Background:  The Western Cape, South Africa has one of the highest burdens of tuberculosis (TB) in the world, with 70% of TB patients co-infected with HIV. ART improves TB treatment outcomes. Absolute Return for Kids (ARK) is a non-governmental organization (NGO) that assists the Department of Health (DOH) in increasing access to ART, primarily by providing clinical and adherence support staff for the ART program. This study aimed to compare TB treatment outcomes between sites in the Western Cape that had ART services enhanced by ARK, and sites that were not enhanced by ARK or had no ART service.

Methods:  A retrospective analysis of routine DOH data from 2004 to 2007 was performed, including multivariable maximum-likelihood logit estimates for grouped data. Outcomes for the periods before and after ARK entry to sites were also compared.

Results:  At 98 non-ART-enhanced sites, 32,853 cases of smear-positive pulmonary TB were treated, of whom 23,232 (70.7%; CI 70.2 to 71.2) were cured. At 11 ART-enhanced sites, 2822 cases were treated, of whom 2222 (78.7%; CI 77.2 to 80.2) were cured, amounting to a relative increase of 11.3% (CI 9.1 to 13.6; p <0.0001). At non-ART-enhanced sites, 3685 patients (11.2%; CI 10.9 to 11.6) defaulted treatment, while 181 (6.4%; CI 5.5 to 7.4; p <0.0001) defaulted at ART-enhanced sites, i.e. a relative decrease of 42.8% (CI 33.9 to 50.5). An additional 6-month period of ARK being present at a site per calendar year was associated with a 23% lower risk of treatment failure (OR 0.77; CI 0.73 to 0.81; p <0.0001) from 2004 to 2007, with a 44% reduced risk (OR 0.56; CI 0.50 to 0.64; p <0.0001) in 2004. Similarly, the risk of defaulting treatment was reduced by 31% (OR 0.69; CI 0.64 to 0.76; p <0.0001) from 2004 to 2007, with a 52% (OR 0.48; CI 0.38 to 0.62; p <0.0001) reduced risk in 2004. Prior to ART enhancement, 60.8% (CI 59.2 to 62.5) of patients were cured, while 74.0% (CI 71.6 to 76.3) were cured following ART enhancement, amounting to a relative increase of 21.6 % (CI 16.7 to 26.7; p <0.0001). Prior to ART enhancement, 13.6% (CI 12.4 to 14.9) patients defaulted treatment, and 9.5% (CI 8.0 to 11.2) defaulted following ART enhancement, i.e. a relative decrease of 30.0% (CI 15.9 to 41.8; p <0.0001).

Conclusions:  ART sites enhanced by ARK had better TB treatment outcomes than non-ART-enhanced sites, treatment outcomes improved following ART enhancement, and the risks of poor outcome were lower for increased duration of ART enhancement. TB treatment outcomes are significantly improved with the scale-up of adherence-supported ART services.