How Can We Boost Colorectal and Hepatocellular Cancer Screening Among Underserved Populations?

Abstract

Colorectal cancer (CRC) and hepatocellular carcinoma (HCC) are common causes of cancer incidence and mortality in the USA, particularly among underserved populations such as racial/ethnic minorities, the under-/uninsured, and individuals with low socioeconomic status. Although screening can reduce cancer-related mortality, participation is suboptimal among underserved populations, likely serving as the largest contributor to observed inequities in HCC and CRC outcomes among US populations. In this narrative review, we highlight inequities across populations in the USA with respect to incidence and mortality for CRC and HCC and highlight potential causes, with a focus on screening rates. In addition, drawing from the recent literature, we highlight promising strategies for increasing screening for HCC and CRC and propose future research and policy solutions to optimize screening rates. With focused implementation of screening strategies and novel research, the burden of HCC and CRC can be reduced among underserved populations.

Introduction

Colorectal and hepatocellular cancers are major public health problems, particularly among underserved populations, such as racial/ethnic minorities, the under-/uninsured, individuals with low income and/or low education, and other groups with disproportionately low cancer screening and surveillance rates. While screening can reduce cancer-related mortality for both colorectal cancer (CRC) and hepatocellular cancer (HCC), and is guideline recommended, participation is suboptimal among underserved populations. In this narrative review, we will highlight the epidemiology and importance of increasing CRC and HCC screening among underserved populations, note current screening rates, identify recently reported evidence-based strategies for increasing screening, and suggest directions for future research. For the purpose of this review, we will focus on CRC and HCC because these are main gastrointestinal cancers for which poor outcomes disproportionately cluster among underserved populations and for which national screening guidelines exist.

Epidemiology of CRC and HCC Among Underserved Populations

CRC

CRC Incidence and Mortality

CRC is the third most common cancer and is the second leading cause of cancer deaths in the USA, with an annual incidence of 39.3 cases per 100,000 and mortality of 15.1 cases per 100,000. Among all racial/ethnic groups, blacks have the highest CRC incidence and mortality, followed by American Indian/Alaska Natives, non-Hispanic whites, Asians, and Hispanics (Fig. 1). Notably, even among groups with lower incidence, CRC is a major cause of cancer incidence and mortality. For example, among Hispanics, CRC is still the third most incident cancer and second leading cause of cancer-related death. In addition to disparities by race/ethnicity, higher CRC incidence has been noted among those with lower socioeconomic status (SES, more recently referred to as socioeconomic position (SEP), lower education, and lower levels of acculturation) [2•, 3].

Fig. 1.

Age-adjusted colorectal cancer incidence and mortality rates per 100,000 by race/ethnicity, SEER 18 database, 2000–2011 [1]

CRC Stage at Presentation and Stage-Specific Survival

Variations in CRC stage at presentation and stage-specific survival have been noted across underserved populations. In a study utilizing the National Cancer Database, Halpern et al. reported that the risk for higher stage CRC was increased 1.2 to 1.4-fold among blacks, Hispanics, and those with other race/ethnicity vs. non-Hispanic whites, and increased 1.8 to 2.2-fold for individuals either without insurance or with only Medicaid compared to those with private insurance [4]. Interestingly, in multi-variable analyses, adjustment for insurance attenuated the observed disparities by race/ethnicity, but adjustment for race/ethnicity did not attenuate observations by insurance status. Having lower education and lower income has also been associated with risk for higher stage at presentation [5]. While disparities in stage at presentation certainly contribute to observed disparities in CRC mortality, disparities in stage-specific survival also likely play an important role. For example, stage-specific mortality is clearly higher among blacks and may be higher among Hispanics compared to non-Hispanic whites [6•, 7, 8]. Low SES, and in particular, under/uninsurance also have been associated with poorer stage-specific mortality [9, 10].

HCC

HCC Incidence and Mortality

HCC is a major cause of cancer incidence and mortality in the USA, with overall incidence and mortality of 8.12 cases and 6.12 cases per 100,000, respectively [1, 11]. The 5-year survival rate is only 16.8 %, and the incidence and mortality for HCC has been increasing since the 1980s. Although HCC is not as common as CRC, it is more likely to be fatal, as highlighted by the smaller difference between incidence and mortality for HCC compared to for CRC (Figs. 1 and 2) [1]. Among racial and ethnic groups, Surveillance Epidemiology and End Results (SEER) registry data suggest Asian/Pacific Islanders have the highest incidence and mortality, followed by American Indian/Alaska Natives, Hispanics, blacks, and non-Hispanic whites (Fig. 2) [1]. National data from the Veterans Affairs administration have also reported higher rates for HCC compared to non-Hispanic whites for hepatitis C-infected veterans. HCC cases are clustered in areas of low SES (e.g., high poverty, high unemployment, and low education areas) [12]. Similarly, mortality rates are higher in racial/ethnic minorities and patients of low SES [13]. Elderly patients, blacks (compared to non-Hispanic whites) and those of low SES all have poorer survival rates than their counterparts. Some studies have hypothesized these differences in survival are related to biologic differences in tumor behavior; however, others have highlighted sociodemographic inequalities across the HCC cancer care continuum as a potential etiology for the health disparities [13–18].

Fig. 2.

Age-adjusted hepatocellular carcinoma incidence rate per 100,000 by race/ethnicity, SEER 18 database, 2000–2011 [1]

HCC Stage and Stage-Specific Survival

Sloane and colleagues used the SEER database to demonstrate that blacks were significantly more likely to have regional and distant metastases at presentation compared to whites (p<0.005) [15]. Evaluating stage-specific treatments, blacks were less likely to have surgical therapy for localized (p=0.001) and regional (p=0.01) HCC. Among patients with localized disease, blacks (HR 1.24, 95 % CI 1.15–1.33) and Hispanics (HR 1.08, 95 % CI 1.01–1.15) both had higher mortality rates than whites, while Asian patients had lower mortality rates (HR 0.87, 95 % CI 0.82–0.93) [19]. After adjusting for treatment, blacks continued to have higher mortality (HR 1.11, 95 % CI 1.03–1.20), and Asians continued to have lower mortality (HR 0.84, 95 % CI 0.79–0.89). The survival difference between Hispanics and non-Hispanic whites was no longer present after adjusting for treatment exposure. There are few published data on variations in HCC stage by income or educational status; however, several studies have found variations in HCC treatment according to race/ethnicity and SES [20–22]. Elderly, racial/ethnic minorities, and patients of low SES had lower treatment rates than their counterparts, even among the subset with early stage HCC [18].

Disparities in Screening for CRC and HCC

CRC Screening

Colorectal cancer (CRC) screening is nationally recommended by the United States Preventive Services Task Force (USPSTF) [23] and has successfully reduced CRC incidence and mortality over the past 30 years [1, 24]. According to 2010 National Health Interview Survey data, the national screening rate is 59 %, defined by completion of a colonoscopy in the last 10 years, sigmoidoscopy in the last 5 years, or fecal occult blood test in the last year among individuals age 50 to 75 [25]. However, disparities in screening rates across race, ethnicity, income, education, and insurance status are profound (Fig. 3). The most striking disparities in proportion of individuals up-to-date with screening have been observed among the uninsured vs. insured (21 vs. 59 %), and disparities are particularly striking in light of the Healthy People 2020 screening rate goal of 70.5 % [26].

Fig. 3.

Up-to-date colorectal cancer screening rates stratified by race, ethnicity, insurance, age, education, and income

HCC Screening

Unlike CRC screening, which is recommended for nearly all 50- to 75-year-old individuals, HCC screening is only recommended for high-risk individuals, primarily for subgroups of patients with hepatitis B virus (HBV) infection and those with liver cirrhosis [27, 28]. The goal of HCC screening is to detect cancers at a stage when curative therapies (such as local resection or liver transplantation) can be offered. It should be noted that for HCC (unlike CRC), there is controversy regarding whether routine screening for HCC should be recommended, because of limitations of currently available studies on the topic [29, 30]. The best evidence for HCC surveillance comes from a large randomized trial among patients with HBV, demonstrating earlier stage of disease and lower mortality rate [31]. However, there is not a randomized trial among patients with cirrhosis, and these data from HBV cannot be directly extrapolated given higher risk of competing liver-related mortality and differential rates of effectiveness. Indeed, some have recommended further randomized-controlled trials prior to wide-spread recommendation for treatment because assessment of benefits and harms has not been rigorously evaluated in multiple trials [30]. In rebuttal, experts point to several cohort studies demonstrating an association between HCC surveillance and better overall survival, even after adjusting for lead time bias [32]. Patients with early HCC can achieve 5-year survival rates near 70 % with resection and liver transplantation, whereas those with advanced HCC have median survival below 1 year [33•]. As a result of such data, national organizations such as the American Association for the Study of Liver Diseases (AASLD), the National Comprehensive Cancer Network, and the Veterans Affairs Healthcare Administration recommend individuals with liver cirrhosis undergo HCC screening every 6 months [28]. At this point, we take the position that until further data are available, underserved populations should have the same access to screening as other populations as part of usual care, unless additional data demonstrating harm or futility become available.

Available data suggest that HCC screening rates are low and may be particularly low among groups who are at increased risk for poor outcomes after HCC diagnosis. A recent meta-analysis reported only 18.4 % of patients with cirrhosis had undergone HCC screening, with a range of 11 to 64 % reported [17]. Rates were lower for studies that primarily sampled patients from primary care vs. subspecialty settings (16.9 vs. 51.7 %). Several sociodemographic factors are associated with HCC screening utilization. Lower screening rates are found among patients older than 65 years, non-Caucasians, and those of low SES [14]. For example in a secondary analysis of SEER-Medicare database, Davilla et al. found the lowest rates of screening among blacks and those with lower income (Fig. 4) [35]. Prior studies have suggested HCC surveillance is highly accepted by patients; in contrast, providers have reported several barriers including insufficient knowledge of guidelines and competing clinical concerns [36]. Currently, the two most notable barriers to HCC surveillance are accurate identification of the at-risk population and providers failing to order HCC surveillance in those with known cirrhosis [37, 38]. It is also possible the controversy surrounding HCC screening implementation may be contributing to low HCC screening rates [36].

Fig. 4.

Variations in HCC screening by sex, race/ethnicity, and income, SEER-Medicare 1994–2002 [34]

Strategies for Increasing CRC and HCC Screening

Strategies for increasing GI cancer screening can be broadly characterized from a health system or public health perspective as either outreach or inreach [35, 39•]. In an outreach strategy, all individuals within a defined population of interest receive a screening promotion intervention, regardless of scheduled health-care visits. An example is mailing fecal immunochemical tests (FIT) to all individuals not up-to-date with CRC screening by a health plan or a health system. An inreach strategy seeks to take advantage of a routine healthcare visit to deliver a screening promotion intervention. An example includes patient navigation or case management to help schedule and complete a colonoscopy after referral for colonoscopy-based CRC screening. The strength of outreach strategies is that they usually can be centralized and delivered with high volume by a small screening promotion team, while the strength of inreach strategies is that they take advantage of in-person interactions and especially health-care provider recommendations to complete screening, which have consistently been shown to be associated with completion of screening tests [40•].

Strategies for CRC Screening

The profound observed disparities in incidence, mortality, stage at presentation, and stage-specific survival, as well as disparities in screening utilization for CRC, have inspired a number of randomized trials of outreach and inreach strategies to promote screening completion among underserved populations; trials from the last 3 years are summarized in Table 1. Among outreach interventions, recent work has confirmed that strategies such as mailed invitation to complete enclosed fecal occult blood tests or FITs, in conjunction with telephone follow-up, as well as mailed invitations to complete colonoscopy with telephone follow-up are highly effective for increasing screening rates; successful interventions generally included ability to provide language concordant informational materials as well as interaction with bilingual program staff [35, 41•, 42, 43, 47, 51]. Among inreach interventions, recent work suggests that providing screening education, enabling patient activation to ask for screening, and addressing barriers to screening through counseling are highly effective. Recent work builds on more remote work. For example, in reach offering choice of screening tests, as well as offering FIT or guaiac fecal occult blood tests at the time of annual flu immunizations have been shown to be effective for boosting screening among underserved populations [35, 52•, 53]. In addition to the interventions highlighted here, we also recommend the reader review the Community Guide to Preventive Services recommendations, which highlight evidence-based strategies for increasing CRC screening [54•].

Table 1.

RCTs for increasing CRC screening published in last 1 to 3 years for underserved populations

Authors	Year published	Outreach/inreach	Description/intervention Type	Underserved groups	Sample size	Key findings
Baker [41]	2014	Outreach	Community health center-based RCT of usual care and intervention group that received (1) mailed FIT with mailed reminder, low literacy instructions, and paid return postage, (2) automated telephone and text message reminders that they were due for screening, (3) automated telephone and text reminders to return the FIT, and (4) personal telephone outreach by patient navigator after 3 months. Patients had prior history of completing guaiac fecal occult blood test but were not up-to-date	Patients of community health center, primarily women (72 %), Latinos (87 %), and uninsured (77 %)	450	Combined telephone and mailed follow-up for patients who are not up-to-date on CRC screening is successful in significantly increasing screening rates for vulnerable populations
Jean-Jacques [42•]	2012	Outreach	Clinic-based RCT of mailed CRC fact sheet and FOBT kit and usual care	Adult patients of community health center ages 50 to 80 years who were not up-to-date on CRC screening	202	Mailed FOBT kits effectively promoted CRC screening among population with high poverty, limited English proficiency, and racial and ethnic diversity
Coronado [43]	2011	Outreach	Community health center clinic-based RCT of usual care, mailed FOBT card with instructions, and mailed FOBT card with instructions, telephone reminders, and home visits	Hispanic patients ages 50–79 years in Seattle-based clinic	501	High intensity intervention involving home-based education resulted in higher rates of CRC screening Differences between high and low intensity intervention arms were not statistically significant
Gupta [44]	2013	Outreach	Urban safety net health system RCT of mailed FIT outreach with telephone follow-up, mailed colonoscopy outreach with telephone follow-up, and usual care	Uninsured male and female patients ages 54–64 years in urban safety net medical assistance program	5970	Mailed outreach invitations for CRC screening resulted in significantly higher screening compared to usual care in underserved patient population Mailed FIT outreach was more effective in increasing CRC screening than colonoscopy outreach
Holt [45]	2012	Outreach	Church-based RCT of spiritually based and non-spiritual community health education on colorectal cancer with surveys at 0, 1, and 12 months	Blacks ages 50–74 years with no history of CRC, with minimum of fifth-grade reading level	285	The spiritually based intervention was not more effective than the non-spiritual intervention in increasing CRC screening No significant difference was reported between the spiritually based intervention and the non-spiritual intervention in pre-post beliefs on CRC screening
Larkey [46]	2012	Outreach	Community-based RCT of one-on-one cancer screening/prevention classes or education delivered in social support groups	Women identifying as Hispanic/Latina aged over 18 years and due for one or more of breast, cervical, or colorectal cancer	1006	Individual education vs. social support group education did not significantly affect cancer screening behavior Both interventions resulted in increased screening rates of previously non-adherent patients Group-based interventions are recommended as a more cost-effective approach to increase cancer screening in Latina women
Jerant [47]	2014	Inreach	Clinic-based RCT of interactive multimedia computer program focused on CRC screening and usual care	Patients ages 50 to 75 years at five UC FQHC sites	1164	Use of the interactive multimedia computer program led to significantly increased CRC screening knowledge, self-efficacy, readiness, test preference specificity, discussion and recommendation Interactive multimedia computer program effects did not significantly increase completion of screening within 12 months
Lasser [48]	2011	Inreach	Clinic-based RCT of patient navigation intervention or usual care.	Patients ages 52–74 who were not up-to-date on CRC screening and who spoke English, Haitian Creole, Portugese, or Spanish as primary language	465	Patient navigation significantly increased CRC screening compared to the usual care group after 1 year Patient navigation was particularly beneficial for patients whose primary language was other than English Patient navigation was more effective in white and black individuals compared to other or unknown race
Raich [49]	2012	Inreach	Safety net health-care system-based RCT of patient navigation and usual care	Safety net health-care system patients with abnormal breast, colorectal, or prostate screening test, or new diagnosis of these cancers.	993	Overall time to resolution was significantly shorter for the patient navigation group Navigation significantly shortened the time for CRC screening resolution Patients in the navigated group were more likely to complete diagnostic follow-up compared to control patients
Katz [50]	2011	Inreach	Clinic-based RCT of CRC screening information, patient activation, and barriers counseling or usual care	Patients of FQHC, 50 years and older with average risk of CRC and not up-to-date on CRC screening	284	Patient activation significantly increased CRC screening among low-income minority patients

Strategies for HCC Screening

Strategies for increasing HCC screening have undergone very limited study overall and have not been tested among underserved populations. Within the Veterans Affairs Healthcare System, a primary care-based reminder for health-care providers to order screening among individuals with cirrhosis not up-to-date resulted in an absolute 10 % increase in the proportion of Veterans completing screening. Additionally, two gastroenterology clinic-based studies successfully used standing protocols and clinical reminders to complete screening resulted in high rates of completion [55–57]. All interventions to date have been inreach, with no attempts at HCC screening outreach to date. Whether these strategies can be effective among underserved populations requires study through both quality improvement programs as well as structured research. We postulate that research and intervention programs in this area might be informed by successful strategies (such as those outlined above) that have been used among underserved populations to promote CRC screening [58, 59].

Aside from interventions to increase participation, we also recognize the need for further evidence to support the effectiveness of HCC screening. We recommend that any studies (including randomized trials) that study outcomes of HCC screening make extended efforts to enroll individuals who are currently underserved with respect to HCC screening or who are at higher risk for HCC screening, so that results can be generalizable to these groups.

Future Directions/Recommendations

CRC and HCC are major public health problems among underserved communities. Despite guidelines recommending screening for CRC, uptake remains suboptimal. For HCC, population-based data to understand whether low screening rates may be contributing to adverse HCC outcomes among underserved populations are lacking, but non-population-based data suggest that low screening rates are likely a problem. Improving screening rates will require different but potentially complimentary strategies for CRC and HCC screening, as the science behind promotion of HCC screening is still in its infancy. In Table 2, we outline key clinical and research strategies that we believe must be addressed to improve CRC and HCC outcomes among underserved populations. Overall, we opine that with focused implementation of screening strategies, and novel research, that the burden of HCC and CRC can be reduced among underserved populations.

Table 2.

Proposed clinical, population, and research strategies for improving CRC and HCC screening among underserved populations

	Clinical and population strategies	Research strategies
CRC	Develop resources to help health systems caring for underserved populations to identify and track individuals not up-to-date Implement evidence-based strategies for increasing screening, such as mailed FIT outreach, offering FIT at time of flu shots, and navigation to aid in colonoscopy completion after diagnostic or screening colonoscopy referral Ensure no cost access to screening (e.g., FIT and colonoscopy) tests and diagnostic colonoscopy	Compare impact of combinations of strategies (such as an inreach plus an outreach approach) for boosting screening Evaluate reasons for non-response to current strategies Consider testing novel approaches, such as financial incentives for increasing screening
HCC	Develop resources to measure and track HCC screening rates Ensure no cost access to HCC screening and diagnostic follow-up Educate providers on importance of HCC screening and how to identify patients with cirrhosis who might benefit	Evaluate differences in stage at presentation and stage-specific survival across race, ethnicity, income, and educational status Evaluate national screening rates, with an eye towards characterizing any disparities across populations Develop and test inreach and outreach strategies for promoting screening, building on previous work in the area of CRC screening Take special steps to enroll populations at higher risk for adverse HCC outcomes into any future prospective studies of screening to ensure generalizability to these populations