University of California, San Francisco Logo

University of California, San Francisco | About UCSF | Search UCSF | UCSF Medical Center

Home > PANCEA > Goals & Methods
Goals and Methods

PANCEA's goal is to provide information that policy makers and program managers can use to improve the allocation of resources for HIV prevention. Thus, it aims to increase the number of cases of HIV prevented for every dollar spent.

PANCEA proceeds by providing new data on the costs and outputs of HIV prevention programs in developing countries. These data will be used in conjunction with epidemiologic models to predict the effect of specific prevention strategies on the number of new HIV infections. Secondarily, PANCEA will identify the conditions that predict high levels of efficiency. Particular attention will be devoted to measuring the relationship between program size and efficiency to the impact of client recruitment strategies, and to the relationship between program efficiency and the enabling environment.

PANCEA's research strategy employs methods for describing cost functions that are new to the field of HIV prevention, though commonly used in private-sector microeconomic analysis. These include two distinct but complementary methods for estimating efficiency: The "case study" method uses in-depth compilation of cost and output data over the program's entire life. This is being performed in 40 prevention programs, eight from each of the five countries. This will allow us to describe both the cost function that relates program costs with outputs, and the underlying production function that describes the particular mix of inputs the program used. The data required will be month-by-month or quarter-by-quarter spending (by budget category) and program outputs (e.g., condoms distributed or counseling sessions).

A second, econometric method will be used to gather more limited cost and output information from 200 programs, typically 40 per country, with a focus on VCT. These data will be used to construct a model that describes variation in program efficiency. This model will regress program efficiency (cost per unit output) against a set of predictors: factor prices (e.g. wages), other selected cost items (e.g., outreach and publicity), demand-side constraints (e.g., population density, percent of population estimated to be at risk), and other variables that may affect efficiency (e.g., program scale, source of funds, private sector versus public sector, and the enabling environment). We use the case study approach to identify important hypotheses to test in the econometrics, and also to understand the reasons for 'outlier' programs that are highly efficient or inefficient.

Based on the findings yielded by these methods, we will develop computer simulation models that estimate program outputs based on prevention program components and key predictors of efficiency. The prevention components will include overall size ("scale") as well as specific activities, such as recruitment and use of volunteers. These models will present results in graphic formats useful for policy and program design. Special attention will be devoted to variables that capture the interactions between targeted interventions and mass media campaigns and variables that reflect the characteristics of the enabling environment.

In order to generate estimates of cost-effectiveness (cost per HIV infection averted), we will translate program outputs into behavior change estimates than can be used as inputs for epidemic models. We will specify the translation from program outputs into behavior change estimates based on existing studies and theory. The model will have the flexibility to portray specific study settings. It will run simulations using the behavioral outputs to generate estimates of change in HIV incidence and cost-effectiveness. It will portray important prevention cost-effectiveness issues within each country, such as comparing efficiency across prevention strategies and scales.

We will disseminate PANCEA's products through several mechanisms intended to facilitate rapid availability to policy makers: HIV InSite will post data collection instruments, selected reports, findings, and simulation tools. We will publish in peer-reviewed journals and present at scientific conferences. We will also meet in-country HIV policy makers, and convene workshops with prevention and modeling experts and policy makers.