




































 
 
 
 
 

 

www.companyofscientists.com/index.php/chd                             e1                                    Cancer Health Disparities 

RESEARCH 

Community-Clinical Linkage Intervention 

to Improve Colorectal Cancer Screening 

Among Underserved Korean Americans 
 

Grace X. Ma1,2 Minsun Lee1  Maayan Beeber,3 Rina Das,4 Ziding Feng,5 Min Qi Wang,6 Yin Tan,1 Lin Zhu,1 

Khursheed Navder,3 Theresa I. Shireman,7 Philip Siu,8 Joanne Rhee,1 Minhhuyen T. Nguyen9 

1 
Center for Asian Health, Lewis Katz School of Medicine, Temple University, Philadelphia, PA. 

2 
Department of Clinical 

Sciences, Lewis Katz School of Medicine, Temple University, Philadelphia, PA. 
3 
Hunter College, City University of New 

York, New York City, New York 
4 

Division of Scientific Programs, Integrative Biological and Behavioral Sciences, 

National Institute of Minority, Health and Health Disparities, National Institutes of Health. 
5 
The University of Texas MD 

Anderson Cancer Center, Houston, TX. 
6 

Department of Behavioral and Community Health, School of Public Health 

(SPH), University of Maryland, College Park, MD 
7 
Department of Health Services, Policy and Practice, School of Public 

Health, Brown University, Providence, RI. 
8 

Greater Philadelphia Health Action, Chinatown Medical Services, 

Philadelphia, PA 
9
 Department of Medicine, Section of Gastroenterology, Fox Chase Cancer Center, Philadelphia, PA. 

 

Corresponding author email: grace.ma@temple.edu 

ABSTRACT 
Korean Americans report the lowest and declined rates of colorectal cancer (CRC) screening, 

compared to general population in the United States. The present study aimed to evaluate the efficacy 

of a community-based multifaceted intervention designed to improve CRC screening among Korean 

Americans. A cluster-randomized trial involving 30 Korean church-based community organizations (n 

= 925) was conducted. Fifteen churches were assigned to intervention (n=470) and the other 15 to 

control (n = 455) groups. Main components of the intervention included interactive group education, 

patient navigation, physician engagement, and provision of fecal immunochemical test (FIT) kit. CRC 

screening rates were assessed at a 12-month follow-up. Participants in the intervention group were 

significantly more likely to receive CRC screening (69.3%) as compared with those in the control group 

(16%). The intervention was particularly effective in promoting FIT among the more disadvantaged 

individuals in the Korean American community. Regression analysis revealed that controlling for the 

intervention effect, male gender, high school education, annual income of $20,000–40,000 were 

significantly associated with increased screening by FIT, whereas English inefficiency was significantly 

and lack of health insurance was marginally significantly associated with decreased screening by 

colonoscopy/sigmoidoscopy. Culturally and linguistically appropriate multifaceted intervention 

combining FIT provision with community-clinical linkage has a potential to be a cost-effective and 

practical approach to effectively targeting hard-to-reach disadvantaged minority populations and 

enhance CRC screening to reduce cancer disparities.   

KEYWORDS: colorectal cancer (CRC) screening, Korean American, Intervention, FIT, community-

based participatory research 

Citation: Ma GX et al (2019). Community-Clinical Linkage Intervention to Improve Colorectal Cancer Screening 

Among Underserved Korean Americans. Cancer Health Disparities 3:e1-e15. doi:10.9777/chd.2019.1001.



 
 
 
 
 

 

www.companyofscientists.com/index.php/chd                             e2                                    Cancer Health Disparities 

RESEARCH 

INTRODUCTION 
Colorectal cancer (CRC) is the second most 

common cancer among Korean Americans, after 

lung cancer for Korean men and breast cancer for 

Korean women (Miller et al., 2008). In the United 

States, CRC incidence rates have decreased over 

the past decade, with the change of rates by more 

than 4% per year in both men and women 

(American Cancer Society, 2017). However, unlike 

the general trend, the CRC incidence rates among 

Korean Americans have increased with the annual 

percent change (APC) of 3.6% between 1988 and 

2007 (Giddings et al., 2012; Lee et al., 2007). 

Despite the high burden from CRC among Korean 

Americans, CRC screening rates in this population 

are significantly lower than the national average 

(Maxwell and Crespi, 2009; Maxwell et al., 2010; 

Hwang, 2013; Oh and Jacobsen, 2014). 

 

Currently, several screening tests can be used to 

find polyps or colorectal cancer. The fecal 

immunochemical test (FIT) uses antibodies to 

detect blood in the stool, which is an early sign of 

cancer. Colonoscopy and sigmoidoscopy both use 

a thin flexible lighted tube with a camera at the 

end to check for polyps or cancer inside the 

rectum and the colon. The difference between the 

two procedures is that a flexible sigmoidoscopy 

examines the left lower third of the colon with a 

short tube, while colonoscopy examines the entire 

colon with a longer tube (Center for Disease 

Control and Prevention, 2017). The prevalence of 

CRC screening in accordance with guidelines 

among the general US population 50 years of age 

and older and overall Asian Americans in 2015 was 

62.6% vs 49.4% (American Cancer Society, 2014). 

However, a systematic review of thirteen studies 

on CRC screening among Korean Americans 

published until May 2013 (Oh and Jacobsen, 2014) 

indicated that only one in four Korean Americans 

aged 50 and older reported having ever had FIT 

and approximately 40% reported having ever had 

a sigmoidoscopy or colonoscopy in different 

studies. In addition, screening rates decreased for 

both colonoscopy and FIT in Korean Americans, 

while the rates increased or were stable in the US 

general population (Maxwell and Crespi, 2009).  

 

The factors contributing to the low rates of and 

declining trend in screening among Korean 

Americans are complex. Inadequate knowledge 

about screening, screening methods, and its 

benefits, as well as low perceived risks of CRC have 

been identified as major factors related to low 

rates of screening among Koreans (Juon et al., 

2003; Ma et al., 2009; Maxwell et al., 2010). These 

studies suggest that Korean Americans may be 

less familiar with utilizing health care system to 

detect health problems before the onset of 

symptoms than other groups, emphasizing the 

need for education of the benefit of screening 

before symptoms develop. In addition, the decline 

of CRC screening rates of this population has been 

associated with limited access to health system 

due to lack of insurance, language inefficiency, 

and lack of transportation (Jo et al., 2008; Lee and 

Im, 2013; Lee and Lee, 2013).  

 

With the high CRC burden combined with low 

rates of screening uptake due to multilevel 

barriers, Korean Americans, particularly, 

disadvantaged groups with limited English 

proficiency and low income are an important 

target group for implementation of innovative and 

culturally tailored programs to enhance CRC 

screening. Collaboration among multisector 

stakeholders, the community members and 

leaders, clinical providers and researchers, has 

great potential to enhance cancer screening for 

this population. Community-based organizations 

(CBOs) who work directly with the target 

population serve as a venue for promotion of 

knowledge and awareness of CRC screening and 

reduction of barriers to screening (Israel et al., 

2005). Engaging CBOs and clinical providers 

(primary care practices, PCPs) can deliver 

screening test more effectively by reaching the 

underserved and hard-to-reach individuals who 

rarely visit doctors for preventive care. Therefore, 

implementation of CRC screening in this 



 
 
 
 
 

 

www.companyofscientists.com/index.php/chd                             e3                                    Cancer Health Disparities 

RESEARCH 

underserved population requires establishing 

strategic partnerships between community and 

clinical settings and sharing their infrastructure and 

capacity to enhance engagement of potential 

patients and PCPs in CRC screening, and follow-up 

care (Ockene et al., 2007; Krist et al., 2013; Persson, 

2016).  

 

However, there is clear paucity of culturally and 

linguistically appropriate CRC community clinical 

linkage intervention (CCL), designed to improve 

CRC screening among Korean Americans. To our 

knowledge, no study has investigated the impact 

of the CCL intervention for CRC screening in 

Korean Americans, except for the one we 

conducted in a small scale quasi-experimental 

design (Ma et al., 2009). In our previous study, we 

reported that a culturally and linguistically tailored 

church-based intervention combined with patient 

navigation assistance and physician engagement 

(including screening reminder, scheduling 

appointment, transportation assistance to clinic 

sites, navigation for screening, follow up care and 

medical record verification) was effective in 

increasing CRC screening rate. Building on our 

preliminary data (Ma et al., 2009), the present 

study was designed to evaluate the impact of a 

large-scale cluster group randomized trial on CRC 

screening.    

 

Regarding CRC screening methods or modalities, 

colonoscopy is most commonly recommended in 

the United States. Colonoscopy has a high 

sensitivity for polyps and cancer, but is invasive, 

cumbersome, expensive, and thus, has limited 

availability in many underserved minority 

populations (Lieberman et al., 2000; Inadomi et al., 

2012; Elmunzer et al., 2015). Although FIT has less 

sensitivity for polyps and cancer than colonoscopy, 

it has advantage of being non-invasive, easy-to-do 

at home, inexpensive, and more readily available 

as a first step of detecting abnormality. Indeed, 

when offered or recommended, individuals, 

particularly from underserved communities, were 

more likely to complete FIT than colonoscopy, 

suggesting that FIT based outreach was more 

effective than colonoscopy-based outreach in 

increasing screening rates among underserved 

populations (Inadomi et al., 2012; Gupta et al., 

2013; Singal et al., 2016). According to these 

studies, a programmatic method with less 

invasiveness and more availability with minimal 

resources and infrastructure such as FIT appears to 

be a promising intervention modality for 

improvement of CRC screening among 

underserved populations.  

 

This article evaluated the primary outcome, 

efficacy of a culturally and linguistically appropriate 

community- clinical linkage intervention that 

included provision of FIT, group education, patient 

navigation, and physician engagement on CRC 

screening.  

 

METHODS                                                                                                                                              
 

Participants 

Korean Americans (N = 925) enrolled in this study 

were members of Korean churches (N = 30) in 

Philadelphia and New Jersey regions that serve a 

predominately underserved immigrant population. 

More than 78% of Koreans in the US are affiliated 

with churches (Chang, 2003; Ma et al., 2009; Min, 

2002) and the churches serve as one of the 

community-based organizations facilitating 

discussion of and participation in various cultural, 

social and personal activities. Thus, Korean church 

is one of the important venues to deliver culturally 

appropriate intervention to provide health 

education and promote CRC screening among 

Korean Americans. Participants were eligible to 

participate in this study if they: (1) were self-

identified Korean Americans; (2) were 50 years and 

older; (3) did not have a colorectal polyp, CRC 

cancer, or a family history of CRC (first degree 

relative); and (4) non-adherent to CRC screening 

guidelines (never had any CRC screening or were 

overdue for screening). Based on the CRC 

screening guidelines recommended by the ACS, 

“overdue” was defined by no FIT in the past year 



 
 
 
 
 

 

www.companyofscientists.com/index.php/chd                             e4                                    Cancer Health Disparities 

RESEARCH 

or no sigmoidoscopy in the past five years, no 

double-contrast barium enema in the past five 

years, or no colonoscopy within the past ten years.   

 

Procedures 

The study was designed as a collaborative 

partnership among an academic institution, 

Korean churches, community-based organizations 

(CBOs), and healthcare providers/organizations. A 

framework for community-clinical linkages was 

used to guide the current study and collaboratively 

established CRC screening program and referral 

systems that connect individuals with abnormal 

screening results to clinical care. As part of 

collaboration, Korean churches provided facilities 

and resources to promote awareness and 

knowledge about CRC screening through 

provision of education sessions and FIT kits. 

Academic researchers coordinated each stage of 

study by linking community and clinical settings 

and provided navigation services as needed.  

 

Participants who performed FIT testing were asked 

to mail the samples in a provided envelope to 

their providers. Participants do not have regular 

physicians or health insurance, were encouraged 

to mail the samples to the partnering clinicians or 

health centers of our network. Participants with an 

abnormal FIT result were notified with a letter 

written by a clinician and referred to diagnostic 

colonoscopy. The CAH research team at Temple 

University has an established network of 

participating clinical partners who provide a range 

of health services either at reduced cost to patients 

who are uninsured or underinsured. Thus, those 

who had no primary care physician or medical 

insurance were sent to the collaborating clinical 

partners for the diagnostic test. This study was 

approved by the Temple University Institutional 

Review Board. 

 

Study design  

The study utilized a two-arm cluster randomized 

design with churches as the unit of randomization 

(Fig 1). Based on organization profile information 

provided by collaborating church leaders, we 

paired the 30 churches by size and geographic 

location. Then, each CBO of the 15 pairs was 

randomly assigned to the intervention or control 

group. We conducted baseline evaluation, post-

intervention assessment, and a 12-month follow-

up on screening outcome. Baseline and post-

intervention assessments were administered in 

person while the 12-month follow-up assessments 

were conducted by telephone interviews and 

primary outcome of screening was verified 

through medical record. 

 

Intervention and control conditions 

The intervention was designed to improve CRC 

screening among participants who are non-

adherent to CRC screening guidelines by 

addressing multilevel barriers. Main intervention 

components included interactive group education, 

navigation services and engagement of health 

care providers for referrals, and linkage to care. 

The interactive group education was aimed to 

increase participant overall understanding of CRC, 

screening methods, and utilization of available 

resources such as home test kit and navigation 

services, with the ultimate goal of increasing 

screening rates. Bilingual health educators 

facilitated discussion on CRC susceptibility and 

severity among Korean Americans and the U.S. 

general population, benefits of CRC screening, and 

CRC screening options and their associated clinical 

procedures, as well as the pros and cons of each 

option. To increase their motivation for CRC 

screening and awareness of the different methods, 

FIT home kit was offered to the participants with 

instructions in Korean. Although it is provided to 

all participants in intervention group, they were 

encouraged to choose any screening methods 

convenient for them. Clinical partners provided 

clinical support and ensured successful screening 

assessment and follow-ups by offering more 

flexible hours of clinic operation with bilingual 

medical staff on site. Patient navigation assistance 

was also offered based on participants’ needs. The 

range of assistance included scheduling 



 
 
 
 
 

 

www.companyofscientists.com/index.php/chd                             e5                                    Cancer Health Disparities 

RESEARCH 

appointments with clinical partners for 

sigmoidoscopy/ colonoscopy for screening and 

diagnosis after a FIT positive result, assisting paper 

work and communication with a physician, and 

arranging transportation. 

 

Figure 1. Study design flow chart.

 

In the control condition, participants (n = 455) 

received a group education session in a similar 

format with that of the intervention group 

delivered by trained Korean community health 

educators. Different from the intervention group, 

the education focused on general health 

education and primary prevention issues, including 

routine health examinations and screening for 

various diseases such as cancer.  Korean version of 

standard printed materials and guidelines related 

to the education contents were also provided, 

including CRC screening guidelines.  

 

Measures 

This study assessment was designed to evaluate 

the multilevel intervention including individual 

participants’ level, CBO organization level and 

clinical linkage to care level, as a result of 



 
 
 
 
 

 

www.companyofscientists.com/index.php/chd                             e6                                    Cancer Health Disparities 

RESEARCH 

education, patient navigation, and physician 

engagement for linkage to care. At participants’ 

level, we assessed participants’ satisfaction on each 

component of intervention, screening behaviors 

and knowledge improvement. In addition, as a 

process outcome, we assessed the adherence of 

main stakeholders (CBOs, Patient Navigators and 

clinical providers) to basic CBPR principles of 

collaboration in various stages of the project 

implementation. Since the main purpose of this 

article was to report the efficacy of the intervention 

on participants’ screening behaviors, we only 

reported CRC screening as primary outcome 

measures and relevant results.      

Participant characteristics. Demographic 

information such as age gender, education, marital 

status, household income, employment was 

collected at baseline. Acculturation related 

information such as years living in the US and the 

English ability was also collected. In addition, 

information on health insurance and regular 

physician status was collected to learn the levels of 

health care access of the participants.  

Barriers to screening. Barriers to CRC screening 

were evaluated by asking them the question: “If 

you have never had any of the screening such as 

stool blood test, sigmoidoscopy or colonoscopy, 

what are the major reasons?” The choices include 

‘Do not know what it is’, ‘Do not know where to 

get it’, ‘Do not have time’, and ‘Feel healthy and 

do not think needing a test.’  

Screening behavior. The primary outcome was CRC 

screening status (completion of either FIT kit, 

sigmoidoscopy, or colonoscopy) at the 12-month 

follow-up. Bilingual interviewers contacted 

participants at 12-month follow-up and assessed 

screening behavior using self-report (“Over the 

past 12 months, have you had a test for colorectal 

cancer?”; yes/no, and “If you had, what is the 

test?”; stool blood test using FIT 

kit/colonoscopy/sigmoidoscopy). Participants who 

reported receiving FIT were verified with the list of 

those who returned the stool sample and received 

the test results from the collaborating clinical 

laboratory. Participants who reported receiving 

colonoscopy or sigmoidoscopy were asked to 

provide a medical release consent to allow project 

staff to contact their health care providers to 

validate self-reported testing. Participants who 

received positive FIT results and were referred to 

diagnostic colonoscopy were also followed-up. 

 

Data analysis 

Participant characteristics were described using 

frequencies and percentages within each of the 

treatment groups (intervention versus control). The 

chi-square test was used to assess whether the 

distributions differed between the treatment 

groups. The chi-square test was also used to 

compare the proportion of participants in the 

intervention and control groups who completed 

CRC screening within 12 months of the 

intervention. The proportion of participants who 

completed screening was compared by different 

CRC screening methods and by recruitment sites. 

To investigate whether the intervention was 

particularly effective in disadvantaged individuals, 

the comparison of screening rates between 

intervention and control groups was conducted 

only among the disadvantaged group (i.e., high 

school or less education, income of <$40,000, no 

insurance etc.). Finally, logistic regression analyses 

were conducted to explore the factors associated 

with FIT and colonoscopy/sigmoidoscopy. In the 

models, the covariance matrix was adjusted to 

account for clustering of individuals at the church 

level. All analyses were performed in SAS version 

9.3 assuming a type 1 error of 0.05. 

RESULTS 



 
 
 
 
 

 

www.companyofscientists.com/index.php/chd                             e7                                    Cancer Health Disparities 

RESEARCH 

Participant characteristics 

Overall, there were slightly more women (59.2%) 

than men enrolled in the study. Majority of 

participants were married (80.4%) and did not 

speak English well (74.1 %). Roughly half were 

college or higher level educated (50.4), insured 

(49.7%), and had a regular physician (55.5%). 

More than half were employed (56.8%) and 

earned less than $40,000 (66.1%). Table 1 describes 

and compares sociodemographic characteristics 

for the intervention and control groups. More 

participants in the control group than in the 

intervention group earned greater than $40,000 in 

annual income (28.1% vs. 21.2%; p = 0.050), and 

had health insurance (55.5% vs. 46%; p = 0.005). 

The distribution of other demographic, 

acculturation, and health access variables were not 

significantly different between groups. 

Barriers of screening were generally similar in 

distribution between the intervention and control 

groups, with the exception of three variables. 

Participants in the intervention group were more 

likely than those in the control group to endorse 

‘not being able to afford it’ (28.4% vs. 22.9%; p = 

0.012), and ‘feeling healthy’ (22.9% vs. 14.3%; p 

=0.001) as the reason of not having had screening. 

However, less participants (16.0%) of the 

intervention group reported ‘not knowing what 

screening is’ compared to those in the control 

group (21.6%; p = 0.035) (Table 1). 

Screening rates by methods and sites  

As shown in Table 2-1, CRC screening behaviors 

were significantly different between intervention 

and control groups. In the intervention group, a 

total of 277 out of 400 participants (69.3%) were 

screened for CRC at 12 months, compared to 61 

out of 382 (16.0%) in the control group (χ2 (1) = 

226.04, p<0.001). In the intervention group, the 

majority (66.8%) screened by FIT, and minority 

screened by colonoscopy (2.0%) and 

sigmoidoscopy (0.3%). In contrast, in the control 

group a minority (1.8%) screened by FIT, and the 

majority screened by colonoscopy (11.5%) and 

sigmoidoscopy (5.8%). 

As shown in Table 2-2, out of the 15 paired church 

sites in the study, the majority (n = 12) had 

screening rates significantly higher by intervention 

group than by control group: the screening rates 

in the intervention group ranged from 60.0% to 

93.8%, while corresponding control group 

screening rates ranged from 0% to 38.9%. 

However, at 3 church pairs (5th, 8th, and 14th), 

screening rates were not significantly different 

between the two treatment groups, with 

substantially low level of screening rates (33.3% - 

45.8%) compared to other sites in the intervention 

group. Compared to the intervention participants 

from the other church sites, those from the 5th, 8th, 

and 14th church pairs were more likely to have 

college or higher education (73.8% vs. 51.2%; p = 

0.005), be employed (73.8 % vs. 55.7%; p = 0.01), 

have greater than $40,000 annual income (49.1% 

vs. 15.9%; p < 0.001), and have health insurance 

coverage (56.5% vs. 42.9%; p = 0.048). 

Screening rates among disadvantaged individuals 

Difference in CRC screening rates between 

intervention and control groups were heightened 

when examined among disadvantaged individuals 

only (Table 3). Specifically, among the participants 

who had an education of high school level or 

below, annual household income of less than 

$40,000, were unemployed, could not speak 

English, and had no health insurance or access to 

a regular physician, the average rates of screening 

were consistently over 70% in the intervention 

group. However, the screening rates in the control 

group among the disadvantaged individuals were 

generally lower than the overall screening rates in 

the control group except for the unemployed 

participants. The difference in screening rates 

between the two groups were even higher among 

those who faced barriers to screening such as not 

knowing what CRC screening is (69.4% vs. 7.8%), 



 
 
 
 
 

 

www.companyofscientists.com/index.php/chd                             e8                                    Cancer Health Disparities 

RESEARCH 

not knowing where to get it (81% vs. 10%), not 

being able to afford it (70.6% vs. 6.3%), having a 

language difficulty (80% vs. 11.1%) and feeling 

healthy (71.4% v. 9.3). 

 

Table 1. Characteristics of Intervention and Control Group Participants: CRC Screening Program (n=925). 

Variable Intervention Group  

(n=470) 

No.  (%) 

Control Group  

(n=455) 

No.  (%) 

 

 

P1 

Age (mean years) 62.28±9.94 62.93±9.27 0.361 

Gender   0.158 

    Male 197 (42.5) 171 (37.8)  

    Female 267 (57.5) 281 (62.2)  

Marital status 0.404 

    Married 378 (81.5) 351 (79.2)  

    Not Married 86 (18.5) 92 (20.8)  

Education 0.104 

    Below high school 55 (12.1) 71 (16.2)  

    High school graduate 150 (32.9) 153 (34.9)  

    Some college or higher 251 (55.0) 215 (49.0)  

Annual household income 0.050 

   Less than $20,000 157 (36.9) 139 (36.2)  

   $20,000 - $40,000 178 (41.9) 137 (35.7)  

   Greater than $40,000 90 (21.2) 108 (28.1)  

Employment 0.736 

   Employed 257 (56.2) 251 (57.4)  

   Unemployed/Retired/Homemaker 200 (43.8) 186 (42.6)  

Year living in US   <0.001 

   Less than 10 yrs   90 (12.2) 104 (15.5)  

   Between 10 and 20 yrs  145 (19.7) 178 (26.6)  

   More than 20 years  501 (68.1) 388 (57.9)  

Ability to speak English 0.296 

   Not at all 43 (9.3) 54 (12.1)  

   Not well 299 (64.7) 289 (64.8)  

   Well/Very well 120 (26.0) 103 (23.1)  

Health insurance status   0.005 

   No 248 (53.9) 199 (44.5)  

   Yes 212 (46.1) 248 (55.5)  

Regular physician status   0.411 

   No 191 (42.7) 171 (40.0)  

   Yes 256 (57.3) 257 (60.0)  

Barriers to screening    

Do not know what it is   0.035 

Unchecked  393 (84.0) 356 (78.4)  



 
 
 
 
 

 

www.companyofscientists.com/index.php/chd                             e9                                    Cancer Health Disparities 

RESEARCH 

Checked 75 (16.0) 98 (21.6)  

Do not know where to get it   1.000 

Unchecked 442 (94.4) 429 (94.5)  

Checked 26 (5.6) 25 (5.5)  

Cannot afford   0.012 

Unchecked   335 (71.6)  358 (78.9)  

Checked 133 (28.4) 96 (21.1)  

Do not have time   0.339 

Unchecked 413 (88.2) 410 (90.3)  

Checked 55 (11.8) 44 (9.7)  

Transportation difficulty   0.579 

Unchecked 460 (98.3) 449 (98.9)  

Checked 8 (1.7) 5 (1.1)  

Language difficulty   0.698 

Unchecked 453 (96.8) 442 (97.4)  

Checked 15 (3.2) 12 (2.6)  

Afraid of pain   0.243 

Unchecked 439 (93.8) 434 (95.6)  

Checked 29 (6.2) 20 (4.4)  

Feel healthy   0.001 

Unchecked 361 (77.1) 389 (85.7)  

Checked 107 (22.9) 65 (14.3)  

1 P-values based on Rao-Scott chi-square test excluding missing values. Rao-Scott chi-square test 

used to account for design effects of clustering at the site level. 

  

 

TABLE 2-1. CRC Screening Rates at 12 Months Follow-up by Screening Methods 

 

Outcome 

Intervention Group 

(n=400) 

No.  (%) 

Control Group 

(n=382) 

No.   (%) 

 

P 

Screened for CRC 277 (69.3) 61 (16.0) < 0.001 

Type of Screening    

     FIT 267(66.8) 7 (1.8) < 0.001 

     Colonoscopy 8 (2.0) 44 (11.5) < 0.001 

     Sigmoidoscopy 1 (0.3) 22 (5.8) < 0.001 

P-values based on chi-square test of Fisher’s exact test excluding missing values. 

 

  



 
 
 
 
 

 

www.companyofscientists.com/index.php/chd                             e10                                    Cancer Health Disparities 

RESEARCH 

 

 

 

TABLE 2-2. CRC Screening Rates at 12 Months Follow-up by Community Sites  

 Intervention (n=400) Control (n=382) P 

Pair N Test rate (%) N Test rate (%)  

1 33/41 80.5 2/20 10.0 <0.001 

2 22/36 61.1 3/17 17.6 0.004 

3 30/37 81.1 7/35 20.0 <0.001 

4 16/31 51.6 0/21 0.00 <0.001 

5 5/15 33.3 10/47 21.3 0.489 

6 15/16 93.8 4/30 13.3 <0.001 

7 13/20 65.0 1/31 3.2 <0.001 

8 11/24 45.8 10/37 27.0 0.171 

9 21/28 75.0 0/20 0.0 <0.001 

10 30/34 88.2 7/18 38.9 0.001 

11 18/21 85.7 4/17 23.5 <0.001 

12 18/22 81.8 4/36 11.1 <0.001 

13 22/31 71.0 3/14 21.4 0.003 

14 11/24 45.8 6/18 33.3 0.530 

15 12/20 60.0 0/21 0.0 <0.001 

P-values based on chi-square test of Fisher’s exact test excluding missing values. 

TABLE 3. CRC Screening Results 12 Months Follow-up Only Among Disadvantaged Individuals 

Variable Intervention 

n (%) 

Control 

n (%) 

p-value 

 

High school or below (n=365) 132/175 (75.4)    29/190 (15.3)  <0.001 

Income < $40,000 (n=516) 211/287 (73.5) 37/229 (16.2) <0.001 

Unemployed (n=310) 117/161 (72.7) 31/149 (20.8) <0.001 

English- not well/not at all (n=580) 207/292 (70.9) 39/288 (13.5) <0.001 

No health insurance (n=390) 154/215 (71.6) 14/175 (8.0) <0.001 

No regular physician (n=310) 114/162 (70.4) 12/148 (8.1) <0.001 

Barriers to screening    

  Do not know what it is (n=139) 43/62 (69.4) 6/77 (7.8) <0.001 

  Do not know where to get it (n=41) 17/21 (81.0) 2/20 (10.0) <0.001 

  Cannot afford (n=199)  84/119 (70.6) 5/80 (6.3) <0.001 

  Do not have time (n=89) 29/49 (59.2) 2/40 (5.0) <0.001 

  Language difficulty (n=24) 12/15 (80.0) 1/9 (11.1) =0.001 

  Afraid of pain (n=44) 17/25 (68.0) 3/19 (15.8) =0.001 

  Feel healthy (n=144) 64/90 (71.1) 5/54 (9.3) <0.001 



 
 
 
 
 

 

www.companyofscientists.com/index.php/chd                             e11                                    Cancer Health Disparities 

RESEARCH 

  

TABLE 4. Factors Associated with FIT and Colono/Sigmoidoscopy at 12 Month Follow-up (n=782) 

 

Variable 

FIT 

OR (95% CI) 

Colono/Sigmoidoscopy  

OR (95% CI) 

Intervention Group   

   Intervention 151.39 (56.97, 402.26)*** 0.10 (0.04, 0.24)*** 

   Control (Ref) 1.00 1.00 

Age 1.03 (0.99, 1.07) 0.99 (0.95, 1.04) 

Years in US 1.05 (0.71, 1.55) 0.84 (0.52, 1.36) 

Gender   

   Male 2.02 (1.17, 3.49)* 1.06 (0.52, 2.15) 

   Female (Ref) 1.00 1.00 

Marital Status   

   Married 0.99 (0.48, 2.04) 0.64 (0.29, 1.41) 

   Not married 1.00 1.00 

Education   

   Below high school  1.50 (0.55, 4.04) 1.80 (0.61, 5.33) 

   High school graduate 2.69 (1.44, 5.03)* 0.92 (0.41, 2.06) 

   Some college or higher (Ref) 1.00 1.00 

Annual household income   

   Less than $20,000 1.80 (0.78, 4.18) 1.11 (0.38, 3.18) 

   $20,000 - $40,000 2.01 (1.02, 3.95)* 1.59 (0.63, 4.00) 

   Greater than $40,000 (Ref) 1.00 1.00 

Employment    

   Employed  0.934 (0.454, 1.92) 0.61 (0.27, 1.38) 

   Unemployed/Retired (Ref) 1.00 1.00 

Ability to speak English   

   Not at all 0.587 (0.16, 2.11) 0.66 (0.16, 2.65) 

   Not well 0.838 (0.45, 1.56) 0.45 (0.20, 0.99)* 

   Well/Very well (Ref) 1.00 1.00 

Health insurance status   

   No 2.54 (1.20, 5.37)* 0.34 (0.10, 1.18)Ɨ 

   Yes (Ref) 1.00 1.00 

Regular physician status   

   No 0.80 (0.40, 1.59) 0.80 (0.24, 2.71) 

   Yes (Ref) 1.00 1.00 

Analysis accounts for clustering at the church level 

Ɨp-value < 0.1, *p-value < 0.05, **p-value <0.01, ***p-value <0.001 

Factors associated with FIT and colonoscopy  

As shown in Table 4, controlling for intervention 

effect, several predictors were significantly 

associated with screening by FIT. Males were twice 

as likely to have screened by FIT (OR = 2.02, 95% 

CI 1.17–3.49, p = 0.012). Individuals who have high 

school education (OR = 2.69, 95% CI 1.44–5.03) 



 
 
 
 
 

 

www.companyofscientists.com/index.php/chd                             e12                                    Cancer Health Disparities 

RESEARCH 

and those who had an annual income of $20,000–

$40,000 (OR = 2.01, 95% CI 1.02–3.95, p=0.046) 

were more likely to be screened by FIT than those 

with college or higher education and those with 

greater than $40,000 income. Compared to 

individuals with health insurance, those without 

health insurance were 2.5 times as likely to screen 

by FIT (OR = 2.54, 95% CI 1.20–5.37, p=0.017). On 

the other hand, English inefficiency was 

significantly (OR = 0.45, 95% CI 0.20–0.99, 

p=0.031) and having no health insurance was 

marginally significantly (OR = 0.34, 95% CI 0.10–

1.18, p=0.081) associated with decreased odds of 

screening by colonoscopy/ sigmoidoscopy.  

Regarding the barriers, controlling for the 

demographic covariates, ‘fear of pain’ was 

associated with decreased odds of screening by 

FIT (OR = 0.29, 95% CI 0.09–0.98, p=0.050), while 

‘not feeling healthy’ was associated with increased 

screening by colonoscopy / sigmoidoscopy (OR = 

3.47, 95% CI 1.13–10.67, p=0.031). 

 

Screening results and follow-up care 

Twenty eight out of 267 (10.2%) who completed 

FIT in the intervention group were screened 

positive. All the 28 participants with abnormal FIT 

results completed diagnostic colonoscopy with 

patient navigation assistance and physician 

engagement in referral and facilitation the linkages 

to follow up care.     

 

DISCUSSION 

This study is the first large-scale cluster 

randomized, multifaceted community-clinical 

linkage intervention trial to improve CRC screening 

among underserved and less acculturated Korean 

Americans. We found that the intervention 

consisting of culturally appropriate education, 

provision of FIT, navigation services, and physician 

engagement was significantly effective in 

increasing CRC screening, particularly among the 

hard-to-reach Korean Americans with limited 

English proficiency. Compared to the previous 

study that 90% of whom had public or private 

insurance (Tu et al., 2006), it is noteworthy that the 

present study achieved high screening rate among 

50% participants were uninsured. The finding of 

the present study demonstrates the efficacy of a 

multilevel intervention through community-clinical 

linkage in this population.  

Korean Americans have reported to have the 

lowest CRC screening rates as US general 

populations with a significant increase in CRC 

screening (Maxwell and Crespi, 2009; Maxwell et 

al., 2010; Hwang, 2013; Oh and Jacobsen, 2014). As 

the contributing factors to the CRC screening 

disparities in Korean Americans, large proportion 

of them are foreign-born, lack of health insurance, 

and have a lower household income, in addition to 

limited awareness and knowledge about CRC 

screening guidelines. Indeed, majority of the 

participants in our study who either never had 

CRC screening or were overdue for screening 

were with low income, unemployment, English 

inefficiency, and no health insurance. Substantial 

number of those who have never had CRC 

screening also endorsed ‘cannot afford’, ‘not 

knowing what it is’, and ‘feeling healthy’ as the 

main barriers to screening. The successful 

implementation and outcome of our intervention 

suggest that enhancing knowledge and access to 

CRC screening through community-clinical linkage 

is the effective way to improve CRC screening, 

particularly among the hard-to-reach underserved 

Korean American population who experience 

multiple barriers to screening.  

Community participation played a crucial role in 

raising awareness, facilitating motivation about 

cancer screening, and reducing logistical 



 
 
 
 
 

 

www.companyofscientists.com/index.php/chd                             e13                                    Cancer Health Disparities 

RESEARCH 

challenges among this population when clinical 

settings have the limitation that they can provide 

services only to individuals who visit the clinical 

settings. Offering education and low-cost FIT kits 

through community settings made it possible to 

reach the underserved Korean Americans and 

engage them in screening behaviors. On the other 

hand, physician engagement compensated the 

limited ability of community settings in delivery of 

clinical care and warranted quality services for the 

participants in need. This community-clinical 

linkage intervention seemed to be an essential 

component in enhancing screening behavior for 

hard-to-reach populations with multiple barriers to 

screening.   

The finding that the intervention was particularly 

effective in promoting FIT is also consistent with 

the previous studies suggesting that when offered, 

FIT is accepted better than colonoscopy among 

the underserved populations (Inadomi et al., 2012; 

Singal et al., 2016). We did not design our study to 

compare FIT versus colonoscopy, but instead 

incorporated low-cost FIT kit provision into a 

multifaceted intervention that included education 

sessions addressing all types of CRC screening and 

navigation assistance to 

colonoscopy/sigmoidoscopy appointments, 

allowing the participants to select the methods of 

CRC screening. Despite the different options, the 

majority of the participants in the intervention 

group selected FIT, confirming the previous 

studies. Moreover, we found that the difference in 

screening rates between intervention and control 

groups were greater among those who were less 

educated, having less income, unemployed, 

unable to speak English, with no health insurance 

or a regular physician, suggesting that our 

intervention is more beneficial among the 

disadvantaged individuals. In contrary, relatively 

better-off participants did not seem to be as 

receptive to the resources offered by intervention. 

For instance, the three paired church sites with the 

lowest screening rates after intervention were 

attended by the participants with generally higher 

education, higher income, and greater health care 

access compared to their counterparts at the other 

paired church sites where the screening rates were 

higher. These findings highlight a promising 

potential of a multifaceted intervention combining 

FIT provision with CBPR approach for enhancing 

screening behavior among the most socially and 

economically disadvantaged members in racial 

minority communities.  

Furthermore, we found that certain demographic 

characteristics and barriers to screening were 

associated with different modalities of CRC 

screening. Multivariate analysis revealed that 

different modalities of CRC screening seemed to 

appeal to different subgroups of participants. 

Controlling for intervention effect, FIT screening 

behavior was associated with high school 

educated, lower income, males, and those without 

health insurance. In contrast, 

colonoscopy/sigmoidoscopy screening behavior 

was associated with participants who have health 

insurance and better English proficiency. Among 

the barriers, ‘being afraid of pain’ was associated 

with increased FIT whereas, ‘not feeling healthy’ 

was associated with increased colonoscopy. These 

results indicate that the less invasive, low-cost, and 

more convenient FIT screening test are preferred 

by disadvantaged groups facing generally more 

barriers to screening. The more invasive but more 

sensitive and specific colonoscopy/sigmoidoscopy 

screening modality seems to be preferred among 

individuals who typically have health care access 

and have health concerns. Future work is needed 

to gain a deeper understanding of the factors that 

predetermine preference of CRC screening 

modality among racial minorities. 

There are several limitations to this study. First, the 

present study had a short follow-up period which 



 
 
 
 
 

 

www.companyofscientists.com/index.php/chd                             e14                                    Cancer Health Disparities 

RESEARCH 

only studied one-time screening behavior at 12-

month follow up. Since FIT screening warrants 

annual testing to be effective in CRC prevention 

(Singal et al., 2017), full understanding about 

screening behavior compliance requires a longer 

follow-up study to examine sustained adherence 

over time. In a previous long term comparative 

outreach study, the rates of screening decreased 

as low as 28% when participants were followed up 

for 3 years (Singal et al., 2017) compared to 1-time 

screening rate of 58.8% (Min, 2002). The authors 

of the study attributed the lower adherence rates 

in the FIT outreach group to the suboptimal 

completion of colonoscopy after a positive FIT 

result, in addition to the failures of completing 

subsequent yearly FIT screens after initial negative 

test results. However, given that 100% of those 

who had positive FIT test results in the intervention 

group underwent diagnostic colonoscopy with 

navigation assistance in our study, compared to 

49.3% in the outreach study for 1-time screening 

(Min, 2002), our study shed the light on the 

potential impact of the intervention incorporating 

CBPR approach and navigation service to FIT 

provision on drawing long-term effect with 

improved sustained adherence to FIT. Second, 

despite the randomization procedures, the two 

groups differed on several demographic variables, 

including household income, health insurance 

status, and barriers to screening. However, 

participants in the control group were more likely 

to have higher income and health insurance, which 

are the factors that generally facilitate access to 

care. Significantly lower screening rates in the 

control group even with this advantage confirm 

the true robustness of the intervention used in the 

present study.  

Despite the limitations, our findings support the 

feasibility and efficacy of a culturally and 

linguistically appropriate multifaceted community-

clinical linkage intervention among the most 

underserved members in a high-risk racial minority 

population. The study offers strong evidence that 

such an intervention approach can be a practical 

intervention method to effectively target hard-to-

reach disadvantaged individuals who are otherwise 

vulnerable to CRC screening disparities. Future 

studies will look to investigate deeper into 

preference of CRC screening modality among 

Korean Americans and evaluate screening 

behavior over a longer follow-up period, as well as 

explore effective strategies to scale up the 

evidence-based community-clinical linkage 

intervention for broad dissemination and 

implementation. 

Acknowledgements 

This research was supported by Center to Reduce 

Cancer Health Disparities, National Cancer 

Institute, National Institutes of Health (NIH) (grant 

U54 CA153513 to PI: Grace X. Ma, PhD, 

Community Cancer Health Disparities Center. The 

authors wish to thank the Asian Community Health 

Coalition and its member community-based 

organizations for their collaboration and 

contributions. The contents of this article are solely 

the responsibility of the authors and do not 

necessarily represent the official views of the 

funding agency, NIH. 

 

Conflict of interest statement 

The author has declared that no competing or 

conflict of interests exist. The funders had no role 

in study design, writing of the manuscript and 

decision to publish.    

 

Authors’ contributions 

Please add. 
 

REFERENCES 
American Cancer Society (2014). Colorectal Cancer Facts & 

Figures 2017-2019 (Atlanta, GA). 

American Cancer Society (2017). Colorectal Cancer Facts & 

Figures | Facts About Colon Cancer. 

Center for Disease Control and Prevention (2017). Colorectal 

Cancer Screening Tests. 



 
 
 
 
 

 

www.companyofscientists.com/index.php/chd                             e15                                    Cancer Health Disparities 

RESEARCH 

Chang, E.T. (2003). Korean Americans. 

Elmunzer, B.J., Singal, A.G., Sussman, J.B., Deshpande, A.R., 

Sussman, D.A., Conte, M.L., Dwamena, B.A., Rogers, 

M.A.M., Schoenfeld, P.S., Inadomi, J.M., et al. (2015). 

Comparing the effectiveness of competing tests for 

reducing colorectal cancer mortality: a network meta-

analysis. Gastrointest. Endosc. 81, 700–709.e3. 

Giddings, B.H., Kwong, S.L., Parikh-Patel, A., Bates, J.H., and 

Snipes, K.P. (2012). Going against the tide: increasing 

incidence of colorectal cancer among Koreans, Filipinos, 

and South Asians in California, 1988-2007. Cancer Causes 

Control CCC 23, 691–702. 

Gupta, S., Halm, E.A., Rockey, D.C., Hammons, M., Koch, M., 

Carter, E., Valdez, L., Tong, L., Ahn, C., Kashner, M., et al. 

(2013). Comparative effectiveness of fecal 

immunochemical test outreach, colonoscopy outreach, 

and usual care for boosting colorectal cancer screening 

among the underserved: a randomized clinical trial. 

JAMA Intern. Med. 173, 1725–1732. 

Hwang, H. (2013). Colorectal cancer screening among Asian 

Americans. Asian Pac. J. Cancer Prev. APJCP 14, 4025–

4032. 

Inadomi, J.M., Vijan, S., Janz, N.K., Fagerlin, A., Thomas, J.P., 

Lin, Y.V., Muñoz, R., Lau, C., Somsouk, M., El-Nachef, N., 

et al. (2012). Adherence to colorectal cancer screening: a 

randomized clinical trial of competing strategies. Arch. 

Intern. Med. 172, 575–582. 

Israel, B.A., Eng, E., Schulz, A.J., Parker, E.A., and Satcher, D. 

(2005). Methods in Community-Based Participatory 

Research for Health (San Francisco, CA: Jossey-Bass). 

Jo, A.M., Maxwell, A.E., Wong, W.K., and Bastani, R. (2008). 

Colorectal Cancer Screening among Underserved Korean 

Americans in Los Angeles County. J. Immigr. Minor. 

Health Cent. Minor. Public Health 10, 119–126. 

Juon, H.-S., Han, W., Shin, H., Kim, K.B., and Kim, M.T. (2003). 

Predictors of older Korean Americans’ participation in 

colorectal cancer screening. J. Cancer Educ. Off. J. Am. 

Assoc. Cancer Educ. 18, 37–42. 

Krist, A.H., Shenson, D., Woolf, S.H., Bradley, C., Liaw, W.R., 

Rothemich, S.F., Slonim, A., Benson, W., and Anderson, 

L.A. (2013). Clinical and Community Delivery Systems for 

Preventive Care: An Integration Framework. Am. J. Prev. 

Med. 45, 508–516. 

Lee, H.Y., and Im, H. (2013). Colorectal Cancer Screening 

among Korean American Immigrants: Unraveling the 

Influence of Culture. J. Health Care Poor Underserved 24, 

579–598. 

Lee, S.-Y., and Lee, E.E. (2013). Korean Americans’ beliefs 

about colorectal cancer screening. Asian Nurs. Res. 7, 

45–52. 

Lee, J., Demissie, K., Lu, S.-E., and Rhoads, G.G. (2007). Cancer 

incidence among Korean-American immigrants in the 

United States and native Koreans in South Korea. Cancer 

Control J. Moffitt Cancer Cent. 14, 78–85. 

Lieberman, D.A., De Garmo, P.L., Fleischer, D.E., Eisen, G.M., 

and Helfand, M. (2000). Patterns of endoscopy use in the 

United States. Gastroenterology 118, 619–624. 

Ma, G.X., Shive, S., Tan, Y., Gao, W., Rhee, J., Park, M., Kim, J., 

and Toubbeh, J.I. (2009). Community-based colorectal 

cancer intervention in underserved Korean Americans. 

Cancer Epidemiol. 33, 381–386. 

Maxwell, A.E., and Crespi, C.M. (2009). Trends in Colorectal 

Cancer Screening Utilization among Ethnic Groups in 

California: Are We Closing the Gap? Cancer Epidemiol. 

Biomark. Prev. Publ. Am. Assoc. Cancer Res. 

Cosponsored Am. Soc. Prev. Oncol. 18, 752–759. 

Maxwell, A.E., Crespi, C.M., Antonio, C.M., and Lu, P. (2010). 

Explaining disparities in colorectal cancer screening 

among five Asian ethnic groups: A population-based 

study in California. BMC Cancer 10, 214. 

Miller, B.A., Chu, K.C., Hankey, B.F., and Ries, L.A.G. (2008). 

Cancer incidence and mortality patterns among specific 

Asian and Pacific Islander populations in the U.S. Cancer 

Causes Control 19, 227–256. 

Min, P.G. (2002). Mass migration to the United States: Classical 

and contemporary periods (Walnut Creek, CA: AltaMira 

Press). 

Ockene, J.K., Edgerton, E.A., Teutsch, S.M., Marion, L.N., Miller, 

T., Genevro, J.L., Loveland-Cherry, C.J., Fielding, J.E., and 

Briss, P.A. (2007). Integrating Evidence-Based Clinical and 

Community Strategies to Improve Health. Am. J. Prev. 

Med. 32, 244–252. 

Oh, K.M., and Jacobsen, K.H. (2014). Colorectal cancer 

screening among Korean Americans: a systematic review. 

J. Community Health 39, 193–200. 

Persson, K. (2016). Integrating Clinical Care with Community 

Health through New Hampshire’s Million Hearts Learning 

Collaborative: A Population Health Case Report. 

Singal, A.G., Gupta, S., Tiro, J.A., Skinner, C.S., McCallister, K., 

Sanders, J.M., Bishop, W.P., Agrawal, D., Mayorga, C.A., 

Ahn, C., et al. (2016). Outreach invitations for FIT and 

colonoscopy improve colorectal cancer screening rates: 

A randomized controlled trial in a safety-net health 

system. Cancer 122, 456–463. 

Singal, A.G., Gupta, S., Skinner, C.S., Ahn, C., Santini, N.O., 

Agrawal, D., Mayorga, C.A., Murphy, C., Tiro, J.A., 

McCallister, K., et al. (2017). Effect of Colonoscopy 

Outreach vs Fecal Immunochemical Test Outreach on 

Colorectal Cancer Screening Completion: A Randomized 

Clinical Trial. JAMA 318, 806–815. 

Tu, S.-P., Taylor, V., Yasui, Y., Chun, A., Yip, M.-P., Acorda, E., 

Li, L., and Bastani, R. (2006). Promoting culturally 

appropriate colorectal cancer screening through a health 

educator. Cancer 107, 959–966. 

 

 


