Social determinants of health and health inequities in breast cancer screening: a scoping review

Introduction This scoping review aims to highlight key social determinants of health associated with breast cancer screening behavior in United States women aged ≥40  years old, identify public and private databases with SDOH data at city, state, and national levels, and share lessons learned from United States based observational studies in addressing SDOH in underserved women influencing breast cancer screening behaviors. Methods The Arksey and O’Malley York methodology was used as guidance for this review: (1) identifying research questions; (2) searching for relevant studies; (3) selecting studies relevant to the research questions; (4) charting the data; and (5) collating, summarizing, and reporting results. Results The 72 included studies were published between 2013 and 2023. Among the various SDOH identified, those related to socioeconomic status (n = 96) exhibited the highest frequency. The Health Care Access and Quality category was reported in the highest number of studies (n = 44; 61%), showing its statistical significance in relation to access to mammography. Insurance status was the most reported sub-categorical factor of Health Care Access and Quality. Discussion Results may inform future evidence-based interventions aiming to address the underlying factors contributing to low screening rates for breast cancer in the United States.


Introduction
The social determinants of health (SDOH) are factors outside of the realm of medicine that impact health outcomes and quality of life on a daily basis (1).According to the World Health Organization (WHO), SDOH are defined as "the conditions in which people are born, grow, work, live, and age, and the wider set of forces and systems shaping the conditions of daily life (1)." These determinants of health can be divided into five categories: economic stability, education access and quality, health care access and quality, neighborhood and built environment, and social and community context (2).While factors within each of these categories can individually impact a different facet of a person's health, these categories often also work collectively to create facilitators and barriers to healthy behaviors and health outcomes (1)(2)(3).Such SDOH play a significant role in creating new and worsening existing Step 1: identifying research questions Three research questions were used for this scoping review: (1) What are the major SDOH hindering breast cancer screening in United States women aged > = 40?; (2) What were the major databases/ data sources used to capture SDOH data to assess its influence on breast cancer screening behaviors in United States women?; and (3) The senior author (LS) reviewed all tabulated data for accuracy and to resolve any discrepancies.Summary tables included an evidence table ( (18).The Healthy People 2030 is a set of science-based objectives with targets to monitor progress and motivate and focus action (18).The Healthy People 2030 first introduced SDOH objectives in 2010, following the World Health Organization's (WHO) call to address SDOH to maintain health and quality of life (18).The five categories listed reflect the social conditions and environments that are shaped by a wider set of forces and influence behavioral outcomes (18).Significance of associations between breast cancer screening as an outcome and identified SDOH were reported (Table 1).Table 2 included a list of databases from where the data was accessed, the availability status of the data (public/private), and the geographical level from where the data was extracted.Basic qualitative content analysis was carried out to identify similar themes in future directions across studies highlighted in Table 3.The three phases of qualitative content analysis for the results of primary qualitative research described by Elo and Kyngas (19) were applied: (i) preparation, (ii) organizing, and (iii) reporting.
Step 4 and 5: charting the data and collation, summarization, and reporting of results Study characteristics were tabulated for primary author/year, study design, sample size, study population, age range, study purpose, type of SDOH, SDOH category based on HP 2030, association between SDOH and outcome (significant/non-significant), and type of methodology/analysis used for data analysis (Table 1).Identified databases were tabulated by primary author/year, database/data source, public availability, and city/state/national level (Table 2).Each database was stratified based on availability (publicly available/not publicly available) and location (city/state/national level).Lessons learned from each relevant study were highlighted in Table 3.

Results
The initial study extraction resulted in 8,124 articles from PubMed (n = 1,293), EMBASE (n = 6,193), Web of Science (n = 527), and Cochrane (n = 111).Studies were excluded due to publication outside of the timeframe (n = 7,775), discussion of all types of cancer rather than focusing on breast cancer (n = 2,349), being a literature review or systematic review (n = 884), lack of focus on breast cancer disparities (n = 717), focusing on big data or no mention of SDOH (n = 124), focusing more on knowledge and attitudes rather than SDOH (n = 112), being an opinion piece or an editorial (n = 25), or emphasizing survival as an outcome rather than treatment (n = 22).Duplicate studies were also excluded (n = 82 from PubMed, n = 60 from EMBASE, n = 20 from Web of Science, and n = 2 from Cochrane).A total of 267 studies met the inclusion criteria from PubMed (n = 222), EMBASE (n = 40), and Web of Science (n = 5).An additional 195 studies were excluded after a full study review due to being an abstract and not a full text (n = 77), having a qualitative or experimental study design (n = 42), having no relation to SDOH (n = 63), and discussing cancer types in general rather than narrowing it down to breast cancer (n = 13).A total of 72 studies were retained for analysis (Figure 1).
The 72 included studies were published between 2013 and 2023.About half of the studies (58%) were published in 2018 or later (n = 42).Study designs included cross-sectional studies (n = 45); cohort studies (n = 18); and case-control studies (n = 9).Sample size ranged from n = 100 to n = 3,821,084 female adults with breast cancer while the age of this target population ranged from 40 to 89 years old (Table 1).

Priority populations
Priority populations who were actively involved (or targeted) in implementation activities were ethnically diverse female patients diagnosed with breast cancer including African American women; Muslim and Christian Arab American; Haitian women; Filipino women; and Korean American women.Another set of studies focused on women from programs, such as women from Geisel School of Medicine (n = 3,413), from the BSPAN program (n = 19,292), women who underwent mammography in Harvard Medical School (n = 9,575), female patients from a single institution undergoing breast radiotherapy (n = 1,057), presenting to radiology department (n = 758), mammogram facilities (n = 1,749), and at a quaternary care academic medical center (n = 738) (Table 1).
Additional studies focused on the characteristics of the women, such as women who have individual subscribers or employer supplemented (n = 95,661), are Medicaid-insured and Medicare fee-for service (n = 11), are insured but have not undergone mammogram in 24 months (n = 47,946), have no history of breast cancer (n = 181,755), have telephone access (n = 169,116), homeless women (n = 100), hospitalized women (n = 250), are medically underserved (n = 518), and have limited accessibility to mammogram (n = 73,718) (Table 1).

Database access and characteristics
Databases with the highest number of occurrences include data from the National Health Interview Survey (n = 8) [over a range of years from 2005 to 2018], the Breast Cancer Surveillance Consortium (n = 4), and the United States Department of Health (n = 2).Other databases used include the National Program of Cancer Registries, the National Assessment of Adult Literacy, and SEER Medicare.Of the 74 databases used, 47% (n = 35) are publicly available.The databases are available at the city (n = 16), county (n = 1), state (n = 28), and national (n = 30) levels (Table 2).

Lessons learned
Using the three phases of qualitative content analysis delineated by Elo and Kyngas (19), qualitative themes were identified.First, data relevant to lessons learned was collected from each of the included studies in the preparation stage (Phase I) (Supplementary material 1).
Second, lessons learned were organized into bullet points and tabulated by primary author to compare data across studies and explore emerging themes (Phase 2) (Supplementary material 1).Major themes were then highlighted in Table 3 (Phase III).
Many of the studies demonstrated a strong association between a lack of health insurance and a lower rate of breast cancer screening (21)(22)(23)(24)(25). Ethnic minority women, with the exception of those identifying as Asian, had a lower likelihood of being screened, and Black women experienced a higher risk of diagnosis upon first screening (25)(26)(27)(28)(29).While few studies analyze the effect of sexual orientation on breast cancer screening, initial insights reveal there are significant differences in mammography between bisexual, lesbian, and heterosexual women regardless of racial/ethnic groups (30).In considering religious values, fatalism-emphasizing religions were associated with less screening adherences and maintenance of modesty did not prove a significant limitation for women receiving mammograms (31)(32)(33).Economic factors present limitations as both high levels of poverty and impoverished rural regions were associated with lower screening rates (27,32,(34)(35)(36)(37).Improving patient-provider communication, addressing perceived discrimination, and improving trust in the health care system is necessary to improve screening rates across all demographics (38)(39)(40)(41)(42). Additionally, structural efforts to improve health insurance coverage, language proficiency, and transportation services could be beneficial (20-110).These steps will need to involve the local community to develop community-tailored educational campaigns to reinforce the importance of establishing yearly mammogram screenings (Table 3) (22, 34,46,49,54,55,70,76,80,86).

Discussion
The purpose of this scoping review was to identify the major SDOH acting as influential factors of breast cancer screening in United States women aged ≥ 40 years old.The analysis of the 72

Lessons learned themes
1 Lack of health insurance was strongly associated with lower breast cancer screening rates across various populations.
2 Functional health literacy was found to be significantly associated with mammography receipt; however, the relationship between health literacy and mammography can be influenced by factors such as ethnicity and language-preference acculturation.
3 Economic factors such as poverty level was a strong indicator of breast cancer screening rates.
4 Geographic factors including regional poverty are associated with increased late-stage breast cancer and lower breast cancer screening rates.
5 Rural areas were associated with less access to on-site breast cancer screening access and had lower overall breast cancer screening rates.
6 Women who identified themselves as nonwhite ethnicity, with the exception of Asians, had a higher likelihood of being unscreened.7 Asian women with less time spent in the U.S. and Korean populations had lower screening rates due to limited acculturation, lack of education surrounding breast cancer screening, and lack of insurance.
8 There is a need to address culturally specific barriers, such as distrust of physicians, which may increase Black women's confidence in breast cancer screenings and motivation to have preventive breast cancer care.9 Methods to enhance patient-provider communication may be important to increasing adherence to mammogram screening guidelines for those reporting less than ideal interactions with healthcare providers.
10 The COVID-19 pandemic was correlated with lower screening rates in women, possibly due to limited healthcare access for individuals.
11 Breast cancer screening and adherence rates differed depending on the religious values of certain populations, more specifically, fatalism-emphasizing religions led to less screening adherence.
12 Cultural efforts include developing culturally appropriate interventions and training health professionals in culturally competent communication skills, while structural efforts include removing barriers to access, improving health insurance coverage, language proficiency, and transportation services.

SDOH factors and healthy people 2030 categories
Of the classifications of SDOH by Healthy People 2030, the Social and Community Context Category was the most prevalent across the included studies (n = 177).However, when looking at the most frequently cited SDOH influential factors of breast cancer screening behaviors, those related to socioeconomic status exhibited the highest frequency.Such factors included income (n = 32), education level (n = 29), employment status (n = 8), birthplace/citizenship (n = 5), acculturation/years lived in the United States (n = 5), marital status (n = 2), social support (n = 2), and number of children (n = 1).Other highly reported factors include insurance status (n = 33) under the Healthcare Access and Quality category, as well as race/ethnicity (n = 79) and age (n = 52) under the Social and Community Context Category.
There is evidence to show the significance of the relationship between socioeconomic factors and breast cancer screening.Over 30 different interventions that address SDOH increased breast cancer screening rates by 12.3% (93).Social determinants such as poverty, lack of education, neighborhood disadvantage, residential segregation, racial discrimination, lack of social support, and social isolation have shown in numerous studies to play a role in the breast cancer stage at diagnosis (94, 95).Gomez et al. (94) highlighted in their review that social and built environments have been shown to factor into cancer diagnoses in 82% of 34 reviewed articles published since 2010, including breast cancer (96).Studies have found that, not only do these factors have a significant association with breast cancer screening individually, but they also work dynamically to impact screening and treatment for breast cancer (97).
Low affordability and healthcare accessibility profoundly impact breast cancer screening, leading to lower adherence in female patients.For instance, Medicaid patients who are required to pay co-payments for preventative services as well as for recommended follow-up visits are less likely to pursue such preventative services and mammograms are included in lost care (96).Co-payments of more than $10 have been associated with reduced rates of mammograms (97).Furthermore, a study investigating breast cancer screening among young military women revealed that, when removing cost and access barriers to obtaining a breast mammography, first-time screening rates were 90% (98).Similar results have been noted when patients were provided free mammograms in underserved areas.The Building Relationships and Initiatives Dedicated to Gaining Equality (BRIDGE) Healthcare Clinic, a free clinic offered by the University of South Florida, provided patients free mammograms and noted that about 84.5% of patients utilized these services (99).

Significance of associations between SDOH factors and breast cancer screening and treatment
The majority of the studies reported a significant association between the SDOH factors under each of the five Healthy People 2030 categories.Insurance status was the most reported sub-categorical factor of Health Care Access and Quality with n = 36 (50%) articles supporting this finding.Insurance status often determines whether patients seek mammography services as they often become costly without robust coverage (93).Despite stable mammography rates among women in the United States between the years 2000 and 2015, women who report being uninsured consistently have the lowest rates of mammography at 35.3% (100).
Moreover, a total of n = 42 (58%) studies showed statistical significance in the social and community context category, with the highest subcategories being age and ethnicity with n = 46 (63%) and n = 40 (55%) articles denoting their significance, respectively.Health disparities in the United States have been consistently associated with delayed screening, which then contributes to higher mortality rates among both Hispanic and Black populations (28).Inequities also exist in mammography rates between patients of different sexual orientations (111).White, bisexual women had significantly lower mammography rates than White, heterosexual women, while mammography rates were significantly higher for bisexual, Black women than for heterosexual, Black women (102).
Income (n = 20; 27%) strongly influences mammography rates since women with estimated household incomes greater than $38,100 have been found to have rates of repeat mammography higher than those of women below $25,399 (109).In addition to household income, food security acts as another influential factor of mammography rates.When patients are forced to choose between feeding their families and pursuing preventative care, mammography becomes more of a luxury than lifesaving care (110).Women facing food insecurity have shown a 54% lower likelihood of obtaining mammography (110).
Language (n = 11; 15%) and availability of translation services, health literacy, and culture also play a strong role in mammography rates since many women with limited English proficiency seek mammography care and receive abnormal results (103).Appropriate, timely follow-up in the correct language is imperative to proper care provision; however, a lack of translation services worsens the language barrier between these patients and their healthcare providers, delaying care (101).Clinics with a patient population that is majority non-English speaking also experience greater follow-up delays than those with a minority of non-English speakers due to language barriers (103).The lower a patient's health literacy, the less likely they are to undergo up-to-date breast cancer screening according to official guidelines (104,105).The cultural and religious beliefs in fatalism have also been continuously found to be associated with lower mammography rates, whereby women with the highest beliefs in fatalism had the lowest breast cancer screening rates (106,107).
Finally, Education Access and Quality sub-categories were significant indicators of mammography rate, with the highest level of education completed acting as the strongest sub-categorical factor in n = 24 (33%) articles.A systematic review by Damiani et al. (109) showed that United States women with the highest level of education were more likely to screen for breast cancer, with a 36% higher rate of adherence to national screening guidelines compared to women with lower levels of education.This finding holds health professionals and community outreach efforts accountable in ensuring that the local patient population is aware of the importance of and has access to breast cancer screening measures (109,110).

Availability of public databases
Of the 74 databases used, only 47% (n = 35) were publicly available.There is a need to establish more widely accessible databases encompassing a routine collection of data on the SDOH to allow for the examination of additional evidence on exiting associations between SDOH and health outcomes.These databases could also inform the development and implementation of longitudinal and experimental studies at the county, city, and national levels to decrease health disparities exacerbated by SDOH factors.

Strengths and limitations
Despite the importance of this study in guiding and informing the development and implementation of future SDOH-oriented evidencebased interventions for breast cancer screening, findings need to take into consideration this study's limitations.First, despite a comprehensive search of the literature in psychosocial databases compatible with the topic at hand, this review did not include gray literature and did not encompass tracing of reference lists in included studies.Second, it also was limited to observational studies to explore SDOH factors acting as factors based on statistical tests looking at significance of reported associations.These observational studies also widely varied in reported sample sizes, ranging from 100 participants to a population of 4 million.Therefore, although statistical significance was reported across different studies, effect sizes, power, and external validity varied greatly.Future systematic reviews should assess the rigor and quality of analysis carried out, evaluate recruitment efforts and data collection methods, and critique analytical tests carried out to account for the difference in sample sizes.Third, the mesh terms included as many technical words and keywords relevant to the SDOH as possible but might have inadvertently omitted some key words due to the continuously evolving and changing definitions related to SDOH.However, the help of an expert research librarian mitigated the impact of this concern by imposing rigor in implemented scoping review protocols when developing the search strategy for this review.Fourth, formal assessment of the methodology and quality of the evidence was beyond the scope of this study and relied on the reported statistical tests to assess significance.Follow-up systematic reviews would help with addressing this limitation by focusing specifically on the analytical proportion of each study.Fifth, although various categorizations exist for SDOH such as the WHO and CDC categories, the Healthy People 2030 taxonomy was adopted for use as it is the most recently updated classification encompassing a wide range of SDOH.Future studies should compare these taxonomies by feasibility, usability, and importance for a more valid and systematic approach to SDOH categorization.

Conclusion
This scoping review describes major SDOH acting as significant influential factors of breast cancer screening behaviors among United States women aged ≥40 years old who are at-risk of the disease.Results may inform future evidence-based interventions aiming to address the underlying factors contributing to low screening rates for breast cancer in the United States.Efforts to integrate SDOH within the different components of intervention planning, implementation, and sustainability are widely gaining recognition, particularly in underserved communities, due to their substantial influence on everyday behaviors.
13 Community-tailored educational campaigns to reinforce the importance of establishing yearly mammogram screening behaviors can be powerful and effective tools for increasing adherence across various populations.Frontiers in Public Health 31 frontiersin.orgincluded studies can inform which SDOH categories to focus on when designing evidence-based interventions for more effective and sustained positive behavior and health outcomes among United States women at-risk of breast cancer.

FIGURE 1 PRISMA-
FIGURE 1 PRISMA-ScR flow chart of study selection process.

Table 1 )
describing study characteristics, types of SDOH, and outcomes.Types of SDOH were first listed and then categorized based on Healthy People 2030 into five categories: Economic Stability, Education Access and Quality, Health Care Access and Quality, Neighborhood and Built Environment, and Social and Community Context

TABLE 1
Study characteristics.

TABLE 1 (
Continued) *Statistical significance was assessed based on the p value (p < 0.05).

TABLE 2
Database availability status and characteristics.
Among the Healthy People 2030 categories, Social and Community Context (n = 177) emerged as the most prevalent, with a striking 177 occurrences of SDOH.Following closely behind were Healthcare Access and Quality (n = 80), Economic Stability (n = 56), Neighborhood and Built Environment (n = 46), and Education Access and Quality (n = 36) (Table ), showing its statistical significance in relation to access to mammography.Insurance status For one control variable, county-level PCP data were obtained across the state from a different database: Area Health Resources Files. *

TABLE 3
Lessons learned identified from thematic analysis across included studies.