Linkage to care following a home-based HIV counselling and testing intervention in rural South Africa

INTRODUCTION Efforts to increase awareness of HIV status have led to growing interest in community-based models of HIV testing. Maximizing the benefits of such programmes requires timely linkage to care and treatment. Thus, an understanding of linkage and its potential barriers is imperative for scale-up. METHODS This study was conducted in rural South Africa. HIV-positive clients (n=492) identified through home-based HIV counselling and testing (HBHCT) were followed up to assess linkage to care, defined as obtaining a CD4 count. Among 359 eligible clients, we calculated the proportion that linked to care within three months. For 226 clients with available data, we calculated the median CD4. To determine factors associated with the rate of linkage, Cox regression was performed on a subsample of 196 clients with additional data on socio-demographic factors and personal characteristics. RESULTS We found that 62.1% (95% CI: 55.7 to 68.5%) of clients from the primary sample (n=359) linked to care within three months of HBHCT. Among those who linked, the median CD4 count was 341 cells/mm(3) (interquartile range [IQR] 224 to 542 cells/mm(3)). In the subsample of 196 clients, factors predictive of increased linkage included the following: believing that drugs/supplies were available at the health facility (adjusted hazard ratio [aHR] 1.78; 95% CI: 1.07 to 2.96); experiencing three or more depression symptoms (aHR 2.09; 95% CI: 1.24 to 3.53); being a caregiver for four or more people (aHR 1.93; 95% CI: 1.07 to 3.47); and knowing someone who died of HIV/AIDS (aHR 1.68; 95% CI: 1.13 to 2.49). Factors predictive of decreased linkage included the following: younger age - 15 to 24 years (aHR 0.50; 95% CI: 0.28 to 0.91); living with two or more adults (aHR 0.52; 95% CI: 0.35 to 0.77); not believing or being unsure about the test results (aHR 0.48; 95% CI: 0.30 to 0.77); difficulty finding time to seek health care (aHR 0.40; 95% CI: 0.24 to 0.67); believing that antiretroviral treatment can make you sick (aHR 0.56; 95% CI: 0.35 to 0.89); and drinking alcohol (aHR 0.52; 95% CI: 0.34 to 0.80). CONCLUSIONS The findings highlight barriers to linkage following an increasingly popular model of HIV testing. Further, they draw attention to ways in which practical interventions and health education strategies could be used to improve linkage to care.