HIV Epidemic Model: NYC HIV Epidemic Model: East Africa
About the Model
This web-enabled version of our published HIV transmission model for New York City allows a user to identify the intervention or combination of interventions which provides the most healthcare benefit per dollar spent. Because budgets may vary from user to user, we present the model’s results as an efficiency frontier. Each point represents a combination of interventions. Points on the frontier deliver the greatest health benefit within the user’s budget and therefore should be considered, while points beneath the frontier deliver less value per unit cost and are not preferred choices. Note that each point may signify one intervention or a combination of interventions.
What's "beneath the hood?"
A deterministic compartmental model of HIV transmission that includes both sexual transmission and transmission through needle-sharing during injection drug use. Additional details and a full technical appendix can be found in the following publication.
Using the Model
The numbered steps below correspond to the steps in the left frame on the model page.
- 1. Enter your annual budget for HIV/AIDS treatment and prevention services.
- Note that the model currently runs in optimization mode. It will determine the most cost effective combination of interventions for your budget.
- Two additional run modes are still under construction (coming soon)!
2. Select a treatment/prevention strategy
- Interventions are grouped into strategies. Choosing an overall strategy will populate the list of interventions in the navigation pane.
- If you wish to choose interventions from across strategies, choose Combine Strategies from the pull-down menu.
3. Choose a level of evidence.
- Interventions will be restricted by your chosen minimum level of evidence. For a glossary of the evidence levels, hover over the informational (?). Level A represents the strongest level of evidence while level D represents the weakest. Choosing D will present interventions at level of evidence D and above (A through D).
4. Choose interventions.
- The interventions displayed match the strategy you chose in Step 2 and the level of evidence you chose in Step 3.
- For each intervention, click the box to the left of the name to set whether you wish to include or exclude each intervention in your optimization. You should include interventions that you have available to implement in your area. If you do not wish to consider an intervention, be sure that the box next to its name is not checked.
- For each intervention, choose a target population from the pull-down menu.
- If there are interventions that you must implement (whether or not they would normally be found to be “optimal”), click the “Require” box. This will ensure that the optimization only considers packages of interventions that contain your required interventions.
5. Choose additional Settings.
- Reach- Choose a value for “Reach” to determine the portion of the target population to whom the intervention is applied.
- Effectiveness- Choose an effectiveness assumption. “Baseline” will use the best estimates from published literature, while “Optimistic” or “Pessimistic” will make the effectiveness assumption stronger or weaker.
Finished entering parameters? Click RUN THE MODEL at the top to see your results.
Understanding the results
Three elements appear...
- A cost effectiveness frontier graph
- A summary of your "Best Intervention Package for Budget" (green window)
- A table of information on each package plotted in the cost effectiveness frontier graph
Cost effectiveness frontier graph
- All packages (combinations) of interventions that are found to be the most cost effective are plotted as blue circles on the graph. The blue line connecting them is the cost effectiveness frontier. These points represent the options that will give you the most healthcare per dollar spent.
- The black dots beneath the frontier line represent packages which are less cost effective than packages on the frontier. These packages yield less healthcare per dollar spent and are not ideal choices.
- The package that gives the most healthcare per dollar within your budget is shown by the green circle on the graph. A vertical dotted line helps to highlight this point. Note, your budget is summed over twenty years, the time horizon of the model, so this value is a multiple of twenty on the annual budget you entered in Step
You can zoom on the graph by clicking and dragging your mouse over an area of the graph. Click “Reset Zoom” in the upper right to zoom out.
Best Intervention Package for Budget
- Details of the best intervention package for your budget are shown in the green window beneath the frontier. All outcomes are shown for a twenty-year time horizon. Review the list of interventions and the relative portion of your budget (over twenty years) that should be spent on each to optimize the number of infections averted.
Data table of frontier packages
- Below the “Best Intervention Package for Budget” window is a data table. The entries are separated by gray lines. Each entry represents a package of interventions plotted on the frontier graph.
- At the top of each entry in bold is the package summary. Beneath the summary are the interventions in the package and the cost of each intervention. Note that all values are over the twenty year time horizon of the model.
- The data packages shown with a light blue background correspond to points on the cost effectiveness frontier (they provide the most healthcare per dollar spent.)
- The data packages not shaded in blue represent the points beneath the frontier which are less optimal- they provide less healthcare per dollar spent.
Tailoring for your project
The complexity and run time of the model limits the features that can be offered in real time on the web. If you have a research question you’re curious to explore or want to partner with us to create a transmission model for your geographical area, please contact TORCH.
Nov 28, 2014 Published in AIDS
Evaluating the impact of prioritization of antiretroviral pre-exposure prophylaxis in New York.
Kessler J, Myers JE, Nucifora KA, Mensah N, Toohey C, Khademi A, Cutler B, Braithwaite S.
September 13, 2013 Published in PLoS One
Averting HIV infections in New York City: a modeling approach estimating the future impact of additional behavioral and biomedical HIV prevention strategies.
Kessler J, Myers JE, Nucifora KA, Mensah N, Kowalski A, Sweeney M, Toohey C, Khademi A, Shepard C, Cutler B, Braithwaite RS.