This study was designed to evaluate the impact of implementing a strategy of routine, provider-initiated (Opt-out) HIV counseling on the uptake of counseling and testing of newly registered TB patients in primary care clinics in the Eastern Cape Province of South Africa. The results
show that the use of an Opt-out strategy was associated with significantly higher HIV counseling and testing rates and that the time to testing tended to be faster in the Opt-out study arm than in the control arm, but the overall proportion of those counseled and tested following training in the Opt-out approach remained unacceptably low at only 21 percent.
The major purpose of HIV testing of TB patients is to ensure that HIV care and treatment are rendered in a timely fashion. In this study, the number of HIV-infected TB patients who were prescribed cotrimoxazole and/or referred for HIV care was small and did not differ between arms. The South African Tuberculosis Control guidelines in place at the time of this study were last updated in 2000, and clinical and treatment guidelines for HIV/AIDS were not included in them[14
]. However, all TB nurses had received specific training about current HIV/AIDS guidelines and should have been aware of the recommendations regarding cotrimoxazole use and referral for HIV care. Thus, while the trial demonstrates the positive impact that training in an Opt-out strategy can have on the uptake of testing, it underscores the difficulty of integrating HIV and TB care in the challenging world of clinical practice in resource poor settings.
This trial was carried out without an influx of additional personnel or resources, in a primary care setting that was already overburdened with a large clinical workload. Although HIV testing for TB patients is recommended by global and South African guidelines, implementation of this mandate into practice depends on the capacity of local health facilities to incorporate additional and time-consuming clinical tasks into their already overstretched workloads. Reducing the effort required by both health workers and patients for providing HIV testing by offering the Opt-out approach makes it more likely that testing will be provided, but this study illustrates that although instituting a strategy of provider-initiated counseling can increase rates of HIV counseling and testing, it is clearly not sufficient. Additional interventions and resources will surely be required to attain high levels of HIV testing referral into HIV care for TB patients in settings such as ours. In order to better understand the reasons for the low uptake of HIV counseling by the staff in the clinics, we have undertaken a qualitative study of staff and administrators, the results of which are being analyzed. The issues of stigma connected to both TB and HIV, experienced by both patients and staff, have been well-documented in several studies and undoubtedly confound an already complex task [16
Pragmatic studies carried out in primary care settings, often have modest results. In Brazil, a cross-sectional study designed to assess the rate of HIV screening of TB patients in primary care clinics found that approximately 23 percent of the patients had been screened, and that perceived risk by the health worker determined who was tested [18
]. A cluster randomized trial that trained clinic TB nurses in Free State Province, South Africa, to use an algorithm for the diagnosis and management of respiratory diseases (including TB), had voluntary HIV counseling and testing rates of only 9.7 percent in the intervention and 7.3 percent in the control study arms [19
Studies that demonstrated high uptake of counseling and testing in TB patients have been carried out in hospitals (Malawi) [6
], in vertical TB programs (Malawi) [1
], during interventions with extra staff dedicated to the process (ProTEST, Côte d’Ivoire and Democratic Republic of Congo) [4
], or by study staff (Haiti and Thailand) [3
]. Recently, TB programs in Rwanda, Malawi and Guyana[22
] have reported substantial increases in HIV testing of TB patients when national initiatives that provided additional training and resources were coupled with use of the provider initiated strategy[22
Our simple and pragmatic cluster randomization design does not permit us to assess a number of qualitative factors that could influence the success of provider-initiated HIV testing as a public health strategy. Randomization based on baseline data permitted us to distribute the diversity of the study clinics evenly between the two study arms. If there were factors that influenced the uptake of HIV counseling and testing, the evidence from our evaluations indicates that they were not differentially represented in one study arm versus the other. A more intensive intervention or an intervention more specifically tailored to the needs of individual study clinics may have resulted in greater numbers of TB patients receiving HIV counseling and testing. However, our intention was to test an intervention that was reproducible in the very real environment of staff shortages, limited privacy and time constraints.
It is possible that our intervention would have been more effective if it had been a Health Department initiative aided by researchers rather than a research project supported by the Health Department. The nurses in the Municipality knew that the interventions were not health department initiatives. Given the high work load, it is perhaps not surprising that an outside intervention was not a high priority for the nurses. Researchers from other pragmatic South African studies of interventions in primary care settings have concluded that the involvement of supervisors and management are essential to the success of interventions [19
]. Although we had the full support of the highest levels of supervision, the concept of provider-initiated HIV counseling and testing was still novel and controversial when this study was begun.
Training in the implementation of routine, provider-initiated HIV counseling increased the proportion of adult TB patients that received HIV counseling and testing by 3-fold, but the absolute magnitude of the effect was small. Given the multitude of restraints that hindered the success of HIV counseling of TB patients, we believe that additional interventions will be necessary to optimize HIV testing and care for TB patients in similar settings in sub-Saharan Africa. Provider-initiated HIV counseling and testing is an important step to ensure that TB patients with HIV-related illnesses are diagnosed and referred into HIV care, but a greater investment in primary care health resources is required to improve comprehensive care of TB and HIV in those areas of the world suffering from the collision of these two epidemics.