Sage Journals: Discover world-class research

Abstract

Objective

The gold standard for breast cancer screening and prevention is regular mammography; thus, understanding what impacts adherence to this standard is essential in limiting cancer-associated costs. We assessed the impact of various understudied sociodemographic factors of interest on adherence to the receipt of regular mammograms.

Methods

A total $N_{c}$ = 14,553 mammography-related claims from $N_{w}$ = 6,336 female Kansas aged between 45 and 54 were utilized from insurance claim databases furnished by multiple providers. Adherence to regular mammography was quantified continuously via a compliance ratio, used to capture the number of eligible years in which at least one mammogram was received, as well as categorically. The relationship between race, ethnicity, rurality, insurance (public/private), screening facility type, and distance to nearest screening facility with both continuous and categorically defined compliance were individually assessed via Kruskal–Wallis one-way ANOVAs, chi-squared tests, multiple linear regression models, and multiple logistic regression, as appropriate. Findings from these individual models were used to inform the construction of a basic, multifaceted prediction model.

Results

Model results demonstrated that all factors race and ethnicity had at least some bearing on compliance with screening guidelines among mid-life female Kansans. The strongest signal was observed in the rurality variable, which demonstrated a significant relationship with compliance regardless of how it was defined.

Conclusion

Understudied factors that are associated with regular mammography adherence, such as rurality and distance to nearest facility, may serve as important considerations when developing intervention strategies for ensuring that female patients stick to prescribed screening regimens.

Keywords

breast cancer urban midwest sociodemographic

Introduction

Breast cancer, as its name suggests, is a cancer that forms in breast cells, most commonly in the ducts, the tubes that transport milk from the lobules to the nipple. While a small proportion of those afflicted are male (<1%), the overwhelming majority of patients are female. To this point, in a typical year in the United States alone, a total of 2400 new cases are diagnosed in men as compared to over a quarter-million in women.^1-3 Thus, while both sexes are affected, the disease is largely a women’s health issue. Indeed, after skin cancer, breast cancer is the most diagnosed cancer in women in the United States, being directly responsible for approximately 42,000 deaths each year.^3,4 While the overall 5-year survival rate is quite high in developed countries such as the United States ( $\approx$ 90%), the relatively high incidence of the disease costs the U.S. economy tens of millions of dollars each year in lost productivity and inflicts an intense emotional burden on the women themselves and their loved ones.^5,6 Therefore, it is imperative that as many breast cancer cases as possible be prevented or caught early to minimize their economic, emotional, and medical impacts.

The current gold standard for early breast cancer detection is the mammogram, which has an estimated sensitivity of 85% and an estimated specificity of 90%.⁷ While current screening guidelines differ by institution, the general consensus is that mid-life women (aged 45 to 74) with average risk should be screened once every 1–2 years.^8,9 The U.S. Center for Disease Control and Prevention and the US Preventive Services Task Force (USPSTF), in particular, recommends that women aged 50 to 74 who are at least average risk for breast cancer receive a mammogram every 2 years. They further recommend that women aged 40–49 speak with their doctor or healthcare provider to determine if starting earlier is beneficial.^10,11 The USPSTF lists adherence to this screening regimen as having a high certainty for moderate to substantial health benefits.

Biennial screening in particular has been demonstrated to avert seven breast cancer deaths for every 1000 women screened, increase overall life expectancy by approximately 1.4 months, and decrease the number of women who are initially diagnosed with a late-stage cancer.¹⁰ Further, it is known that late-stage breast cancer is generally more expensive to treat than early-stage breast cancer; thus, it stands to reason that mammography is an effective tool in not only improving the health prospects of those screened but also in saving potentially large amounts of money.¹² Given the considerable evidence regarding the benefits of routine mammography, it is paramount that mid-life women adhere to a regular screening regimen.

Recent research concerning mammography screening adherence has largely been focused on racial or ethnic disparities, but these variables do not fully capture the issue.^13,14 Rural, densely settled rural, and frontier counties make up 85% of the state of Kansas.¹⁵ Consequently, as a state, Kansas has a sizable rural population, with 25.8% of its total population coming from these areas.¹⁶ This makes Kansas screening data well equipped to investigate the effects of understudied factors, like rurality, on screening adherence. While rurality is clearly of great importance in Kansas, the Sunflower state is not alone in this regard. Indeed, in ten states, it has been shown that more than 26% of women had limited accessibility to get mammography’s in 2022, in part due to lack of facilities, distance required to get to facilities, and time required to get facilities.¹⁷ Therefore, the purpose of the present study is to determine how several sociodemographic factors of interest including rurality impact adherence in the state of Kansas, with the hope that sub-populations at risk of engaging in less frequent screening may be intervened upon (via additional reminders, for instance) and the aforementioned benefits of routine screening may be enjoyed by a greater share of women.

Methods

Data for this project were collected by Centrus Health, an Accountable Care Organization (ACO) and participant in the Medicare Shared Savings Program (MSSP).¹⁸ The earliest mammography claims in this dataset were filed in 2014 and originate from Aetna, BlueCross BlueShield, the Centers for Medicaid and Medicare Services (CMS), and the Great Plains Health Alliance (GPHA). To keep our analytic sample in line with screening guidelines while making use of as much data as possible, women younger than 45 and older than 54 were excluded from the study. The upper bound of 54 was chosen due to the fact that women 55 and older could potentially have uncaptured mammograms from 2013 and earlier. Additionally, to protect privacy, randomly generated identifiers were used in lieu of medical record numbers to group claims.

The sociodemographic factors of interest for this study included race (White, Black, AAPI, Native American, and other), ethnicity (Hispanic vs non-Hispanic), rurality (rural vs urban), health insurance type (public vs private), mammogram screening facility type (critical access hospital, mobile unit, clinic office, outpatient hospital, and other), and distance from the nearest mammography screening facility. Race is a categorical variable in which only one value can be assumed; thus, multi-racial assessments were not possible in the present study. Rurality was defined using the Rural-Urban Commuting Area Codes (RUCA),¹⁹ which we mapped to the zip codes present in our claims dataset. RUCA codes of 1–3 were classified as “urban,” and codes of 4–10 were classified as “rural.” Aetna, BlueCross BlueShield, and the GPHA were defined as “private” health insurances, and the CMS was defined as “public” health insurance. A complete listing of mammogram screening facilities in the state of Kansas was obtained via a publicly available U.S. Food and Drug Administration (FDA) database;²⁰ distance was then computed using the zip_distance function from the zipcodeR package.²¹ All other variables were used as parameterized in the original dataset.

To quantify a given woman’s adherence to the prescribed regimen of routine mammograms, we created a compliance ratio (CR), defined as follows

C R_{i} = \frac{Σ_{j = 1}^{9} S_{i j}}{A_{i} - 45 + 1}, i \in {1,2, \dots, N}

(1)

where

A_{i}

is the

i

^th participant’s age, and

S_{i j}

represents whether a screening was (

= 1

) or was not (

= 0

) received in year

j

. Note that

j = 1

corresponds to the year 2014, and

j = 9

corresponds to the year 2022. Thus,

C R_{i} \in (0,1]

, where a value of 0 indicates that no mammograms were received from 2014 to 2022, and a value of 1 indicates that at least one mammogram was received each year beginning at age 45. Therefore, this compliance ratio allows us to directly assess the strength of adherence to a given screening recommendation. Note that since only women with at least a single claim were present in the dataset, we did not observe any women with a

C R_{i}

of 0.

We then created four binary compliance variables, each indicating whether a given woman possessed a

C R_{i}

of at least X, where

X \in {

1/4, 1/2, 3/4, 1 (full)

}

. If we assume the compliance ratio for the i^th woman is

C R_{i}

such that

C R_{i} \geq X

, where

X = g

, then the i^th woman belongs to compliance group g and each group below it. For example, if

C R_{i} = 0.32

, then the i^th woman belongs to the compliance group denoted by

g = 1 / 4

. Similarly, if

C R_{i} = 0.64

, then the i^th woman belongs to the compliance groups denoted by g = 1/4 and g = 1/2. Thus, a single individual can belong to multiple compliance groups. Table 1 summarizes the compliance groups and the status of the compliance (met or not).

Table 1.

Summary of Each Compliance Group.

Compliance Ratio Category	Total Meeting Threshold
Compliance Ratio Category	0 (Not Met)	1 (Met)
¼	1494	4842
½	3170	3166
¾	4506	1830
Full	4768	1568

The nature of the compliance ratio (CR) allowed us to assess “compliance” in a multifaceted manner—both in terms of a raw, continuous “score” and in terms of a “threshold” to be reached. For categorical predictors, Kruskal–Wallis one-way ANOVAs were used for the continuous parameterization of compliance, and chi-squared tests were used for the categorical parameterization. In this context, both tests assess whether compliance differs across all levels of a given predictor. For the lone continuous predictor (distance), multiple linear regression models and multiple logistic regression models were used, respectively. Given the expected high correlation between rurality and distance from nearest facility, these regression models additionally controlled for rurality (urban/rural). A parameter $β$ significantly different from 0 would serve as evidence of a linear relationship between distance and compliance.

To get a sense of the likelihood that a future woman would comply with mammography screening recommendations given her sociodemographic profile, we also fit a series of multiple logistic regression models (one for each categorical compliance threshold) and a multiple linear regression model (where the outcome is the continuous compliance ratio), each of which contained the significant factors under study. Together, the parameter estimates can be used to generate a rough guess for a given profile.

All analyses were performed in R v. 4.1.0. To account for the testing of multiple hypotheses, the Benjamini-Hochberg procedure for multiple hypothesis testing²² was used to control the type I error to $α$ = .05, while giving us an individual-model significance threshold of $α^{*}$ = .017.

Results and Discussion

After data cleaning, a total of $N_{c}$ = 14,553 claims from $N_{w}$ = 6336 unique women remained, meaning that each woman had an average of 2.30 (SD = 1.32) mammography claims during the observation period (2014–2022). The vast majority of women in the analytic sample were White (80%), non-Hispanic (96%), resided in urban areas (91%), and had private insurance (96%). The average age was around 50 years old. A plurality of women (39%) utilized outpatient hospitals for their mammograms. A complete breakdown of all demographic characteristics surveyed can be found in Appendix A.

Table 2 presents the odds ratios and corresponding P-value for each compliance group when the distance is the only predictor variable in the logistic regression model. When assessed as a categorical measure, compliance differed by rurality (all compliance proportions), facility type (1/2 and 1/4 compliance proportions only), insurance type (1/4 compliance proportion only), and distance (3/4 and 1/2 compliance proportions only). Generally speaking, for the significance compliance proportions, a greater proportion of urban women adhered to regular mammography screening than rural women, a greater proportion of women on public insurance adhered to regular mammography screening than women on private insurance, and a smaller proportion of women utilizing a critical access hospital adhered to regular mammography screening than women utilizing a different facility type. We also observed that each additional mile of distance from the nearest mammography facility was associated with a decrease of about 2% in the odds of adherence for 1/2 and 3/4 compliance; however, these effects were not significant after application of the Benjamini-Hochberg procedure. Racial and ethnic categories did not differ in their rates of adherence across any of the four compliance proportions assessed. A full listing of the exact values used to generate Figure 1 can be found in Appendix B.

Table 2.

Compliance (Categorical) by Distance.

Compliance Group	Estimated per-Mile OR^a	P-Value
¼	.983	.155
½	.979	.045
¾	.977	.047
Full (1)	.982	.141

^aOR = odds ratio.

Figure 1.

Categorical compliance by sociodemographic factor.

Table 3 presents the mean compliance ratios for each level of the categorical predictor variables and the distance variable. The levels of the categorical predictor variables (i.e., race, ethnicity, rurality, insurance type, and facility type) are compared to test whether there is a significant difference in average compliance ratios. Additionally, linear association between the compliance and distance variable is presented in this table. When assessed as a continuous measure, significant differences in (mean) compliance were observed among the same three factors. Largely aligning with the previously stated results, rural women (.43) had smaller mean compliance ratios than urban women (.53), women on public insurance (.48) had smaller mean compliance ratios than women on private insurance (.52), and women utilizing critical access hospitals (.37) had smaller mean compliance ratios than women utilizing a different facility type (all >.50). Similar to results obtained in Table 2, each additional mile of distance from the nearest facility was associated with a decrease in compliance of approximately .005, although this effect was not significant after accounting for multiple testing.

Table 3.

Continuous Compliance by Sociodemographic Factor.

Factor	Mean Compliance Ratio	P-Value
Race		.326
AAPI	.54 (.33)
Black	.52 (.32)
Native American	.48 (.35)
Other	.51 (.32)
White	.54 (.32)
Ethnicity		.988
Non-Hispanic	.52 (.32)
Hispanic	.52 (.33)
Rurality		<.001
Urban	.53 (.32)
Rural	.43 (.32)
Insurance type		.016
Public	.48 (.32)
Private	.52 (.32)
Facility type		<.001
Critical access hospital	.37 (.31)
Mobile unit	.52 (.34)
Office	.52 (.32)
Other	.55 (.32)
Outpatient hospital	.51 (.33)
Distance	−.004 (.002)	.027

*All results are given as mean (SD) except for distance, which is $\hat{β} (s e (\hat{β}))$ .

**Distance model controls for rurality (urban/rural).

Ultimately, the results from the generalized linear models (GLMs) with multiple predictor variables underscore the outsize importance of rurality in adhering to screening guidelines in this population (See Table 4 for a summary of findings and Appendix C for statistical model). Indeed, when treated continuously, compliance is significantly associated with only rurality; those who live in such areas have a compliance approximately .09 points lower than those who live in urban areas. This is roughly equivalent to a 54-year-old woman having one fewer year in which she receives a mammogram.

Table 4.

Compliance Prediction Model Results.

	Continuous	Categorical (CR = 1)	Categorical (CR = 3/4)	Categorical (CR = 1/2)	Categorical (CR = 1/4)
Rural = yes	$β_{R}$ = −.089 (P < .001)	OR = .673 (P = .004)	OR = .648 (P = .001)	OR = .616 (P < .001)	OR = .501 (P < .001)
Insurance = public	$β_{I}$ = −.033 (P = .112)	OR = .731 (P = .045)	OR = .910 (P = .510)	OR = .838 (P = .171)	OR = .852 (P = .264)
Facility type	(P = .095)	(P = .293)	(P = .278)	(P = .073)	(P = .025)
Critical access hospital	REF	REF	REF	REF	REF
Mobile unit	$β_{M U}$ = .045	OR = 1.302	OR = 1.484	OR = 1.465	OR = 1.026
Office	$β_{O f f i c e}$ = .057	OR = 1.292	OR = 1.119	OR = 1.701	OR = 1.533
Other	$β_{O t h e r}$ = .082	OR = 1.472	OR = 1.289	OR = 1.945	OR = 1.829
Outpatient hospital	$β_{O H}$ = .065	OR = 1.486	OR = 1.253	OR = 1.706	OR = 1.553
Distance	$β_{D}$ = −.001 (P = .504)	OR = .997 (P = .715)	OR = .997 (P = .706)	OR = .996 (P = .480)	OR = .999 (P = .924)

*REF = reference category.

** OR = odds ratio, model-based point estimate.

When parameterized categorically, compliance continues to have a noticeable, significant association with rurality. More specifically, those who live in rural areas have anywhere from a 33% to a nearly 50% decrease in the odds of being compliant with screening recommendations (depending on which threshold is used). Of the other variables included in the prediction models, only insurance type and facility type displayed any indication of a signal, and only when full compliance (CR = 1) (for the former) and 1/4 compliance (for the latter) were analyzed. More specifically, those with public insurance had about 27% decreased odds of being completely compliant with screening recommendations as compared to those with private insurance, and women utilizing offices, outpatient hospitals, and other facilities had at least 50% greater odds of having a CR $\geq$ .25. However, neither of these two effects were statistically significant after accounting for multiple testing.

This study sought to determine the impact of six sociodemographic factors (race, ethnicity, rurality, insurance type, facility type, and distance from nearest mammography facility) on a sample of mid-life female Kansans’ adherence to national breast cancer screening guidelines. We defined compliance two different ways: (1) continuously, where the number of years in which a woman received a mammogram was divided by the number of years she was recommended to have received a mammogram, and (2) categorically, where the continuous formulation of compliance was used to create binary variables indicating compliance was greater than or equal to $X$ , $X \in {1 / 4, 1 / 2, 3 / 4, F u l l}$ .

The findings suggest that all factors save race and ethnicity have at least some bearing on compliance with screening guidelines among mid-life female Kansans. The strongest signal was observed in the rurality variable, which demonstrated a significant relationship with compliance regardless of how it was defined. This supports the initial claim that various factors impact adherence in a multifaceted manner and rurality in particular is of considerable importance in the state. One study found that in Indiana and Ohio mailing a DVD and having a patient navigator call a week later resulted in rural women who were five times more likely to become up to date with mammography screening.²³ An outreach program like this may be beneficial from improving regular screening among rural women in the state of Kansas.

Strengths of our study include the large overall sample size, the use of objectively measured claims data to capture screening frequency (as opposed to self-report measures), and the use of a variety of sociodemographic factors as opposed to a single construct.

This study also contained some notable limitations. For instance, given that a total of twenty-five Native Americans were included in this study, we concede that findings related to this group may not be representative of the group as a whole; this is regrettable, especially given Kansas’ sizeable Native American population (estimated at just over 60,000, or 2.1%, of the state in 2017).²⁴ As described earlier, the database used for the study also necessitated that the ages under study be restricted to a relatively small range, thereby further limiting the generalizability of our findings. The database also does not constitute a complete listing of a woman’s medical history, for instance, it is entirely feasible that a mammogram could have been received and not recorded with Centrus Health. Additionally, the RUCA was used to define a zip code as rural or urban; however, this definition is likely stricter than is necessary. Further, the arguably rudimentary nature of the prediction model is another limitation; further work is needed to refine and assess the model’s utility before recommending it be used by clinicians and other health professionals.

Conclusion

Given its large rural population, Kansas is a feasible environment for studying the impacts of rurality; in this case, on adherence to and compliance with various mammography screening guidelines. Our findings suggest that improving rural women’s access to (and awareness of) mammography may be important targets for improving overall adherence to screening regimens, particularly in regions with large rural populations.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethics Statement

The present study does not meet the definition of either an animal or a human study.

ORCID iDs

Benjamin Brewer

Sam Pepper

Appendix A: Detailed Demographics of Analytic Sample

Description	Summary Statistics
Age	49.11 (2.62)
Ethnicity = Hispanic	253 (4.0)
Insurance = public	270 (4.3)
Rural = yes	547 (8.6)
Race
AAPI	176 (2.8)
Black	626 (9.9)
Native American	25 (.4)
Other	449 (7.1)
White	5060 (79.9)
Facility type
Critical access hospital	38 (.6)
Mobile unit	72 (1.1)
Office	1776 (28.0)
Other	2013 (31.8)
Outpatient hospital	2437 (38.5)
Number of unique women	6336
Average number of claims	2.30 (1.32)
Average compliance ratio	.52 (.32)

* Descriptive statistics presented as mean (SD) for continuous variables and N (%) for categorical.

Appendix B: Compliance (Categorical) by Sociodemographic Factor (Row Values)

	Compliance Proportion
	Full	3/4	1/2	1/4
Race
AAPI	.292	.315	.528	.753
Black	.233	.286	.503	.758
Native American	.25	.25	.417	.833
Other	.232	.281	.491	.776
White	.259	.306	.532	.799
Ethnicity
Non-Hispanic	.246	.287	.496	.761
Hispanic	.267	.297	.496	.733
Rurality
Urban	.254	.296	.512	.778
Rural	.183	.21	.366	.618
Insurance type
Public	.25	.29	.503	.767
Private	.193	.263	.433	.707
Facility
Critical access hospital	.132	.184	.263	.526
Mobile unit	.25	.347	.472	.694
Office	.236	.276	.499	.766
Other	.261	.307	.535	.798
Outpatient hospital	.246	.283	.476	.741

* Results given as proportion of sampled group meeting given compliance; values can range from 0 to 1.

Appendix C

References

Nazari

Mukherjee

. An overview of mammographic density and its association with breast cancer. Breast Cancer. 2018 May;25(3):259-267. Epub 2018 Apr 12. PMID: 29651637; PMCID: PMC5906528. doi:10.1007/s12282-018-0857-5

Center for Disease Control and Prevention [Internet] . Basic Information About Breast Cancer. [cited 2023 February 9] Available from: https://www.cdc.gov/cancer/breast/basic_info/index.html

Siegel

Miller

Wagle

Jemal

. Cancer statistics, 2023. CA Cancer J Clin. 2023 Jan;73(1):17-48. PMID: 36633525. doi:10.3322/caac.21763

The University of Kansas Cancer Center [Internet] . Most Common Cancers in Women and How to Detect Them Early. [cited 2023 February 9] Available from: https://www.kucancercenter.org/news-room/blog/2020/08/most-common-cancers-women-how-to-detect-them-early

National Cancer Institute Surveillance, Epidemiology, and End Results Program [Internet]. Cancer Stat Facts: Female Breast Cancer . [cited 2023 February 9] Available from: https://seer.cancer.gov/statfacts/html/breast.html https://seer.cancer.gov/statfacts/html/breast.html

Kaur

Jajoo

Naman

Kandwal

Singh Brar

Garg

, et al. Identifying barriers to early diagnosis of breast cancer and perception of women in Malwa region of Punjab, India. Global Health Journal. 2023;7(1):34-42. DOI: 10.1016/j.glohj.2023.02.006

Zeeshan

Salam

Khalid

QSB

Alam

Sayani

. Diagnostic Accuracy of Digital Mammography in the Detection of Breast Cancer. Cureus. 2018 Apr 8;10(4):e2448. PMID: 29888152; PMCID: PMC5991925. doi:10.7759/cureus.2448

Center for Disease Control and Prevention [Internet] . What Is Breast Cancer Screening? [cited 2023 February 9]. Available from: https://www.cdc.gov/cancer/breast/basic_info/screening.html

Guan

Haardörfer

McBride

Lipscomb

Escoffery

. Factors Associated with Mammography Screening Choices by Women Aged 40–49 at Average Risk. J Wom Health. (Aug 2022);31(8):1120-1126. DOI: 10.1089/jwh.2021.0232

10.

Center for Disease Control and Prevention [Internet] . What Is Breast Cancer Screening? [cited 2023 May 3]. Available from:. https://www.cdc.gov/cancer/breast/basic_info/screening.htm#:∼:text=The_USPSTF_recommends_that_women,often_to_get_a_mammogram

11.

Breast Cancer . Screening.” Recommendation: Breast Cancer: Screening | United States Preventive Services Taskforce, US Preventive Services Taskforce. (11 Jan. 2016). https://www.uspreventiveservicestaskforce.org/uspstf/recommendation/breast-cancer-screening.

12.

Center for Disease Control and Prevention [Internet] . Health and Economic Benefits of Breast Cancer Interventions. [cited 2023 February 9]. Available from: https://www.cdc.gov/chronicdisease/programs-impact/pop/breast-cancer.html

13.

Ponce-Chazarri

Ponce-Blandón

Immordino

Giordano

Morales

. Barriers to Breast Cancer-Screening Adherence in Vulnerable Populations. Cancers. 2023;15(3):604. https://www.mdpi.com/2072-6694/15/3/604

14.

Advani

Nayak

VonVille

Diamond

Burnett

, et al. Racial/ethnic disparities in use of surveillance mammogram among breast cancer survivors: a systematic review. J Cancer Surviv. 2022 Jun;16(3):514-530. Epub 2021 May 13. PMID: 33982233. doi:10.1007/s11764-021-01046-2

15.

Issues in Rural Service Provision, Kansas Children’s Cabinet and Trust Fund , Nov. 2017, https://kschildrenscabinet.org/wp-content/uploads/2018/03/Rural-Access-Kansas.pdf. Accessed 09 May 2023.

16.

Weeks

. Kansas, a Rural State?” Voice For Liberty. Wichita Liberty. July 4, 2017. https://www.wichitaliberty.org/economics/kansas-a-rural-state/

17.

Wiese

Islami

Henry

. Changes in geographic accessibility to mammography by state and rural-urban status, united states, 2006-2022. JNCI. Journal of the National Cancer Institute. 2022 Dec 14.

18.

Centers for Medicare and Medicaid Services [Internet] . Accountable Care Organizations (ACOs). [cited 2023 February 9]. Available from: https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/ACO

19.

Economic Research Service U.S. Department of Agriculture [Internet] . Rural-Urban Commuting Area Codes. [cited 2023 February 9]. Available from: https://www.ers.usda.gov/data-products/rural-urban-commuting-area-codes.aspx

20.

U.S. Food and Drug Administration [Internet] . Mammography Facilities. [cited 2023 February 9]. Available from:. https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfMQSA/mqsa.cfm

21.

Gavin

. ZipcodeR: Advancing the analysis of spatial data at the ZIP code level in R, Software Impacts. 2021;Volume 9:100099, ISSN 2665-9638, DOI: 10.1016/j.simpa.2021.100099

22.

Haynes

. Benjamini–Hochberg Method. In: Dubitzky

Wolkenhauer

Cho

Yokota

, eds Encyclopedia of Systems Biology. New York, NY: Springer; 2013. DOI: 10.1007/978-1-4419-9863-7_1215.

23.

Champion

Monahan

Stump

Biederman

Vachon

Katz

, et al. The Effect of Two Interventions to Increase Breast Cancer Screening in Rural Women. Cancers. 2022;14(18):4354.

24.

Kansas Department of Health and Environment [Internet] . Document Center. [cited 2023 May 3]. Available from:. https://www.kdhe.ks.gov/DocumentCenter/View/13567/Status-of-American-Indian-Alaska-Native-Population-Death-and-Birth-in-Kansas-2011-to-2015-PDF#:∼:text=An_estimated_61%2C363_American_Indians,with_one_or_more_races

Rurality Impacts Mammography Screening Adherence Among Mid-life Women in the Kansas Region

Abstract

Objective

Methods

Results

Conclusion

Keywords

Introduction

Methods

Results and Discussion

Conclusion

Footnotes

Declaration of Conflicting Interests

Funding

Ethics Statement

ORCID iDs

Appendix A: Detailed Demographics of Analytic Sample

Appendix B: Compliance (Categorical) by Sociodemographic Factor (Row Values)

Appendix C

References