Reducing disparities and achieving health equity in colorectal cancer screening

Abstract

Increases in colorectal cancer screening are linked to the declining incidence of the disease over the past three decades. These favorable trends, however, are not observed in marginalized racial and ethnic populations with disproportionately lower rates of screening, higher disease incidence, and increased mortality despite advances in health technology and policy. This review describes the differences in screening uptake and test selection amongst racial and ethnic groups, discusses known obstacles and facilitators that impact screening, and highlights existing frameworks developed to achieve health equity in colorectal cancer screening.

Introduction:

Colorectal cancer (CRC) is the third most common malignancy and the third leading cause of cancer death in the United States.¹ While longitudinal studies have demonstrated an overall decrease in CRC incidence in the past three decades predominantly due to increased screening, these strides are not equitably distributed across racial and ethnic groups. ^{2, 3} CRC incidence remains high in Black and Hispanic populations, and mortality is disproportionately higher in Black Americans.^4–6 These health disparities serve as a strong motivator to identify and address modifiable healthcare risk factors along the CRC care spectrum that reduce disease burden on racial and ethnic minorities in the United States.

Advances in research, technology, and policy have been introduced to promote CRC screening.^7–9 The United States Preventative Services Task Force (USPSTF) recently released updated guidelines outlining recommended testing strategies and timeframes to inform patient-provider decision making.¹⁰ Ongoing research initiatives have resulted in new strategies that help patients understand the importance of CRC screening and improve access to testing.^{11, 12} Additionally, favorable health policy decisions, such as the Affordable Care Act and a proposal by the Centers for Medicare and Medicaid Services (CMS) to eliminate colonoscopy cost-sharing after positive FIT testing, may promote screening uptake.^{13, 14}

Despite these advances, significant discrepancies in CRC screening exist amongst racial and ethnic groups. For example, national and state-wide health surveys demonstrate screening rates remain low in Hispanic Americans, and suggest there are signs of a widening gap in screening rates among Black and American Indian/Alaskan Natives compared to the non-Hispanic White population.^{15, 16} Such disparities shed light on the increased numbers of advanced stage cancers at initial diagnosis and higher rates of cancer-related mortality amongst Black individuals.^17–19

Nearly four in five physicians in the United States believe unmet patients’ social needs negatively impact quality of care and lead to worse health outcomes.^20,21 This has spurred interest in health disparities research designed to identify root causes of healthcare disparities and inform organizational and governmental policy. Given the notable disparities in screening and in disease outcomes, there is ongoing work focused on unraveling racial and ethnic differences in the CRC continuum of care. Core terminology in this field have been previously defined in the literature and are summarized in Table 1.^{21, 22} Throughout this review, we describe racial and ethnic differences as disparities rather than inequities because, when synthesizing a wide variety of studies that differ in specific details and contexts, “disparities” serves as a more inclusive and therefore accurate summary-level description.

Table 1:

Key terms in health disparities^21,22

Key term	Definition
Health disparity	A difference in a measure of health, such as disease burden or outcomes, between racial and ethnic populations.
Healthcare disparity	Differences in care received by patients, including aspects such as delivery, quality, and treatment, between racial and ethnic populations after controlling for access, clinical need, and patient preference.
Inequity	Disparities that result from structural, institutional, and/or interpersonal biases that contribute to broad imbalances in power, justice, social structures, or resources.
Health equality	A strategy that aims to optimize health by providing equal resources to all groups without addressing health disparities.
Health equity	A strategy that aims to optimize health for all by recognizing and eliminating health disparities.

The goal of this review is to outline the current state of CRC screening in the United States as viewed through the lens of health equity. First, we will briefly outline the current model of CRC screening in the United States and describe the differences in screening uptake and test selection in racial and ethnic populations. Second, we will describe known barriers and facilitators that may explain observed differences in screening, and outline policy and research gaps. Finally, we will highlight existing experimental and in-practice models directed at eliminating inequities in CRC screening.

Colorectal cancer screening tests: Differences in efficacy and utilization by racial and ethnic groups

Current USPSTF guidelines recommend CRC screening for all adults aged 45 to 75 years old. Approved screening strategies include: fecal immunohistochemical test (FIT), guaiac fecal occult blood test (gFOBT), multi-targeted stool DNA testing (DNA-FIT) every, virtual computed tomographic (CT) colonography, flexible sigmoidoscopy, sigmoidoscopy with FIT, and colonoscopy.¹⁰ Although these methods can diagnose and prevent CRC deaths, there are differences in test performance and also in uptake amongst different racial and ethnic populations. Below, we review data on efficacy and utilization amongst racial and ethnic groups (Table 2). We report the mean number of CRC cases and deaths prevented as estimates based on simulations conducted by the Cancer Intervention and Surveillance Modeling Network (CISNET).²³

Table 2:

types of CRC screening tests recommended by USPSTF

Screening Strategy	Mean CRC cases and deaths prevented per 1000 screened when starting at 45 y agea	Test utilization by ethnicity/race (%, [year of data])	p-value (ref. NHW)
Direct visualization
Colonoscopy every 10 y24	Cases: 61 Deaths: 28	non-Hispanic White: 64.2% (2018) non-Hispanic Black: 60.0% (2018) Hispanic: 49.2% (2018) Non-Hispanic Asian: 49.3% (2018) Other non-Hispanic: 54.5% (2018)	<0.001
Sigmoidoscopy every 5 y30	Cases: 51 Deaths: 24	non-Hispanic White: 2.4% (2008) non-Hispanic Black: 1.4% (2008) Hispanic: 2.9% (2008) Non-Hispanic Asian: 3.1% (2008) Other non-Hispanic: N/A	NS
Sigmoidoscopy every 5–10 y + FIT every 1 y25	Cases: 57 Deaths: 28	non-Hispanic White: 0.6% (2016) non-Hispanic Black: N/A Hispanic: 1.1% (2016) Non-Hispanic Asian: N/A Other non-Hispanic: N/A	0.005
CT colonography every 5 y24	Cases: 55 Deaths: 26	non-Hispanic White: 1.1% (2018) non-Hispanic Black: 2.2% (2018) Hispanic: 1.8% (2018) Non-Hispanic Asian: N/A Other: 1.3% (2018)	<0.05
Stool-based
Guaiac FOBT every 1 y30	Cases: 42 Deaths: 24	non-Hispanic White: 8.9% (2010) non-Hispanic Black: 8.2% (2010) Hispanic: 5.6% (2010) Non-Hispanic Asian: 8.0% (2010) Other non-Hispanic: 7.1% (2010)	NS
FIT every 1 y24	Cases: 50 Deaths: 26	non-Hispanic White: 10.0% (2018) non-Hispanic Black: 12.6% (2018) Hispanic: 15.1% (2018) Non-Hispanic Asian: 15.5% (2018) Other non-Hispanic: N/A	0.002
DNA-FIT every 1 to 3 y24	Cases: 51 to 57 Deaths: 25 to 28	non-Hispanic White: n/a non-Hispanic Black: n/a Hispanic: n/a Non-Hispanic Asian: n/a Other non-Hispanic: n/a	n/a

^aestimated based on simulations conducted by the Cancer Intervention and Surveillance Modeling Network (CISNET)

Colonoscopy

Colonoscopy remains the gold standard for CRC screening in the United States due to its ability to detect early cancers and prevent malignancy via endoscopic removal of precancerous lesions. Four cohort studies found that colonoscopy has a sensitivity for detecting adenomas larger than 10mm ranging from 0.89 (95% CI = 0.78–0.96) to 0.95 (95% CI, 0.74–0.99) with specificity as high as 0.89 (95% CI = 0.86–0.91).⁹ Two prospective cohort studies also demonstrated a lower risk of developing CRC in adults 8 years after screening colonoscopy (standardized risk, 0.42% [95% CI, 0.24%–0.63%]), and a mortality benefit in persons undergoing colonoscopy (adjusted hazard ratio = 0.32, 95% CI = 0.24–0.45) compared to those that did not undergo colonoscopy. Simulations carried out by CISNET demonstrated that, on average, screening colonoscopy starting at age 45 would result in an estimated 61 CRC cases prevented, and 28 CRC deaths prevented per 1,000 individuals screened.²³ Additionally, these models approximate a lifetime risk of 16 complications and a 784 additional lifetime tests per 1,000 individuals screened.¹⁰

Data from the Behavioral Risk Factor Surveillance System (BRFSS) and the National Health Interview Survey (NHIS) estimate that colonoscopy is the test of choice in approximately 80% of adults aged 50–75 that are up to date with CRC screening.^{16, 24} Shapiro et al. further demonstrated that significant racial and ethnic differences in colonoscopy utilization exist. Non-Hispanic White and non-Hispanic black individuals had the highest colonoscopy rates in 2018 at 64.2% and 60.0%, respectively, while Hispanic and non-Hispanic Asian groups had considerably lower rates at 49.2% at 49.3% respectively. When adjusted for other sociodemographic factors and access, colonoscopy rates increased in all racial and ethnic groups indicating that unmet social needs may contribute to disparities in colonoscopy use.²⁴ Additionally, English-speaking Hispanic individuals were significantly more likely than Spanish-speaking Hispanic people to choose colonoscopy as a screening modality, suggesting language and culture contribute to CRC screening behaviors.²⁵ Differences in colonoscopy completion rates also exist between racial and ethnic groups. In a randomized trial comparing clinician recommendation versus patient choice, non-Hispanic White individuals who voluntarily selected colonoscopy had three times the odds of completing screening compared to Black, Hispanic, or Asian participants in the patient choice arm (odds ratio = 3.2, 95% = CI 1.6–6.1).²⁶ Therefore, notable racial and ethnic disparities exist in selection of colonoscopy as a screening method and in its subsequent completion.

Flexible sigmoidoscopy

Flexible sigmoidoscopy has been considered a viable screening modality for several decades. Several large, randomized control trials conducted in the 1990s, including the Prostate, Lung, Colorectal, and Ovarian (PLCO) trial, demonstrated that screening with sigmoidoscopy resulted in lower CRC incidence and decreased mortality due to CRC. The PLCO trial demonstrated that in those screened with sigmoidoscopy, CRC incidence was 13% lower (relative risk = 0.87, 95% CI = 0.76–0.88]) and CRC-related mortality was 25% lower (relative risk = 0.75, 95% CI = 0.66–0.85]) compared to the usual care arm after a median 15 year follow up period.²⁷ A hybrid strategy of sigmoidoscopy plus FIT has also been proposed as a way of addressing concerns about missing proximal CRC. Holme et al. evaluated this method and demonstrated in a randomized trial with 11 years of follow up that there was no significant difference in 15-year CRC incidence or mortality between sigmoidoscopy alone versus sigmoidoscopy plus a single round of FIT.²⁸ No prior studies have directly assessed the efficacy of sigmoidoscopy with yearly FIT, though it is assumed that parallel screening would increase test sensitivity. Additionally, no studies have assessed the accuracy of sigmoidoscopy to detect malignant or pre-malignant lesions, thus sensitivity and specificity are imputed from colonoscopy test performance data.²⁹ CISNET modeling estimates that sigmoidoscopy alone would avert 51 CRC cases and 24 CRC deaths per 1000 screened persons, and sigmoidoscopy with FIT would prevent 57 CRC cases and 28 CRC deaths per 1000 screened persons.^{10, 23}

Use of sigmoidoscopy with or without FIT testing has declined over time and now accounts for approximately 2% of CRC screening among adults aged 50 to 75.^{24, 30} Current information on sigmoidoscopy use by ethnicity and race is limited due to its infrequent use. The last published study examining sigmoidoscopy use by race and ethnicity uses data from 2000 to 2008.³⁰ Although the study’s point estimates are no longer valid, trends in use may reveal persistent disparities. In non-Hispanic White individuals, sigmoidoscopy rates declined by 7.0% and colonoscopy rates increased by 30.2%; however, amongst Hispanic persons, sigmoidoscopy decreased by 2.6% while colonoscopy use increased by only 17.8%. Viramontes et al. evaluated screening patterns using data from 2016 and found a small but significant difference in sigmoidoscopy use between Hispanic and non-Hispanic White groups (1.1% vs 0.6%, p < 0.005), suggesting that sigmoidoscopy use remains more common amongst the Hispanic population.²⁵ PLCO data also shows that amongst those with an abnormal result, Black Americans were less likely than non-Hispanic White participants to undergo follow-up diagnostic colonoscopy (adjusted risk ratio = 0.88, 95% CI = 0.83 to 0.93), highlighting racial and ethnic disparities.³¹

CT colonography

CT colonography, also known as virtual colonoscopy, uses tomography to generate two- and three-dimensional images of the entire colon and rectum to identify cancerous lesions and pre-cancerous polyps. Several studies have shown CT colonography as an accurate means of detecting precancerous lesions, though there is insufficient data to evaluate its impact on CRC incidence or mortality. Lin et al. pooled data from 7 studies evaluating CT colonography with bowel preparation and calculated the sensitivity and specificity to detect adenomas larger than 10 mm was 0.89 (95% CI = 0.83–0.96) and 0.94 (95% CI = 0.89–1.0).⁹ Modeling using available data estimates that CT colonography starting at age 45 would result in 55 fewer cases and 26 fewer deaths attributed to CRC per 1000 persons screened.¹⁰ Current society guidelines recommend CT colonography in patients with contraindications to endoscopy, as an adjunct following incomplete or inadequate colonoscopy, or in patients who decline colonoscopy or FIT testing. ^{32, 33}

Overall, CT colonography is the least common CRC screening method used in the United States.²⁴ Data obtained from the NHIS survey, however, indicate that CT colonography use has increased, from 0.79% to 1.33% (p < 0.001) between 2015 and 2018.³⁴ When stratified by racial groups, Hong et al. showed that in 2018 CT colonography use was 1.1% for non-Hispanic White, 2.2% for non-Hispanic Black, and 1.8% Hispanic individuals.³⁴ This indicates racial and ethnic differences in CT colonography utilization exist and necessitates further study.

Guaiac fecal occult blood test and fecal immunohistochemical test

Guaiac fecal occult blood testing (gFOBT) has been commercially available since the mid-20^th century and used for CRC screening since the 1970s.³⁵ Several large trials in the US and in Europe have demonstrated that, compared to non-screened individuals, yearly or biennial gFOBT is associated reduction in CRC mortality of 15 to 33%.^{36, 37} Shaukat et al. further demonstrated that the mortality benefits of gFOBT persisted even 30 years after years of follow up (relative risk with annual screening = 0.68, 95% CI = 0.56–0.82; relative risk with biennial screening = 0.78, 95% CI = 0.65–0.93]).³⁸ However, accuracy studies demonstrate gFOBT is imprecise in detecting CRC (sensitivity 0.5–0.75 [95% CI range 0.09–1.0]; specificity 0.96–0.98 [95% CI range 0.95–0.99]) and insensitive in detecting advanced adenoma (sensitivity 0.06–0.17 [95% CI range 0.02–0.23]).^{39, 40} CISNET modeling estimates that yearly gFOBT beginning at age 45 would result in 42 fewer cases and 24 fewer deaths due to CRC, though imprecision adds considerable uncertainty to model predictions.^{9, 10}

Fecal immunohistochemistry testing (FIT), unlike gFOBT, uses an anti-human globin antibody to qualitatively or quantitatively detect human hemoglobin in feces, which greatly increases both sensitivity and specificity.⁴¹ Although large efficacy trials have not been completed in the United States, encouraging data from other nationwide programs support the use of FIT. In a large prospective cohort study of over 5 million Taiwanese individuals, those undergoing screening with biennial FIT had a 10% reduction in CRC mortality versus controls not undergoing screening (adjusted relative risk = 0.90 [95% CI 0.84–0.95]) with a follow up of 6 years.⁴² FIT has the ability to detect CRC (sensitivity = 0.74–0.81, 95% CI range = 0.64–0.91; specificity = 0.93–0.94, 95% CI range = 0.91–0.96) and advanced adenoma (sensitivity = 0.23–0.28, 95% CI range = 0.19–0.37]) with greater confidence compared to gFOBT.⁹ Modest differences in sensitivity and specificity exist between qualitative and quantitative FIT, with qualitative methods generally having higher sensitivity and lower specificity than qualitative testing.⁴³ Prediction modeling shows that annual FIT beginning at age 45 would prevent 50 cases of CRC and avoid 26 deaths due to CRC.²³

Stool based gFOBT and FIT are the second-most common modalities used for CRC screening in the United States. In 2010, Shapiro et al. estimated using NHIS data that approximately 8.8% of screening aged adults had used a home FOBT test in the past 1 year, with small differences between ethnicity and race (White = 8.9%, 95% CI = 8.1–9.8%; Black = 8.2%, 95% CI = 6.6–10.0%; Hispanic = 5.6%, 95% CI = 4.3–7.3%; Asian = 8.0%, 95% CI = 5.4–11.6%; American Indian/Alaskan Native = 7.1%, 95% CI = 3.3–14.6).³⁰ Data from 2018 revealed an increase in FOBT or FIT use overall and in all ethnic and racial groups, with the highest increases seen in the Hispanic and Asian population (White = 10.0%, 95% CI = 9.2–10.9%; Black = 12.6%, 95% CI = 10.3–15.4%; Hispanic = 15.1%, 95% CI = 12.5–18.2%; Asian = 15.5%, 95% CI = 11.8–20.0%; Indigenous American = 7.1%, 95% CI = 3.3–14.6])²⁴. After 2010, the NHIS combined gFOBT and FIT into one variable, which may limit evaluation of trends during this time.

Given that current guidelines recommend yearly gFOBT or FIT testing, there is concern for disproportionate non-adherence over time between ethnic and racial populations. Inadomi et al. evaluated CRC screening adherence in a study of a diverse, urban population randomized to colonoscopy, FOBT, or a patient choice (colonoscopy or FOBT) and found that Hispanic and Asian groups were more likely to complete and adhere to FOBT compared to the non-Hispanic White population.²⁶ Additionally, there is evidence suggesting follow-up colonoscopy rates for abnormal FIT or FOBT testing differs between racial and ethnic groups. Data from the PROSPR Consortium indicate that Asian and American Indian/Alaskan Native groups were less likely to complete follow up colonoscopy within 90 days after an abnormal FIT or FOBT test.⁴⁴ Non-invasive FIT and FOBT are useful tools for preventing CRC, with national patterns in test usage suggest this method may be preferred by certain groups. Future research should focus on optimizing adherence across all racial and ethnic populations.

DNA-FIT

Multitarget stool DNA testing (DNA-FIT), marketed as Cologuard^™, was approved for use in the US in 2014 after parallel review by the FDA and CMS.⁴⁵ The test analyzes stool for specific, methylated DNA fragments that have been associated with malignancy. Four, large cross-sectional studies evaluated test performance, with pooled estimates showing a sensitivity of 0.93 (95% CI 0.87–1.0) and specificity of 0.85 (95% CI 0.84–0.86) for CRC, and a sensitivity of 0.43 (95% CI 0.40–0.46) and specificity of 0.89 (95% CI 0.86–0.92) for advanced adenoma.^{39, 46}At this time, there are no available studies evaluating the efficacy of DNA-FIT on reducing CRC incidence or mortality nor significant comparative data on accuracy between DNA-FIT and other stool-based or direct visualization methods.

DNA-FIT accounts for about 2.7% of CRC screening based on estimates from the 2018 NHIS, the first national survey to collect information on DNA stool test usage.²⁴ At this time, there is limited information on DNA-FIT uptake and adherence by race or ethnicity. However, the cost of DNA-FIT, nearly 15 times that of gFOBT, may be prohibitively expensive for many Black, Hispanic, and Indigenous Americans, groups that have disproportionately higher rates of poverty.⁴⁷ Further study is needed to identify disparities in DNA-FIT testing among race and ethnic groups.

Structural racism as the root cause of inequities in colorectal cancer screening

It is well established that social and environmental needs, also referred to as social determinants of health (SDOH), and wellbeing are inextricably linked. Structural racism, however, has historically limited access to opportunities, resources, and power on the basis of race and ethnicity.^{48, 49} In the United States, this is largely attributed to redlining, which is the policy of withholding financial services from customers based on neighborhood of residence that unjustly targets racial and ethnic groups, notably Black and Hispanic urban communities. This has resulted in entrenched financial exclusion, neighborhood disinvestment, and limited socio-financial mobility. Families living in these neighborhoods are, thus, less likely to attain adequate housing or meaningful generational wealth, and more likely have inadequate access to food, education, and healthcare services, factors that have been linked to increased mortality to chronic illness such as diabetes and heart disease.⁵⁰ Healthy People 2030, spearheaded by the US Department of Health and Human Services, recognizes the role of SDOH on health outcomes and has mobilized efforts to reduce disparities in numerous areas, including CRC screening. In this section, we will review known barriers and facilitators – organized into five SDOH domains – that impact equity in CRC screening.⁵¹

Barriers in Healthcare access and quality

To elect and undergo CRC screening, patients must interact with the healthcare system in a favorable environment that promotes preventive health. Prior studies, however, show that non-White racial and ethnic groups experience suboptimal access and quality with regards to CRC screening. Black, Hispanic, and Asian individuals are less likely have a usual source of care and less likely to report having a primary care provider, which limits entry into healthcare system.^{52, 53} Lack of health insurance is considered the biggest obstacle to accessing preventive care. Uninsurance rates analogously are higher in Black and Hispanic populations, with Hispanic patients uninsured at a rate more than double than that of non-Hispanic White individuals.^{54, 55} Data from the 2018 NHIS shows that, for adults aged 50 to 64, those uninsured had lower CRC screening rates compared to adults with any insurance, and those with Medicaid coverage had lower screening rates compared to adults with private insurance, strongly suggesting that lack of insurance independently results in less CRC screening behavior.²⁴

Patients with access to healthcare report that physician recommendation is the most important motivator to complete CRC screening. However, Black and Hispanic persons are less likely to receive counseling regarding CRC screening.⁵⁶ There is also evidence showing that quality of communication varies between racial and ethnic groups, with Black and Hispanic patients reporting less patient centered discussion and more dissatisfaction.^{57, 58} The role of race and ethnicity in patient-provider communication is complex; however prior work has identified potentially modifiable factors that may impact this relationship. Guerra et al. identified that the high number of acute patient visits, and lack of time dedicated to preventive care as the most common barriers to preventive care for outpatient primary care physicians.⁵⁹ Further studies indicate that such time constraints are exacerbated in non-White racial and ethnic groups. Dickman et al. demonstrated that between 1963 to 2009, Black Americans have fewer ambulatory visits per year versus White patients, thus have less opportunity and time to discuss preventive care.⁶⁰ Physicians similarly reported difficulty recommending CRC screening to non-English speaking Hispanic patients, who have among the lowest rates, given the amount of time needed to translate during the encounter.^{59, 61}

Additionally, implicit bias from healthcare workers is known to impact quality of care. Systematic reviews have shown implicit bias against Black and Hispanic individuals is present amongst many healthcare workers at varying levels of experience.⁶² Research has shown that Black patients treated by physicians with higher levels of bias against persons of color report lower quality of patient-provider interactions and lower levels of perceived respect.^{63, 64} Other studies have demonstrated Black patients were less likely to receive thrombolytics for myocardial infarction, and that Black and Hispanic patients with diagnosed with acute coronary syndrome were more likely to be discharged home from the emergency department.^{65, 66} Although implicit bias has not been evaluated in the context of CRC screening, the literature suggests it is pervasive throughout all of medicine and likely contributes to disparities, meriting further study in our field.

Barriers in education access and quality

Health literacy and level of educational attainment are considered a positive predictors for CRC screening behaviors.^{67, 68} Broadly defined as the skills and knowledge necessary to make favorable health decisions, health literacy is influenced by access to relevant and culturally appropriate medical information.⁶⁹ Regional telephone surveys have shown a modest yet significant association between CRC knowledge and participation in CRC screening (odds ratio = 1.05, 95% = CI 1.02–1.41).⁷⁰ Low health literacy, moreover, disproportionately affects racial and ethnic minorities and results in low awareness and knowledge about CRC screening.^{71, 72} US Census data demonstrates that educational attainment is lower in ethnic and racial groups, which may exacerbate disparities in health literacy.⁷³ National data on CRC screening utilization patterns, however, have not reliably demonstrated a significant association between educational attainment and screening prevalence or adherence to CRC screening.^{24, 30} Additionally, in a study of urban-dwelling Hispanic patients, health literacy was not a risk factor for CRC screening non-adherence.⁷⁴ This suggests that clinical comprehension of CRC is important in decision making that favors screening behaviors, and that this knowledge may be attained outside of traditional education.

Barriers in economic stability

Poverty in the US has experienced an overall decrease since the early 1990s, though the highest rates remain concentrated in Black, Hispanic, and Indigenous American groups.⁷⁵ Income and net worth, markers of economic stability, are positively associated with educational attainment, which is lower in persons from traditionally marginalized groups.⁷⁶ People with low socioeconomic status (SES) have higher rates of colorectal cancer incidence, even after controlling for age, sex and family history, have lower rates of CRC screening, and are less likely to undergo colonoscopy compared to those with higher SES indices.^{77, 78} There is also evidence that low SES may negatively impact quality of CRC screening. In a retrospective cohort study of over 12,000 participants, Lebwohl et al. demonstrated that patients with Medicaid had rates of suboptimal bowel preparation quality that were nearly twice that of the non-Medicaid group (odds ratio, 1.84 [CI 95% 1.61–2.11]).⁷⁹ Although the relationship between SES and CRC screening is complex, current evidence suggests that economic instability negatively impacts uptake and quality of CRC screening and must be considered in marginalized groups that disproportionately bear this burden.

Barriers in neighborhood and built environment

The built environment of neighborhoods has been shown to have an impact on well-being and health outcomes, including in CRC care.⁸⁰ Living in neighborhoods classified as food deserts or with low landscape vegetation have been shown to have a modest, yet significant, increase on CRC mortality.^{81, 82} People living in neighborhoods with high racial segregation or high area deprivation indices, which are composite scores that include census variables relating to education, income, housing, and household characteristics, have also been shown to have lower rates of CRC screening.^{83, 84} Although the reasons for this are multifactorial, transportation barriers such as cost, access, and time required to travel have been identified as significant and modifiable risk factors for suboptimal CRC care.⁸⁵ Amongst patients referred for initial CRC screening, an estimated 41% report lack of transportation as a barrier in completing recommended testing.⁸⁶ Lack of transportation has also been implicated in non-adherence to follow up colonoscopy after a positive stool based screening test.⁸⁷ Nonemergency medical transportation (NEMT) was developed to increase transportation access for Medicaid eligible patients; however these services require advanced scheduling, and have long travel and wait times that limit efficacy.⁸⁸ Marginalized racial and ethnic groups are more likely to live in areas with fewer available resources and incur more barriers to transportation, which may contribute to observed disparities in CRC screening and outcomes.⁸⁹

Barriers in social and community context

Social context has also been demonstrated to influence overall wellbeing and preventative healthcare behaviors, including those related to CRC screening. Data from the NHIS in 2009 demonstrated lack of family or friends to discuss health issues significantly decreased the probability of completing CRC screening, particularly in Black men.⁹⁰ Conversely, family and friend support were associated with increased odds of discussing CRC screening at primary care visits and with increased completion of colonoscopy among urban-dwelling Black individuals.^{91, 92} Knowing social isolation disproportionately affects all racial and ethnic groups, particularly those of older age and lower SES, it is prudent to develop CRC screening interventions that promote human bonding or rely on existing social networks.⁹³

Trust between patients and their healthcare providers is recognized as an important social mediator that promotes use of preventative services, including CRC screening.⁹⁴ Long-standing conscious and implicit racism, differences in communication, and socioeconomic context, however, engender medical mistrust among racial and ethnic groups.^{95, 96} Qualitative and quantitative studies have illustrated a negative association between mistrust and rates of stool-based and endoscopic forms of screening in Black men.⁹⁷ The root cause is complex and has been attributed, in part, to increased prejudice in daily life, such as being followed around stores or targeted by persons in power, and evidenced racism in healthcare.⁹⁸ Implicit racial bias has also been shown to result in lower quality and less patient-centered communication that further erodes trust.^{63, 99} Within socioeconomic context, educational attainment and economic stability have been implicated as factors that influence trust in the healthcare system among major racial and ethnic groups.^{96, 99} Undoing this form of structural racism ultimately requires vast societal change, and we as physicians can facilitate this by encouraging patient-centered discussions that humanize and empower traditionally marginalized populations.

Current healthcare strategies aimed at achieving equity in CRC screening Strategies to improve healthcare access and quality

The Patient Protection and Affordable Care Act (ACA) of 2010 represents the most significant overhaul and expansion of insurance coverage in the US since Medicare and Medicaid were established in 1965.¹⁰⁰ Nearly 20 million Americans gained health coverage by 2016 as a result of the ACA (Table 3).¹⁰¹ Its impact on CRC screening is known, with data showing an overall increase in the prevalence up to date CRC screening prevalence after ratification of the ACA, with the highest increases seen in states with early expansion of Medicaid.¹⁰² Despite these promising trends, disparities amongst racial and ethnic groups remain, suggesting further expansion of Medicaid is needed to impart further structural change.^{16, 61}

Table 3:

existing strategies aimed at achieving equity in CRC screening

SDOH domain	Disparity identified	Intervention(s)	Impact of Intervention
Healthcare access and quality	• Lack of health insurance54,55 • Lack of time per health encounter59	• Medicaid expansion: Protection & Affordable Care Act (ACA)100–102 • Educational text-based messaging outreach with opt-out mailed FIT104	• Nearly 20 million Americans gained health coverage. • Intervention increased screening by 17.3% versus text message alone.
Education access and quality	• Low health literacy71,72	• Decision tool and patient navigation (PN) for a Hispanic population111	• Intervention resulted in 40% higher completion of CRC screening by 6 months compared to usual care.
Economic stability	• High out-of-pocket cost 77,78,115	• Eliminating cost-sharing of colonoscopies following noninvasive CRC screening tests14	• Oregonians had 6% higher odds of receiving any CRC screening, 35% higher odds of undergoing a noninvasive screening compared to neighboring states without this policy.
Neighborhood and built environment	• Inadequate transportation85,86	• Ride-share based nonemergency medical transport (NEMT) to and from colonoscopy appointments86	• Stakeholder analysis identified key barriers (NEMT contractual differences, lack of chaperone, lack of door-to-door service) and provides framework for future implementation.
Social and community context	• High social isolation in medical settings90 • High medical mistrust94,97	• CRC screening PN provided to Black men that were initially recruited at barbershops in New York City119 • Faith-based education delivered during church services120,121	• In Black men recruited at barbershops, those receiving PN for CRC screening were more likely to be screened for CRC within 6 months versus controls that did not receive PN for CRC (17.5% vs 8.4%, p<.01) • Among Black churchgoers, 25% of those receiving spiritual-based CRC education completed colonoscopy within 3 months, compared to 4% of those who did not receive intervention. • No difference in CRC screening between spiritual and non-spiritual education delivered in church.

Open access colonoscopy has been proposed to increase availability, though data indicates this method results in patients receiving fewer colonoscopies than traditional methods due to high no-show and cancellation rates.¹⁰³ Text-based messaging outreach with opt-out mailed home FIT kits may also increase test access by decreasing provider burden by providing education outside of the medical encounter, and leveraging reciprocity to encourage participation in screening. Huf et al. demonstrated in a predominantly Black community that a strategy of behaviorally informed and educational text messages plus a mailed home FIT kit presented in an opt-out fashion significantly increased CRC screening by in a 17.3% compared to text messaging outreach alone (Table 3).¹⁰⁴ Further study is needed to determine if this strategy remains efficacious in other groups or when the outreach scale is expanded.

Telehealth services, prolific in the wake of the COVID-19 pandemic, present a unique opportunity and challenge to develop equitable strategies for increased healthcare access. In April 2020, telehealth usage for outpatient medical care soared to nearly 78 times higher than in February 2020. Usage rates subsequently declined but remained stably elevated compared to pre-pandemic levels, with telehealth visits accounting for 13% of claims in gastroenterology, suggesting lasting patient interest.¹⁰⁵ Telehealth may promote CRC screening behaviors, with one randomized trial showing that telephonic and web-based outreach increased stool-based screening by about 11%.¹⁰⁶ However, there is decreased use of these services among Hispanic, Black, Asian, and American Indian/Alaskan Native groups, and care must be taken to avoid exacerbate existing racial and ethnic disparities as telehealth expands.¹⁰⁷

Mitigating implicit bias in healthcare remains challenging, though one potential framework consists of increasing bias awareness and developing tools that eliminates discriminatory behavior. Colleagues at the University of Washington have shown online education on implicit bias can lead to significant and lasting increases in bias awareness among faculty, an important first step towards health equity.¹⁰⁸ Other organizational strategies include developing diversity, equity, and inclusion programs that foster structured bias training, professionalism, and workforce diversity, and anonymous, online tools that empower observers to report discrimination.¹⁰⁹

Strategies to improve education and quality

Patient navigation (PN) and education initiatives, such as print materials or online education tools, have been implemented to increase medical awareness and knowledge. PN has been demonstrated as an effective means of increasing CRC screening in vulnerable and low income populations.¹¹⁰ Reuland et al. showed that a clinical decision tool plus PN increased completion of CRC screening completion at 6 months by 40% in a predominantly Hispanic population (Table 3).¹¹¹ Other studies, however, demonstrate that PN has no significant effect on CRC screening rates in Black communities, suggesting other cultural factors may impact this relationship.¹¹² There is limited evidence that PN may promote adherence to gFOBT-based screening within community health centers, though further study is needed to clarify this relationship in different healthcare settings, with different screening tests, and amongst different racial and ethnic groups.¹¹³ Social media has been postulated as another method of disseminating relevant education on preventative care to improve preventative care, and there is a scientific need to assess its impact on CRC screening.¹¹⁴

Strategies to resolve economic barriers

Addressing the socioeconomic disparities that result in healthcare disparities among racial and ethnic groups in the United States requires intentional and meaningful legislative action. The ACA alleviated some economic burden of healthcare by removing barriers to obtaining health insurance, as described above, and eliminating cost-sharing of preventative care measures. The effect of eliminating cost-sharing on CRC screening modalities has been studied. Currently, it is estimated that 48.2% of privately insured patients and 77.9% of Medicare beneficiaries incur out-of-pocket costs for CRC screening.¹¹⁵ Barthold et al. demonstrated that rates of stool-based screening increased in Oregon after ending cost-sharing for follow up colonoscopy, suggesting that reducing out-of-pocket costs along the CRC screening spectrum can have positive downstream effects (Table 3).¹⁴ In response to widespread advocacy, CMS has proposed expanding the regulatory definition of “colorectal cancer screening tests” to waive cost-sharing of follow up colonoscopy after positive stool-based testing starting in 2023.¹¹⁶ While further study is necessary, this is a valuable example of medical professionals collaborating with legislators to advance health policy.

Strategies to improve the neighborhood and built environment

Overcoming disparities related to the physical living environment ultimately requires extensive regulatory overhaul to enact durable governmental programs that that promote development of equitable infrastructure. There is, however, ongoing research to study the impact of non-policy measures, namely rideshare-based NEMT on access to medical care. Pilot programs in primary care and radiology settings demonstrated that ridesharing NEMT programs modestly increased the show rate for primary care visits and timeliness for outpatient MRI studies in Medicaid patients, though uptake was low.^{117, 118} To evaluate issues in rideshare uptake, Issaka and colleagues engaged key stakeholders in the CRC screening pathway, including patients, clinical providers, and non-clinical staff, and identified contractual differences among existing NEMT providers, lack of door-to-door service, and lack of chaperone as major issues impeding successful implementation of rideshare-based NEMT as a part of CRC screening outreach (Table 3).⁸⁶ This study additionally provides a clear framework for implementing rideshare-based NEMT, and to study its impact on CRC screening.

Strategies to improve the social and community context

Traditional patient navigation models for CRC screening initially engage patients in an isolated, clinical setting and may fail to capture patients without a regular medical provider. Several studies have evaluated the impact of community-based interventions to improve CRC screening uptake. Cole and colleagues conducted a randomized trial in which Black men aged 50 years and older recruited from barbershops were randomized to receive CRC screening PN, blood pressure counseling only, and both CRC screening PN and blood pressure counseling. They demonstrated that those receiving CRC PN in barbershops had more than twice the odds of completing screening than those that did not receive CRC screening guidance (Table 3), suggesting culturally-based PN may encourage preventative care.¹¹⁹ Faith-based initiatives have leveraged local churches, often a stable communal organization in Black communities, to promote CRC screening with success. For example, the F.A.I.T.H. study demonstrated that faith-based education delivered in a church increased 3-month colonoscopy rates by about 19% among Black attendees (Table 3).¹²⁰ Holt et al. subsequently compared spiritual versus non-spiritual education on CRC screening delivered in churches and showed both interventions resulted in increased screening, with no significant difference in sigmoidoscopy or colonoscopy.¹²¹ These examples underscore the influence culturally-sensitive PN and social networks have on individual healthcare choices, and why future screening initiatives must partner with community organizations.

Considerations for future research investigating disparities in CRC screening

Since the release of the Heckler Report, the last 40 years have ushered considerable scientific progress in health inequities research; yet there are still gaps in methodology and knowledge. First, the collection and use of data on race and ethnicity must be optimized and standardized to ensure all groups are adequately captured. Missing or misclassified race/ethnicity data can result underestimating the magnitude of disparities.¹²² Merging race and ethnicity as “race/ethnicity” is no longer recommended, as this oversimplification increases confusion and misclassification.¹²³ Additionally, relying on race or ethnicity assigned by healthcare systems can introduce significant inaccuracy. Instead, standardizing self-reporting of race and ethnicity, and providing comprehensive categories to promote inclusivity and dignity are needed.

Second, longer follow-up time is needed in observational and intervention studies. Adequate CRC screening requires a multi-year investment beyond index screening, and we must better understand the mechanisms of long-term adherence. Additional research is needed to evaluate the efficacy of older strategies, such as church-based or barbershop outreach, after societal changes resulting from the COVID-19 pandemic. Third, efforts to increase the number of Black, Hispanic, Asian, and Alaskan Native/American Indian groups in CRC screening interventions and studies must be prioritized.

Conclusion

CRC screening remains a powerful strategy to prevent the development of colon cancer when patients are given access and resources to adhere to recommended strategies. Colonoscopy remains the gold standard test given its efficacy, and all recommended strategies are proven to save lives. However, health disparities in CRC screening stemming from systemic racism represent a serious public health challenge. As gastroenterologists, we must continue to recognize the social barriers and facilitators of optimal CRC care and collaborate with patients, healthcare systems, and legislators to eliminate disparities and achieve health equity.

What You Need to Know.

Background

Colorectal cancer screening saves lives yet marginalized racial and ethnic populations experience lower rates of screening and, consequently, higher disease burden.

Findings

We describe existing disparities among racial and ethnic groups by type of colorectal cancer screening test, identify barriers to screening behavior, and highlight strategies designed to overcome these obstacles.

Implications for Patient Care

Identifying critical disparities in the existing continuum of care is necessary to develop effective patient-centered initiatives and inform health policy that helps achieve equity in colorectal cancer screening.