to be improved, as less than half of eligible people living with HIV were initiated on IPT, and only a small proportion of newly registered TB patients newly diagnosed as HIV-infected in February 2011 were initiated on ART. Addressing those gaps is a priority and future interventions should build on existing efforts to support current national policies of routine TB screening of all HIV patients, initiation of all eligible HIV-infected patients on IPT and early ART initiation of eligible TB patients irrespective of CD4 T-cell count 
. Our findings represent estimates of point prevalence and repeated surveys should be conducted to monitor trends over time.