Persisting Racial Disparities in Colonoscopy

J. Racial and Ethnic Health Disparities DOI 10.1007/s40615-017-0418-1

Persisting Racial Disparities in Colonoscopy Screening of Persons with a Family History of Colorectal Cancer Meng-Han Tsai 1 & Sudha Xirasagar 1 & Piet C. de Groen 2

Received: 17 April 2017 / Revised: 19 July 2017 / Accepted: 25 July 2017 # W. Montague Cobb-NMA Health Institute 2017

Abstract With 23 and 47% higher colorectal cancer (CRC) incidence and mortality, respectively, among African Americans vs. Whites, CRC screening studies are important. Screening guidelines recommend 5-yearly colonoscopy screening of persons with a family history of CRC (first-degree relatives, FDRs), beginning at 40 years of age. For this elevated-risk group, colonoscopy screening is preferred because of the risk of more aggressive cancer that may elude early detection by other methods. African Americans with a family history of CRC are at the intersection of two elevated risk demographics, race and FDR status. This study explored racial disparities in colonoscopy screening of FDRs using 2005, 2010, and 2015 national survey data on 3220 Whites and 466 African Americans. Despite increasing colonoscopy rates among FDRs (72.3 and 62.2% in 2015 among Whites and African Americans, respectively), the 40–49 age group showed substantial racial disparities each year, persisting through 2015 (58.8, 31.7, and 35.3% lower among African Americans in 2005, 2010, and 2015, respectively). Adjusted analysis of the pooled 3-year sample showed that FDRs aged 40–49 years had one-third the colonoscopy likelihood of the 50-plus age group. African Americans without college education were 40 and 60% less likely than

* Meng-Han Tsai [email protected]

1

Department of Health Services Policy and Management, Arnold School of Public Health, University of South Carolina, Columbia, SC 29208, USA

2

Department of Medicine, Division of Gastroenterology, Hepatology and Nutrition, University of Minnesota, 406 Delaware St. SE MMC 36, Minneapolis, MN 55455, USA

Whites without college and with college education, respectively, to have had a colonoscopy. The sustained, high screening disparity, and low colonoscopy rates in the 40– 49 age group overall, call for novel approaches to reduce CRC mortality disparities, such as, patient navigation programs to reach out to younger FDRs, particularly, less educated African Americans. Keywords Colorectal cancer . African American . Colonoscopy . Family history . Cancer prevention . Health disparities

Introduction First-degree relatives of sporadic colorectal cancer (CRC) patients suffer higher rates of CRC [1]. A first-degree relative (FDR) is a biological parent, sibling, or child of persons diagnosed with CRC, and they account for 25% of incident CRC cases [1]. Sporadic CRC accounts for 70% of total CRC incidence, arising in persons without a known high-risk factor, such as inflammatory bowel disease or a family history of hereditary genetic syndromes such as hereditary nonpolyposis colorectal cancer (HNPCC) or familial multiple polyposis. The latter are high-risk sub-populations, usually placed under life-long colonoscopy surveillance. Of total CRC incidence, about 5% is contributed by the high-risk sub-population, 70% by sporadic CRC arising among the average-risk population, and 25% by the elevated-risk population consisting of FDRs of sporadic CRC cases [1]. Tracking screening compliance among the FDR population is important, given the absence of screening guidelines for this group by the US Preventive Services Task Force (USPSTF). Most insurers including the Center for Medicare and Medicaid

J. Racial and Ethnic Health Disparities

Services benchmark their coverage policies to the USPSTF guidelines. Surveys show that about 9.4% of the US population consists of FDRs with a family history of sporadic CRC [2]. The CRC risk of FDRs increases with increasingly younger age of the CRC-affected relative, and with multiple family members affected [1, 3, 4]. Outcomes of screening colonoscopy show that FDRs are diagnosed with cancer or high-risk adenomas at double the rates of the average-risk population [5–7]. Therefore, racial disparities in FDRs’ screening rates have important implications for racial disparities in CRC incidence and mortality. Screening colonoscopy is the preferred screening method for FDRs because it enables the removal of premalignant polyps that cause most CRCs, and because of the elevated CRC risk of FDRs [8, 9]. Screening colonoscopy is documented to reduce CRC incidence and mortality by 76–83% and 53–89%, respectively, among the general population [10–13]. The USPSTF acknowledges FDRs as an elevatedrisk group, but stops short of specifying a screening guideline [14]. For the general population, the USPSTF recommends 10-yearly colonoscopy screening beginning at 50 years of age, and defers to the specialty society guidelines for the elevated-risk and high-risk populations. The USPSTF requires evidence with a high level of certainty for issuing guidelines. Such evidence is not available on early screening of FDRs. This population is falling through the cracks despite the issuance of specific FDR screening guidelines by the gastroenterology societies since 2008, and reiterated in 2009 and 2017, one of them being a joint statement by the US Multi-Society Task force (USMSTF) on Colorectal Cancer, the American Cancer Society (ACS), and American College of Radiology (ACR) [8, 9, 15]. These guidelines recommend 5-yearly screening beginning at 40 years of age for persons with one family member diagnosed with CRC or advanced adenoma before 60 years of age, or with two relatives diagnosed with CRC or advanced adenomas [8, 9]. However, patients’ screening compliance is driven, not by professional society guidelines, but by each individual’s physician recommendations [9]. Physicians’ recommendations are mostly driven by insurance coverage of the service, and insurers rely on USPSTF recommendations for making coverage policy decisions. Given the focus of the USPSTF guidelines on the averagerisk population, insurers, primary care providers, and CRC researchers have mostly focused on this population. There is little clinical or research documentation on screening of FDRs. FDRs’ elevated CRC risk may be accentuated among African Americans. African Americans suffer 47% higher CRC mortality than Whites, and 22.7% higher CRC incidence [16, 17]. Despite minimal racial differences in cancer stage at diagnosis, the disproportionate mortality disparity is due to African Americans’ lower 5-year survival rates compared to Whites within each cancer stage [16, 18]. One factor driving the mortality disparity may be the higher mortality with

younger age at diagnosis within each cancer stage. African American CRC patients are, on average, about 5 years younger than Whites at diagnosis [19, 20]. When CRC occurs among FDRs, it is typically diagnosed at a younger age than among sporadic CRC patients [20]. Therefore, AfricanAmerican FDRs are at the intersection of two high-risk demographics (African-American and FDR). Therefore they are a high-priority population to be studied for potential screening gaps. Among the average-risk population, racial differences in screening have steadily narrowed since 2000, when Medicare approved screening colonoscopy coverage [16, 21]. A study of CRC screening rates in the National Health Interview Surveys (NHIS) from 2000 through 2013, showed that the screening disparity between average-risk African Americans and Whites was almost eliminated by 2013 [16]. Screening compliance among African Americans increased from 32% in 2000 to 59% in 2013, compared to 40 and 61%, respectively, among Whites. Yet, the latest CRC statistics for 2013 continue to show 47% excess CRC mortality among African Americans, stagnant since the 1990s. This calls for new avenues of policy action to address the mortality disparity [16, 17, 21, 22]. Few studies have assessed screening disparities among FDRs. The Southern Community Cohort Study reported that AfricanAmerican FDRs were 50% less likely than White FDRs to have completed colonoscopy screening in the southern states [23]. Other FDR studies have not examined screening disparities by race. They compared persons with and without a CRC family history, stratified on race [24, 25]. The most recent national study using 2007 data, reported that among African Americans aged over 50 years, those with a family history had higher screening rates than those without a family history by 5.3 percentage points, and both categories lagged behind Whites by 8–12 percentage points [25]. A California study of first-degree relatives using 2009 data reported a 23.3% lower screening rate among African Americans than Whites [24]. African Americans are documented to be less likely to be aware of family members diagnosed with CRC: even among families affected by hereditary colorectal cancer syndromes, African Americans were about a third as likely to be aware of having a CRC-affected relative compared to Whites [26]. There is no documentation of the factors driving screening disparities among the FDR population, a knowledge gap that is addressed by this study using a population-based dataset.

Methods Study Population Data from the 2005, 2010, and 2015, NHIS surveys were pooled to study racial differences in FDR screening and the


trends over time. Pooling also powered the study to explore socio-demographic factors driving the disparities. The NHIS is a cross-sectional household survey, covering a nationally representative sample of the US population, with an additional module on cancer family history and cancer screening (Family History and Cancer Screening section) fielded every 5 years. A family history of CRC is defined as an affirmative response to questions asking if their biological parent, sibling, or child had a cancer diagnosis, and affirming that the cancer type was CRC [27]. To establish respondent eligibility for early CRC screening, the joint USMSTF-ACS-ACR guideline was used [9]. Adults aged ≥ 40 years reporting a first-degree relative diagnosed with CRC were defined as FDRs and eligible for inclusion in the study. This inclusion criterion is more general than the nuanced guideline noted earlier due to lack of data on the number of affected relatives and their age at CRC diagnosis. As a result, the study sample includes some screeningineligible persons. Our inclusion criteria are consistent with the documented studies based on self-reported family history [23, 24, 28]. Measures The main outcome measure was completion of a screening colonoscopy in the previous 5 years. A secondary outcome measure was any CRC screening (colonoscopy or sigmoidoscopy in the previous 5 years, or a fecal blood test in the prior year) [8, 9]. The key dependent variables of interest were colonoscopy screening, and any CRC screening. The key independent variable of interest was race (White, African American). The co-variates studied are documented predictors of the likelihood of CRC screening among average-risk population [29]. Co-variates studied were age (40–49, 50–64, ≥ 65 years), sex, education (high school or less, any college education), insurance type (private, public including Medicare and Medicaid, uninsured), and survey year (2005, 2010, and 2015). We adjusted for marital status (currently married, not currently married), consistent with the literature [30, 31]. Statistical Analysis To account for the stratified sampling scheme, weighted analysis was carried out per the National Center for Health Statistics recommendations for pooled analysis of the 2005, 2010, and 2015 NHIS samples [27]. We studied the sample distributions for each year and used logistic regression analysis of the pooled sample to model the unadjusted and adjusted likelihood of colonoscopy screening and any screening. The associations of sex, age, race, education, insurance and marital status, and the change in screening likelihood over time (using survey year as the indicator variable) were estimated. We tested interactions of race with all other independent variables.

The only interaction that was statistically significant was the race-education interaction. To generate estimates of the combined effect of race and education, a composite variable was created to classify all study subjects into race-education categories. Such estimates enable a policy-relevant interpretation of the associations. A two-sided test of statistical significance at the 0.01 level was used, as recommended by the National Center for Health Statistics [27]. The SAS statistical software (v.9.4) was used.

Results Figure 1 shows the sample exclusions and the final analytic sample of FDRs pooled from the 2005, 2010, and 2015 surveys. Of total 58,903 respondents aged over 40 years, we excluded 54,768 persons (32,572 who reported not having a CRC family history and 22,196 non-respondents to the family history question). We also excluded 449 FDRs of ethnicities other than African American and Whites (due to small numbers). A comparison of non-respondents within the study sample showed that non-respondents were significantly more likely to be African American (17.6 vs. 12.4%), younger (mean age, 57.7 vs. 60.4 years), less educated (51.6% with high school education or less vs. 42.8%), and much less likely to have completed colonoscopy screening (16.9 vs. 40.4%) (Table not presented). The final study sample consisted of 3686 first-degree relatives of CRC-affected persons (1221, 987, and 1478 from the 2005, 2010, and 2015 surveys, respectively). Of them, 13.0% were African American. In 2015, 72.3% of Whites and 62.2% of African Americans had completed a screening colonoscopy. Sample Demographic Characteristics and Screening Status Table 1 shows the sample demographics by study year. In 2005, sample African Americans were, on average, younger than Whites, but similar on education. In 2010, African Americans were less educated than Whites, but of similar age. In 2015, sample African Americans were, on average, more educated, but of similar age. The unadjusted racial difference in education level attained statistical significance in 2010 (p < 0.01). More African Americans had a high school education or less than Whites (55.4 vs. 42.3%). Colonoscopy rates almost tripled in 2010 compared to 2005 among Whites, but increased only slightly in 2015. The increases over time were smaller among African Americans. The colonoscopy screening disparity by race widened from 2005 to 2010, and narrowed in 2015: from a 16% higher colonoscopy rate among Whites in 2005, to 23 and 14% higher in 2010 and 2015, respectively. Sigmoidoscopy and fecal blood test rates were negligible in both racial groups in all survey rounds.


Table 2 presents the screening status of the sample classified by race and age. The racial disparity was consistently high among younger FDRs aged 40–49 years, with 31.7 and 32.5% lower colonoscopy rates among African Americans in 2010 and 2015, respectively (29.2 and 45.1% among African Americans and Whites in 2015). The racial disparities in colonoscopy rates were stable among persons aged 50 years or older (Table 2).

Adjusted Analysis Table 3 shows the results of multiple regression analysis. The odds of colonoscopy were almost 6-fold and 11-fold higher in 2010 and 2015, respectively, relative to 2005 (AOR for 2010, 5.8; 95%CI, 4.7–7.0; AOR for 2015, 10.9; 95%CI, 9.1–13.2). The interaction of race with education was significant, i.e., the association of race with colonoscopy differed by educational level. Interactions of race with age, sex, health insurance, marital status, and year were not significant. Using a composite variable of race and education, we observed that African Americans with a high school education or less were 40 and 60% less likely than Whites with similar education and Whites with college education, respectively, to have completed colonoscopy (AORs, 0.6; and 0.4 respectively). Among Whites, there was no difference in colonoscopy likelihood by educational level, and both educational groups were similar to college-educated African Americans. Adults aged 40–49 years Fig. 1 Sample selection from the National Health Interview Survey (NHIS) 2005, 2010, and 2015 data—screening-eligible adults aged 40 years or more with a firstdegree relative with colorectal cancer

were about a third as likely to have screening as those aged 50 or older (AORs, 2.6, 95%CI: 2.1–3.4 and 3.3, 95%CI: 2.3– 4.6, respectively, for 50–64 years and ≥ 65 years, relative to 40–49 years). Health insurance was associated with a fourfold higher likelihood of screening relative to the uninsured (AOR 4.0, 95%CI: 2.7–5.8 and 3.7, 95%CI: 2.4–5.8 for public and private insurance, respectively). Currently married persons were more likely to have completed colonoscopy (AOR 1.5, 95% CI, 1.3–1.8). The results for any CRC screening were similar to those observed for colonoscopy screening (Table 4). African Americans with a high school education or less were about half as likely to have to have completed any CRC screening.

Discussion Our study makes a contribution by exploring racial disparities in screening of FDRs, a particularly high-risk factor for CRC among African Americans. After adjusting for sociodemographic factors known to impact colonoscopy screening, we find that African Americans with lower education (high school or less) are 40% less likely than Whites with similar education to have had a colonoscopy, and 60% less likely than Whites with college education to have completed colonoscopy. Among Whites, education makes no difference to colonoscopy rates, and college-educated African Americans are

Adults aged ≥ 40 in NHIS 2005, 2010, 2015, n=58,903 (2005, 19,618; 2010, 16,957; 2015, 22,328) Exclusions: No response to the family history question (n=22,196) No CRC family history (n=32,572)

With family history (First degree relatives) (n=4,135) 2005: n=1,346 / 2010: n=1,124 / 2015: n=1,665

Other race/ethnicity (n=449) 2005: n=125 / 2010: n=137 / 2015: n=187

Study sample of FDRs (n=3,686) 2005: n=1,221 / 2010: n=987/ 2015: n=1,478

White (n=3,220) 2005: n=1,083 2010: n=839 2015: n=1,298

African American (n=466) 2005: n=138 2010: n=148 2015: n=180

J. Racial and Ethnic Health Disparities Table 1 Demographic distribution of screening-eligible US adult respondents aged 40 years or more with a CRC family history in the 2005, 2010, and 2015 National Health Interview Surveys (n = 3686)

2005

2010

White % (n = 1083)

African American % (n = 138)

White % (n = 839)

2015 White % (n = 1298)



Sex Male

37.2

Female Age (years) 40–49 50–64 ≥ 65 Education High school or less Any college education Marital status Currently married Not currently married Insurance types Private Public Uninsured CRC screening Colonoscopy Sigmoidoscopy Fecal tests No screening

32.6

36.8

41.2

42.0a a

62.8

67.4

63.2

58.8

58.0

33.3 66.7

16.3a 38.1a 45.5a

29.7 36.2 34.1

15.9 40.9 43.3

20.3 41.2 38.5

11.8 37.7 50.5

13.3 41.1 45.6

45.3

50.7

42.3a

55.4

39.3

36.7

54.1

49.3

57.6a

43.9

60.3

63.3

52.3a

33.3

49.2a

35.8

50.2a

27.2

47.5a

66.7

50.5a

64.2

49.4a

72.8

45.1 49.9 5.1

42.0 47.8 10.1

45.2 47.3 7.2

39.9 50.0 9.5

38.5 56.5 4.9

33.9 60.0 5.6

19.9 0 0.5 79.6

16.7 0 0.7 82.6

57.1a 0a 2.5a 40.4a

43.9 0 5.4 50.7

72.3a 0.2a 0.6a 26.9a

62.2 2.2 1.1 34.4

All p < 0.01 between White and African American in 2005, 2010, and 2015. p value of 0.01 is recommended by the NCHS for statistical significance tests

a

Table 2

Colonoscopy screening completion status by age group and race (n = 3686) Noa.

2005

2010

Colonoscopy Otherb

Whites (n = 3220)

No screening

40–49 yrs. 50–64 yrs. ≥ 65 yrs African American (n = 466)

177/133/153 11.9% 413/343/489 17.9% 493/363/656 24.5% Colonoscopy No.

0.6% 0.2% 0.6% Others

87.6% 81.8% 74.9% No screening

40–49 yrs. 50–64 yrs ≥ 65 yrs

41/30/24 50/61/74 47/57/82

0 0 2.1%

95.1% 82.0% 72.3%

4.9% 18.0% 25.5%

a

Total respondents in the category in 2005, 2010, and 2015, respectively

b

Others include fecal occult blood test and flexible sigmoidoscopy

yrs. years.

2015

Colonoscopy Other

No screening

39.1% 60.9% 60.1% Colonoscopy

1.5% 1.8% 3.6% Others

59.4% 37.3% 36.4% No screening

26.7% 49.2% 47.3%

3.3% 6.6% 5.3%

70.0% 44.3% 47.4%

Colonoscopy Others

45.1% 74.9% 76.7% Colonoscopy

0.7% 0.6% 1.1% Others

29.2% 60.8% 73.2%

4.2% 4.1% 2.4%

No screening 54.3% 24.5% 22.3% No screening 66.7% 35.1% 24.4%

J. Racial and Ethnic Health Disparities Table 3 Regression analysis results showing the factors associated with colonoscopy screening (vs. no screening) among persons aged ≥ 40 years with a family history of CRC (n = 3665)a

Sex Male Femalec Age (years) 40–49c 50–64 ≥ 65 Race-education status White and high school or less White and any college educationc African American and high school or lessd African American and any college education Marital status Currently married Not currently marriedc Insurance types Private Public Uninsuredc Survey year 2005c 2010 2015

Unadjusted model

Adjusted model b

OR(95%CI)

OR(95%CI)

1.0(0.8–1.1) –

0.8(0.7–1.0) –

–

–

2.6(2.1–3.2) 3.1(2.6–3.9)

2.6(2.1–3.4) 3.3(2.3–4.6)

0.7(0.6–0.8) – 0.4(0.3–0.6)

0.7(0.6–0.8) – 0.4(0.3–0.6)

0.9(0.7–1.1)

0.8(0.6–1.1)

1.3(1.2–1.5) –

1.5(1.3–1.8) –

3.8(2.7–5.4)

4.0(2.7–5.8)

4.9(3.4–6.9) –

3.7(2.4–5.8) –

– 5.0(4.2–6.1)

– 5.8(4.7–7.0)

10.1(8.5–12.1)

10.9(9.1–13.2)

a

Education data were missing for 21 respondents. Income was not significant, therefore it is excluded from the models

b

Adjusted for gender, age, marital status, insurance, and survey year

c

Reference group

d

Changing the reference group to Whites with high school or lower education provided an AOR of 0.6; 95% CI: 0.4–0.9 for African Americans with high school or less education

similar to Whites of any educational level in having a colonoscopy. The findings should be viewed in context of a significant uptick in colonoscopy rates among FDRs in 2015 among both racial groups aged over 50 years (76.0 and 67.3% among Whites and African Americans, respectively, in 2015), a documented finding by other authors [32]. Our adjusted analyses show 11-fold higher colonoscopy odds among FDRs in 2015 relative to 2005, and 6-fold higher odds in 2010 relative to 2005. Despite much increased screening rates among FDRs, the screening disparity between Whites and African Americans in 2015 has been sustained at the same level as in the prior decade (interaction between race and year is not significant). A previous study documented a significant narrowing of the racial disparity in colonoscopy rates of the average-risk population aged 50 years or older in 2008 (a 12-percentage point difference in 2005, narrowing to a 4.6 point difference in 2008) [21]. Another study, which examined the outpatient

population of federally qualified community health centers in 12 southern states, reported that African-American FDRs aged over 50 years had lower colonoscopy rates than Whites (33 vs. 47%). The same study also reported a larger difference among those aged below 50 years, 27.3 vs. 43.1%, respectively [23, 28]. In interpreting these studies, it should be noted that the unadjusted colonoscopy rates are driven by the composition of the surveyed samples by education and other confounding factors that may inflate or deflate the disparities. The role of sample composition in unadjusted rates is illustrated in our study. Our study showed significant variations in the sample composition by age, education, and marital status in the three survey years (important variables driving the colonoscopy disparities between Whites and African Americans in adjusted analysis). Previous studies were under-powered to identify the socio-demographic factors driving the screening gap among FDRs or limited to a certain regions. Our study leveraged pooled samples from the 2005, 2010, and 2015 NHIS surveys to document the role of education in racial

J. Racial and Ethnic Health Disparities Table 4 Regression analysis results showing the factors associated with any CRC screening (vs. no screening) among persons aged ≥ 40 years with a family history of CRC (n = 3665)a

Sex Male Female c Age (years) 40–49c 50–64 ≥ 65 Race/education sub-groups White and high school or less White and any college educationc African American and high school or lessd African American and any college education Marital status Currently married Not currently married c Insurance types Private Public Uninsured c Survey year 2005 c 2010 2015 a

Unadjusted model

Adjusted model b

OR(95%CI)

OR(95%CI)

1.0(0.9–1.1) –

0.9(0.7–1.0) –

–

–

2.6(2.1–3.2) 3.2(2.6–3.9)

2.7(2.1–3.4) 3.4(2.4–4.8)

0.7(0.6–0.8) – 0.5(0.4–0.6)

0.7(0.6–0.8) – 0.5(0.3–0.7)

1.0(0.7–1.2)

1.0(0.7–1.3)

1.3(1.2–1.5) –

1.5(1.3–1.8) –

3.6(2.6–5.1)

3.9(2.6–5.6)

4.7(3.4–6.7) –

3.6(2.3–5.5) –

– 5.5(4.5–6.6)

– 6.3(5.2–7.7)

10.4(8.7–12.5)

11.2(9.3–13.5)

Education data were missing for 21 respondents. Income was not significant, and is excluded from the models

b

Adjusted for gender, age, marital status, insurance, and survey year

c

Reference group

d

Changing the reference group to Whites with high school or lower education provided an AOR of 0.7; 95% CI: 0.5–1.0 for African Americans with high school or less education

disparities in screening, and the findings generalize to the US population. The repeated finding of lower colonoscopy rates among African-American FDRs relative to Whites in the literature is consistent with the documented racial difference in awareness of the heightened CRC risk of relatives [26]. Another important study finding is the lower colonoscopy likelihood among younger FDRs of both races, consistent with earlier studies [24, 28, 30]. The Southern Community Cohort Study reported 38% higher colonoscopy odds among White relatives to African Americans within the 40–49 age group, similar to a 2002 survey of Maryland residents [28, 33]. Our study adds to the evidence by showing persistent racial disparity of the same magnitude even in the most recent year of data, 2015, and demonstrating it at the national level. This finding is consistent with another study that used the 2009 California Health Interview Survey data, albeit based on a small sample of 107 African Americans in one state [24]. Both the California study and our study findings are concerning, because the joint professional society-ACS

guidelines recommend even average-risk African Americans to commence CRC screening at 45 years of age [8, 9]. Low screening among younger FDRs, both African Americans and Whites, is attributable to the lack of public health messaging directed to the younger-than-conventional screening age group, combined with widespread anchoring of primary care providers to the USPSTF guidelines that are limited to the average-risk population [14]. The USPSTF defers to the specialty societies to recommend guidelines for elevated-risk population. This nuance appears to be lost in translation into practice. Indeed, major insurers and primary care quality monitoring agencies (e.g., NCQA) benchmark their policies and indictors to the USPSTF guidelines. Addressing the younger FDR sub-population may be particularly important for reducing racial CRC disparities. This is because of disproportionately higher CRC incidence among younger African Americans, and higher mortality among CRC patients diagnosed before the age of 60 years [18, 34]. Further, younger age at CRC diagnosis is more likely to be


associated with a family history of CRC [1, 35]. And survival of younger CRC patients is worse among African Americans [36]. All these factors place African Americans at a higher risk of family history-driven CRC mortality than Whites. Policy interventions to increase the salience and accuracy of public health messaging regarding FDR screening will be a major step forward to reduce CRC mortality disparities affecting African Americans. The magnitude of disparity, 50% lower screening odds for less educated African Americans (even after adjusting for health insurance), indicates an emphatic need for policy interventions to address less educated African Americans. One policy implication is the need to prioritize the identification and screening navigation of African-American FDRs of CRC patients, perhaps starting with recently diagnosed CRC patients, encouraging engagement of their first-degree relatives in CRC screening conversations. An active role of healthcare providers should be included as part of the FDR navigation program, because personal physicians’ recommendation is a key factor in patient decisions on colonoscopy [24, 37]. In California, 52% of screening-eligible FDRs who were aware of their CRC family history had not completed a colonoscopy because of not receiving their physician’s recommendation, and those who were recommended screening were 89% more likely to have completed colonoscopy than those who were not recommended to undergo screening [24]. In other studies, it is noted that racial disparities in CRC screening of the average-risk population are significantly driven by racial differences in receiving physician recommendations [38]. Other qualitative and interventional studies also support an active role of physicians in screening expansion efforts [39–43]. Another policy-relevant finding, consistent with the published literature, is the role of health insurance [43]. In our study, health insurance was associated with four-fold higher odds of colonoscopy. Higher colonoscopy odds for public insurance relative to private insurance may be reflecting the variations in private insurer policies regarding covering colonoscopy among younger FDRs. Other authors have reported that health insurance is one major factor driving lower screening rates among African Americans [32]. Our finding regarding higher colonoscopy likelihood among currently married persons is consistent with the documented literature [44]. The study had some limitations. Family history data were self-reported, which may have disproportionately excluded screening-eligible African American respondents. Lower rates of awareness of a CRC family history among African Americans are documented [26, 45]. This may bias our results. Our study excluded 46.3% of African American respondents (vs. 33.3% of Whites), because of either lack of awareness of a CRC family history or non-response to the question. Univariate comparisons showed that respondents were significantly different from non-respondents on race (non-respondents were disproportionately African Americans), age, and

education (non-respondents were younger and less educated, on average). Given that our adjusted analysis show that the latter two factors are significant drivers of screening rates, the sample composition may have produced underestimates of the effect sizes for age and education. Potential recall bias may favor under-reporting of care procedures, although studies suggest that colonoscopy recall is likely to be accurate [46–48]. Another limitation is an assumption that all FDRs aged over 40 years were screening-eligible (consistent with the documented literature) [23, 24, 28]. We had no data on the age at diagnosis and number of CRC-affected relative(s). Persons aged between 40 and 50 years of age are screening-eligible if the CRC-affected relative was diagnosed before 60 years of age, or if they had at least two CRC-affected relatives. Mitigating this bias is the fact that African Americans are diagnosed with CRC about 5 years earlier than Whites. Therefore, on this count also, our study may under-estimate the racial disparity because the sample may have included disproportionately more screening-ineligible Whites than African Americans.

Conclusion The study leveraged the most current population-based data on CRC screening, the 2015 National Health Interview Survey. Given the rapid increases in colonoscopy screening in the US population in recent years, there has been much optimism that racial disparities in CRC mortality will be rapidly mitigated. However, our study finds that FDR screening among African Americans has not kept pace, despite their elevated CRC risk. Despite 11-fold higher colonoscopy likelihood of FDRs in 2015 relative to 2005, racial disparities in screening are persisting at levels similar to 2005 and 2010. Further, the racial disparity is driven largely by lower educational levels among African Americans and younger age. The findings suggest the need for a multi-pronged approach to reduce CRC racial disparities: outreach and navigation of African Americans diagnosed with CRC to encourage their first-degree relatives to seek CRC screening advice from their physicians. The findings also call for targeted CRC education of less educated African Americans. The impact of such programs could be reinforced by professional society efforts to increase primary care physicians’ awareness regarding FDRs’ special screening needs. Provider-initiated conversations to seek information on potential family history of CRC, and to educate first-degree relatives about the need and timing of screening are essential to increase FDR screening rates among the primary care patient population [49]. FDRs constitute about 10% of the US adult population but contribute 25% of incident CRC cases. Prioritizing them for screening expansion efforts may help address CRC racial disparities.

J. Racial and Ethnic Health Disparities Compliance with Ethical Standards Conflict of Interest Authors Meng-Han Tsai, Sudha Xirasagar, and Piet C. de Groen declare that they have no conflict of interest.

16.

Ethical Approval This article does not contain any studies with human participants or animals performed by any of the authors. 17.

References 1. 2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

Jasperson KW, Tuohy TM, Neklason DW, Burt RW. Hereditary and familial colon cancer. Gastroenterology. 2010;138(6):2044–58. Fuchs CS, Giovannucci EL, Colditz GA, Hunter DJ, Speizer FE, Willett WC. A prospective study of family history and the risk of colorectal cancer. N Engl J Med. 1994;331(25):1669–74. Ng SC, Lau JY, Chan FK, Suen BY, Leung WK, Tse YK, et al. Increased risk of advanced neoplasms among asymptomatic siblings of patients with colorectal cancer. Gastroenterology. 2012;144(3):544–50. Johns LE, Houlston RS. A systematic review and meta-analysis of familial colorectal cancer risk. Am J Gastroenterol. 2001;96(10): 2992–3003. Imperiale TF, Ransohoff DF. Risk for colorectal cancer in persons with a family history of adenomatous polyps: a systematic review. Ann Intern Med. 2012;156(10):703–9. Cottet V, Pariente A, Nalet B, Lafon J, Milan C, Olschwang S, et al. Colonoscopic screening of first-degree relatives of patients with large adenomas: increased risk of colorectal tumors. Gastroenterology. 2007;133(4):1086–92. Butterworth AS, Higgins JPT, Pharoah P. Relative and absolute risk of colorectal cancer for individuals with a family history: a metaanalysis. Eur J Cancer. 2006;42(2):216–27. Levin B, Lieberman DA, McFarland B, Andrews KS, Brooks D, Bond J, et al. Screening and surveillance for the early detection of colorectal cancer and adenomatous polyps, 2008: a joint guideline from the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and the American College of Radiology. Gastroenterology. 2008;134(5):1570–95. Rex DK, Johnson DA, Anderson JC, Schoenfeld PS, Burke CA, Inadomi JM, et al. American College of Gastroenterology guidelines for colorectal cancer screening 2009 [corrected]. Am J Gastroenterol. 2009;104(3):739–50. Winawer SJ, Zauber AG, Ho MN, O’Brien MJ, Gottlieb LS, Sternberg SS, et al. Prevention of colorectal cancer by colonoscopic polypectomy. The National Polyp Study Workgroup. N Engl J Med. 1993;329(27):1977–81. Xirasagar S, Li YJ, Hurley TG, Tsai MH, Hardin JW, Hurley DM, et al. Colorectal cancer prevention by an optimized colonoscopy protocol in routine practice. Int J Cancer. 2015;136(6):E731–42. Lieberman DA, Weiss DG, Bond JH, Ahnen DJ. Garewal H, Chejfec G. Use of colonoscopy to screen asymptomatic adults for colorectal cancer. Veterans Affairs Cooperative Study Group 380. N Engl J Med. 2000;343(3):162–8. Zauber AG, Winawer SJ, O’Brien MJ, Lansdorp-Vogelaar I, van Ballegooijen M, Hankey BF, et al. Colonoscopic polypectomy and long-term prevention of colorectal-cancer deaths. N Engl J Med. 2012;366(8):687–96. US Preventive Services Task Force (USPSTF). Screening for Colorectal Cancer: US Preventive Services Task Force Recommendation Statement. JAMA. 2016;315(23):2564–75. Rex DK, Boland CR, Dominitz JA, Giardiello FM, Johnson DA, Kaltenbach T, et al. Colorectal cancer screening: recommendations

18.

19. 20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

for physicians and patients from the U.S. multi-society task force on colorectal cancer. Gastroenterology. 2017;153(1):307–23. American Cancer Society (ACS). Cancer facts & figures for African Americans 2016-2018. Atlanta: American Cancer Society. 2016. https://www.cancer.org/content/dam/cancer-org/research/cancer-factsand-statistics/cancer-facts-and-figures-for-african-americans/cancerfacts-and-figures-for-african-americans-2016-2018.pdf. Accessed 1 April 2017. Howlader N, Noone AM, Krapcho M, Miller D, Bishop K, Altekruse SF, et al., 2015. SEER cancer statistics review, 19752012. http://seer.Cancer.Gov/csr/1975_2012/, based on November 2014 SEER data submission, posted to the SEER web site Bethesda, MD: National Cancer Institute, 2015. Jemal A, Simard EP, Dorell C, Noone AM, Markowitz LE, Kohler B, et al. Annual Report to the Nation on the Status of Cancer, 1975-2009, featuring the burden and trends in human papillomavirus(HPV)-associated cancers and HPV vaccination coverage levels. J Natl Cancer Inst. 2013;105(3):175–201. Karami S, Young HA, Henson DE. Earlier age at diagnosis: another dimension in cancer disparity? Cancer Detect Prev. 2007;31(1):29–34. Kerber RA, Neklason DW, Samowitz WS, Burt RW. Frequency of familial colon cancer and hereditary nonpolyposis colorectal cancer (Lynch syndrome) in a large population database. Familial Cancer. 2005;4(3):239–44. Klabunde CN, Cronin KA, Breen N, Waldron WR, Ambs AH, Nadel MR. Trends in colorectal cancer test use among vulnerable populations in the U.S. Cancer Epidemiol Biomark Prev. 2011;20(8):1611–21. Jemal A, Ward EM, Johnson CJ, Cronin KA, Ma J, Ryerson B, et al. Annual Report to the Nation on the Status of Cancer, 1975-2014, Featuring survival. J Natl Cancer Inst. 2017. doi:10.1093/jnci/djx030. Murff HJ, Peterson NB, Fowke JH, Hargreaves M, Signorello LB, Dittus RS, et al. Colonoscopy screening in African Americans and Whites with affected first-degree relatives. Arch Intern Med. 2008;168(6):625–31. Almario CV, May FP, Ponce NA, Spiegel BM. Racial and ethnic disparities in colonoscopic examination of individuals with a family history of colorectal cancer. Clin Gastroenterol Hepatol. 2015;13(8): 1487–95. Laiyemo AO, Thompson N, Williams CD, Idowu KA, Bull-Henry K, Sherif ZA, et al. Race and colorectal cancer screening compliance among persons with a family history of cancer. World J Gastrointest Endosc. 2015;7(18):1300–5. Kupfer SS, McCaffrey S, Kim KE. Racial and gender disparities in hereditary colorectal cancer risk assessment: the role of family history. J Cancer Educ. 2006;21(1 Suppl):S32–6. National Center for Health Statistics (NCHS), Division of Health Interview Statistics, Centers for Disease Control and Prevention. National Health Interview Survey Public Use Data File 2010. Hyattsville: National Center for Health Statistics; 2011. Murff HJ, Peterson NB, Greevy RA, Shrubsole MJ, Zheng W. Early initiation of colorectal cancer screening in individuals with affected first-degree relatives. J Gen Intern Med. 2007;22(1):121–6. Shapiro JA, Klabunde CN, Thompson TD, Nadel MR, Seeff LC, White A. Patterns of colorectal cancer test use, including CT colonography, in the 2010 National Health Interview Survey. Cancer Epidemiol Biomark Prev. 2012;21(6):895–904. James TM, Greiner KA, Ellerbeck EF, Feng C, Ahluwalia JS. Disparities in colorectal cancer screening: a guideline-based analysis of adherence. Ethn Dis. 2006;16(1):228–33. Longacre AV, Cramer LD, Gross CP. Screening colonoscopy use among individuals at higher colorectal cancer risk. J Clin Gastroenterol. 2006;40(6):490–6. White A, Thompson TD, White MC, Sabatino SA, de Moor J, Doria-Rose PV, et al. Cancer screening test use—United States, 2015. MMWR Morb Mortal Wkly Rep. 2017 Mar 3;66(8):201–6.

J. Racial and Ethnic Health Disparities 33.

34. 35.

36.

37.

38.

39.

40.

41.

Griffith KA, McGuire DB, Royak-Schaler R, Plowden KO, Steinberger EK. Influence of family history and preventive health behaviors on colorectal cancer screening in African Americans. Cancer. 2008;113(2):276–85. Siegel R, Desantis C, Jemal A. Colorectal cancer statistics, 2014. CA Cancer J Clin. 2014;64(2):104–17. Plath J, Siebenhofer A, Koné I, Hechtner M, Schulz-Rothe S, Beyer M, et al. Frequency of a positive family history of colorectal cancer in general practice: a cross-sectional study. Fam Pract. 2017;34(1):30–5. Holowatyj AN, Ruterbusch JJ, Rozek LS, Cote ML, Stoffel EM. Racial/ethnic disparities in survival among patients with youngonset colorectal cancer. J Clin Oncol. 2016;34(18):2148–56. May FP, Almario CV, Ponce N, Spiegel BM. Racial minorities are more likely than Whites to report lack of provider recommendation for colon cancer screening. Am J Gastroenterol. 2015;110(10): 1388–94. Coleman Wallace DA, Baltrus PT, Wallace TC, Blumenthal DS, Rust GS. Black white disparities in receiving a physician recommendation for colorectal cancer screening and reasons for not undergoing screening. J Health Care Poor Underserved. 2013;24(3): 1115–24. May FP, Whitman CB, Varlyguina K, Bromley EG, Spiegel BM. Addressing low colorectal cancer screening in African Americans: using focus groups to inform the development of effective interventions. J Cancer Educ. 2016;31(3):567–74. Jackson CS, Oman M, Patel AM, Vega KJ. Health disparities in colorectal cancer among racial and ethnic minorities in the United States. J Gastrointest Oncol. 2016;7(Suppl 1):S32–43. Horne HN, Phelan-Emrick DF, Pollack CE, Markakis D, Wenzel J, Ahmed S, et al. Effect of patient navigation on colorectal cancer

screening in a community-based randomized controlled trial of urban African American adults. Cancer Causes Control. 2015;26(2): 239–46. 42. Pelto DJ, Sly JR, Winkel G, Redd WH, Thompson HS, Itzkowitz SH, et al. Predicting colonoscopy completion among African American and Latino/a participants in a patient navigation program. J Racial Ethn Health Disparities. 2015;2(1):101–11. 43. Zlot AI, Silvey K, Newell N, Coates RJ, Leman R. Family history of colorectal cancer: clinicians’ preventive recommendations and patient behavior. Prev Chronic Dis. 2012;9:E21. 44. El-Haddad B, Dong F, Kallail KJ, Hine RB, Ablah E. Association of marital status and colorectal cancer screening participation in the USA. Color Dis. 2015;17(5):O108–14. 45. Mitchell RJ, Brewster D, Campbell H, Porteous ME, Wyllie AH, Bird CC, et al. Accuracy of reporting of family history of colorectal cancer. Gut. 2004;53(2):291–5. 46. Partin MR, Grill J, Noorbaloochi S, Powell AA, Burgess DJ, Vernon SW, et al. Validation of self-reported colorectal cancer screening behavior from a mixed-mode survey of veterans. Cancer Epidemiol Biomark Prev. 2008;17(4):768–76. 47. Khoja S, McGregor SE, Hilsden RJ. Validation of self-reported history of colorectal cancer screening. Can Fam Physician. 2007;53(7):1192–7. 48. Madlensky L, McLaughlin J, Goel V. A comparison of self-reported colorectal cancer screening with medical records. Cancer Epidemiol Biomark Prev. 2003;12(7):656–9. 49. Katz ML, Fisher JL, Fleming K, Paskett ED. Patient activation increases colorectal cancer screening rates: a randomized trial among low-income minority patients. Cancer Epidemiol Biomarkers Prev. 2012;21(1):45–52.