Central Data Catalog

Citation Information

Type Journal Article - PloS One
Title Rates and Predictors of Non-Adherence to Antiretroviral Therapy among HIV-Positive Individuals in Kenya: Results from the Second Kenya AIDS Indicator Survey, 2012
Volume 11
Issue 12
Publication (Day/Month/Year) 2016
Page numbers e0167465
URL http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0167465

Understanding the levels and associated factors of non-adherence to antiretroviral therapy (ART) is crucial in designing interventions to improve adherence and health outcomes of ART. We assessed non-adherence to ART among HIV-infected persons reporting ART use in a nationally representative survey in Kenya.


The Kenya AIDS Indicator Survey 2012 was a population-based, household survey of persons aged 18 months-64 years conducted in 2012–2013. Self-reported information was collected on demographics, sexual behaviour, HIV status, and ART use. Blood was collected for HIV testing, and if HIV infected, CD4 and viral load testing. HIV-positive specimens were tested for the presence of antiretroviral (ARV) drugs using a qualitative ARV assay using liquid chromatography-tandem mass spectrometry. HIV-positive persons who reported receiving ART but did not have the ARV biomarker present were defined as being non-adherent to their ARV medication. We restricted our analysis to HIV-infected persons aged 15–64 years who reported receiving ART and had laboratory-confirmed results from ARV testing. Multivariate logistic regression was used to identify variables associated with non-adherence.


A total of 648 (5.6%; CI 4.9–6.3) tested HIV-positive of whom 559 (86.3%) had sufficient volume of blood to be tested for ARV drugs. Of those, 271 (47.7%; CI 41.8–53.6) self-reported HIV-positive status during the interview and 186 (69.1%; CI 62.2–76.0) of those reported taking ART. The ARV biomarker was absent in 18 of 186 individuals (9.4%; CI 4.9–13.8) who thus were defined as being non-adherent to ART. Non-adherence was associated with being aged 15–29 years (AOR 8.39; CI 2.26–31.22, p = 0.002) compared to aged 30–64 years, rural residence (AOR 5.87; CI 1.39–25.61, p = 0.016) compared with urban residence and taking recreational drugs in the past 30 days (AOR 5.89; CI 1.30–26.70, p = 0.022).


Overall, less than 10% of Kenyans aged 15–64 years on ART were not adhering to their HIV medication, highlighting the success of the Kenyan national ART program. Our findings, however, point to the need for targeted interventions particularly for young persons, those in rural areas to improve adherence outcomes, as well as delivery of treatment programs that include psychosocial support as a preventative measure to minimize substance abuse and the risk of treatment failure.

Related studies