Introduction
All members of the Cancer Intervention and Surveillance Modeling Network (CISNET) for colorectal cancer are
collaborating on a set of common simulations to gain insight in similarities and differences between the models.
The CISNET-Colon group is taking a graduated approach to its base cases, proceeding in sequential fashion, with
each one becoming more complex and more reflective of the complex dynamics actually occurring in the US population.
By doing this, the group is trying to understand differences at each level before proceeding to the next one.
About the Colorectal Base Cases
Base Case I is an evaluation of how the natural history models (i.e., the adenoma-carcinoma sequence) compare
under a standardized set of conditions. These conditions simulate 1978, and then the model from 1978 forward
in time assuming that conditions are frozen as they were in this base year. Model outputs include incidence,
mortality, adenoma prevalence all by gender, age, stage of disease and localization of the cancer in the colorectum.
Base Case II compares 3 screening tests - colonoscopy, flexible sigmoidoscopy, and Fecal Occult Blood Test
(FOBT), with and without surveillance for those with positive findings. The comparison of the models is made
under the simplified assumption of a single screening event at age 65 with 100% compliance beginning in 1980.
The base case also considers the hypothetical case if colonoscopy had zero sensitivity for adenomas and another
run where the sensitivity for cancer is assumed to be zero.
This sequential approach to base cases should prove to be very fruitful in pinpointing model differences and
similarities, and future ones will address more realistic and policy-related issues. Some differences between
the models may continue to persist, because data is lacking to support specific assumptions. We can use our
different models to inform policy makers about the effects of different model assumptions on policy issues.
| 
