Stigma and attrition within an HIV treatment and care programme in rural KwaZulu-Natal, South Africa

Background: Attrition from HIV treatment and care programmes presents significant public health problems in high HIV prevalence areas. This study explores the association between psychosocial variables, including stigma, and attrition from a programme in KwaZulu-Natal, South Africa.Methods: Psychosocial and demographic data were collected at baseline in a prospective cohort sexual behaviour study (January 2009 to June 2010) among patients in an HIV treatment and care programme. Psychosocial exposures included questions on stigma, social capital, HIV disclosure, reason for testing, ART knowledge and social support. Demographic exposures included ART eligibility status, marriage, TB history, age and religion. Attrition was the outcome if patients had (a) not been seen in the clinic within one month of the last scheduled appointment (b) died (c) transferred out. Analysis used Cox Regression.Results: 456 individuals were recruited (313 eligible to initiate antiretroviral therapy (ART), 143 not eligible). The overall attrition rate was 22.4 per 100 p/y for those eligible to initiate ART (95% CI:17.5-28.6) and 109.7 per 100 p/y (95% CI: 88.2-136.4) for those not eligble to initiate ART. A multivariate analysis of the pooled cohort that included group as a covariate in the model showed that stigma was unrelated to attrition. Greater HIV treatment and care programme attrition was significantly associated with not being eligible for ART, greater reliance on family and friends, more time spent with friends and younger age.Conclusions: There was little evidence that stigma was related to attrition in the pooled analysis. There was a very strong relationship between not being eligible for ART and programme attrition. The results also suggested that greater social capital and social support may present barriers to retention. Public health implications are outlined