Disparities in screening mammography utilization among Hispanic women in rural Texas from 2002 to 2018

Abstract

Purpose

To examine the trends of racial/ethnic and urban–rural disparities in screening mammography use with a focus on Hispanic women in rural Texas, as well as to further investigate barriers to mammography screening practices.

Methods

A serial cross-sectional study was conducted on screening mammography including eligible female respondents (≥ 40 years) from the Texas Behavioral Risk Factor Surveillance System survey from 2002 to 2018.

Findings

Weighted descriptive analyses showed persistent racial/ethnic and urban–rural disparities in mammography screening rates among eligible women (≥ 40 years) in Texas. Overall, the mammography screening rates for women in rural areas were significantly lower than women in urban areas with a mean rate of 64.09% versus 70.89% (p < 0.001). Rural Hispanic women had the lowest mean mammography screening rate (55.98%) among all eligible women which is 16.27% below the mean mammography screening rate of non-Hispanic white women in urban areas. Weighted logistic regression model revealed that women with no health insurance or primary care providers were 52% (95% Confidence Interval [CI] 0.36–0.63, p < 0.001) or 54% (95% CI 0.35–0.6, p < 0.001) less likely having an up-to-date mammography screening compared with women with health insurance or primary care providers, respectively.

Conclusions

Our study demonstrated significant and persistent racial and urban–rural disparities in screening mammography utilization among Hispanic women compared with non-Hispanic white women from 2002 to 2018. Healthcare access is a major contributor to these disparities. It highlights the need for wide-scale interventions from public health and policymakers targeting under screened racial minorities and rural regions population to promote screening mammography services among disadvantaged population.

Introduction

Breast cancer is the most commonly diagnosed cancer and the leading cause of cancer death among Hispanic women in the United States (US) [1, 2]. Despite a lower lifetime risk compared to non-Hispanic white women (10% vs 12.5%), the rate of breast cancer incidence among Hispanic women is increasing at a faster rate [1, 3]. Studies have also shown that the annual breast cancer death rate declined slower in Hispanic women than non-Hispanic white women from 2010 to 2019 (0.8% vs 1.3%) [4]. These factors may contribute to the overall breast cancer racial disparities between Hispanic women and non-Hispanic white women [1, 5]. The Hispanic population is the fastest growing population in the US and is projected to increase to 111.22 million by 2060 [6]. As the population of Hispanics increases, breast cancer incidence and mortality are anticipated to increase as well. Texas is the second largest state in the nation with a large Hispanic population (40.2%), of which about 27% (3,204,962, US Census Bureau, 2021) live in rural areas. Given the evidence that detecting and treating breast cancer at a more favorable stage is associated with reduced morbidity and mortality [1], a better understanding of disparities and barriers in screening mammography utilization among Hispanic women is needed which can help to develop tailored interventions to reduce and ultimately eliminate those healthcare disparities.

Mammogram (low-dose X-rays of the breast) is a well-established evidence-based screening method for breast cancer early detection. For women at average risk, the US Preventive Services Task Force (USPSTF) recommends biennial mammography screening in women aged 50–74 years, every 1–2 years in women aged 40–49 based on shared decision making, and optional for women older than 74 years [4]. The mammography screening goal of Healthy People 2030 is to increase the proportion of women who get screened for breast cancer to 77.1% [7]. The prevalence of breast mammography screening has remained unchanged over a decade which was about 5% below the desired target [8–10]. However, breast cancer screening rates vary substantially between racial/ethnic as well as urban and rural groups within the US [10–12]. The 2022 Cancer Disparity Report by the American Association for Cancer Research showed persistent disparities in cancer screening among racial and ethnic minorities and other medically underserved populations [13]. Furthermore, the disparities in cancer screening have not been equally distributed across racial groups, between urban and rural regions, and across the states of the US [14–17]. To date, there have been limited studies focusing on the impact of interaction of race and geographic areas on breast cancer screening disparities. The trends and disparities in mammography use among Hispanic women in rural Texas remain unclear.

Lower socioeconomic status and structural barriers have been shown to contribute to the lower rates of breast cancer screening [10, 13, 18, 19]. National data suggests that Hispanic women are less likely to have had an up-to-date mammogram when compared with non-Hispanic white women [1, 2]. Hispanic women and those with less than a high school education have the lowest rates of breast cancer screening in the US [2]. Differences in the use of recommended mammography screening may contribute to the worse outcomes observed in Hispanic women who tend to be diagnosed with more advanced breast cancers than white women [20, 21] Additionally, rural Critical Access Hospital (CAH) closures have been shown to affect rural residents’ access to healthcare [22]. Mammography is one of the most commonly available imaging services at CAHs [23]. Texas has the most CAH closures since 2005 among all states in the US which may exacerbate the existing urban and rural disparities in breast cancer screening [24]. Understanding mammography screening barriers among Hispanic women in rural Texas is critical prior to implementing any tailored interventions.

The purpose of this study was to examine the trends and existing racial/ethnic and urban–rural disparities in screening mammography use with a focus on Hispanic women in rural Texas, and to further investigate the barriers of mammography screening among women in Texas using the Texas Behavior Risk Factor Surveillance System (BRFSS) data from 2002 to 2018. We hypothesized that Hispanic women in rural Texas would have lower up-to-date mammography screening rates compared to their counterparts in urban areas, as well as non-Hispanic white women in both urban and rural areas. The aim was to provide the most up-to-date evidence in order to help guide public health officials and policymakers to address any disparities in breast cancer screening among minority and disadvantaged population.

Methods

Sample

We used data from the Texas Behavioral Risk Factor Surveillance System (BRFSS) from 2002 to 2018 [25]. Initiated in 1987, the Texas BRFSS is federally supported telephone survey conducted annually to collect information on demographics, health-related risk behaviors, chronic health conditions, and use of preventive services of randomly selected Texas adult residents aged 18 and older. Mammography screening behavior data were collected every other year on even years only. Texas BRFSS Data were weighted to adjust for the probabilities of selection and a post-stratification weighting factor that adjusted estimates to the geographic stratification population by sex and age [25]. The inclusion criteria for this study were female and aged 40 years or older. We excluded people with missing value on mammography screening status, race, urban–rural status, resulting in a final sample size of 39,955 (42,898,194 weighted population).

To assess barriers and facilitators associated with screening mammography use in Texas, we combined data from the years of 2014, 2016, and 2018 for regression analyses. Due to methodologic changes in the BRFSS in 2011 by adding cellular telephone survey, data prior to 2011 were unable to be included in the regression analysis [26]. The year 2012 data were not included because a calculated variable of Metropolitan Statistical Areas was missing. We also excluded respondents with missing values on covariates. The final sample size for regression analysis was 11,454 (3,780,132 weighted population). The institutional review board of the University of Texas at Arlington reviewed the study and exempted it from a full review because the study utilized publicly available data.

Measures

The dependent variable was whether a respondent had an up-to-date screening mammography (yes and no). The respondents were asked the following two questions to assess their mammography screening status: “Have you ever had a mammogram?” “How long has it been since you had your last mammogram?” The respondents were grouped into two categories: “yes” if the respondents had screening mammogram within the last two years, and “no” if the respondents never had a mammogram or had a screening mammogram but more than two years ago. The primary independent variables were race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, and other), geographic areas (urban and rural), and their interactions. According to the BRFSS methodology reports, individuals were classified as Hispanic if they reported their Hispanic or Latino origin [25]. We defined urban and rural areas in our study using the metropolitan statistical areas (MSA), which is a variable utilized for defining rurality in the BRFSS’s Metropolitan Status Codes (MSCODE) [27, 28]. The U.S. Office of Management and Budget (OMB) defines MSA as areas having at least one urbanized area with a minimum population of 50,000, while all counties not included in an MSA are classified as rural [29]. Covariates included education (Less than high school graduate, high school graduate, some college, and college graduate and above), income (< $25,000, $25,000–$75,000, and ≥ $75,000), insurance status (yes or no), whether having a primary provider (yes or no), marital status: married (married or living as married), and unmarried (single, widowed, or divorced), body mass index (BMI): “normal” if BMI < 25, “overweight” if BMI between 25 and 30, and “obese” if BMI ≥ 30, and age groups (40–49, 50–74, and ≥ 75).

Statistical analysis

Weighted descriptive analyses were used to summarize the screening mammography uptake rates for all respondents, then further stratified by age groups, race/ethnicity, urban–rural residential areas, and the interaction of race/ethnicity and geographic areas. Weighted binary logistic regression was used to assess the association of race/ ethnicity, urban–rural residency, and covariates on mammography screening behavior. Adjusted odds ratios (OR) with 95% confidence interval (CI) were reported. Statistical significance was set at p < 0.05, and all regression tests were two-tailed. Above analyses were conducted using Stata 17.0 (https://www.stata.com).

Results

Our study consisted of 39,955 screening mammography eligible (≥ 40) female respondents in Texas from 2002 to 2018. Overall, the average mammography screening rates were 69.89% (Relative Standard Deviation [RSD] 2.25%) for all eligible women (≥ 40), 61.13% (RSD 4.07%) for women aged 40–49, 75.28% (RSD 2.03%) for women aged 50–74, and 67.52% (RSD 5.87%) for women greater than 74 years old (Fig. 1a and Supplemental Table S1). Hispanic women had significantly lower average mammography screening rates than non-Hispanic white women (65.31%, RSD 3.22% vs 71.45%, RSD 2.47%, p = 0.0015) (Fig. 1b and Supplemental Table S1). About 64.09% (RSD 2.61%) eligible women in rural areas had an up-to-date screening mammography which was significantly less than 70.89% (RSD 2.39%) eligible women in urban areas (p < 0.001) (Fig. 1c and Supplemental Table S1). Rural Hispanic women had persistent lower mammography screening rates from 2002 to 2018 compared with urban Hispanic women and non-Hispanic white women in both urban and rural areas (Fig. 2). Hispanic women in rural areas had a mean of 55.98% (RSD 5.68%) mammography screening rate compared with Hispanic women in urban (66.74%, RSD 4.26%, p = 0.002), non-Hispanic white women in rural (67.62, RSD 2.48%, p = 0.0011) and urban (72.25%, RSD 2.63%, p = 0.0001) areas, respectively (Fig. 2 and Supplemental Table S2).

Fig. 1.

Weighted trends of mammography screening rates for eligible women in Texas from 2002 to 2018: a for all eligible women and different age groups (40–49, 50–74, and > 74), b for non-Hispanic white, nonHispanic black, and Hispanic women, and c for urban–rural residents

Fig. 2.

Weighted trends of mammography screening rates among eligible non-Hispanic white and Hispanic women in Texas 2002–2018 by the interaction of race/ethnicity and urban–rural areas

In the combined years of 2014, 2016, and 2018 Texas BRFSS survey, there were a total of 3,780,132 weighted population. The respondents were predominantly non-Hispanic white women (55.66%), were between ages 50–74 (60.56%), had an income between $25,000 and $75,000 (37.3%), had some college education (31.2%), were married (57.53%), had a BMI greater than 30 (38.1%), and were unemployed (54.42%) (Table 1). The majority lived in urban areas (85.07%), had health insurance coverage (85.23%) and a personal health care provider (85.5%). The urban and rural population was significantly different in race/ethnicity population distribution (p < 0.001). Overall, female population (≥ 40 years) in rural areas were poorer (p < 0.001) and less educated (p < 0.001) than their counterparts in urban areas.

Table 1.

Weighted characteristics of screening mammography for eligible women (≥ 40 years old), Texas Behavioral Risk Factor Surveillance System 2014–2018

Variables	Urban	Rural	Total	p
	(3,215,758; 85.07%)	(564,374; 14.93%)	(3,780,132)
	%[95% CI]	%[95% CI]	%[95% CI]
Race/ethnicity				< 0.001
Non-Hispanic White	53.68 [51.57,55.77]	66.94 [62.25,71.32]	55.66 [53.72,57.58]
Non-Hispanic Black	14.01 [12.41,15.78]	7.22 [5.119,10.09]	13 [11.58,14.56]
Hispanic	28.14 [26.22,30.14]	23.7 [19.74,28.17]	27.47 [25.73,29.28]
Other	4.174 [3.366,5.165]	2.139 [1.353,3.365]	3.87 [3.166,4.723]
Insurance				0.0643
Yes	85.82 [84.21,87.29]	81.89 [77.33,85.71]	85.23 [83.72,86.63]
No	14.18 [12.71,15.79]	18.11 [14.29,22.67]	14.77 [13.37,16.28]
PCP				0.4622
Yes	85.29 [83.6,86.83]	86.68 [83.11,89.59]	85.5 [83.98,86.89]
No	14.71 [13.17,16.4]	13.32 [10.41,16.89]	14.5 [13.11,16.02]
Education				< 0.001
Less than High School	16.58 [14.93,18.39]	19.93 [16.14,24.36]	17.08 [15.55,18.74]
High School Graduate	24.26 [22.52,26.09]	35.8 [31.13,40.75]	25.98 [24.32,27.72]
Some College	31.45 [29.48,33.49]	29.81 [25.77,34.18]	31.2 [29.42,33.05]
College Graduate	27.71 [26.05,29.42]	14.46 [12.19,17.08]	25.73 [24.27,27.24]
Income				< 0.001
< $25,000	32.11 [30.17,34.11]	39.71 [35.16,44.44]	33.25 [31.46,35.09]
$25,000 to < $75,000	36.48 [34.48,38.54]	41.93 [37.27,46.75]	37.3 [35.45,39.18]
$75,000 +	31.4 [29.54,33.33]	18.36 [14.95,22.33]	29.46 [27.78,31.19]
Marital status				0.4842
Married	57.26 [55.19,59.31]	59.07 [54.4,63.58]	57.53 [55.64,59.4]
Unmarried	42.74 [40.69,44.81]	40.93 [36.42,45.6]	42.47 [40.6,44.36]
BMI				0.0619
< 25	31.27 [29.41,33.19]	25.55 [21.84,29.66]	30.41 [28.73,32.16]
25–30	31.07 [29.21,33]	33.84 [29.42,38.56]	31.49 [29.76,33.26]
> 30	37.66 [35.59,39.77]	40.6 [35.92,45.47]	38.1 [36.2,40.03]
Age				0.4842
40–49	28.3 [26.37,30.3]	25.21 [20.9,30.05]	27.84 [26.07,29.67]
50–74	60.42 [58.34,62.45]	61.4 [56.53,66.06]	60.56 [58.66,62.43]
≥ 75	11.29 [10.27,12.4]	13.39 [10.77,16.53]	11.6 [10.64,12.64]
Employment status				0.2653
Yes	46.03 [43.96,48.11]	43.03 [38.31,47.89]	45.58 [43.68,47.49]
No	53.97 [51.89,56.04]	56.97 [52.11,61.69]	54.42 [52.51,56.32]

Abbreviations: CI confidence interval; PCP, primary care provider; BMI, body mass index

Significance: p <0.05

Hispanic women in rural areas were 55% (95% confidence interval [CI] 0.29–0.7, p < 0.001) less likely to have an upto-date screening mammography compared to non-Hispanic white women in urban areas prior to adjusting for covariates (Table 2). Although they were still 14% (95% CI 0.54–1.36) less likely to have an up-to-date screening mammography compared with non-Hispanic white women in urban areas after adjusting for covariates, the difference was no longer statistically significant (p = 0.512) (Table 3). Hispanic and non-Hispanic black women were 1.8 (95% CI 1.32–2.46, p < 0.001), and 1.82 (95% CI 1.28–2.59, p = 0.001) times more likely to have an up-to-date screening mammography compared with non-Hispanic white women after adjusting for covariates, respectively, if they all lived in urban areas. Of the potentially preventable covariates, mammogram eligible women without health insurance or primary care provider were 52% (95% CI 0.36–0.63, p < 0.001) and 54% (95% CI 0.35–0.6, p < 0.001) less likely to have an up-to-date screening mammogram. Of the sociodemographic covariates, women with annual income between $25,000 and $75,000 or more than $75,000 were 1.35 (95% CI 1.06–1.71, p = 0.013) and 2.03 (95% CI 1.48–2.8, p < 0.001) times more likely to have an up-to-date screening mammography than women with annual income less than $25,000. Women aged 50–74 years were 2.03 (95% CI 1.6–2.58, p < 0.001) times more likely to have an up-to-date mammogram than women aged 40–49 years.

Table 2.

Weighted logistic regression model assessing mammography screening behavior from the interaction of race/ethnicity and urban-rural residency without adjusting for covariates

Variables	OR	95% CI		p
Urban
Non-Hispanic White	Ref.
Non-Hispanic Black	1.21	0.87	1.68	0.26
Hispanic	0.93	0.75	1.17	0.555
Other	0.87	0.55	1.36	0.532
Rural
Non-Hispanic White	0.85	0.65	1.1	0.218
Non-Hispanic Black	0.97	0.47	1.97	0.928
Hispanic	0.45	0.29	0.7	< 0.001
Other	1.12	0.45	2.8	0.81

Abbreviations: OR, odds ratio; CI confidence interval; Ref, reference

Table 3.

Weighted logistic regression model assessing mammography screening behavior from the interaction or race/ethnicity and urban-rural residency after adjusting for covariates

Variables	OR	95% CI		p
Urban
Non-Hispanic White	Ref.
Non-Hispanic Black	1.82	1.28	2.59	0.001
Hispanic	1.8	1.32	2.46	< 0.001
Other	0.91	0.57	1.47	0.502
Rural
Non-Hispanic White	0.98	0.75	1.28	0.892
Non-Hispanic Black	1.7	0.84	3.46	0.143
Hispanic	0.86	0.54	1.36	0.512
Other	1.46	0.58	3.66	0.425
Insurance
Yes	Ref.
No	0.48	0.36	0.63	< 0.001
PCP
Yes	Ref.
No	0.46	0.35	0.6	< 0.001
Education
< High School	Ref.
High School Graduate	1	0.73	1.37	0.995
Some College	0.93	0.66	1.29	0.647
College Graduate	1.15	0.8	1.65	0.455
Income
< $25,000	Ref.
$25,000 to < $75,000	1.35	1.06	1.71	0.013
$75,000 +	2.03	1.48	2.8	< 0.001
Marital status
Married	Ref.
Unmarried	0.86	0.71	1.06	0.162
BMI
< 25	Ref.
25–30	1.1	0.89	1.36	0.392
> 30	0.97	0.77	1.22	0.774
Age
40–49	Ref.
50–74	2.03	1.6	2.58	< 0.001
≥ 75	1.16	0.83	1.64	0.386
Employment status
Yes	Ref.
No	1.09	0.88	1.34	0.435

Abbreviations: OR, odds ratio; CI, confidence interval; Ref, reference

Discussion

This study revealed persistent racial/ethnic and urban–rural disparities in screening mammography utilization among eligible women in Texas from 2002 to 2018. Women (≥ 40) in rural areas were about 7% less likely to have an up-todate screening mammography than their counterparts in urban areas. Rural Hispanic women had lower odds of having an up-to-date screening mammography compared to their counterparts in urban areas and non-Hispanic white women in both urban and rural areas prior to adjusting the covariates. These findings are in line with our hypotheses and previous research on racial and geographic disparities in breast cancer screening practices in the US [17, 30–32]. A systematic review and meta-analysis showed that Hispanic populations had lower odds (0.83, 95% CI 0.74–0.93) of utilizing screening mammography when compared with the white population [14]. Using the national BRFSS data from 2012 to 2016, Tran et al. found that urban women had small but significantly higher adjusted probabilities (81.1%) compared to rural women (80.2%) regarding having an up-to-date mammogram [17]. Berkowitz et al. found that there was about 2% difference between urban and rural mammogram eligible women (≥ 40) who had a mammogram within 2 years [33]. Among Medicare beneficiaries, the national mammographic screening uptake is generally lowest in rural counties [31]. However, these reported urban–rural differences in screening mammography rates were smaller than the gaps we found between urban (70.89%, RSD 2.39%) and rural (64.09%, RSD 2.61%) women in Texas. Contrary to our results, Thomson et al. found no difference in biennial mammography screening between urban and rural residents, as well as by race in Virginia [21]. Mottram et al.’s systematic and meta-analysis also indicated no diffidence in mammography screening by rural versus urban residence [19]. These differences in screening mammography practices are likely due to states variations in the US which had been reported by previous studies [17, 33, 34].

In the present study, we found that Hispanic women in rural Texas had the significantly lowest mean mammography screening rate among all eligible women. After adjusting for covariates, these differences in screening mammography use among rural Hispanic women were no longer significant different from urban non-Hispanic white women. The logistic regression analyses revealed that healthcare access including insurance and primary care access are the most significant and potentially preventable barriers of screening mammography utilization for rural Hispanic women. These findings are consistent with prior reports that healthcare access can contribute to the lower rates of screening mammography utilization in minority population [35–38]. Access to breast cancer screening services is impaired by shortages of primary care and specialist providers, rural hospital closures and geographic distance from medical facilities for rural residents [22, 39]. Additionally, although rural patterns of primary care deliveries are comparable to urban, beneficiaries cared for in more rural settings received fewer recommended mammograms [40]. Among sociodemographic factors, we found that income had a significant impact on screening mammography practice among women in Texas. These findings support the evidence that financial barriers are significant predictors of mammography uptake in populations of low-income women including Hispanics [35, 38, 39, 41]. Rural residents reported significantly lower household income and greater use of Medicaid compared to urban residents [21]. Texas is one of the states that has not adopted the Medicaid expansion plan, which may contribute to the widening racial and geographic disparities in preventive breast cancer screening uptakes in Texas. Interestingly, we noted a relative increase in mammography utilization among non-Hispanic black women compared with Hispanic and non-Hispanic white women since 2010 (Fig. 1b). Although selection bias could potentially account for this discrepancy, recent public health interventions (e.g., the Affordable Care Act) could be resulting in higher proportions of non-Hispanic black women to receive timely breast cancer screening [42]. Several recent studies using national level data also showed similar trends in breast cancer screening among non-Hispanic black women in the US [17, 43].

The eligible Hispanic women in rural Texas who had an up-to-date screening mammography is about 21% below the breast cancer screening goal of Healthy People 2030. These findings have practical implications for healthcare practice, public health, and policymakers in promoting mammography screening among minority and/or geographically disadvantaged populations. It is especially important given the well-established evidence that screening mammography utilization has resulted in significantly lower breast cancer-caused mortality rates. Healthcare professionals and public health workers must educate eligible women on the importance of screening and early detection of breast cancer. Policymakers should ensure that all eligible women in the US have access to this preventive screening tool.

Limitations of this study should be considered. First, BRFSS data are self-reported which could result in recall bias and misclassification in mammography screening timeframe. Second, the focus on women in Texas limits the generalizability of this study. Third, methodology changes in BRFSS data collection in 2011 make it a challenge to study the trends of screening mammography utilization. However, we assumed that it is affecting all racial groups and both urban and rural groups in a comparable way. In the present study, we compared the relative mammography screening behavior within the same racial and/or geographic groups. Tailored interventional studies to reduce the racial and geographic disparities in mammography screening are needed in the future. Furthermore, considering the diversity of Hispanic groups based on acculturation, further research is needed to examine mammography utilization within subgroups based on nativity and other factors (e.g., immigration status) among Hispanics [44].

Conclusions

This study provides a more up-to-date assessment of the trends, degrees, and barriers of racial and geographic disparities in screening mammography utilization among all eligible women with a focus on Hispanic women in rural Texas. Our findings suggest that healthcare access and low-income status are major barriers to the persistent racial and geographic disparities in screening mammography utilization. Given screening mammography is still the most effective strategy to reduce breast cancer-caused mortality among women in the US, it highlights the need for wide-scale interventions from public health and policymakers targeting under screened racial minorities and rural regions population to promote screening mammography services among disadvantaged populations.

Data availability

Data are publicly available upon request from Texas Health and Human Services. Here is the link for data request: https://www.dshs.texas.gov/texas-behavioral-risk-factor-surveillance-system-brfss.

Code availability

Code is available upon request from the authors.