The mammography dissemination model generates screening histories representative of the US population for the years 1975 - 2000. These generated histories were used by CISNET groups to model the impact of mammography and treatment on incidence and mortality over the same time period. More information on the CISNET breast base case effort is available on the CISNET website.

Two distinct statistical modeling efforts were performed and combined through simulation to generate screening life histories for women. Cross-sectional survey data from the NHIS survey was used to estimate the cumulative distribution for the time to first mammography. Longitudinal data from the Breast Cancer Surveillance Consortium was used to model repeat screening behavior. The two pieces were combined in the simulation program posted to generate screening exam histories for women. Note that the simulation does not model diagnosis of cancer or death, so the user of the simulation must truncate the screening history at the time of diagnosis or death.

Description of simulation

This simulation generates individual level mammography histories that are representative of the US population over the years 1975-2000. Given a year of birth, the program will generate an example series of mammography exams for a woman in that birth cohort. In some cases, women may receive no mammography exams. At the other extreme, some women may continue to receive mammograms approximately annually after an initial exam. This demonstration on this web site will simulate up to 9 women at a time.

The details of the simulation and the statistical model that it is based on can be found in Cronin et al. (1). In brief, the time until a first mammography exam was based on a series of national surveys that asked women if they ever had a mammogram (2). Intervals between subsequent mammography exams were based on longitudinal data collected by consortium of mammography registries (3). These two sources of information are combined by the simulator to generate screening histories between the years 1975 and 2000 for women born before 1965.

Once a women receives an initial mammography, she is assigned a Selection Parameter value that represents her tendency to return for a subsequent screening mammogram. This value determines a general screening category of annual screener, biennial screener or an irregular screener for the woman. The value is kept constant over the course of her lifetime, although the classification could change at different ages, with the highest probability of being a regular (either annual or biennial) screener in the years 50-70 and the higher probability of being an irregular screener between 40 and 50 years and after 70 years.

Each woman is also assigned a frailty parameter that represents individual variation within screening category. Within each screening category (annual, biennial, and irregular screening) each individual has different gap time probabilities. For example some woman may return each year on their birthday and always have exactly 1 year gap times. Others may still be classified as an annual screening, but really return somewhere between 12 to 18 months.

The simulation does not truncate a woman's mammogram history at a point of diagnosis of breast cancer or at the time of death. For example a generated screening history for women born in 1900 will assume that she lives until age 100 (calendar year 2000). Censoring due to death or breast cancer diagnosis must be done outside this simulation program.

Contact Information

If you have questions on the modeling used in the simulation or if you would like access to the simulation source code, please send an email to cronink@mail.nih.gov.

References

1. Cronin KA, Yu B, Krapcho M, Miglioretti DL, Fay MP, Izmirlian G et al. Modeling the dissemination of mammography in the United States. Cancer Causes Control. In press
2. National Health Interview Surveys http://www.cdc.gov/nchs/nhis.htm
3. National Cancer Institute. Breast Cancer Surveillance Consortium: Evaluating Screening Performance in Practice. NIH Publication No. 04-5490. Bethesda, MD: National Cancer Institute, National Institutes of Health, U.S. Department of Health and Human Services, April 2004. Available at: http://breastscreening.cancer.gov/espp.pdf