Colorectal cancer screening: low health literacy and limited English proficiency among Asians and Whites in California.

The authors examined the relationship between low health literacy (LHL), limited English proficiency (LEP), and meeting current U.S. Preventive Service Task Force colorectal cancer (CRC) screening guidelines for Asians and Whites in California. For 1,478 Asian and 14,410 White respondents 50-75 years of age in the 2007 California Health Interview Survey, the authors examined meeting CRC screening guidelines using multivariable logistic models by LEP and LHL separately and in combination. Analyses were run with the full sample, then separately for Whites and Asians controlling for demographics and insurance. For those with LEP, patient-provider language concordance and CRC screening was examined. Overall, respondents with LEP and LHL were the least likely to meet CRC screening guidelines (36%) followed by LEP-only (45%), LHL-only (51%), and those with neither LHL nor LEP (59%), a hierarchy that remained significant in multivariable models. For Whites, LHL-only was associated with screening, whereas LEP-only and LEP and LHL were significant for Asians. Having a language concordant provider was not significantly associated with CRC screening among those with LEP. Health literacy is associated with CRC screening, but English proficiency is also critical to consider. Asians with both LEP and LHL appear particularly vulnerable to cancer screening disparities.

Cancer is a leading cause of death in Asian Americans (Centers for Disease Control and Prevention, 2008). This is an increasing public health problem given that the Asian American population has almost doubled in the past 10 years and continues to grow largely because of immigration patterns (U.S. Census Bureau, 2002, 2009). Immigration to the United States is associated with increasing cancer risk for Asians (Gomez et al., 2010), indicating that cancer morbidity and mortality are likely to rise in this group. Colorectal cancer (CRC), a largely preventable cancer, is the second-leading cause of cancer for Asian American/Pacific Islander men and the third among Asian American/Pacific Islander women (McCracken et al., 2007). CRC is highly treatable with early detection, which can significantly reduce morbidity and mortality. Yet, Asian Americans have some of the lowest CRC screening rates in the United States (Etzioni et al., 2004; Wong, Gildengorin, Nguyen, & Mock, 2005).

CRC screening places particularly high demands on individuals, including scheduling beyond a regular doctors' appointment, independent test preparation, and/or complex completion instructions (Arnold et al., 2012; Vernon & Meissner, 2008). Screening rates may thus be strongly affected by low health literacy (LHL). Although the relationship between cancer screening and LHL specifically among Asian Americans is not well studied, LHL has been associated with less cancer knowledge, negative attitudes toward cancer screening, lower self-efficacy, and less likelihood of completing screening in other racial/ethnic groups (Bennett et al., 1998; Berkman, Sheridan, Donahue, Halpern, & Crotty, 2011; Davis, Williams, Marin, Parker, & Glass, 2002; Davis et al., 2001; Dolan et al., 2004; Dwyer, Mulvaney, Dietrich, & Rothman, 2007; Lindau et al., 2002; Miller, Brownlee, McCoy, & Pingone, 2007; Peterson, Guerra, Dominguez, & Shea, 2005).

Individuals with limited English proficiency (LEP) face additional barriers in understanding and accessing health information and services (Jacobs, Karavolos, Rathouz, Ferris, & Powell, 2005; Tu et al., 2008). Although LHL and LEP are related, they represent distinct health communication barriers (Leyva, Sharif, & Ozuah, 2005; Sudore et al., 2009). As one study of health literacy among LEP Spanish speakers concluded, one should not assume that “Spanish-speaking Latino parents who are comfortable speaking English will understand a prescription label written in English, or that Latino parents who speak Spanish will understand drug information written in Spanish” (Leyva et al., 2005, p. 56). Even in a state such as California, which has strong and innovative language access policies (Perkins & Youdleman, 2008; Snowden, Masland, Peng, Wei-Mien Lou, & Wallace, 2011), significant gaps exist by English proficiency in health outcomes (Kim et al., 2011), cancer screening (Johnson-Kozlow, Roussos, Rovniak, & Hovell, 2009) and access to linguistically appropriate health information (Hadler, Chen, Gonzalez, & Roby, 2013; Tanjasiri et al., 2004).

LEP is common in many Asian American populations. For example, approximately 60% of Chinese, Korean, and Vietnamese speakers have LEP (U.S. Census Bureau, 2003). LEP may escalate the health-related consequences of LHL generally and in Asian populations specifically. Previous work has found that individuals with both LEP and LHL are at particular risk for poor health (Sentell & Braun, 2012). The relationships among cancer screening, LEP, and LHL are unknown.

Patient-provider language concordance has been found to predict both access to health care and health status, without fully mitigating all language-based health disparities (August, Nguyen, Ngo-Metzger, Sorkin, 2011; Wilson, Chen, Grumbach, Wang, & Fernandez, 2005). Findings regarding language concordance and CRC screening specifically are mixed. Eamranond and colleagues found that LEP individuals in patient-provider language-concordant relationships had higher rates of CRC screening compared with individuals in language-discordant relationships (Eamranond, Davis, Phillips, & Wee, 2011). However, Linsky and colleagues found that LEP individuals with language concordant providers had lower screening prevalence than those with a language-discordant provider (Linsky, McIntosh, Cabral, & Kazis, 2011). The relationships among language, health literacy, and provider language concordance are similarly complex and understudied (Sudore et al., 2009).

Particularly among Asian Americans who have notable cancer screening and mortality disparities, it is important to consider the distinct relationships between LHL and LEP in CRC screening. These findings could play a role in designing effective interventions, along with furthering our understanding of distinct predictors of CRC screening. The purpose of this study was to examine the relationships among LHL, LEP, and CRC screening for Asian Americans and Whites. For those with LEP, we also examined the role of patient-provider language concordance to investigate whether this factor might play a role in mitigating CRC screening disparities.

Method

The 2007 California Health Interview Survey (CHIS; CHIS, 2010a) public-use data file was used in this analysis. The CHIS is a random digit dialing telephone survey administered by the University of California, Los Angeles, Center for Health Policy Research. The CHIS is a representative sample of California, home to 4 out of every 10 Asian Americans in the United States (U.S. Census Bureau, 2002). The CHIS specifically oversamples underrepresented ethnic/racial groups, including Asians (CHIS, 2010b). Along with Spanish, the 2007 CHIS was administered in Mandarin, Cantonese, Korean, and Vietnamese, allowing participation by Asian American individuals with LEP (CHIS, 2010b).

CRC Screening

In the 2007 CHIS, compliance with colorectal screening guidelines was obtained for all participants 50–75 years of age based on a series of cancer screening methods (i.e., having a fecal occult blood test in the past year, a flexible sigmidoscopy or double-contrast barium enema in the past 5 years, or a colonoscopy in the past 10 years), per the U.S. Preventive Services Task Force's guidelines (U.S. Preventive Services Task Force, 2003, 2008, 2009). Meeting colorectal cancer screening guidelines was defined as yes for this variable if timing of the test was as recommended.

Respondents who reported a CRC test were asked whether they completed the test as a result of a problem. Individuals who took the test for diagnostic purposes (n = 3,425) were excluded from this study because of the concern that their utilization may be subject to different factors.

Race/Ethnicity

Race/ethnicity was self-reported and defined as Asian (including Chinese, Filipino, Japanese, Korean, Vietnamese, and other Asian American) or White.

LHL

We assessed health literacy using two questions:

“When you get written information at a doctor's office, would you say that it is very easy, somewhat easy, somewhat difficult, or very difficult to understand?”

“When you read the instructions on a prescription bottle, would you say that it is very easy, somewhat easy, somewhat difficult, or very difficult to understand?”

LHL was defined as responding that (a) written information at the doctor's office is “somewhat” or “very difficult” to understand or (b) instructions on a prescription bottle are “somewhat difficult” or “very difficult” to understand (Health Research for Action, 2009; Sentell, 2012; Sentell & Braun, 2012). In these two questions, there was no indication that the language of written materials or prescription drugs had to be English. Thus, a non-English speaker could be answering this question regarding in-language materials. Individuals who reported both not getting written information and not using prescription medicine (n = 166) were excluded because their health literacy could not be determined.

LEP

Respondents who spoke any languages at home besides English were asked, “Since you speak a language other than English at home, we are interested in your own opinion of how well you speak English. Would you say you speak English very well, well, not well, not at all.” LEP is defined as self-reporting speaking English “not well” and “not at all.” This LEP variable is considered a superior measure for language barriers in health (Flores, Abreu, & Tomany-Korman, 2005), has strong face validity, and is standard in many studies (Cordasco, Ponce, Gatchell, Traudt, & Escarce, 2011; Ponce, Hays, & Cunningham, 2006).

LHL and LEP

On the basis of previous research that found distinct relationships with health status for combinations of LHL and LEP (Sentell & Braun, 2012), we compared four possible combinations of LEP and LHL to CRC screening: (1) individuals with both LEP and LHL, (2) individuals with LEP-only, (3) individuals with LHL only, and (4) individuals with neither LEP nor LHL. Group 4, those with neither LEP nor LHL, served as the reference group.

Sample

As described earlier, we included only individuals who were White or Asian, 50–75 years of age, who did not take a CRC test for a diagnostic purpose, and who were not missing health literacy assessment. The final, unweighted sample size was 15,888 individuals, including 14,410 Whites and 1,478 Asians.

Statistical Methods

All data were analyzed in Stata 12 (Stata Corp., 2011) using appropriate weighting methods to both correct for the complex sample design and to provide population-level estimates (CHIS, 2011). We first compared descriptive statistics for Whites and Asians overall using Chi-square analyses.

We then considered meeting CRC screening guidelines by the four LHL and LEP combinations using chi-square and then multivariable analyses. We performed these analyses overall with the full sample, then separately for Whites and Asians. In these subgroup analyses, we dropped any LEP or LHL combination group with a sample size <45 because these lacked sufficient (<79%) power to find a 20% screening difference in CRC screening for that group compared with the reference group (neither LEP nor LHL). The 20% difference (specifically 60% vs. 40%) was based on previous research of CRC screening by English proficiency (Diaz et al., 2102) and among Asians (Lee, Lundquist, Ju, Luo, & Townsend, 2011).

In multivariable models, control variables were age (continuous 50–75 years), sex, education (less than high school, high school graduate, college graduate, more than college), poverty (≤100% of the federal poverty level vs. not), living in a rural area (vs. not), current insurance (vs. none), being born in the United States (vs. elsewhere), and marital status (married vs. other). These variables are associated with health literacy and self-reported health status in other studies (Berkman et al., 2011). Age is presented dichotomously as middle age (50–64 years) and older age (65–75 years) in the descriptive tables for ease of interpretation.

For those with LEP, we also considered language concordance. The language concordance variable was defined based on previous work with the CHIS (August et al., 2011). The two groups possible for this variable were (a) other-language concordant (patients and providers spoke the same non-English language) or (b) language discordant (patient did not share a language with their provider).

To test the importance of the language concordance variable to our study findings, we first considered whether language concordance had a significant relationship with CRC screening in bivariate models for those with LEP (overall, Whites, and Asians). Only groups with a significant relationship for language concordance and CRC screening in bivariate models would include this variable in final multivariable models to investigate whether language concordance affected the relationship between LEP and/ or LHL to CRC screening.

Results

Descriptive Results

Table 1 shows the demographic and CRC screening comparisons of Asians and Whites, revealing significantly lower CRC screening for Asians (51.1%) compared with Whites (57.7%). Distinct relationships with health communication variables also were seen. Overall, 21.2% Asians reported LHL compared with 11.2% of Whites. For LEP, 33.5% of Asians reported this, compared with 1% of Whites. Considered together, 13.8% of Asians reported both LEP and LHL, compared with <1% of Whites. Asians and Whites also varied significantly for all control variables, including education, age, U.S. nativity, and insurance.

Table 1.

Demographics in the 2007 California Health Interview Survey overall and for Whites and Asians for those 50–75 years of age and eligible for colorectal cancer screening

	Total (15,888a) %	White (14,410a) %	Asian (1,478a) %	Whites vs. Asians p
Met colorectal cancer screening	56.6	57.7	51.1	<.001
LHL	12.8	11.2	21.2	<.001
LEP	6.3	1.0	33.5	<.001
LEP and LHL combinations
LEP + LHL	2.4	<1	13.8	<.001
LEP-only	3.9	<1	19.7
LHL-only	10.5	11.1	7.5
Neither LEP or LHL	83.2	88.1	59.1
Education				<.001
Less than high school	6.8	5.8	11.6
High school graduate	48.1	50.6	35.4
College graduate	26.4	25.0	36.7
More than college degree	18.2	18.6	16.3
Age group				<.001
50–64 years	74.1	75.4	68.0
65–75 years	25.9	24.6	32.0
Control variables
Not born in the United States	21.7	8.7	87.1	<.001
Female	50.9	56.2	54.3	.03
≤ 100% federal poverty level	6.0	4.2	14.9	<.001
Rural	14.8	17.0	3.3	<.001
Insured	91.4	92.2	87.5	<.001
Married	71.4	69.4	81.7	<.001
Only those with LEP	539b	59b	480b
Language concordance	54.5%	25.3%	58.4%	.002

Note. Numbers are weighted to account for the complex survey design and to present a representative sample of the California population. Totals may not equal 100% because of rounding.

Unweighted N.

Unweighted n.

LEP = Limited English proficiency; LHL = low health literacy.

Overall, respondents with LEP and LHL were the least likely to meet CRC screening guidelines (36%) followed by LEP-only (45%), LHL-only (51%), and those with neither LHL nor LEP (59%) (design-corrected chi-square: 14.27; p < .001).

Multivariable Models

Multivariable models are shown in Table 2. For the full sample, the hierarchy seen earlier in CRC screening by LEP and LHL combinations remained significant even when controlling for associated factors. Compared with the reference group (individuals with neither LEP nor LHL), having both LEP and LHL was associated with lowest odds of timely CRC screening (OR: 0.52; 95% CI [0.31, 0.98]), followed by those with LEP-only (OR: 0.60; 95% CI [0.40, 0.91]) and those with LHL-only (OR: 0.77; 95% CI [0.65, 0.91]).

Table 2.

Logistic models predicting odds of meeting colorectal cancer screening guidelines by health literacy and limited English proficiency in 2007 California Health Interview Survey 50–75 years of age, overall, for Whites, and for Asians (unweighted N = 15,888)

	Total OR (95% CI)	White OR (95%o CI)	Asian OR (95%o CI)
LHL and LEP	0.52 (0.31–0.89)	N/A	0.50 (0.28–0.89)
LEP only	0.60 (0.40–0.91)	0.53 (0.20–1.39)	0.62 (0.38–0.99)
LHL only	0.77 (0.65–0.91)	0.78 (0.64–0.93)	0.71 (0.39–1.28)
Neither	Reference	Reference	Reference
Education
Less than high school	0.62 (0.48–0.81)	0.61 (0.47–0.81)	0.66 (0.31–1.43)
High school graduate	0.67 (0.59–0.76)	0.66 (0.58–0.75)	0.73 (0.46–1.14)
College graduate	0.76 (0.66–0.88)	0.74 (0.64–0.86)	0.89 (0.57–1.39)
More than college degree	Reference	Reference	Reference
Other control variables
Age	1.06 (1.05–1.07)	1.06 (1.05–1.07)	1.04 (1.02–1.07)
Born in the United States	1.05 (0.84–1.30)	1.05 (0.81–1.37)	1.02 (0.63–1.65)
Female	0.86 (0.77–0.95)	0.82 (0.73–0.91)	1.08 (0.78–1.50)
≤ 100% federal poverty level	0.65 (0.49–0.86)	0.65 (0.46–0.91)	0.67 (0.39–1.14)
Rural	0.85 (0.76–0.95)	0.87 (0.77–0.98)	0.52 (0.27–1.02)
Insured	3.77 (2.98–4.76)	3.90 (2.98–5.11)	3.33 (2.08–5.33)
Married	1.24 (1.10–1.39)	1.26 (1.12–1.43)	1.03 (0.74–1.44)
Race/ethnicity
Chinese	0.92 (0.68–1.25)	—	0.88 (0.51–1.53)
Filipino	0.75 (0.51–1.12)	—	0.68 (0.35–1.30)
Korean	1.25 (0.81–1.94)	—	1.10 (0.57–2.11)
Vietnamese	1.43 (0.83–2.46)	—	1.28 (0.59–2.80)
Other Asian	0.76 (0.44–1.30)	—	0.73 (0.39–1.37)
Japanese	1.02 (0.67–1.55)	—	Reference
White	Reference	—	—

Note. LEP = Limited English proficiency; LHL = low health literacy; N/A = variable could not be included due to insufficient cases for analysis.

The analyses for Asians and Whites are separately shown in Table 2. In the multivariable model for Whites, the key health communication variable predicting CRC screening was LHL-only (OR: 0.78; 95% CI [0.64, 0.93]). LEP-only was not significant for Whites. (The group with both LHL and LEP was not included in the White analyses due to the small sample size: <45.) However, among Asians, LEP was a critical variable. For Asians, having both LEP and LHL was associated with having lower likelihood of cancer screening (OR: 0.50; 95% CI [0.28, 0.89]), as was LEP-only to a slightly less degree (OR: 0.62; 95% CI [0.38, 0.99]). LHL-only was not a statistically significant variable among Asians in multivariate analyses.

Among the control variables, health insurance was the strongest variable across all multivariate analyses (OR > 3.00 across all groups). Other significant control variables associated with meeting CRC screening recommendations in the full sample were older age, higher education, being male, living above poverty, living in a nonrural area, and being married.

Role of Doctor-Patient Language Concordance

Among the 539 individuals in the sample with LEP, 54.5% had a language-concordant provider (Table 1). Specifically, 25% of Whites with LEP had a language concordant provider compared with 58% of Asians with LEP. However, in bivariate analysis, language concordance was not significantly associated with timely CRC screening overall or specifically for Whites or Asians. Thus, we did not include this variable in further analyses.

Discussion

This study has three key findings. First, LHL was associated with CRC screening overall. However, to fully understand CRC screening disparities, our findings emphasizes that it is also important to consider English proficiency. Considering health literacy along with LEP in CRC screening was particularly important for Asians, a group with substantial numbers (34%) reporting LEP.

Second, we note that those with both LEP and LHL were particularly vulnerable groups for cancer screening disparities. In unadjusted models, only 39% of those with LEP and LHL met CRC screening guidelines, compared with 59% of those with neither health communication disadvantage. Even in fully adjusted models, CRC screening rates were significantly lower for those with both LEP and LHL overall and among Asians. This issue is also particularly important among Asians, as almost 14% reported both LEP and LHL, indicating significant self-reported health communication challenges and likely vulnerability to cancer screening disparities in this group.

However, it is important to note that not all individuals with LEP reported LHL. Those with LEP and/or LHL may not realize, or wish to admit, their own challenges (Parikh, Parker, Nurss, Baker, & Williams, 1996). These individuals may comprise another particularly vulnerable group for cancer screening disparities, although one that cannot be identified with self-reported health literacy measurement used in this study.

Last, this study found that language concordant providers do not appear to play a strong role in CRC screening among those with LEP. Previous research has found that language-concordant providers can reduce confusion about health information among those with LEP, though they do not fully mitigate language barriers (Wilson et al., 2005). This suggests that language concordant providers are not sufficient to resolve language-based disparities in cancer screening.

This speaks to the importance of also considering other interventions, including tailoring and disseminating education materials and messages to different cultural and language groups. Tips on developing effective cancer education materials for consumers are available in the literature (Kulukulualani, Braun, & Tsark, 2008) and online (National Cancer Institute, 2008; Susan G. Komen Foundation, 2007), including checklists and protocol for pretesting early versions of educational pieces. Low reading level text and congruent and clear graphics are essential for individuals with LEP and LHL. Other research has found that interpreters are effective in increasing cancer screening in Asian and Pacific Islander populations with LEP (Dang et al., 2010).

California has strong linguistic access polices, such as mandating that certain health materials and language assistance services be provided in a non-English language when a particular linguistic group surpasses a county percentage threshold (Perkins & Youdleman, 2008; Snowden et al., 2011). However, access to both in-language materials and programs in California vary across factors such as geography and racial/ethnic subgroup (Tanjasiri et al., 2004; Zargarzadeh & Law, 2011), and existing language assistance regulation have not resolved all communication barriers (Hadler et al., 2013). Furthermore, some individuals may also access in-language health education materials independently. For example, educational materials on colorectal cancer are available in some Asian languages (including Chinese, Hindi, Hmong, Khmer, Korean, Ilocano, Tagalog, and Vietnamese) through a website maintained by the Asian American Network for Cancer Awareness, Research, and Training (http://www.aancart.org/apicem-web-tool). Because there is no CHIS question asking if non-English speakers received linguistically appropriate health information from their providers or on their own, we could not test these issues in this analysis. We think that the relationship between LEP and LHL in cancer screening generally is important to consider, yet it is likely that specific relationships between LHL and LEP in CRC screening may vary by contextual factors (particularly interacting with demographic characteristics). While specifying these multilevel relationships are beyond the scope of the present study, future research should consider these issues in order to design the most effective interventions.

Outside of language, known barriers to CRC screening include diet restrictions associated with the guaiac-based fecal occult blood test, multiday sampling usually associated with both the guaiac and immunochemical fecal occult blood test, and diet restrictions and cleansing requirements associated with sigmoidoscopy and colonoscopy. LHL may add comprehension challenges to these instructions. LEP and LHL together may further compromise comprehension of and adherence to CRC screening procedures.

Another known barrier to CRC screening is lack of insurance, a particularly important and significant predictor of CRC screening compliance in this study. It is hoped that the Affordable Care Act will reduce the number of Americans without health insurance, mitigating this barrier. Also, health care facilities are now being monitored on CRC screening rates through the Healthcare Effectiveness Data and Information Set, a tool used by more than 90% of America's health plans to measure performance on important dimensions of care and service. Ideally, this change will help motivate health care organizations to increase adherence to CRC screening among all Americans, including those with LEP and LHL.

Limitations

This study has many strengths, including a population-based sample with both substantial numbers with LEP and a large Asian sample. However, this study does have some limitations.

The 2007 CHIS health literacy items are not standardized in-person measures, which is a limitation for direct comparison with other studies. Also, health literacy is self-reported. Despite these limitations, there is strong justification for the use of these health literacy items. First, they are similar to items with known validity in health literacy measurement (Chew, Bradley, & Boyko, 2004; Chew el al., 2008) and have been used previously (Health Research for Action, 2009; Sentell, 2012; Sentell & Braun, 2012). Second, they have strong face validity as direct measures of health communication challenges. Third, to our knowledge, no other US, population-based study includes health literacy and cancer screening data for those who are LEP and speak multiple Asian languages. Fourth, only with findings of disparities in these factors are there likely to be justification for adding more detailed health literacy measures into population-based health surveys, including those with substantial Asian American samples. As these efforts are expensive, it is important to use rich existing data sources to consider if there is sufficient justification for the costly undertaking of such in-person health literacy measures.

Our other items are self-reported as well. Respondents may not be fully aware of their English proficiency or their cancer screening history. Furthermore, racial/ethnic groups respond differently to self-reported items (Kandula, Lauderdale, & Baker, 2007). The CHIS data also lack some variables (health beliefs and previous cancer history) that have been associated with cancer screening in previous research (Breslau, 2010; McCracken et al., 2007; Maxwell, Crespi, Antonio, & Lu, 2010). Also, because our health literacy measure does not distinguish between difficulty reading materials in the patient's native tongue versus difficulty reading English, as discussed above, it is possible that some of the health literacy barriers for LEP patients could be overcome by having materials available in the patient's native tongue.

The American Asian label encompasses at least 50 ethnic groups with distinct cultures and languages, which may have distinct relationships to cancer screening and associated factors (McCracken et al., 2007). However, as other researchers have found, small numbers of Asians in population-based samples often require combining American Asian groups for statistical analysis (Ghosh, 2003), an approach which we followed to assure adequate power for our analyses of LEP and LHL. Future research with larger sample sizes of Asian subgroups should consider how LEP and LHL might differentially impact diverse Asian subgroups. The small sample sizes of our LEP populations, particularly in Whites, may also have impacted our analyses, including the language concordance analyses. Future work including larger sample sizes would be useful for this topic as well.

Last, these findings are based on California data. Our findings may not be fully relevant to Asian American populations in other states. They also may apply differently to non-Asian American LEP populations in California or elsewhere, including Latinos. Asians and other racial/ethnic groups may have unique pathways to cancer screening in California, particularly in light of the innovative language polices there. Yet, even in California data, we found significant disparities. Only 42% of those with LEP overall received CRC cancer screening, compared with 58% of those with English proficiency. Also, only 51% of Asians received screening compared with 58% of Whites. We might expect disparities to be larger in states with less access to Asian language materials and less innovative linguistic policies.

Conclusions

This study confirms that health literacy is associated with CRC screening in a large population-based sample. We also provide insight to significant health communication predictors for CRC screening among Asian Americans generally, a group with notable cancer screening and mortality disparities. Unlike race/ethnicity and many other factors associated with cancer disparities, health literacy is amenable to direct intervention at the individual, patient/provider, and systems levels and may be a fruitful focus for intervention. However, this study also confirms that English proficiency is a critical health communication variable to consider along with health literacy in understanding CRC screening disparities. Further research should consider which health literacy and language-based policies and interventions might explain and/ or mitigate these differences. We note that provider language concordance does not appear to either explain or mitigate CRC screening disparities in the groups studied. Future research should also consider the role of health communication barriers on other points along the cancer care continuum in which health disparities are perpetuated and might be decreased. The role of language and health literacy should also be considered for other racial/ethnic groups and in screening for other cancers, including those with screening tests that place fewer health literacy demands on individuals. Overall, it appears that addressing health communication factors, such as language and health literacy, will be important to addressing critical cancer screening disparities in the United States.