Risk management in HIV/AIDS: ethical and economic issues associated with restricting HAART access only to adherent patients

Abstract

Like many other developing nations, South Africa faces the challenge of mobilising resources in response to the HIV pandemic. There is a large budget gap  between the ideal and the actual amount of funding needed to achieve universal  access to highly active antiretroviral therapy (HAART). In addition to financial demands, new burdens are being placed on HAART programmes with the emergence of HIV drug resistance (HIVDR). Thus, a major threat to successful HAART rollout is HIVDR due to non-adherence to HAART. The use of HAART as a primary and secondary HIV-prevention strategy could be ineffective in situations characterised by high rates of non-adherence. In this context, the research looked at issues related to adherence and non-adherence to HAART from the perspective of the provider. Using the software TreeAge Pro 2009, we developed  a Markov model to project economic outcomes for a hypothetical cohort of HIV/AIDS patients on HAART. The model compared two scenarios: adherence and non-adherence to HAART. Input data for the model was obtained from existing literature on HAART uptake in South Africa. Moral arguments were  analysed and managed through moral reasoning and critical thinking. Discounted lifetime costs for adherent and non-adherent HAART patients in South Africa  were estimated at US$9 771 and US$14 762, respectively. The model showed the  loss of 4.55 quality-adjusted life years (QALYs) for non-adherent patients, which could be otherwise gained through improved adherence. The incremental  cost-effectiveness ratio (ICER) indicated that restricting HAART access only to  adherent patients was the dominant strategy. We suggest that, although not a panacea, the withholding or withdrawal of treatment from non-adherent individuals as a precautionary intervention has economic and moral merit.

Keywords: adherence, cost-effectiveness, economic evaluations, healthcare, medical ethics, modelling, moral reasoning, quality of life, simulation models, South Africa

African Journal of AIDS Research 2011, 10(supplement): 369–380