OBJECTIVE: Long-term retention in care for people living with HIV and AIDS (PLH) remains a public health challenge in South Africa.

PATIENTS AND METHODS: The study used the clinical records of people living with HIV and AIDS attending a health clinic in the Mpumalanga province of South Africa to examine the effect of selected sociodemographic factors on patients’ non-adherence to anti-retroviral therapy (ART). Pearson Chi-square test was used to examine treatment default proportion. Generalized estimating equation (GEE) was used for the bivariate and multivariate analysis. Statistical analyses were performed using SAS (SAS Institute, Cary, NC, USA) version 9.4.

RESULTS: Of 777 participants selected, more than half were female (56%, n=435) and over 60% were aged between 24 and 44 years old (69.6%, n=541), 62.7% (n=487) had never been married, 67% (n=521) were unemployed, almost half had an incomplete high school education (44%, n=342), and 56.5% (n=439) lived in an urban area. The mean age and standard deviation of all enrolled patients were 37 and 10. Mean and standard deviation of patients who defaulted on treatment was 35 and 9. Overall, 293 (37.7%) defaulted on ART between 2010 and 2014. Sociodemographic predictors significantly associated with non-adherence to ART were age, education, employment status and place of residence. The multivariate analysis shows that patients between 20 and 24 years old were five times more likely to default on treatment and those between 25 and 34 years old [AOR=2.23 (95%) CI, 1.32-3.76] and those from 35 to 44 years old [AOR=2.30 (95%) CI, 1.37-3.85] were twice as likely to default compared to the older age groups (from 45 and above). Patients with no formal education were four times more likely to default on treatment [AOR=4.32, (95%) CI, 2.57-7.24] than those who had completed high school. Those unemployed at the time of the study were twice as likely to default on treatment [AOR=2.83, (95%) CI, 1.85-4.32] as to those employed. Patients residing in a rural area were nine times more likely to default on ART [AOR=9.47, (95%) CI, 6.53-13.7] compared to those residing in an urban area.

CONCLUSIONS: Strategies and intervention should be patient-focused, with comprehensive individualized programs and interventions. Socio-demographic factors should be targeted to improve ART adherence.