Antiretroviral therapy and HIV prevention in India: Modeling costs and consequences of policy options

Mead Over, Elliot Marseille, Kurapati Sudhakar, Julian Gold, Indrani Gupta, Abhaya Indrayan, Subhash Hira, Nico Nagelkerke, Arni S.R.Srinivasa Rao, Peter Heywood

Research output: Contribution to journalArticlepeer-review

31 Citations (Scopus)


OBJECTIVES: The objective of this study is to assess the costs, cost-effectiveness, and HIV epidemic impact of 3 antiretroviral therapy (ART) policy options. STUDY DESIGN: We constructed an epidemiologic model to predict the course of the HIV epidemic in the absence of expanded ART availability. Based on background studies of the willingness to pay for ART among patients with AIDS, of the costs to the government of the alternative treatment interventions, and of ART's likely effects on HIV transmission, we simulated the consequences of 3 possible alternative government ART policies. RESULTS: A program to reduce the negative consequences of the currently unstructured private-sector provision of ART is the most cost-effective of the 3 options at a 10% discount rate and least cost-effective at a 3% rate. The costs and cost-effectiveness of all options are highly sensitive to the effect of ART on condom use. CONCLUSION: The design of ART policy should capitalize on the potential of ART to decrease HIV transmission through institutional arrangements that reward effective prevention programs, thereby raising the likelihood that treatment has beneficial rather than negative external effects.

Original languageEnglish
Pages (from-to)S145-S152
JournalSexually Transmitted Diseases
Issue number10 SUPPL.
Publication statusPublished - Oct 2006
Externally publishedYes

ASJC Scopus subject areas

  • Dermatology
  • Public Health, Environmental and Occupational Health
  • Microbiology (medical)
  • Infectious Diseases


Dive into the research topics of 'Antiretroviral therapy and HIV prevention in India: Modeling costs and consequences of policy options'. Together they form a unique fingerprint.

Cite this