Towards more cost-effective HIV prevention planning

Some background: The Philadelphia Prevention Planning Group (PPG) and the City of Philadelphia's AIDS Activity Coordinating Office (AACO) work together to decide which communities have the greatest need for HIV prevention activities, and which interventions work best for those communities.  This is part of what we call "community planning."  AACO decides how the money gets spent and which organizations do the work, keeping in mind the community's needs and what is proven to work best. The PPG provides the community's voice to the process.

In 2011, the Centers for Disease Control and Prevention (CDC) approached the Philadelphia Department of Public Health (PDPH) to participate in the design and testing of a model for how to best spend HIV prevention funds to prevent the spread of HIV. The model was supposed to help health departments make cost-effective decisions to meet the goals of the National HIV/AIDS Strategy.  You can read more about the data used in the model and how the model influenced HIV prevention planning in the 2012 Philadelphia Jurisdictional Plan  (pages 93-95). 

AIDS.Gov recently released a podcast explaining this resource allocation model.  It also talks about the pilot project in Chicago, Nebraska, and Alabama, which is supposed to determine if the model works for those communities and to see if their experiences are similar to Philadelphia's.

The AIDS.Gov blog post explains more about the pilot project, HIV RAMP:

"That model was developed to help Philadelphia’s health department leadership identify the optimal allocation of HIV prevention funds across populations and prevention interventions that would prevent the most new HIV infections. Using local demographic and HIV epidemiological data inputs, and based on calculations of the cost per new infection averted, the model recommended the optimal allocation of the jurisdiction’s HIV prevention budget among several evidence-based interventions for the populations most impacted. Feedback from Philadelphia indicated that the model’s outputs served as a very useful “roadmap” to inform decision making for HIV prevention resource allocation within the jurisdiction, informing both applications for federal funds as well as efforts to strengthen local HIV prevention activities.

The current project, HIV RAMP, involves refining and piloting the original Philadelphia model in three additional jurisdictions that have different local profiles and HIV prevalence rates (Chicago, Nebraska, and Alabama), testing a technical assistance (TA) protocol to support jurisdictions in using the model, and assessing the feasibility of a software or online version of the model that could be more broadly used by other health departments.

“Because making decisions about how to spend HIV prevention funding is never ‘black and white,’ health departments need tools to help them decide how to support the best combinations of effective, evidence-based prevention strategies in their communities,” said Dr. Ronald Valdiserri, Deputy Assistant Secretary for Health and Director of OHAIDP. “Through this pilot project, we hope to develop a practical tool that can help communities apply the principles of the NHAS on the ground.”

Here are some highlights from the Philadelphia jurisdictional plan's discussion of the model development, focusing on the results:

"Cost per new infection averted is an integral part of this resource allocation model. For HIV testing in a clinical setting, the cost per infection averted is $51,293, making it the second most cost-effective intervention for Philadelphia. Testing in non clinical settings for IDU (3) and MSM (1) rounded out the top three most cost-effective interventions for averting HIV transmission at $53,935 and $17,965, respectively. The least cost-effective interventions used in the model were behavioral interventions for HIV- high risk heterosexuals ($15,642,127) and IDU ($2,931,406). Behavioral interventions were not found to be as cost-effective as HIV testing, even for HIV+ individuals. Adherence to ART (4), retention in care (5), partner services, and linkage to care (8) were ranked in the middle of the list of interventions. 

The optimal resource allocations according to those calculations would avert 72 infections in the first year and 245 infections within five years. Testing in clinical settings would receive 39% of the resources and avert 20 infections in one year, 93 infections in five years. Retention in care would receive 29% of the resources and avert 27 infections in year one and 52 infections within five years....Two-thirds of the resources would be allocated to HIV testing in this optimal model, because of the cost-effectiveness of HIV testing (non-targeted), particularly for MSM and IDU communities.  These allocations would serve 1,930 HIV+ individuals and result in 792 new diagnoses. Return on investment ranges (prevention dollars only) from 1.21 in year one to 4.42 in year five....

Recommendations from the model include allocations decisions should be made by both cost and effectiveness. Testing should be prioritized in clinical and non-clinical settings. More resources should be allocated to interventions located in care-settings. Behavioral interventions for HIV- individuals are not allocated resources. Behavioral interventions for MSM can be allocated resources under certain conditions. "

If you want to learn more about the community planning of HIV prevention in Philadelphia, check out the Prevention Planning Group or come to a meeting.