To Register      SMDM Homepage

Tuesday, 19 October 2004 - 10:45 AM

This presentation is part of: Oral Concurrent Session A - Simulation

EXPLAINING THE SECULAR TRENDS IN COLORECTAL CANCER INCIDENCE AND MORTALITY USING A POPULATION-BASED MICROSIMULATION MODEL

Amy B. Knudsen, BS1, Y. Claire Wang, MD, MS2, Milton C. Weinstein, PhD1, Jane C. Weeks, MD2, and Karen M. Kuntz, ScD1. (1) Harvard School of Public Health, Harvard Center for Risk Analysis, Boston, MA, (2) Harvard School of Public Health, Health Policy and Management, Boston, MA

Purpose: To examine the relative contribution of changes in risk factors, screening and treatment on the overall population trends in colorectal cancer (CRC) incidence and mortality.

Methods: We developed a dynamic first-order Monte Carlo model to simulate the US population aged 25 years and older from 1978 to 2000. The model tracks the development of adenomas and their progression to invasive CRC as a function of demographic characteristics and risk factors (e.g., smoking history, body mass index, and folate intake). Secular trends in risk factors were estimated from the National Health and Nutrition Examination Surveys, dissemination of CRC screening was derived from the National Health Interview Surveys, and advancements in treatment were obtained from SEER and SEER-Medicare linked data. We used the model to simulate the observed trends in CRC incidence and mortality from 1978 to 2000 and generate estimates of the number of cases and deaths that would have occurred under alternative scenarios regarding changes in risk factors, screening, and treatment over this time period. We then compared the results of these scenarios and decomposed the overall changes in incidence and mortality into the proportions explained by each factor.

Results: During the period 1978 to 2000, approximately 2.54 million Americans were diagnosed with CRC and 1.28 million died from the disease. The number of cases represents a 4.7% reduction from the estimated number that would have occurred if there had been no secular trends in risk factors and no dissemination of screening over this time period (8.6% reduction in 2000). Changes in risk factors alone account for 31.5% of the overall reduction, and 67.3% is attributable to screening. The number of cancer deaths represents a 13.1% reduction from the estimated number that would have occurred in the absence of changes in risk factors, screening, or treatment (22.0% reduction in 2000). Advancements in treatment alone account for 59.9% of the reduction, while risk factors and screening account for 7.8% and 30.7% of the decline, respectively.

Conclusions: Screening and advancements in treatment have played significant roles in the declines in CRC incidence and mortality. Population-based models can provide insight into observed disease trends and shed light on areas in which public health interventions could have a large effect.


See more of Oral Concurrent Session A - Simulation
See more of The 26th Annual Meeting of the Society for Medical Decision Making (October 17-20, 2004)