Understanding Poor and Minority Women who are Unscreened or Underscreened for Breast Cancer

Date of Award

Summer 8-15-2012

Author's School

Graduate School of Arts and Sciences

Author's Department

Social Work

Additional Affiliations

Brown School of Social Work

Degree Name

Doctor of Philosophy (PhD)

Degree Type



Despite progress in increasing breast cancer screening rates for more than two decades, disparities in screening mammography persist. Women who are unscreened (i.e., never had a mammogram) or underscreened (i.e., last mammogram was more than two years ago) are disproportionately poor and from racial/ethnic minority groups. Targeted interventions are badly needed as behavioral interventions promoting use of mammography have not been particularly effective among unscreened and underscreened women. Little is known about the broader context of these women's lives, yet this information seems particularly promising for identifying the intervention needs and capacities of these populations.

Using a sample of predominantly poor and racial/ethnic minority women who are age-eligible for mammography, the study aims to (1) identify distinct subgroups of women based on patterns of variation in demographic and health-related characteristics associated with screening behavior and (2) identify contextual factors associated with screening behavior by examining the extent to which contextual variables distinguish membership in the identified subgroups. As part of a larger ongoing randomized intervention trial, callers to United Way 2-1-1 Missouri complete a cancer risk assessment and baseline survey at the end of their standard 2-1-1 call. Over a 20-month period during 2010-2012, 711 female callers ages 40 and older answered questions related to use of mammography, other health behaviors, provider recommendation for mammography, social norms for mammography, health insurance coverage, self-rated general health, and individual contextual issues (social capital, perceived stress, unmet basic needs, sense of coherence, and children living in the home). Data from the 2010 U.S. Census assessed area-level characteristics of participants' residential areas. Data were analyzed using Chi-squared Automatic Interaction Detection (CHAID) and multinomial logistic regression.

The study found six distinct subgroups with varying patterns of screening (p<0.001). After taking area-level characteristics into account and controlling for demographics, contextual determinants of being in the subgroup with the lowest screening rate were having a child in the home (aOR 2.70, 95% CI 1.75-4.16) and not having enough to eat in the next month (aOR 1.76, 95% CI 1.17-2.70). Contextual determinants of being in the subgroup with the highest screening rate included having high social capital (aOR 1.97, 95% CI 1.18-3.23) and a weaker sense of coherence (aOR 0.97, 95% CI 0.93-0.99).

The study demonstrates the utility of applying a subgroup-based approach to categorizing and studying breast cancer screening disparity populations. Findings provide an initial basis for better understanding the heterogeneity of these groups and can be used to inform the development of more targeted and appropriate interventions promoting use of mammography in these populations.


English (en)

Chair and Committee

Matthew Kreuter

Committee Members

Melody Goodman, Darrell Hudson, Jason Purnell, Douglas Luke, Vetta Sanders Thompson


Permanent URL: https://doi.org/10.7936/K7KS6PJ5

This document is currently not available here.