Dissertation Abstract

Distance to Mammography Facilities as a Predictor of Breast Cancer Screening and Stage at Diagnosis

The current study used Geographical Information Systems (GIS) to link health data geographically to the location of mammography facilities to examine mammography utilization and breast cancer stage at diagnosis. The current study examined data from Colorado Cancer Registry included 2042 women who were diagnosed with breast cancer during the period of 1999 to 2001. The majority of the study populations were White, and non-Hispanic in origin, ages ranged from 25 to 98 years. Almost three-quarters of the women had private health insurance and one-fifth had Medicaid and/or Medicare. Most of the women had their cancer diagnosed at a non-advanced stage and almost 90% of the women had a previous mammogram. Distance to a mammography facility was measured in two different ways using Geographic Information System (GIS). First, a straight-line distance was measured with ZipFind Deluxe 5.0 software between the centroids of two Zip codes (woman's residence Zip code and mammography facility Zip code where the index examination was done). Second, the Floating Catchment Area (FCA) Method was used which considered all the available facilities within an arbitrary radius using Arc GIS 9.0 software that measured access to mammography facilities. Moreover, access to mammography facilities was measured for several different radii. Several different radii were used since it is not well established in the literature what minimum distance a woman would be willing to travel to get a mammogram.

Two different dependent variables, 'breast cancer stage at diagnosis' and 'previous mammogram' were regressed separately with a number of independent variables including age, race/ethnicity, ethnicity, education, insurance status, family history, hormone replacement therapy, and physician recommendation, along with distance to mammography facility where the index examination was done, and then by access to mammography facilities considering all available facilities that woman might use. Results from logistic regressions found age, family history, hormone replacement therapy, physician recommendation, and breast cancer stage at diagnosis were significant predictors of having had a previous mammogram in both models (distance measure and access measure). In the logistic regression models, women who had a high access to mammography facilities were statistically significantly less likely to have had a previous mammogram compared to women who had low access, when the minimum distance a woman would be willing to travel was considered 10 miles, 30 miles and 40 miles. (Abstract shortened by UMI.)

This dissertation citation and abstract are published with permission of ProQuest Information and Learning. Further reproduction is prohibited without permission.