Lay Abstract
To date, there are no computer models available to patients, providers, or researchers that are up-to-date and customizable to fit the needs of the user. Every person or group has to recreate the system.
Computer Models Can Show:
- Where more research is necessary
- The long-term effectiveness of early intervention (such as weight loss or glucose control)
- Where policy makers should focus efforts for early treatment and prevention
Diabetes is a complex disease with many stages and complications. Much research focuses on a single aspect of diabetes to better understand that one piece of the puzzle. There is little research done to understand the entire process of diabetes because such studies are expensive and long. Often, by the time the study is finished, there are new standards of treatment that were not considered by the study. Computer models of the entire process can put all the pieces of the puzzle together to provide a complete picture.
Goals:
- To improve estimates of progression rates
between stages of chronic diseases.
- To develop user-friendly software that helps researchers to model chronic diseases
This modeling software will be open source for the convenience of other researchers. The second goal requires continual updating of the parameter estimates and comparison of those estimates to many different populations to verify that the estimates provided with the model are appropriate for all Americans, not just a few.
Technical Abstract
Computer modeling of chronic diseases allows a more comprehensive view of the disease process than a single study since the computer model can be comprised of the accumulated knowledge in the field. For a complicated disease like Non-Insulin Dependent Diabetes Mellitus, (NIDDM), which develops over an extended period of time, a single randomized clinical trial provides a fragmentary view of progression. However, implementation of a longitudinal study of disease progression is expensive, time-consuming, and rapidly out-dated.
Limitations of current methods:
- Current methods of estimation do not accommodate many aspects of the study design
- There are no disease modeling tools available to clinical researchers
Almost all of the existing diabetes models are not available to persons outside a proprietary health system. The few remaining models are limited in scope and applicability. Thus each group has to develop their own engine, interface, and output mechanisms in addition to the parameters. This replication of effort limits the progress and accessibility to diabetes researchers.
Expected Benefits:
With disease modeling techniques, clinical researchers can place their results in the context of an entire disease process and simulate the long-term effects of early intervention. Moreover, diabetes models can inform policy makers of key areas to target for prevention or guide reimbursement policies for preventive care.