Identifying Targeted Health Communication Strategies to Reduce Disparities: Understanding Contextual Determinants of Mammography Use Among Subgroups of Poor and Minority Women

Theoretical Background and research questions/hypothesis: Despite progress in increasing breast cancer screening rates, disparities in screening mammography persist. Women who underutilize mammography are disproportionately poor and from racial/ethnic minority groups, and current evidence-based interventions promoting use of mammography are not particularly effective among these women. The general approach of segmenting populations or audiences into distinct subgroups is used widely in marketing and communication to design and better target products and messages; however, its use in public health to develop disparity-reducing interventions is rare. This study applies segmentation theory and methods to health disparities research for the purpose of developing targeted health communication strategies. The study had two aims: (1) to identify subgroups of poor and minority women with differential patterns of breast cancer screening; and (2) to identify contextual determinants of screening that may be important or unique to each subgroup.

Methods: Data were collected as part of a large RCT of callers to United Way 2-1-1 Missouri, a telephone information and referral system that connects callers to resources addressing their unmet basic needs (e.g., housing). During 2010-2012, 711 female 2-1-1 callers ages ≥40 answered questions related to health behavior, demographics, and contextual issues (social capital, perceived stress, unmet basic needs, life outlook, and children living in the home). Census 2010 data on neighborhood characteristics were linked to participant survey data. CHAID was used to identify subgroups of women with differential screening patterns, and multilevel multinomial logistic regression was used to examine contextual determinants of screening by subgroup.

Results: The study found six distinct subgroups of women with varying patterns of screening (p<0.001). After taking neighborhood characteristics into account and controlling for demographics, contextual determinants of being in subgroups with the lowest screening rates included having a child living in the home (aOR 2.7, 95% CI 1.8–4.2), not having enough food to eat (aOR 1.8, 95% CI 1.2–2.7), and having higher perceived stress (aOR 1.1, 95% CI 1.1–1.2). Being in the subgroup with the highest screening rate was associated with having high social capital (aOR 2.0, 95% CI 1.2–3.2). Notably, many contextual variables were stronger determinants of screening than other, widely-known determinants such as receiving a provider recommendation for mammography.

Conclusions: The study identified distinct subgroups of poor and minority women and contextual determinants of mammography use by subgroup. As life burdens can interfere with preventive health behavior, current evidence-based health communication interventions (e.g., education, screening reminders) are minimally effective with these populations. Strategies are needed to connect women to resources that enable them to make preventive health a priority in their lives. For example, health communication interventions that link women to social services as a necessary precursor to behavioral change seem promising. This study begins to provide a foundation for such work.

Implications for research and/or practice: Findings can aid in developing targeted communication strategies to help reduce disparities in breast cancer screening. For populations suffering persistent disparities, health communication interventions that take contextual determinants into account may be more appropriate and, ultimately, more effective.