HIV prevention costs and program scale: data from the PANCEA project in five low and middle-income countries
1 Institute of Health Policy Studies, University of California, San Francisco, USA
2 George Institute for International Health – India, Hyderabad, India; Health Studies Area, Centre for Human Development, Administrative Staff College of India, Hyderabad, India; School of Public Health and George Institute for International Health, University of Sydney, Sydney, Australia
3 Office of AIDS Research, National Institutes of Health, Bethesda, USA
4 Instituto Nacional de Salud Pública, Cuernavaca, Mexico
5 HIVAN(Centre for HIV/AIDS Networking), Durban, South Africa
6 Axios International, Paris, France
7 St. Petersburg Pavlov State Medical University, St. Petersburg, Russia
8 Elizabeth Glaser Pediatric AIDS Foundation, Washington, D.C., USA
9 Center for Global Development, Washington, D.C., USA
10 AIDS Infoshare, Moscow, Russia
11 George Institute for International Health – India, Hyderabad, India; Health Studies Area, Centre for Human Development, Administrative Staff College of India, Hyderabad, India
12 Axios International, Kampala, Uganda
13 World Bank, Washington, D.C., USA
Citation and License
BMC Health Services Research 2007, 7:108 doi:10.1186/1472-6963-7-108Published: 12 July 2007
Economic theory and limited empirical data suggest that costs per unit of HIV prevention program output (unit costs) will initially decrease as small programs expand. Unit costs may then reach a nadir and start to increase if expansion continues beyond the economically optimal size. Information on the relationship between scale and unit costs is critical to project the cost of global HIV prevention efforts and to allocate prevention resources efficiently.
The "Prevent AIDS: Network for Cost-Effectiveness Analysis" (PANCEA) project collected 2003 and 2004 cost and output data from 206 HIV prevention programs of six types in five countries. The association between scale and efficiency for each intervention type was examined for each country. Our team characterized the direction, shape, and strength of this association by fitting bivariate regression lines to scatter plots of output levels and unit costs. We chose the regression forms with the highest explanatory power (R2).
Efficiency increased with scale, across all countries and interventions. This association varied within intervention and within country, in terms of the range in scale and efficiency, the best fitting regression form, and the slope of the regression. The fraction of variation in efficiency explained by scale ranged from 26% – 96%. Doubling in scale resulted in reductions in unit costs averaging 34.2% (ranging from 2.4% to 58.0%). Two regression trends, in India, suggested an inflection point beyond which unit costs increased.
Unit costs decrease with scale across a wide range of service types and volumes. These country and intervention-specific findings can inform projections of the global cost of scaling up HIV prevention efforts.