Date on Master's Thesis/Doctoral Dissertation


Document Type

Master's Thesis

Degree Name



Bioinformatics and Biostatistics

Committee Chair

Wu, Dongfeng

Author's Keywords

Breast cancer; Cancer screening; Disparities; Trend; Mammography; Pathology result


Breast--Cancer--Diagnosis; Medical screening; Breast--Radiography; Mobile hospitals


This study used the de-identified breast cancer screen data which recorded the information of 58231 screens and 22534 participants with different background. The analysis of disparities on age, ethnic, health insurance and screen location has been done based on screen and participant number. Women younger than 50 or white, or the women without insurance are most likely to take the screen. The percentages are 40%,64% and 40% under each category. The most popular location for screen is partnership clinic, which comprise 43% of screens. The year 2006 has the most screens which is 6861, and 2001 has the least which is 4813. The trend over the years and the disparities of pathology results were assessed. Chi-square test was performed to check the association. The odds ratios based on participants number are calculated. For race/insurance, in age groups "50-", "50-59" and "70+", the odds of white people with insurance is about 1.7, 1.9 and 0.4 of black people with insurance. No difference was found in age group "60-69'. For race/pathology result, the significant result is only for age group "70+", and the odds of white women with "Negative" pathology result is 4.4 times of black women with "Negative" pathology result. Besides, for insurance/pathology result, the odds of women with insurance "Yes" having "Negative" pathology result are 2.26 and 2,08 times of the women with insurance "Yes" having "Negative" pathology result for age group "50-" and "50-59".The follow-up study was presented in this thesis. The maximum number that each person took the screen is 17. The majority (45%) only take screen once. About 19% of women take twice, 12% take triple. Only 1 woman takes 17 times. Pearson and Spearman correlation coefficients were calculated to check the association.