Framework and Strategies to Eliminate Disparities in Colorectal Cancer Screening Outcomes

Abstract

Preventable differences in colorectal cancer (CRC) mortality across racial/ethnic, economic, geographic, and other groups can be eliminated by assuring equitable access and quality across the care continuum, but few interventions have been demonstrated to do so. Multicomponent strategies designed with a health equity framework may be effective. A health equity framework takes into account social determinants of health, multilevel influences (policy, community, delivery, and individual levels), screening processes, and community engagement. Effective strategies for increasing screening uptake include patient navigation and other interventions for structural barriers, reminders and clinical decision support, and data to continuously track metrics and guide targets for improvement. Community resource gaps should be addressed to assure high-quality services irrespective of racial/ethnic and socioeconomic status. One model combines population-based proactive outreach screening with delivery at in-person or virtual points of contact, as well as community engagement. Patient- and provider-based behavioral interventions may be considered for increasing screening demand and delivery. Providing a choice of screening tests is recommended for CRC screening and access to colonoscopy is required for completion of the CRC screening process.

INTRODUCTION

In the United States, there are persistent, avoidable differences in rates of mortality from colorectal cancer (CRC) across racial/ethnic, economic, geographic, and other stratifications (1-3). Screening is an established strategy for reducing the risk of death from CRC, and there is convincing evidence that disparities in CRC death rates are driven by differences in access, use, and quality of screening (1, 4-8). However, screening involves a series of care processes that are each influenced by factors at intersections of social determinants of health (SDHs) (7, 9-13), with health behaviors, and design of screening delivery. This paper summarizes the evidence on disparities in CRC health outcomes and discusses a health equity framework to guide the use of evidence-based interventions (1, 7, 12, 14). Mortality difference is considered the primary measure of disparities in CRC screening outcomes, with other disease-specific outcomes being overall and late-stage CRC incidence. Differences in delivery of care along the CRC screening continuum are on the causal pathway of disparities in mortality and are important process outcomes for guiding quality improvement activities.

COLORECTAL CANCER AS A PROMISING MODEL FOR ELIMINATING DISPARITIES IN HEALTH OUTCOMES

Cancers of the colon and rectum are currently the second leading cause of death from cancer among men and women combined and are projected to cause an estimated 53,200 deaths in 2020 (15). The annual incidence of CRC dropped from 66 per 100,000 population in 1985 to 36.3 per 100,000 in 2017, and the age-adjusted mortality rate dropped from 28.1 per 100,000 in 1980 to 13.5 per 100,000 in 2017 (16). These improvements are attributed primarily to increasing uptake of screening in the population (17). Carcinogenetic and molecular pathways that are well-described for CRC (18), along with strong evidence on risk factors, has enabled development of preventive interventions such as lifestyle modification, chemoprevention, and interception of precancerous lesions (19). Screening has been shown to be effective in reducing both the incidence and mortality of CRC (19-22).

CRC is a prototypical disease for the classic Wilson & Jungner screening criteria (23). Most CRCs are believed to arise in adenomas, the removal of which can prevent CRC (18). The long estimated dwell time from adenoma to detectable cancer and sojourn time form invasive cancer to clinical detection of >10 years and ~5 years, respectively (24, 25), provide opportunities for detection and treatment of precursor and early cancerous lesions before they become fatal. Evidence shows that detection at precursor adenoma stages reduces incidence, and detection at early cancer stages improves curative treatment. Current 5-year survival rates for CRC are 90%, 72%, and 14% when detected at localized, regional, and distant stages, respectively (16). A number of screening tests are recommended in various national guidelines. Tests included in the US Preventive Services Task Force (USPSTF) recommendation include colonoscopy, flexible sigmoidoscopy, computed tomographic (CT) colonography, fecal immunochemical test (FIT), multitarget stool-DNA (MT-sDNA) test, and guaiac-based fecal occult blood tests (26). The tests vary in ease of use, sensitivity, specificity, ability to detect precancerous adenomas and early cancers, and the need for follow-up testing. Some of the tests may be used in combination, but the evidence basis for such approaches is limited (27). For practical purposes, colonoscopy and stool-based tests (such as FIT and MT-sDNA) form the core for the delivery of CRC screening in the United States.

The majority of CRCs occur in people at least 50 years old (28). Besides older age, the principal risk factors for CRC are family history of CRC or inherited genetic disorders, inflammatory bowel disease, male sex, unhealthful dietary patterns, physical inactivity, obesity, smoking, and high alcohol consumption (1, 28-30). Nonsteroidal anti-inflammatory agents, including aspirin, reduce CRC risk (19). The USPSTF recommends the use of low-dose aspirin for the primary prevention of CRC in adults aged 50–59 years (Grade B) or 60–69 years (Grade C) who have a 10-year cardiovascular disease risk of ≥10%, a life expectancy of ≥10 years, and no increased risk for bleeding.

DISPARITIES IN COLORECTAL CANCER OUTCOMES

Barriers to screening in underserved populations are well described (1, 7). Disparities in health outcomes across many sociodemographic and clinical factors, including sex, race/ethnicity, socioeconomic status (SES), and geographic location have persisted for decades (1, 7, 15, 31-33). In one of the best-known stratifications for disparities, men of African descent have among highest CRC incidence and mortality rates of any racial/ethnic groups and have the highest proportion of cases diagnosed with metastasis (1, 15, 16). During the 2012–2016 period, the CRC incidence among non-Hispanic Black/African American (Black) men was 53.8 per 100,000 population compared with 44.0 per 100,000 among non-Hispanic white men and 35.3 per 100,000 among Asian/Pacific Islander men (15). Similarly, the CRC mortality rate during 2013–2017 was 23.9 per 100,000 among Black men compared with 14.1 and 11.4 among non-Hispanic white men and Asian/Pacific Islander men, respectively. With the decrease in CRC cases in people over the age of 50, an increasing proportion of cases is being diagnosed in those under 50, and in that group, too, Blacks have the highest rates (15). CRC mortality rates are also highest in communities with high levels of poverty and limited healthcare resources, such as limited endoscopic services (1, 7, 9, 15, 31, 33, 34). As an example of geographic disparities, Alaskan Natives have one of the highest incidence rates (89 per 100,000) for CRC, likely due to a combination of genetics, dietary patterns, and poor access to screening services (15).

The causes of CRC disparities are complex, but the evidence shows that they result primarily from differences in access to preventive and treatment interventions that are driven by SDHs (1, 35). Social risk factors that are ubiquitous in underserved populations, as well as services that are not culturally aligned or create structural barriers, limit access to the preventive and treatment services needed for CRC. The evidence suggests that racial and socioeconomic disparities in CRC can be readily addressed by assuring equitable access to care along the CRC care continuum (1, 12, 14, 36), if strategies are designed and implemented with health equity as the goal (35).

HEALTH EQUITY INTERVENTION FRAMEWORK

There are a number of health equity frameworks in the extant literature, such as R4P (remove, repair, restructure, remediate, and provide) and ConNECT (integration of context) (37). Domains of the framework used by the Centers for Disease Control and Prevention (CDC) include data and measurement, program implementation, policy implementation, and Health In All Policies approaches (37). Here we refine a previously described approach for CRC by drawing on elements of the CDC, ConNECT, and R4P frameworks (1, 37). The CRC health equity framework has four elements: SDHs (3, 13), multilevel influences (38, 39), multistep screening processes (21), and community integration (Figure 1) (40). A foundational element for the framework is accurate and complete data, both qualitative and quantitative, for each step of the screening process and at multiple levels of influence to iteratively understand, measure, and set priorities to guide strategies.

Figure 1.

Conceptual framework for health equity for addressing social determinants of health and structural barriers to eliminate disparities in colorectal cancer mortality.

Footnote: Social determinants of health and structural barriers to eliminate disparities in colorectal cancer mortality. SDH exert multilevel influences of care on a multistep screening progress. Community integration is essential to understand these influences and improve care.

Social Determinants of Health

Extensive direct and indirect evidence support the role of social factors as primary drivers of disparities across the CRC care continuum (see sidebar titled Major Stratifications of Disparities in Colorectal Cancer Screening Outcomes) (1, 21). SDHs include healthcare access, SES and financial strain, employment, education, housing and food insecurity, transportation, neighborhood factors, health literacy, language, and community and social connections (13), some of which are collected in electronic health record (EHR) data. SES data may be available at the individual level or from area-level measures at different levels of aggregation, such as the census tract, based on address information (41). The pervasive effects of SDHs were catalogued in a 2003 Institute of Medicine report, which concluded that factors related to social injustice, bias, and discrimination create differences in the quality of care and risk of chronic disease and premature death (42). SDHs influence multiple levels on the social-ecological model (38, 39) and shape access to resources, including those needed to realize benefit from CRC screening (2, 3, 7, 42), but may also have biological effects (1). Factors such as low educational level, low income, inadequate insurance coverage, and jobs that do not afford time off reduce access to screening, including colonoscopy. Medicare beneficiaries without supplemental insurance may forgo screening or follow-up colonoscopy due to cost (43).

BOX 1: MAJOR STRATIFICATIONS OF DISPARITIES IN COLORECTAL CANCER SCREENING OUTCOMES.

1. Race/ethnicity

2. English proficiency/Language

3. Immigrant status

4. Educational level

5. Income

6. Insurance coverage

7. Occupation

8. Age

9. Sex/Gender

10. Geography (neighborhoods, county, state, rural vs. urban, etc.)

11. Behavioral risk factors (e.g., obesity)

SDHs should be considered in measurement, prioritization, intervention design, and evaluation (13). Specific factors that are considered may vary depending on the population, but should include race/ethnicity, SES, education, health insurance, transportation, language and literacy, communication preferences and access to broadband technology, and employment status. Transportation is relevant in both urban and rural areas to get to and from colonoscopy appointments. Reliable contact information is necessary for appointments, reminders, and outreach strategies needed to increase the reach of screening in both rural and urban areas.

Multilevel Assessment of Barriers and Facilitators Across the Care Continuum

Stratifying the influences on CRC screening delivery and use of services across the screening continuum at appropriate levels (38, 39) helps to delineate barriers and promoters of care at policy, community, delivery (system, facility, and provider), and person levels (1). In this context, the individual or person level should be distinguished from social factors and focus on modifiable health behaviors. For instance, a lower percentage of patients who receive care in facilities that serve underserved populations receive follow-up colonoscopy for positive FIT compared with those receiving care in other facilities (44-46). Stratifying the barriers and facilitators to CRC screening at the different levels helps inform the selection, adaptation, implementation, and evaluation of interventions.

Policies and incentives.

Policies and incentives affect design, delivery, and reach of screening. Affordable Care Act (ACA) provisions include coverage for CRC screening by private insurers without cost-sharing because of the USPSTF Grade A recommendation (47). Despite ACA provisions, in Medicare, a screening colonoscopy is reclassified as diagnostic if a polyp is found (so-called surprise billing) or if the colonoscopy is performed for follow-up of a positive non-colonoscopy screening test, which then requires a coinsurance (43). Recommendations in major guidelines, except the American Cancer Society’s (48), do not define screening as inclusive of follow-up colonoscopy when a non-colonoscopy test is positive (43). National Committee for Quality Assurance measures on CRC screening promote provider delivery but do not include follow-up colonoscopy for positive non-colonoscopy tests, nor are there separate measures for other steps in the process. These barriers to colonoscopy are added to the inconvenience of bowel preparation, time off work, and need for an accompanying adult (49). Payers may invest less in improving CRC delivery because of concerns that the benefits of the investment will accrue to another payer (the so-called wrong-pocket problem) (1). One policy initiative that has been advocated for addressing disparities is to lower the recommended age of initial screening to 45 years, but with limited or fixed resources, especially colonoscopy capacity. However, lowering the age to start screening may divert resources to individuals with greater access and proclivity to screen than those burdened by social barriers.

Community-level factors.

Community-level factors, such as availability of healthcare facilities that provide the full range of screening options, are critical to screening participation (1, 15). Some facilities (such as federally qualified health centers) that provide care in hard-to-reach or remote communities may lack resources to provide the full continuum of screening for CRC (15, 50). Community asset mapping can help identify relevant partners and resources. Stakeholder mapping may clarify existing and needed relationships to facilitate care delivery. Stakeholder interviews may provide insights about historical injustices that may impact trust and fatalism as well as the feasibility, acceptability, and adaptations of interventions. Understanding historical contexts, such as racism and other forms of discrimination, may improve understanding of the underlying reasons for disparities in communities of color.

Delivery.

Provider, facility, and systems factors influence CRC delivery. Provider and facility factors include provider awareness, preferences, and attitudes; patient–physician communication and trust; recommendation of screening; designated staff to oversee screening; and team-based approaches. Systems factors include the use of population health management tools in the EHR to identify screening gaps and track outcomes, reminder systems, use of social incentives or audit and feedback, and policies or procedures on screening. The delivery model used is an important variable. The two primary models of CRC screening delivery are visit-based alone or visit-based plus proactive population-based screening (usually referred to as organized screening) (1, 51-53).

Individual health behavior.

A person’s health behavior, in the absence of structural barriers, determines screening uptake. Thus, in addition to social and demographic characteristics, health behavior (such as preventive care, physical activity, dietary pattern, and body mass index) and health status should be evaluated because of their influence on both the uptake of screening and the risk of CRC.

Screening Process

Screening is a multistep process, and its effectiveness depends on completing all appropriate steps (21, 54-57). The process involves risk assessment, screening initiation, regular rescreening, follow-up or diagnostic testing when results are abnormal, and treatment and surveillance (Figure 2; Table 1) (57). Evidence shows that potentially modifiable failures of the screening process, including failure of diagnostic testing, are common, particularly in underserved populations (6, 21, 54, 55). The quality of screening also affects outcomes (58), and there is evidence that the quality of screening varies by race/ethnicity (6). Receipt of CRC treatment, including the adequacy of adenoma removal and staging and treatment for cancer (26), is the final common pathway for screening benefit, but a full discussion of that step is outside the scope of this review.

Figure 2.

Colorectal cancer screening process and metrics for tracking delivery and quality based on a health equity framework.

Footnote: Various outcomes are possible for persons eligible for CRC screening. While the most common quality metric is the percentage of persons eligible who are up to date on screening, other screening failures are also associated with increased CRC mortality.

Table 1.

Colorectal cancer screening outcome measures and denominator definitions

Screening outcomes	Denominator(s)
Screening process outcomes
Percent eligible for routine screening	Universe (entire population)
Percent with orders or referral	All eligible people
Percent of completed orders	People with orders or referrals
Percent ever screened	All eligible people
Percent who did not screen at appropriate intervals	All eligible people
Incomplete test ratea	Tests performed by type
Percent who did not receive follow-up for abnormal resultsb	People with abnormal screening result
Percent who did not receive surveillance	All patients who had adenoma detected
Percent up to date	All eligible people
Screening intermediate health outcomes
Adenoma detection rate	All physicians performing colonoscopy
Receipt of surveillance	People with adenomas/polyps (based on guidelines)
Interval cancer rate	People undergoing screening by test
Receipt of treatment	People diagnosed with cancer
Screening health outcomes
CRC-specific mortality	Universe, all screening-eligible, or incident cancers
All-cause mortality	Universe or all screening-eligible

^aMay also be applied to completion of orders.

^bMost relevant for non-colonoscopy tests

Risk assessment.

The first step in the screening process is risk assessment. Age, family history, personal history of high-risk conditions (e.g., inflammatory bowel disease, Lynch syndrome, familial adenomatous polyposis, CRC), and the absence of symptoms of CRC are used to assess eligibility for average-risk screening (26). These are used in EHR algorithms to identify eligible people for reminders, provider dashboards, and delivery. Patients with a history of colectomy should be excluded from the population of eligible people.

Participation in screening.

Participation in CRC screening has been studied extensively using surveys (e.g., the National Health Interview Survey, the Behavioral Risk Factor Surveillance System, and the Medicare Current Beneficiary Survey) and EHR data, which have shown consistent disparities in uptake by race/ethnicity, SES, and geography (9-11). Many measures of uptake are based on being up-to-date or ever-exposed. However, persistence or regular rescreening for those with negative results should also be tracked (21, 59), because failure to screen at appropriate intervals is associated with higher risk of death from CRC (21). Up-to-date status considers guideline-concordant screening as described according to race/ethnicity by Mehta et al. (52). Being up-to-date means receiving colonoscopy within 10 years, sigmoidoscopy within 5 years, CT colonography within 5 years, MT-sDNA within 3 years, and FOBT/FIT within 12 months (less frequently in some countries) (52). One study found that being up-to-date on screening reduces the risk of death from CRC by 63% (21).

Follow-up testing.

Follow-up testing for abnormal screening, including colonoscopy for positive non-colonoscopy tests, is needed for the benefits of screening (21, 54-56), but delays in follow-up are commonly observed (21, 44, 46), especially in underserved populations (44, 46). Relatively few interventions have been shown to improve follow-up of abnormal result (60) or surveillance after adenoma detection (61-63). All screening efforts should include standardized measures of completion of colonoscopy in those with a positive result.

Community Integration

It is not always the case, contra-Field of Dreams, that “if you build it, they will come.” Simply creating and implementing an intervention does not make it accessible to underserved communities. That is why, in the implementation of evidence-based interventions in communities with health disparities, community engagement is critical to participation—so that when it is built, they will come, because they helped to build it (1). Although few empirical data exist on reductions in health disparities from community outreach and engagement (COE), COE enables understanding, prioritization, and integration of unmet social and clinical needs into strategy identification and implementation (see sidebar titled Principles of Community Engagement) (40). Meaningful COE builds and maintains trust and helps create interventions that are aligned with community values and norms (40).

BOX 2: PRINCIPLES OF COMMUNITY ENGAGEMENT.

1. Have shared goals with the community

2. Understand the community and its history of engagement

3. Build trust and seek commitment from stakeholders

4. Respect diverse perspectives within the community

5. Identify and mobilize community assets

6. Partner with the community

7. Assure community ownership and control of actions

8. Long-term commitment

DESIGN AND IMPLEMENTATION OF INTERVENTIONS TO ELIMINATE DISPARITIES

Few interventions have been specifically tested with comparators to demonstrate reductions or elimination of disparities in screening outcomes (51, 64). However, individual studies and natural experiments provide support for some strategies (7, 36, 52, 65). The health equity framework helps define principles for designing the delivery of CRC screening in underserved populations (see Table 2).

Table 2:

Colorectal cancer screening outcomes measures and denominator definitions

Social-ecological domains	Health equity applications	Intervention Strategies*
Personal/interpersonal	Social determinants of health Perceptions or experiences of social injustice (mistrust or fatalism) Individual health behaviors	Increase demand: Reminders Incentives Small media Mass media Education (group vs. one-on-one) Increase access: Support (e.g., navigation, transportation, and language) Remove cost-sharing
Delivery Provider Facility System	Goal-setting for health equity Cognitive bias and discrimination Structural barriers (e.g., design of services) Data on gaps population and community disparities Population health management tools with tracking systems to identify, report, and monitor gaps Clinical decision support (defaults in workflows) Choice of screening test Colonoscopy capacity Access to treatment Quality of care	Increase delivery: Social and financial incentives Reminders and clinical decision support Practice facilitation Increase access: Population-based approaches (e.g., mailed FIT or MT-sDNA) Navigation or support Language assistance
Community	Resources Access to full continuum of screening (risk assessment through screening Colonoscopy capacity Data on disparities within and between communities	Community engagement and outreach Asset mapping Stakeholder mapping Partnerships Outreach Mass media
Policy and Incentives (public and private)	Out of pocket cost (re-classification of screening) Screening options “Wrong pocket problem”(50)	Increase access: Remove structural barriers Remove cost-sharing Measure/metrics Definition of screening Quality metrics/incentive alignments

^aAdapted from the Community Guide with integration of core elements of effective implementation strategies (52, 71).

The most important is to intentionally set health equity as the goal and outcome for developing strategies, similar to the approach in creating patient navigation (66). One approach that has been recommended is to treat health disparities as a crisis (33, 66). Health disparities related to the coronavirus of 2019 (COVID-19) pandemic is a reminder of why that approach is needed (67). CRC and other disparities can be hidden in plain sight and discovered only when sought (68). Thus, quantitative and qualitative data should be used to identify disparities and relevant drivers and inform the choice of candidate intervention strategies. Rates of screening participation, follow-up testing for abnormal results, and incidence and mortality should be analyzed according to race/ethnicity, income, education, geography, and other means of stratification (2, 3). This should be followed by goal setting and/or policy-making to eliminate differences in outcomes (69). A large organized screening program that adopted mailed FIT and increased screening rates in all groups initially widened differences in screening rates between non-Hispanic whites and Blacks (52). After conducting focus groups, the program leaders created tailored invitations for people identifying as Blacks; racial differences have since narrowed.

Second, SDHs, mistrust and fatalism, and social injustice should be central elements in the heuristics of intervention planning (66). Fear and initial disgust may be especially prevalent in CRC screening tests, explaining lower uptake compared to other less effective cancer screening tests (70). Screening is associated with upfront cost or opportunity outlay, can be complex to navigate, and is subject to bias or temporal discounting (49). It may also be counterintuitive for underserved populations, because benefits of averted cancers or death in the future are not directly linkable to prior screening, and knowledge of people who died despite screening may create false association with getting screened. Meaningful engagement strategies such as story-telling and endorsement by community leaders may create trust and address misperceptions to help increase community demand.

Third, structural barriers should be addressed at all levels (71). A major barrier to screening participation in underserved communities is delivery models that are based solely on face-to-face encounters, which is exacerbated if tracking systems are not in place to proactively identify gaps in the screening process and provide feedback and reminders. Another is missed opportunities during face-to-face visits due to competing medical needs or failure to recommend or offer screening that may occur from cognitive bias. This underscores the importance of practice or system transformation to provide proactive outreach such as mailed FIT or MT-sDNA (1), in combination with transformation of visit-based delivery of screening.

Fourth, interventions should consider the resource needs of health facilities and communities as well as the community partnerships to address unmet needs related to the delivery of the full continuum of screening. Evidence suggests that lower-resourced health centers, small practices, and remote communities experience difficulty in delivering screening. Organizational interventions that decrease reliance on rushed interactions between patients and primary care providers are especially helpful. Examples include having medical assistants address screening while they are welcoming and rooming patients or mailed outreach outside of primary care visits (72).

Fifth, the best test is the one that gets done well, and a choice of tests should be available (26). The options can be offered sequentially or contemporaneously in an active choice format (73, 74). For one-time screening, offering the choice of FIT or colonoscopy increases screening participation, with strong evidence in underserved communities; challenges for FIT come with maintaining annual participation (75). Successful FIT screening programs use proactive, annual outreach. Access to colonoscopy is needed even if it is not used as the primary modality.

These principles should be incorporated in multicomponent interventions (71). An approach may encompass COE with data analysis to identify gaps in screening, set goals, and provide mailed outreach FIT or MT-sDNA along with workflows to offer FIT, colonoscopy, and other options at all feasible encounters. It requires population tracking systems to flag patients with gaps in screening, and regular audits should be conducted to support monitoring of performance across the screening continuum. Support or navigation is needed irrespective of the test offered. While mailed or automated outreach can reduce several structural barriers (76), many participants will need personalized support to participate. Specific support and appropriate navigation should be provided to reduce no-shows and obtain high-quality examinations. For example, studies show that Blacks and low-SES patients are more likely to no-show for colonoscopy appointments (77, 78). Studies also show that Blacks may have higher rates of inadequate bowel preparation (79). Therefore, colonoscopy should be performed with the intent to “get it right the first time.” Further, procedures should be in place to complete diagnostic testing within 90 days of a positive non-colonoscopy test (6, 21, 54, 55). Those diagnosed with cancer should have a source of care to receive the full work-up and course of cancer treatment appropriate for the stage at diagnosis. The program should be evaluated, including analysis of outcomes, and reported and reviewed with community partners on a regular basis to help guide iterations (1).

Successful programs that have achieved high screening rates use comprehensive, system-wide, long-term efforts and a population health management approach. Leadership alignment, goal setting, and quality assurance using directly mailed FIT outreach all contribute to success, along with offering screening during visits (virtual or in-person) (52, 80). Screening is provided at no cost. That approach enabled Kaiser Permanente Northern California to increase screening participation to 81% in non-Hispanic whites and about 78% in Blacks and Hispanics/Latinos in 2013 (52). Follow-up colonoscopy was completed within 90 days of a positive FIT in >80% of patients across all racial/ethnic groups (52). Community-wide efforts with alignment across stakeholders have also demonstrated improved uptake of screening in low-SES populations (36, 65), but sustainability and long-term effects on health outcomes (mortality) are unclear.

Thus, for eliminating disparities in CRC screening outcomes, direct and indirect evidence supports the use of multicomponent strategies that address multiple barriers and enable promoters at multiple levels and for all steps of the screening process. Mailed outreach and automated reminders can remove multiple structural barriers; personalized outreach can help overcome fear and mistrust. The Community Guide provides a set of curated interventions for structural barriers and SDH (71).

The Community Guide framework addresses the need to increase community demand (reminders, incentives, small media, mass media, group education, and one-on-one education); community access (reducing structural barriers and reducing client out-of-pocket costs); and delivery of screening (provider assessment and feedback, provider incentives, and provider reminders) (71). Rather than a kitchen-sink approach, interventions should be combined on the basis of identified needs. Tailored navigation or support is an important component and varies from low- to high-touch support, including community health workers (1). Because the cost-effectiveness and feasibility of high-touch support is not clear, creating services with multiple touchpoints, such as community health workers trained to support people across multiple care needs, rather than CRC screening alone, appears more feasible (1), but this requires comparative economic analysis. Specific behavioral tactics such as opt-out approaches may be effective, but evidence on the effectiveness of financial incentives is mixed (81). mHealth solutions such as mobile apps and interactive texting, plus information on websites including online instructional videos and wordless instructions, can help improve access to screening.

CONCLUSION

Disparities in CRC screening outcomes are well documented, but evidence suggests that they can be eliminated with appropriately designed multicomponent interventions. However, relatively few studies have specifically evaluated the impact of interventions on CRC health disparities using either process or health outcomes. A series of four large, cluster-randomized trials in the United Kingdom may provide examples of pragmatic interventions to specifically target screening disparities (82). Strategies should be designed with the intention of eliminating disparities by addressing structural barriers at each step of the screening continuum, addressing barriers at all levels of influence, and integrating community insights in the design and implementation of screening delivery strategies. Factors such as mistrust, fatalism, misperceptions, and specific resource gaps in the community, including transportation to facilities, should be considered. Outcomes of screening should be measured for each screening step to identify where gaps exist and interventions are needed, but the evidence base to inform strategies is scant or nonexistent because of the lack of large-scale efforts in affected populations (21, 60). Importantly, addressing CRC disparities requires an explicit understanding that extra resources will be required to achieve health equity in screening uptake and follow through.

Research is needed on whether using a health equity framework in the design, implementation, and evaluation of interventions eliminates health disparities. Accurate disaggregated data are needed to continuously monitor progress for all screening process steps as well as the quality (not just receipt) of screening. Evidence is needed on the relative contributions of gaps in different steps of the screening continuum to disparities in CRC death. Although a number of interventions have been shown to improve the uptake and persistence of screening, including reminders and small media (7, 51, 64, 71, 74, 80, 83, 84), there remains uncertainty as to effective interventions for closing gaps across the entire screening continuum. Scalable evidence-based interventions are needed for resource-limited facilities in underserved communities. Empirical evidence is needed on whether lowering the age to start screening narrows or widens CRC disparities screening outcomes. It is encouraging that CRC screening programs can have meaningful impacts on CRC morbidity and mortality within 5–10 years (85). After years of carefully documented differences in CRC screening rates and mortality across racial/ethnic, economic, geographic, and other groups, models of success in mitigating CRC screening disparities can help create and maintain momentum on reducing and ultimately eliminating them.