![[ Visit Client Website ]](images/banner.gif)
Schedule:
Thursday, June 2, 2016
Pacific N/O (Hyatt Regency San Francisco)
* noted as presenting author
There are a variety of approaches for representing age and time dependent attributes in computer simulation modeling of population health. Demographers commonly use annual discrete time steps to predict population dynamics, while continuous time methods such as system dynamics have typically relied on an “aging chain” where people move through age categories as a series of first-order delays. While providing a reasonable approximation in many situations, first-order time delays introduce distortions in the age distribution that can confound analysis of simulation results.
Over the years, there have been a variety of ways to work around this limitation in modeling population health dynamics. In computer simulation of cancer trends in incidence and mortality at the population level, best practices developed by the Cancer Intervention and Surveillance Modeling Network (CISNET) emphasize the importance of tracking birth cohorts due to the time variant risk exposures for different birth cohorts. However, this approach has not been used in system dynamics population health modeling.
This poster presents an innovative and generalizable approach to modeling obesity and cancer incidence using birth cohorts and takes into consideration time-variant risk factors such as smoking and HIV/AIDS that impact the incidence of diffuse large B-cell lymphoma, the most common type of non-Hodgkin lymphoma. The method developed successfully replicates population dynamics for the United States from 1975 to 2010.