WK 6 DIS DATA
Walden University Walden University
ScholarWorks ScholarWorks
Walden Dissertations and Doctoral Studies Walden Dissertations and Doctoral Studies Collection
2023
Language Barrier: An Unmet Challenge for Low Screening of Language Barrier: An Unmet Challenge for Low Screening of
Colorectal Cancer Among Hispanic Americans in Texas Colorectal Cancer Among Hispanic Americans in Texas
Moses Owusu Walden University
Follow this and additional works at: https://scholarworks.waldenu.edu/dissertations
Part of the Epidemiology Commons, and the Public Health Education and Promotion Commons
This Dissertation is brought to you for free and open access by the Walden Dissertations and Doctoral Studies Collection at ScholarWorks. It has been accepted for inclusion in Walden Dissertations and Doctoral Studies by an authorized administrator of ScholarWorks. For more information, please contact [email protected].
Walden University
College of Health Sciences and Public Policy
This is to certify that the doctoral dissertation by
Moses Owusu
has been found to be complete and satisfactory in all respects,
and that any and all revisions required by
the review committee have been made.
Review Committee
Dr. Simone Salandy, Committee Chairperson, Public Health Faculty
Dr. Howell Sasser, Committee Member, Public Health Faculty
Dr. Loretta Shields, University Reviewer, Public Health Faculty
Chief Academic Officer and Provost
Sue Subocz, Ph.D.
Walden University
2023
Abstract
Language Barrier: An Unmet Challenge for Low Screening of Colorectal Cancer Among
Hispanic Americans in Texas
by
Moses Owusu
MBA, Plymouth State University, 2015
MS, New School University, 2000
B.Sc., Kwame Nkrumah University of Science and Technology, 1994
Dissertation Submitted in Partial Fulfillment
of the Requirements for the Degree of
Doctor of Philosophy
Public Health: Epidemiology
Walden University
August 2023
Abstract
Hispanic population represents the fastest growing minority group in the United States,
and only California surpasses Texas with Hispanic residents in the United States. Overall,
non-Hispanic Whites have higher rates of colorectal cancer (CRC) incidence than
Hispanics; however, Hispanic Americans have lower survival rates than non-Hispanic
Whites. Using cross-sectional analysis, CRC screening modalities were examined to
assess disparities among White only, non-Hispanic, Black only, non-Hispanic, and
Hispanic populations in Texas to evaluate the impact of limited English proficiency
(LEP) on CRC screening among Hispanic Americans residents. The age of study
participants ranged from 50 to 79 (mean age = 65.8 years), and the data were obtained
from the 2020 Texas Behavioral Risk Factor Surveillance System (BRFSS) survey. The
primary outcome was self-reported CRC screening status by the respondents. The study
had a sample population, N =766, chosen randomly with 68.5% White only, non-
Hispanic, 10.1% Black only, non-Hispanic, and 21.4% Hispanic participants. Pearson
Chi-square test was used to compare CRC screening rates and modalities and bivariate
and multivariate logistic regression to determine predictors of CRC screening among
participants in the study. The Chi-square tests indicated that there was a statistically
significant association between LEP and non-LEP respondents. The findings showed that
Hispanics with LEP have low CRC screening rates. Suggestions for positive social
change include improvement in CRC screening using social centers to promote CRC
screening, promotion of health literacy and transportation accessibility for vulnerable
communities without access to public transport.
Language Barrier: An Unmet Challenge for Low Screening of Colorectal Cancer Among
Hispanic Americans in Texas
by
Moses Owusu
MBA, Plymouth State University, 2015
MS, New School University, 2000
B.Sc., Kwame Nkrumah University of Science and Technology, 1994
Dissertation Submitted in Partial Fulfillment
of the Requirements for the Degree of
Doctor of Philosophy
Public Health: Epidemiology
Walden University
August 2023
Dedication
This dissertation is dedicated to my late mother, Madam Deborah Akosua
Ankamaa; my wife, Rosemary Barney Owusu; and our children, Patricia, Mimi, Nana,
Adwoa, Brenda, and David. I also appreciate the unshakable support from Mr. Kofi
Fofie, Elder Francis Koosono, Mr. Augustine Amankona (Peace) (brothers), Auntie
Agnes Mmuah, and my sisters in Berekum, Ghana. I am also highly indebted to Dr.
Francis Asomah (Australia) who mentored me in academic pursuits during my high
school days and remains a bedrock supporter in this journey. Similarly, I am grateful to
Dr. Godfred Owusu-Boateng of Kwame Nkrumah University of Science and Technology
(Kumasi, Ghana); Mr. Abor Yeboah (Walter Sisulu University, South Africa); and
nieces, Salome (Mayo Clinic, Rochester, MN) and Francisca (Berekum, Ghana); Mad.
Charlotte Awuku Boateng (mother-in-law); Ms. Stella Green; Dr. Elizabeth Hagan
Asamoah; Ms. Leslie Gillman (Walden University); Mr. Patrick Obeng; Ms. Wilhelmina
Asabea Adu-Darko; Rev. James Acquaah and Ms. Lydia Acquaah; and my nephews,
nieces, family members, distant relatives, and colleagues who encouraged me along the
way. Finally, I understand that I could not have completed this journey without the Lord
Almighty and His strength and grace. All the glory is accredited unto Him for making
this journey fruitful to its successful end.
Acknowledgments
I wish to express my profound gratitude to special people who contributed in
numerous ways to make my dissertation a success. I have successfully come to the
pinnacle of this journey because the Lord Almighty granted me an aura of reassuring
solidity through many people who held my hand in friendship and support. Still, many
others selflessly availed themselves as a springboard to carry me through the journey. My
deepest appreciation goes to my dissertation chair, Dr. Simone W. Salandy for guidance,
patience, encouragement, and prompt responses to all dissertation concerns and
commitment to my success in excelling through the thick and thin of academic trajectory
that has finally become a reality of success. In a similar way, Dr. Howard C. Sasser, my
second committee member, whom I was blessed to have in a pre-dissertation course,
played an irreplaceable role, forming a triad together with my chair and me to ensure that
this journey was destined for a successful completion. Dr. Sasser, thank you for being
around both day and night with an unending patience and boundless epidemiological
knowledge that sailed me through quantitative methodology and its intricacies and
nuances. My gratitude also goes to Dr. Loretta R. Shields, the URR, whose contributions
and expertise ensured that this dissertation met Walden University’s highest doctoral
standard. I would also like to thank my academic advisor, Mr. Garth Woodbury, whose
prompt emails always alerted me of important deadlines. Equally, my sincere thanks goes
to Dr. Christy Fraenza (Lead, Instructional Design at Walden University Peer Mentors
program); Ms. Cassie Fox (MPH) of Texas Department of State Health Services for
BRFSS data for analysis; Ms. Libby Munson, Research Ethics Support Specialist at IRB;
and Ms. Bethany G. Duarte, Senior Editor, Form and Style of the Office of Research and
Doctoral Services, both of Walden University.
I am very grateful to my wife, Rosemary Barney Owusu, whose support and
sacrifices in diverse ways, eased a lot of family stress and ensured a surgical focus on
academic work. Importantly, I’m equally grateful for the patience and understanding
exhibited by my children, Mimi, Nana, Adwoa, and Patricia for all their concerns and
encouragements. I am particularly indebted to Mr. Isaac Boateng and his wife, Mrs. Lily
Gaba Boateng, for their boundless brotherly love, motivation, and support that calmed
tumultuous and stormy periods, making this journey a successful one. My gratitude also
goes to the Rt. Rev. Dr. Yaw Frimpong-Manso (Papa), former Moderator of the
Presbyterian Church of Ghana, and his wife, Mrs. Lucy Frimpong-Manso whose
towering presence of motivation served as an illuminating power of hope during this
academic journey. Papa, thank you! My deepest appreciation also goes to Rev. Ebenezer
Darko Boateng, Mrs. Ellen Boateng, and the entire congregation of the Presbyterian
Church of the Redeemer in Houston, USA, for all their motivational support. Special
thanks to Mr. Robert Kyeremateng, Mrs. Jane Seweh Kyeremateng, Mr. Joseph Danquah,
Mrs. Cecilia Amponsa Danquah, Mr. Yaw Adjei, Mr. James Brobbey, Mrs. Grace Asiedu
Mensah, Ms. Doris Tachie, Ms. Florence Serwaa, Mr. Isaac O. Boateng, Mrs. Emma
Boateng, Mr. Aurelio Lozano (Paul, Weiss, Rifkind, Wharton & Garrison, LLP, NYC),
Nana Yaw Ogyiri, Mr. Samuel Kwabena Britwum Carr, and Mrs. Mary Grant Carr for all
their moral support.
i
Table of Contents
List of Tables .......................................................................................................................v
Chapter 1: Introduction to the Study ....................................................................................1
Background of the Study ...............................................................................................4
Problem Statement .........................................................................................................6
Purpose of the Study ......................................................................................................9
Research Questions and Hypothesis ............................................................................10
Framework: Conceptual or Theoretical .......................................................................12
Nature of the Study ......................................................................................................15
Definitions of Study Variables .....................................................................................17
Assumptions .................................................................................................................19
Scope and Delimitations ..............................................................................................19
Limitations ...................................................................................................................20
Significance of the Study .............................................................................................21
Summary ......................................................................................................................22
Chapter 2: Literature Review .............................................................................................24
Introduction ..................................................................................................................24
Theoretical Foundation ................................................................................................26
Conceptual Framework: Health Belief Model .............................................................27
Literature Review Related to Key Variables and/or Concepts ....................................36
CRC Screening...................................................................................................... 36
Health-related LEP................................................................................................ 52
ii
Gender and CRC Screening .................................................................................. 57
Cultural Influences on CRC Screening ................................................................. 59
Methodology ......................................................................................................... 60
Summary and Conclusions ..........................................................................................72
Chapter 3: Research Method ..............................................................................................75
Overview ......................................................................................................................75
Instrument to Measure Limited English Proficiency ............................................ 76
Research Design and Rationale ...................................................................................79
Population ............................................................................................................. 81
Sampling Design ................................................................................................... 82
Data Collection ............................................................................................................84
Variables ............................................................................................................... 86
Threats to Validity .......................................................................................................87
Descriptive Statistical Analysis ...................................................................................90
Inferential Statistics .....................................................................................................91
Test of Assumption ......................................................................................................92
Continuous and Categorical Variables.........................................................................93
Inferential Statistical Analysis .....................................................................................93
Ethical Procedures .......................................................................................................94
Summary ......................................................................................................................96
Chapter 4: Results ..............................................................................................................98
Introduction ..................................................................................................................98
iii
Research Questions and Hypotheses ...........................................................................99
Data Analysis .............................................................................................................101
Descriptive Statistics ..................................................................................................101
Sociodemographic Factors .................................................................................. 101
Inferential Statistical Analysis ............................................................................ 149
Model with Interaction Effects ........................................................................... 150
Interaction of the Independent Variable ............................................................. 154
Summary ....................................................................................................................164
Chapter 5: Discussion, Conclusions, and Recommendations ..........................................166
Introduction ................................................................................................................166
Sample Description Summary ...................................................................................167
Interpretation of Findings ..........................................................................................168
Participants Grouped by Age .............................................................................. 168
Sex…................................................................................................................... 169
Race…................................................................................................................. 170
Language ............................................................................................................. 171
Income (Annual Household) ............................................................................... 171
Employment Status ............................................................................................. 172
Healthcare Access ............................................................................................... 173
Education Status.................................................................................................. 173
Colorectal Screening Types .......................................................................................174
Colonoscopy ....................................................................................................... 174
iv
Sigmoidoscopy .................................................................................................... 180
Fecal Occult Blood Test...................................................................................... 182
Sensitivity, Compliance, and Access of CRC Screening Types ................................183
Gender Influences on CRC Screening ................................................................ 184
Research Questions and Hypotheses .........................................................................188
Analysis: The Health Belief Model Conceptual Framework .....................................192
Mediating Effects of Limited English Proficiency ....................................................194
Summary ....................................................................................................................197
Strengths and Limitations ..........................................................................................198
Recommendations ......................................................................................................200
Implications................................................................................................................202
Social Change ............................................................................................................204
Conclusion .................................................................................................................205
References ........................................................................................................................206
Appendix: Abbreviations .................................................................................................252
v
List of Tables
Table 1 Sample Size per Instrumental Variables ........................................................... 101
Table 2 Sex of Participants ............................................................................................ 102
Table 3 Frequency Table 1: Five-year Groups, Sex, Race, Language .......................... 106
Table 4 ............................................................................................................................ 107
Table 5 Descriptive Statistics of Study Population Based on Screened and Unscreened
for Colorectal Cancer (N=766). .............................................................................. 108
Table 6 Five-year Age Groups and Colorectal Screening Status .................................. 109
Table 7 Chi-Square Tests on Five-Year Age Groups .................................................... 110
Table 8 Symmetric Measures on Five-year Age Groups ............................................... 110
Table 9 Sex of Participants and Colorectal Screening Status ........................................ 111
Table 10 Chi-Square Tests on Sex of Participants and Colorectal Screening Status .... 111
Table 11 Symmetric Measures on Sex of Participants and Colorectal Screening Status
................................................................................................................................. 112
Table 12 Participant Race and Colorectal Screening Status .......................................... 113
Table 13 Chi-Square Tests on Race of Participants and Colorectal Screening Status .. 113
Table 14 Symmetric Measures on Participant Race and Colorectal Status ................... 114
Table 15 Language of Participants and Colorectal Screening Status ............................ 114
Table 16 Chi-Square Tests on Participants’ Language and Colorectal Screening Status
................................................................................................................................. 115
Table 17 Symmetric Measures on Language of Participants and Colorectal Screening
Status ....................................................................................................................... 115
vi
Table 18 Employment Status and Colorectal Cancer Screening ................................... 116
Table 19 Chi-Square Tests on Employment of Participants and Colorectal Screening
Status ....................................................................................................................... 117
Table 20 Symmetric Measures on Employment of Participants and Colorectal Screening
Status ....................................................................................................................... 117
Table 21 Healthcare Access and Colorectal Cancer Screening ..................................... 118
Table 22 Chi-Square Tests on Healthcare Access of Participants and Colorectal
Screening Status ...................................................................................................... 118
Table 23 Symmetric Measures on Health Care Access of Participants and Colorectal
Screening Status ...................................................................................................... 119
Table 24 Group Statistics for English Speaking and Spanish Speaking Participants in
CRC Screening........................................................................................................ 120
Table 25 One Sample Statistics 1: Independent Samples Test on Colorectal Cancer
Screening and Age of Participants .......................................................................... 120
Table 26 One Sample Statistics 2: Independent Samples Test on Colorectal Cancer
Screening and Age of Participants .......................................................................... 121
Table 27 Number of Participants (N), Mean, Std Deviation, and Std Error Mean and CRC
Screening and Participants with Income and Education ......................................... 122
Table 28 One-Sample T Test for CRC Screening with Income and Education of
Participants .............................................................................................................. 122
Table 29 Crosstabulation of Five-year Age Groups and Colonoscopy Screening Status
................................................................................................................................. 123
vii
Table 30 Chi-Square Tests on Five-year Age Groups and Colonoscopy Screening Status
................................................................................................................................. 123
Table 31 Symmetric Measures on Five-year Age Groups and Colorectal Screening Status
................................................................................................................................. 124
Table 32 Sex of Participants and Colonoscopy Screening Status .................................. 124
Table 33 Chi-Square Tests on Males and Females and Colonoscopy Screening Status 125
Table 34 Symmetric Measures on Sex of Participants and Colorectal Screening Status
................................................................................................................................. 125
Table 35 Race of Participants and Colonoscopy Screening Status................................ 126
Table 36 Chi-Square Tests on Race of Participants and Colonoscopy Screening Status
................................................................................................................................. 127
Table 37 Symmetric Measures on Race of Participants and Colonoscopy Screening
Status ....................................................................................................................... 127
Table 38 Language of Participants and Colonoscopy Screening Status ........................ 128
Table 39 Chi-square Tests on Language of Participants and Colonoscopy Screening
Status ....................................................................................................................... 128
Table 40 Symmetric Measures on Language of Participants and Colonoscopy Screening
Status ....................................................................................................................... 129
Table 41 Five-year Age Groups and Colonoscopy Screening Status Within the Past 10
Years ....................................................................................................................... 129
Table 42 Chi-square Tests on Five-year Age Groups and Colonoscopy Screening Status
Within the Past 10 Years......................................................................................... 130
viii
Table 43 Symmetric Measures on Five-year Age Groups and Colonoscopy Screening
Status Within the Past 10 Years .............................................................................. 130
Table 44 Sex of Participants and Colonoscopy Screening Status Within the Past 10 Years
................................................................................................................................. 131
Table 45 Chi-Square Tests on Sex of Participants and Colonoscopy Screening Status
Within the Past 10 Years......................................................................................... 131
Table 46 Symmetric Measures on Sex of Participants and Colonoscopy Screening Status
Within the Past 10 Years......................................................................................... 132
Table 47 Race of Participants and Colonoscopy Screening Status Within the Past 10
Years ....................................................................................................................... 133
Table 48 Chi-Square Tests on Race of Participants and Colonoscopy Screening Status
Within the Past 10 Years......................................................................................... 133
Table 49 Symmetric Measures on Race of Participants and Colonoscopy Screening
Status Within the Past 10 Years .............................................................................. 134
Table 50 Language of Participants and Colonoscopy Screening Status Within the Past 10
Years ....................................................................................................................... 134
Table 51 Chi-Square Tests on Language of Participants and Colonoscopy Screening
Status Within the Past 10 Years .............................................................................. 135
Table 52 Symmetric Measures on Language of Participants and Colonoscopy Screening
Status Within the Past 10 Years .............................................................................. 135
Table 53 Five-year Age Groups and Sigmoidoscopy Screening Status Within the Past
Five Years ............................................................................................................... 136
ix
Table 54 Chi-Square Tests on Five-year Age Groups and Sigmoidoscopy Screening
Status Within the Past Five Years ........................................................................... 137
Table 55 Symmetric Measures on Five-year Age Groups and Sigmoidoscopy Screening
Status Within the Past Five Years ........................................................................... 137
Table 56 Sex of Participants and Sigmoidoscopy Screening Status Within the Past Five
Years ....................................................................................................................... 138
Table 57 Chi-Square Tests on Sex of Participants and Sigmoidoscopy Screening Status
Within the Past Five Years ..................................................................................... 138
Table 58 Symmetric Measures on Sex of Participants and Sigmoidoscopy Screening
Status Within the Past Five Years ........................................................................... 139
Table 59 Race of Participants and Sigmoidoscopy Screening Status Within the Past Five
Years ....................................................................................................................... 140
Table 60 Chi-Square Tests on Race of Participants and Sigmoidoscopy Screening Status
Within the Past Five Years ..................................................................................... 140
Table 61 Symmetric Measures on Race of Participants and Sigmoidoscopy Screening
Status Within the Past Five Years ........................................................................... 141
Table 62 Language of Participants and Sigmoidoscopy Screening Status Within the Past
Five Years ............................................................................................................... 141
Table 63 Chi-Square Tests on Language of Participants and Sigmoidoscopy Screening
Status Within the Past Five Years ........................................................................... 142
Table 64 Symmetric Measures on Language of Participants and Sigmoidoscopy
Screening Status Within the Past Five Years .......................................................... 142
x
Table 65 Five-year Age Groups and Blood Stool Test Within the Past Year ............... 143
Table 66 Chi-Square Tests for Five-year Age Groups and Blood Stool Test Within the
Past Year ................................................................................................................. 144
Table 67 Symmetric Measures on Five-year Age Groups and Blood Stool Test Within
the Past Year ........................................................................................................... 144
Table 68 Sex of Participants and Blood Stool Test Within the Past Year ..................... 145
Table 69 Chi-Square Tests on Sex of Participants and Blood Stool Test Within the Past
Year ......................................................................................................................... 145
Table 70 Symmetric Measures on Sex and Blood Stool Test Within the Past Year ..... 146
Table 71 Race of Participants and Blood Stool Test Within the Past Year ................... 146
Table 72 Chi-Square Tests on Race of Participants and Blood Stool Test Within the Past
Year ......................................................................................................................... 147
Table 73 Symmetric Measures on Race and Blood Stool Test Within the Past Year ... 147
Table 74 Language of Participants and Blood Stool Test Within the Past Year ........... 148
Table 75 Chi-Square Tests on Language of Participants and Blood Stool Test Within the
Past Year ................................................................................................................. 148
Table 76 Symmetric Measures on Language of Participants and Blood Stool Test Within
the Past Year ........................................................................................................... 149
Table 77 DV Encoding .................................................................................................. 150
Table 78 Classification Table: Colorectal Cancer Status .............................................. 150
Table 79 Omnibus Tests of Model Coefficients ............................................................ 151
Table 80 Hosmer and Lemeshow Test........................................................................... 151
xi
Table 81 Model Summary ............................................................................................. 151
Table 82 Improved Predictor Variables ......................................................................... 152
Table 83 Variables in the Equation ................................................................................ 153
Table 84 Parameter Estimates of Variables on Colorectal Cancer: Yes - Screened for
Colorectal Status ..................................................................................................... 155
Table 85 Parameter Estimates of Variables on Colorectal Cancer: Yes - Had a
Colonoscopy ........................................................................................................... 156
Table 86 Parameter Estimates of Variables on Colorectal Cancer: Yes - Colonoscopy
Within Past 10 Years .............................................................................................. 157
Table 87 Parameter Estimates of Variables on Colorectal Cancer: Yes - Sigmoidoscopy
Within Past 5 Years ................................................................................................ 158
Table 88 Parameter Estimates of Variables on Colorectal Cancer: Yes – Blood Stool Test
Within the Past Year ............................................................................................... 159
Table 89 Case Processing Summary .............................................................................. 160
Table 90 Observations and Predicted Frequencies of Males and Females Among the
Racial Groups in the Study ..................................................................................... 162
1
Chapter 1: Introduction to the Study
Colorectal cancer (CRC) is the third leading cause of cancer mortality in both
men and women in the United States but assumes a second position when men and
women are put together (Siegel et al., 2020). Although CRC morbidity and mortality
could be reduced through relevant screening and surveillance measures, such preventive
steps are often lacking, particularly among minorities in the United States (Jackson et al,
2016). For 2021, the American Cancer Society (ACS) estimated that CRC incidence
cases were estimated to be 149,500, and 52,980 were expected to die from it. Researchers
argue that 68% of deaths could be prevented with screening, and CRC at its most
treatable state (stage I), it is 90% curable. However, only 38% of CRCs are diagnosed at
stage I; that number is further reduced for individuals diagnosed before age 50, whereas
approximately 10% often receive late-stage disease (stages III and IV), even though the
ACS guidelines recommend screening to be started at 45 years of age (Siegel et al.,
2021).
Nationwide, data indicate that African Americans have the highest increased
burden of CRC compared to other ethnic groups such as Whites and Asians, with higher
incidence, worse outcomes, and earlier-onset of disease (Siegel et al., 2017). However, in
recent years, data demonstrate that like African Americans, the proportion of the
Hispanic population younger than age 50 and diagnosed with CRC is nearly double the
rate seen in Whites (12% vs. 6.7%; Rahman et al., 2015). This is the case even though
Hispanics have lower overall CRC incidence than Whites across all age groups, including
for individuals younger than 50 years of age (Muller et al., 2021). Yet, recent
2
epidemiological data show that the incidence of early-onset CRC is rising at a faster rate
among Hispanic people (Ellis et al., 2018), with Hispanic men among the fastest growing
demographic of early-onset CRC with an annual increase of 2.7% from 1992 to 2005
(Siegel et al., 2009). In contrast, the incidence for Hispanic women increased only 1.2%
annually during the same time (Siegel et al., 2009). In total, analysis indicates that annual
incidence of early-onset CRC among Hispanics has increased 2.35%, as compared to
2.02% for Whites, and these data reflect similar trends in statewide data obtained from
California and Texas (Wang et al., 2017).
According to the Census Bureau, between 2000 and 2010, the United States total
population grew steadily at a rate of 9.7% (United States Census Bureau, 2011). More
than half of the growth was attributed to the rising population of Hispanics, growing from
35.3 million in 2000 to 50.5 million in 2010, a soaring rate of 43%. The increase in the
Hispanic population accounted for over half the 27.3 million increase in the total United
States population within the period. In 2010, Hispanics comprised 16% of the total
United States population of 308.7 million (United States Census Bureau, 2011),
ballooning to 18.5% or 60.6 million in 2020 (United States Census Bureau, 2020). A
significant point reveals that in 2019, 30.8% of Hispanics were under the age 18, in
comparison to 18.6% of non-Hispanic whites, based on data available at the Office of
Minority Health (OMH) at the United States Department of Health and Human Services
(HHS; 2021).
Notwithstanding their soaring population growth, English language fluency is a
major challenge for many Hispanics. Analysis from the Census Bureau (2019) data on the
3
profile of Hispanic and Latino Americans demonstrated that English language fluency
varies among Hispanic subgroups who reside within the continental United States. The
data indicate that 71.1% of Hispanics speak a language other than English at home
(OMH, 2021). For example, 70.4% of Mexicans, 58.9% of Puerto Ricans, 77.7% of
Cubans, and 86.2% of Central Americans speak different languages at home, while
28.4% of the overall Hispanics state that they are not fluent in English (OMH, 2021).
Studies indicate that limited English proficiency (LEP) individuals have difficulty
reading, writing, and understanding English (Genoff et al., 2016), which creates
hindrances to participation in the English-language dominant healthcare system in the
United States. Language impediments contribute to poor health processes and outcomes
(Al Shamsi et al., 2020), such as reduced access of preventive services and cancer
screening rates among LEP patients (Tatari et al., 2020).
As the largest ethnic minority group in the United States, the rising CRC
incidence among the Hispanic group needs a serious public health attention that could
influence policymaking to reduce its impact on the population. Barriers, such as language
and other cultural considerations, need to be addressed to enhance screening of CRC in
this population as a means of addressing its adverse effects on the population. This study
focuses on language as an obstacle that leads to challenges for low screening of CRC
among Hispanic Americans in Texas. With the limitations of language on health
outcomes, this study seeks to understand the impact of the language barrier in CRC
screening and treatment among Hispanics in Texas, which ranks only second to
California as the largest Hispanic population in the United States.
4
Background of the Study
Language barrier describes a roadblock to communication between people who
are unable to speak a common language, such as English. In public health sense,
language barrier is more than merely speaking limitation of a language. Because LEP
individuals have difficulty reading, writing, and understanding English (Genoff et al,
2016), they struggle to participate in the English-dominant healthcare system in the
United States. Thus, language barriers contribute significantly to poor health processes
and outcomes (Fernandez et al., 2011; Stephen & Zoucha, 2020), including limited ability
to access preventive services (Tseng et al., 2008), low cancer screening rates among LEP
patients (Busch et al., 2015; Griffey et al., 2015), nonadherence to medication schedules,
and skipping physician office visits due to misunderstanding physician instructions (Al
Shamsi et al., 2020; Brown & Bussell, 2011).
Clinicians suggest that several types of CRC screening exist, and their respective
procedures differ from one to another. While most of them are performed in clinical
settings, others could be done at home. Notable recommended tests performed by health
care providers in clinical settings include colonoscopy, virtual colonoscopy, and
sigmoidoscopy, whereas Cologuard, which is an at-home fecal immunochemical test-
DNA (FIT-DNA) test, is conducted by patients themselves (Butterly, 2020). These
screening tests are important in detecting CRC in its rudimentary stages when prognosis
often leads to positive outcomes because precancerous polyps could be removed before
they turn into cancer. Without screening, CRC is often detected late because early stages
tend to be asymptomatic (Siegel et al., 2021).
5
A lack of symptoms at its initial stages keeps CRC hidden where it could only be
found out with screening at that time. The instructions that patients must follow to
undertake CRC screening could be difficult to understand if individuals have limitations
in their knowledge of the English language, which is the main medium of communication
in the United States. Literature demonstrates that individuals with LEP are often reluctant
to seek health care, particularly preventive care, such as CRC screening, owing to
difficulties they experience in communicating or following health care provider
instructions (Sentell et al., 2013). As a result of LEP challenges, prevalence of patients
with cancer and other chronic diseases tends to be significantly higher than in the general
population (Gunn et al., 2020).
Similarly, low health literacy (LHL) negatively affects CRC screening due to
unmet informational needs over the relevance of interventions, such as CRC screening.
Effects of social determinants of health (SDH) like low socioeconomic status (SES)
coupled with LHL could potentially exacerbate the effects of LEP on CRC screening
(Schillinger, 2020). Studies indicate that while each of these factors (SDH, LHL, and
LEP) independently decreases preventive care measures like CRC screening, their
combined effects could make the situation worse for individuals identified with all those
factors (Sentell et al., 2015). Most immigrants, particularly those whose primary
language is not English such as Hispanic Americans, tend to suffer from SDH, LEP, and
LHL (Jacobson et al., 2016). Consequently, these factors pose significant public health
challenges on health care preventive interventions. With a high Hispanic population in
Texas, which ranks only second to California in size, this study sought to understand the
6
impact of LEP on CRC screening among Hispanic residents in the state. The study also
assessed whether gender differences lead to any variations in CRC rates among Hispanics
in the state.
Problem Statement
Studies indicate that the United States Hispanic population, identified by
individuals of Mexican, Puerto Rican, Cuban, Dominican, and additional Central/South
American as well as other Spanish ancestry based on self-identification (Jackson et al.,
2016), particularly reflects low rates of CRC screening, which may be blamed on certain
barriers, especially deficits in their ability to use the English language (Wang et al.,
2013). While not all individuals suffering from LEP may also have LHL, investigators
argue that a vast majority of individuals with LEP also have LHL status (Schillinger,
2020). Because LHL is associated with less cancer knowledge, negative attitudes toward
cancer screening, lower self-efficacy, and less likelihood of completing screening,
individuals suffering from LEP are confronted with compounded barriers in
comprehending and accessing health information and services (Arnold et al., 2016).
Researchers suggest that LEP individuals in patient-provider language-concordant
relationships experience increased rates of CRC screening as compared to individuals in
language-discordant relationships (Hsueh et al., 2021), thereby making LEP a key
component of effective health management (Schillinger, 2020). Similarly, other studies
indicate that LEP status is associated with multiple suboptimal health outcomes, partly
because of misinterpretation of patient complaints. Although hospitals often make
provisions for language translations, such provisions do not necessarily enhance the
7
understanding of patients to appreciate the services available to them, including effective
preventive measures like screening for CRC (Al Shamsi et al., 2020).
Because CRC screening involves individuals scheduling beyond a regular
doctors’ appointment, independent test preparation, and/or complex completion
instructions, it is often adversely influenced by LEP (Hill et al., 2021). Consequently,
most individuals who are diagnosed with CRC, particularly those with LEP challenges,
are found in late stages of the disease (Andrew et al., 2018). To gain a comprehensive
understanding of obstacles to CRC screening in the Hispanic population, an integrative
analysis of possible factors such as low literacy/educational levels, lack of provider
recommendations primarily due to lack of health insurance, cost of screening, and fear of
colonoscopy procedure is necessary. Most of these factors may be associated with LEP,
which constrains immigrants in the larger American society and culture (Zhang et al.,
2012). Researchers suggest that no matter the stage at which CRC is found, challenges
remain for patients, even after successful treatments, including surgery, radiotherapy,
and/or chemotherapy (Xie et al., 2020).
Late-stage CRC often leads to poor prognosis due to complications in treatment
and disease management because the cancer is no longer resectable when it has
metastasized to other organs such as the liver and lungs; this leaves chemotherapy as the
mainstay treatment option, which involves frequent provider office visits (Huang et al.,
2020). Even though the overall incidence of CRC among Hispanics compared to non-
Hispanic Whites is lower, the former’s incidence rate of 35.5 per 100,000 population is
still significant, and overall five-year CRC survival rates are equivalent between
8
Hispanics and non-Hispanic Whites (Jackson et al., 2016). However, researchers have not
observed increases in survival among Hispanics in comparison with non-Hispanic Whites
for metastatic CRC, raising questions about nonmedical factors like language barrier and
other cultural impediments (Sineshaw et al., 2014).
These findings indicate that the Hispanic population in the United States is
confronted with problems associated with CRC screening for early detection and poor
management when diagnosed late with CRC. Investigators suggest that several
randomized screening trials have shown a decrease in CRC mortality by repeated fecal
occult blood test (FOBT) testing annually or biannually, followed by colonoscopy for
participants with positive test results (Lin et al., 2021). Whether the low screening rates
for CRC among the Hispanic population in Texas are due to their deficiencies in the
English language remains unknown. Researchers have sought to understand the impact of
language barrier on health outcomes of many diseases elsewhere in the United States and
beyond; however, no attempt has been made to study the impact of language barrier on
CRC screening among communities of Hispanic origin in the state of Texas. The gap
makes this study, which assesses the impact of language barrier as a potential unmet
challenge in screening for early detection of CRC among Hispanics in Texas, important
to evaluate whether LEP contributes to low screening with its concomitant elevated
morbidity and mortality among Hispanic residents in Texas. The study also evaluated
whether LEP effects on CRC screening affect both male and female genders differently.
9
Purpose of the Study
The purpose of this study was to determine the impact of LEP on CRC screening
among Hispanic Americans in Texas. The study also sought to explore the differences in
the effect of language proficiency on CRC screening by their male and female
counterparts. I compared the CRC screening rate differences between English-speaking
and non-English speaking populations in Texas using a quantitative method to analyze
data obtained from the 2020 Texas Behavioral Risk Factor Surveillance System (BRFSS,
2020; the latest Texas BRFSS with data on CRC at the time of the study) on the general
population to deduce general health information of noninstitutionalized civilian
(Hispanic) residents of Texas. The study findings may assist in understanding the impact
of LEP and gender differences and CRC screening among Hispanic Americans in Texas.
Importantly, it would inform policymakers about any improvements they need to
undertake to promote and enhance CRC screening by initiating programs that would
reduce LEP in the target population. It would also highlight gender influences on
differences in CRC screening within the target community and find means to address any
hindrances that suppress CRC screening rates by gender.
Language barrier was the independent variable (IV) of the study. Language ability
may be defined by grouping study participants into language barrier or no language
barrier categories based on responses to the question, “Which language(s), English or
Spanish, do the study participants speak?” Given that English is the main language
spoken in Texas, “Spanish only” respondents were categorized as having language
barriers, while respondents who spoke “only English” or “both English and Spanish”
10
were categorized as having no language barriers. CRC screening was the dependent
variable (DV), or outcome variable of this study. Potential confounding variables in this
study included age and SES, such as income and educational levels of participants, which
were controlled in the data analysis. To absolve the confounding variables, age and other
factors associated with SES such as education, migration background, marital status,
statistical area of residence, employment, and income were considered in the computation
to assess LEP as a barrier to low CRC screening among the target population of the
study. The research questions were designed to measure the impact of language barrier on
screening rates for CRC among the Hispanic population in Texas. By understanding the
impact of the English language deficit on the Hispanic population, effective public health
measures could be developed to improve upon their CRC screening rates.
Research Questions and Hypothesis
The main independent variable (IV) was LEP. Because LEP individuals could not
communicate well in English, they attempted to respond to the BRFSS questionnaire in
Spanish. The IV was divided into eight categories: Non-Hispanic White Men, Non-
Hispanic White Women, Non- Hispanic Black Men, Non- Hispanic Black women,
Hispanic Men Responding in English, Hispanic Women Responding in English, Hispanic
Men Responding in Spanish, and Hispanic Women Responding in Spanish. The
dependent variable (DV) was based on CRC screening tests and described as reporting
FOBT within the past year, and/or sigmoidoscopy within the past 5 years, and/or
colonoscopy within the past 10 years. The research questions (RQs) and hypotheses were
as follows.
11
RQ1: Are CRC screening rates different between residents in Texas with and
without proficiency in English, when potential confounding variables including age,
income, occupation, health care access, and educational levels of participants are
controlled?
H01: There is no significant relationship between language barrier and CRC
screening rates among Hispanic and non-Hispanic populations in Texas, when potential
confounding variables including age, income, occupation, health care access, and
educational levels of participants are controlled.
Ha1: There is a significant relationship between language barrier and CRC
screening rates among Hispanic and non-Hispanic populations in Texas, when potential
confounding variables including age, income, occupation, health care access, and
educational levels of participants are controlled.
RQ2: Are CRC screening rates different between male and female residents in
Texas with and without proficiency in English, when potential confounding variables
including age, income, occupation, health care access, and educational levels of
participants are controlled?
H02: There is no significant relationship between gender and CRC screening of
Hispanic and non-Hispanic populations in Texas, when potential confounding variables
including age, income, occupation, health care access, and educational levels of
participants are controlled.
Ha2: There is a significant relationship between gender and CRC screening of
Hispanic and non-Hispanic populations in Texas, when potential confounding variables
12
including age, income, occupation, health care access, and educational levels of
participants are controlled.
Framework: Conceptual or Theoretical
Developed initially in the 1950s by social psychologists employed by the United
States Public Health Service, the health belief model (HBM) has assumed a central
position among theoretical frameworks that seek to account for the broad failure of
individuals to participate in events to avert or detect asymptomatic disease (Hochbaum,
1958; Rosenstock, 1966, 1974). In addition, HBM seeks to explain individuals’ responses
based on their experienced symptoms (Kirscht, 1974) and their behavior in response to
clinically diagnosed illnesses, especially in compliance to medical regimens (Becker,
1974). Social psychologists developed the HBM due to significant limitations of success
regarding various programs within the public health system in the 1950s and thereafter.
Consequently, efforts were committed to developing a theory that would delineate public
acceptance (or lack thereof) of programs to screen for disease and to vaccinate against
viral diseases like poliomyelitis and influenza, as well as attempts to improve on
compliance with medical advice regarding diabetes, hypertension, cancer, obesity,
exercise, seat-belt use, and HIV-risk behavior (Janz & Becker, 1984; Stretcher &
Rosenstock, 1997). Being one of the most broadly applied theories of health behavior,
HBM postulates six constructs, including risk susceptibility, risk severity, benefits to
action, barriers to action, perceived self-efficacy, and cues to action predict health
behavior (Becker, 1974).
13
The logical connections between HBM and the nature of this study are explained
by the HBM postulates. These constructs are relevant to understanding the impact of
language barrier in identifying asymptomatic CRC patients through screening for early
detection, which leads to positive disease prognosis. The constructs also relate to clinical
instructions of managing CRC symptoms and improve on core determinants of disease
prognosis (Hochbaum, 1958; Rosenstock, 1966, 1974).
The risk susceptibility describes an individual’s subjective awareness of the risk
of acquiring an illness or disease. The risk severity deals with an individual’s
assumptions on the concerns of contracting CRC, and the possibility of not receiving an
appropriate treatment, thereby leading them to a wide variation of apprehensiveness,
including medical consequences like death or disability, and social consequences that
may adversely affect their families with economic hardships. The benefits to action
explain an individual’s understanding of the effectiveness of available actions to
minimize the threats of CRC (or to cure CRC). Individuals who perceive these dangers
take actions to prevent or cure the cancer based on their understanding and evaluation of
both perceived susceptibility and perceived benefit as a motivation to accept the
recommended health action they consider beneficial.
Barriers to action refers to an individual’s recognition of impediments to
undertake a recommended health action on CRC, which may involve a broad variation of
barriers, such as costs, time-consumption, and inconvenience associated with health
actions they accept in dealing with CRC. As a result, individuals perform cost and benefit
analysis and weigh the effectiveness of decisions they embrace. Perceived self-efficacy
14
describes individuals’ confidence in their ability to appropriately perform a behavior,
which was a construct that was added to the model in mid-1980s; it is also a construct in
many behavioral theories as it is directly associated with whether an individual takes up
the desired behavior. The cues to action, which could be internal factors like fatigue due
to anemia, lack of appetite and weight loss, are often associated with CRC residual
symptoms; external cues such as advice from close relatives and friends, illness of family
member, and newspaper articles serve as a stimulus that influence decision-making
process to embrace a recommended health action.
To appreciate the relevance of these constructs, individuals suffering from or
prone to acquiring CRC need to overcome the barrier of language limitations to
understand the American healthcare delivery system and be able to communicate
effectively with healthcare providers. Thus, the fundamental importance of HBM
constructs makes it an appropriate public health theory in appreciating the unmet
challenges, like language barrier, for individuals in identifying asymptomatic CRC
patients and the ability to relate to clinical instructions of managing CRC residual
symptoms (Rakhshanderou et al., 2020).
This indicates that HBM demonstrates a value-expectancy theory, where behavior
assumes the function of the subjective value of an outcome and the subjective probability,
or expectation, where a specific action would lead to the expected outcome (Lewin et al.,
1944). In the context of health-associated behavior, the value-expectancy theory
demonstrates the willingness to prevent illness or to get well (value), which also signifies
the belief that a particular health action available to individuals would avoid or
15
ameliorate illness (expectancy), where individual’s estimate of personal susceptibility to
and the severity of an illness could be linked to the likelihood of being able to reduce that
threat through personal action.
Like other cognitive theories, HBM underscores those mental processes, such as
thinking, reasoning, hypothesizing, or expecting as fundamental components of the
model. In the same way, behaviorists like Skinner (1938) maintain that reinforcements or
consequences of behavior impact behavior directly, whereas cognitive theorists assume
that reinforcements function by influencing expectations (or hypotheses) about situations
instead of affecting behaviors directly (Bandura, 1965). The primary tenet in the HBM
reveals that people would make sound decisions to ward off, screen for, or control ill-
health conditions, such as CRC, if they consider themselves as susceptible to the
detrimental effects of the disease; if they perceive that an accessible course of action to
them would be helpful in lowering either their susceptibility to or the severity of the
condition; and if they perceived that the predicted obstacles or costs related to taking the
action are overridden by its benefits (Lin et al., 2019). Thus, the fundamental importance
of HBM makes it an appropriate public health theory in understanding language barrier
for individuals screening for CRC (Lau et al., 2020).
Nature of the Study
I employed quantitative method and cross-sectional design to examine the
relationships in the RQs and hypothesis between and among the variables using
secondary data from the Texas BRFSS (2020) on the general population to deduce
general health information of noninstitutionalized civilian residents of Texas. Individuals
16
younger than 50 years of age were excluded from the study because the United States
Preventive Services Task Force (USPSTF) recommendation for CRC screening begins at
age 50 to 75 years. Although recommendations from the ACS include individuals who
are at least 45 years of age in CRC screenings, because this study used data from BRFSS,
which is prepared by the CDC and follows recommendations from the USPSTS, I set the
baseline for this study at 50 years of age for participants. The Texas BRFSS provides
information on all races and ethnic groups for analyses of cancer incidence to support
health care assessment, evaluation, and planning, identifying populations at increased risk
of cancer, improving research associated with cancer etiology, prevention, and control
with appropriate interventions. The IV, which was represented by LEP in this study, was
dichotomous (language as a barrier to screening—yes/no), and the DVs (screening rates
among ethnic groups and between male and female gender differences) were categorical
variables. The DVs, which were based on CRC screening tests and described as reporting
FOBT within the past year, and/or sigmoidoscopy within the past 5 years, and/or
colonoscopy within the past 10 years, could be described as high screening rate, average
screening rate and low screening rate, indicating that these categorical variables were
ordinal variables.
The secondary data contained sociodemographic information on subjects who
participated in the study. To compute for language barrier levels, I employed
sociodemographic features such as age, gender, Hispanic origin (place of birth), Spanish
as first language, highest educational attainment, income level, and English fluency. To
find variables that could promote screening for CRC or improve on poor quality of life
17
after diagnosis with CRC, I evaluated cancer screening knowledge, accessibility and
utilization of health care services, health literacy, and environmental barriers, such as
legal status and preparation for and fear of colonoscopy procedure. The Texas BRFSS
database provides all sociodemographic characteristics needed to analyze language
barrier as an unmet challenge to screen for CRC and improve upon the quality of life
after diagnosis.
Definitions of Study Variables
Language barrier, which in this study is cited as limited English proficiency
(LEP): Refers to the inability of a health care provider and individuals they serve to
communicate because the former and latter speak different languages. When health care
providers and patients speak different languages, quality of care is adversely affected
because language concordance between patients and providers is essential for patients to
participate in effective preventive medicine, including screening for CRC. Investigators
note that language-incongruent encounters within the United States healthcare system
suppress individuals whose primary language is not English from participating in
preventive measures, such as CRC screening (Cano-Ibáñez et al., 2021). These
individuals who suffer from LEP may have difficulty reading, writing, and understanding
English, which creates hindrances to participation in the English-language dominant
healthcare system in the United States (Genoff et al., 2016).
Health literacy: The Centers for Disease Control and Prevention (CDC) defines
health literacy as “the degree to which individuals have the ability to find, understand,
and use information and services to inform health-related decisions and actions for
18
themselves and others” (Office of Disease Prevention and Health Promotion, 2021).
Researchers state that while literacy and health literacy are different, both factors are
interconnected to influence people’s eagerness to undertake preventive health measures,
such as CRC screening. Thus, health literacy is connected to literacy and implies
individuals’ knowledge, motivation, competences to access, comprehend, appraise, and
apply health information to make informed evaluations that result in suitable decision
making on their health. Low literacy promotes nonparticipation, which is influenced by
poor SES, defined by factors like income, educational level, and employment status
(Horshauge et al., 2020).
CRC screening: refers to medical procedures used to detect polyps and early
cancers in the large intestine (colon and rectum). CRC screening allows providers to
identify such abnormalities and treat them before cancer spreads or metastasizes. Studies
indicate that regular CRC screening prevents late detection of CRC and often make it
treatable with good prognostic outcomes (Issa & Noureddine, 2017).
Hispanic American: The United States Office of Management and Budget (OMB)
defines “Hispanic or Latino” as a person of Cuban, Mexican, Puerto Rican, South or
Central American, or other Spanish culture or origin regardless of race. Individuals who
trace their background origins to a Spanish-speaking country are referred to as Hispanic,
whereas those described as Latino refer to individuals who identify themselves with the
background origin of a Latin American country.
19
Assumptions
Costas-Muñiz et al. (2106) proposed that low rates of CRC screening among
immigrants, such as Hispanics, in relation to United States natives may be due in part to
the different values and beliefs exhibited by immigrants about health services used in the
United States. It may also be due to acculturation resistance to adapting the American
social norms on CRC screening. While previous studies among Hispanic and Asian
Americans found mixed results, it is important to assume that healthy migrant effects
could also influence low rates of CRC screening among Hispanic immigrants in the
United States; although, established facts indicate that Hispanic Americans tend to
improve on CRC screening when socioeconomic factors and access to care are addressed
(Velasco-Mondragon et al., 2016). Another important assumption among Hispanic
Americans is that this segment of American immigrants come from different countries
with varied cultures that may potentially influence their approach to the American native
culture on healthcare (Castañeda et al., 2019), an indication that sociocultural factors that
influence CRC screening may differ from one country to the other.
Scope and Delimitations
I focused my research on the noninstitutionalized, civilian, Hispanic American
population in Texas at the time of the study. The study focused on individuals who were
at least 50 years of age, as recommended by the USPSTF (United States Preventive
Services Task Force, 2021). Although ACS recommends that individuals need to be
screened at 45 years of age with increased potential benefits to detect early-onset disease
and minimize higher incidence and mortality from CRC (Abualkhair et al., 2020), the
20
BRFSS data are prepared by the CDC, which depends on the recommendations by the
USPSTF. Hispanic Americans residing in Texas but younger than 50 years of age were
not included in the study because the USPSTF recommendation for CRC screening
begins at age 50 to 75 years. Since preparation and screening procedures like a
colonoscopy could be complicated and cumbersome, they may pose a health risk for
individuals older than 75 years old. For that reason, for individuals aged 76 years or
more, USPSTF suggests making decision based on individuals overall health and history
of CRC screening (United States Preventive Services Task Force, 2021).
Limitations
Because my study used secondary data, I was unable to establish the validity of
the method used to collect the data. Any errors made in the collection of the data, such as
how questions were worded, may elicit responses in a certain way and could lead to a
measurement error, which could have introduced misinformation bias or miscalculation
into this study. BRFSS data are a probability sample of United States households with
both landline and cellphone telephones. Since telephone coverage differs by state and
subpopulations, selection bias exists in BRFSS data collection. This means that
individuals whose numbers are not selected, and those who do not have telephones could
not be covered by BRFSS sampling. Also, the participants’ ability to recall details of
responses they provide relies on how much time has elapsed since the event, leading to
response errors or recall bias in their answers. Notwithstanding its limitations, BRFSS
remains the best estimated source of health data for assessment at geographic levels
smaller than state employing limited community analysis techniques (Iachan et al., 2016).
21
Significance of the Study
This study sought to explore the effects of LEP on CRC screening among
Hispanic Americans in Texas and assess whether variations in their gender have any
influences on CRC screening. To date, review of literature on CRC health outcomes
among Hispanics primarily focuses on the Hispanic population around the United States,
with little to no research available on more than 11.5 million Hispanic residents, or nearly
40% of the 29 million people living in Texas (OMH, 2021). This study would offer clues
to understanding the impact of the English language as a barrier to low screening rates for
early CRC detection. Such clues would contribute to the knowledge in public health
about the adverse effects of lack of CRC screening and nonadherence to physician
instructions, particularly among migrants whose understanding of the English language is
limited, such as the Hispanic population in Texas.
Furthermore, the study findings may assist in understanding the impact of LEP
and gender differences on CRC. Importantly, it would inform policymakers on public
health policy about any improvements they need to undertake to promote and enhance
CRC screening by initiating programs that would reduce LEP in the target population. It
would also highlight gender influences on differences in CRC screening within the target
community and find means to address perceived hindrances that suppress CRC screening
rates by gender. Such a transformation would likely enhance positive social change in
those communities.
22
Summary
Timely CRC screening leads to early detection of CRC and better treatment
outcomes. Although nationwide, Hispanics have lower overall CRC incidence than
Whites across all age groups, including for individuals younger than 50 years of age
(Muller et al., 2021), recent epidemiological data show that the incidence of early-onset
CRC is rising at a faster rate among Hispanic people (Ellis et al., 2018), with Hispanic
men among the fastest growing demographic of early-onset CRC, with an annual increase
of 2.7% from 1992 to 2005 (Siegel et al., 2009). Whereas the incidence for Hispanic
women increased only 1.2% annually during the same time (Siegel et al., 2009), in total,
analysis indicates that annual incidence of early-onset CRC among Hispanics has
increased 2.35%, compared to 2.02% for Whites; these data reflect similar trends in
statewide data obtained from California and Texas (Wang et al., 2017).
As the largest ethnic minority group and fastest growing population in the United
States, the rising CRC incidence among the Hispanic group needs serious public health
attention that would influence policymaking on applications to reducing its impact on the
population. Investigators have studied the impact of language barrier on many diseases,
including CRC; however, no attempt has been made to study its impact on the overall
Hispanic population in the in the state of Texas. The gap makes this study, which seeks to
evaluate the impact of language barrier as a potential unmet challenge in screening for
early detection of CRC among Hispanics in Texas important to evaluate whether LEP
contributes to low screening with its concomitant elevated morbidity and mortality
among Hispanic residents in Texas. The study also evaluates whether LEP effects on
23
CRC screening affect male and female genders differently. Using HBM, the purpose of
this study was to evaluate secondary quantitative data from the Texas BRFSS (2020) to
analyze whether there was a significant relationship between language barrier and low
CRC screening rate among the target population. Similarly, the study sought to assess
whether there was a significant relationship between male and female genders in the
target population. Knowledge garnered from the study could contribute to the body of
knowledge to guide health enhancement programs within the ethnic group.
To further understand challenges that may confront the Hispanic Americans in
Texas about the impact of LEP and gender differences on CRC screening, I undertook an
extensive literature review from peer-reviewed journals. I also performed quantitative
data analyses using cross-sectional design on publicly available data from the Texas
BRFSS (2020) to evaluate the relationships between the variables. The results from this
investigation were then compared to the current body of knowledge to examine whether
they were in harmony with the literature and offered possible explanation regarding any
observed discrepancies. Chapter 2 focuses on the body of literature on CRC, LEP, and
sociodemographic data of the population of interest. Chapter 3 includes the methodology,
Chapter 4 deals with the results, and Chapter 5 focuses on discussion and analysis of the
results, and recommendations.
24
Chapter 2: Literature Review
Introduction
Researchers understand that irrespective of different racial and ethnic
backgrounds, health concerns are important to people of all backgrounds (Williams et al.,
2016). Significantly, investigators argue that sociocultural differences and socioeconomic
disparities in the immigrant communities play a leading role in limiting their access to
healthcare in the United States. The complex interplay between numerous factors such as
LEP, literacy skills, health knowledge, sociocultural factors, and the established
healthcare system, which embodies health care practitioners, health care infrastructure,
and quality of health care workforce that work together to promote health access have not
been adequately explored (Schillinger, 2020). This is particularly significant in the
Hispanic population whose primary language is often different from English, the main
language of the United States. Singularly, various factors that influence the LEP are
directly related to the effectiveness of CRC screening among Hispanic people in the
United States (Schillinger, 2020).
These factors are not limited to basic literacy but are made up of complex
interactions of individual limitations and varied sociocultural deficiencies that fail to
promote screening for CRC in its rudimentary, localized state. For example, minority
status, young age, less education, recent immigration to the United States, or being a
foreign born, low knowledge about CRC screening and reduced contact with the United
States healthcare system contribute to the poor participation in the CRC screening among
Hispanic Americans. The underutilization of CRC screening leads to lost opportunities
25
for CRC prevention and control (Savas et al., 2015). Overall, Hispanic Americans run
into a higher burden of access-associated screening limitations, such as living below the
federal poverty level (FPL) and reduced insurance coverage, which could also be blamed
on their limitations in English proficiency. Lack of health insurance is consistently
related to underutilization of CRC screening, particularly among Hispanic Americans
(Ou, et al., 2019). Investigators that have considered individual constituents such as LEP
and CRC screening knowledge, awareness, attitudes, beliefs, and availability of health
care have not exhaustively accounted for low CRC screening among Hispanic Americans
in Texas.
Chapter 2 highlights on peer-reviewed articles exploring the main IVs of
ethnicity, LEP and its determinants, while the DV was based on CRC screening, and the
association between and among the variables. Most of the literature used include articles
and other sources published in English since 2016. The articles in this review were
chosen using PubMed/Medline, Biological Abstracts, Biosis Citation Index, Cumulative
Index to Nursing and Allied Health, Cochrane Library, Embase Classic, SAGE,
ScienceDirect, and Web of Science. Relevant keywords/text words employed in the
search for appropriate literature were Texas Hispanics OR Latinos OR
Hispanic/Latino/Latina; Language barrier AND Mass screening OR Screening OR
Prevention; Colorectal cancer OR Colorectal neoplasms OR Colonic cancer OR Colonic
neoplasm OR CRC OR Colorectal carcinoma OR Colon carcinogenesis OR Sigmoid
carcinoma or Colon adenocarcinoma OR Colon carcinogenesis. Studies indicate that
26
these wide range of unique choices of research instruments often provide clear and
concise literature on target populations in investigations (Charrois, 2015).
Theoretical Foundation
It is well established that early detection of CRC improves prognosis, and many
screening tests are available to most of the general population in the United States. Yet,
CRC screening is suboptimal, particularly among individuals within the lower SES
brackets (Li, 2018). In view of lack of effective participation in CRC screening, various
models of public health have been developed by behavioral scientists to identify and
assess the elements that contribute to people’s participation of CRC screening (Topaloğlu
& Aydoğdu, 2021). Researchers suggest that HBM is one of the most important model-
based interventions that explains behaviors toward cancer screening, particularly among
subjects of low SES (Zare et al., 2016). The HBM could provide guidance for researchers
to assess screening behaviors of people and to appreciate their decision to accept
screening as a preventive mechanism against cancer. Health analysts indicate that HBM,
theory of justified action, theory of planned behavior, social cognitive theory,
transtheoretical model and health promotion model assume positions among the
models/theories often employed in the literature to understanding adherence with CRC
screening. Researchers have also employed other models, such as health behavior
framework (HBF), sociocultural health behavior model (SCHBM), and behavioral model
of health services use (BMHSU) to study CRC screening. While all these models are
approvingly useful in estimating CRC screening behaviors, it is demonstrated that
models/theories other than HBM are relatively deficient in predicting these relationships
27
(Topaloğlu & Aydoğdu, 2021), because the HBM essentially illustrates the association
between both internal and external health beliefs and health behaviors or intentions (Feng
et al., 2021). In this study, I used the HBM to ascertain the impact of its six constructs on
preventive care and how it influences individuals to seek CRC screening.
Conceptual Framework: Health Belief Model
Health investigators use the HBM as a model to evaluate cancer screenings and
other preventive health behaviors, thereby assessing the willingness of individuals to take
action to avoid, control, or screen for disease. Such an action leads to identifying
particular constructs that alter the individual’s behavior. For instance, if people become
vulnerable to the negative effects of CRC, they are more likely to accept screening
behavior, appreciate the benefits of CRC screening, and exhibit less resistance to
screening. As one of the most broadly applied theories of health behavior, the HBM
postulates six constructs, including risk susceptibility, risk severity, benefits to action,
barriers to action, perceived self-efficacy, and cues to action predict health behavior
(Becker, 1974). The logical connections between the HBM and the nature of this study
are explained by the HBM postulates. These constructs are relevant to understanding the
impact of language barrier in identifying asymptomatic CRC patients through screening
for early detection, which leads to positive disease prognosis. The constructs also relate
to clinical instructions of managing CRC symptoms and improve on core determinants of
disease prognosis (Hochbaum, 1958; Rosenstock, 1966, 1974).
The risk susceptibility describes an individual’s subjective awareness of the risk
of acquiring an illness or disease. The risk severity deals with an individual’s
28
assumptions on the concerns of contracting CRC, and the possibility of not receiving an
appropriate treatment, thereby leading them to a wide variation of apprehensiveness,
including medical consequences like death or disability, and social consequences that
may adversely affect their families with economic hardships. The benefits to action
explain an individual’s understanding of the effectiveness of available actions to
minimize the threats of CRC (or to cure CRC). Individuals who perceive these dangers
take actions to prevent or cure the cancer based on their understanding and evaluation of
both perceived susceptibility and perceived benefit as a motivation to accept the
recommended health action they consider beneficial.
Barriers to action refers to an individual’s recognition of impediments to
undertake a recommended health action on CRC, which may involve a broad variation of
barriers, such as costs, time-consumption, and inconvenience associated with health
actions they accept in dealing with CRC. As a result, individuals perform cost and benefit
analysis and weigh the effectiveness of decisions they embrace. Perceived self-efficacy
describes individuals’ confidence in their ability to appropriately perform a behavior,
which was a construct that was added to the model in mid-1980s; it is also a construct in
many behavioral theories as it is directly associated with whether an individual takes up
the desired behavior. The cues to action, which could be internal factors like fatigue due
to anemia, lack of appetite and weight loss, are often associated with CRC residual
symptoms; external cues such as advice from close relatives and friends, illness of family
member, and newspaper articles serve as a stimulus that influence decision-making
process to embrace a recommended health action.
29
In view of its practical implications on screening, researchers suggest that HBM
could be used as an effective theory-based educational intervention model to screen
people who are vulnerable to chronic diseases, such as cancer, due to increased factors
like unhealthy diet, smoking, and physical inactivity (Rakhshanderou et al., 2020). The
HBM proposes that messages like public health campaigns would attain suitable changes
in behavior if the message was individualized appropriately to address perceived barriers,
benefits, self-efficacy, and threat (Jones et al., 2015).
Applying an interventional study, Rakhshanderou et al. (2020) used a researcher-
made questionnaire to recruit 110 employees of Shahid Beheshti University of Medical
Sciences in Iran. The participants were randomly grouped into intervention and control
groups with cluster sampling. The questionnaire was made up of two portions of 10-
dimensional information and HBM constructs, which was administered for 1 month in
four sessions. Each session took the form of classroom lecture, pamphlet, educational text
messages received on mobile phones, and educational pamphlets via the office
automation system. The two groups were assessed at pre-test and post-test levels. The
researchers evaluated the data using SPSS-18 software, analysis of covariance
(ANCOVA) and independent t-test for intergroup comparisons (Rakhshanderou et al.,
2020).
Assessing with variables such as age, sex, education level, and family history of
CRC, the researchers found no significant variation between the two groups (p > 0.05).
Aside from the mean score of perceived barriers, which showed no remarkable change
after the intervention, the mean scores of knowledge, perceived susceptibility, perceived
30
severity, perceived benefits, perceived self-efficacy, behavioral intention, and preventive
behaviors rose remarkably in the post-intervention analysis in the intervention group as
compared to the control group (p < 0.05). The outcome demonstrated that administration
of the educational intervention using HBM was effective for the personnel and could
potentially promote the preventive nutritional behaviors associated with CRC
(Rakhshanderou et al., 2020).
In another investigation, Lau et al. (2020) undertook a systemic review using
HBM to evaluate CRC screening in the general population. In 2019, the researchers used
four databases to assess the impact of sociobehavioral factors on screening participation
in line with behavior change, specifically with respect to HBM constructs associated with
CRC screening. Reviewing a total of 30 studies that satisfied their criteria for inclusion,
the researchers used quantitative observational studies in line with HBM to evaluate CRC
screening history, intention, or behavior. All the studies explored used cross-sectional
design. The researchers found out that perceived susceptibility, benefits, and cues to
action had a direct relationship with screening history or intention, whereas perceived
barriers were found to be negatively related to screening history or intention (Lau et al.,
2020). Other modifying factors that had influence on the study outcome included
sociodemographic and cultural norms. Limitations found by the researchers were self-
reporting of screening history, intention or behavior, convenience sampling, and lack of
temporality. The researchers concluded that HBM’s associations with CRC screening
uptake were associated with preventive health behaviors; however, they also noted that
31
future investigations that explore impact of socioecological factors would present a more
concrete understanding of theory-based behavioral interventions (Lau et al., 2020).
To elucidate the efficacy of HBM in comparison with the theory of reasoned
action (TRA) in screening practices, Firouzbakht et al. (2021) measured competitive
cognitive models through the exploration of women breast cancer screening behaviors
using structural equation modelling. The researchers performed population-based cross-
sectional study in northern Iran with a sample of 500 women aged 35-85 years. The
demographic data features included were awareness, health belief, subjective norms, and
screening behaviors gathered with standard instruments. The investigators used SEM to
predict the pathways of regression coefficients. The outcome differed between the HBM
and TRA model. For the HBM, the standardized coefficient of the knowledge scores
showed a remarkable impact on the health belief perception (beta = 0.375); hence, health
belief directly influenced screening behaviors (beta = 0.73). On the other hand, in the
TRA model, although the direct impact of knowledge on intention was insignificant, the
researchers noted a significant inverse association on health belief and subjective norms
(indirect beta = 0.35) on behavior intention. Also, an increased coefficient of intention
was found by subjective norms (beta = 0.626), and the intention had a higher direct effect
on screening behavior (beta = 0.601), indicating that all fitting indexes improved in the
TRA model in comparison with HBM.
In another study, researchers blended two health promotion theories, the HBM
and the transtheoretical model (TTM) to understand the context of their studies and
selected measures. Both models, which overlap considerably with respect to specific
32
beliefs that impact on decisions, fundamentally posit that individuals’ beliefs and
characters are determinants of their decisions, and therefore, their behavior (Jandorf et al.,
2010). Based on these assessments, the researchers recruited and evaluated patients,
health care, and cultural elements that influence colonoscopy screening among Hispanics.
A total of 400 men (28%) and women (72%) were selected and interviewed, all of them
from East Harlem in New York City where the Hispanic population is dominant. They
assessed five hypothesis, including (1) older people and individuals with higher income
or highly educated would likely undertake colonoscopy screening (p < 0.01); (2)
individuals who have lived in the United States for a considerable period of time, and
have access to Medicaid or Medicare, or obtained their health care at an academic versus
a community health clinic, would likely undertake colonoscopy screening; (3) individuals
whose primary care givers motivated or encouraged them to undertake CRC screening
would more likely choose to be screened; (4) individuals who exhibit positive attitudes
about screening would likely participate in screening; and (5) individuals who
demonstrate high sense of medical mistrust, fatalism, fear, and worry would show
adverse correlation with cancer screening (p < 0.05) (Jandorf et al., 2010).
Using multivariate analysis, Jandorf et al. (2010) showed that colonoscopy
participation was inversely related to Medicaid and positively correlated with English as
preferred language, physician recommendation, and motivation for CRC screening and
less fear. The multivariate analysis revealed that individuals who opted to participate in
the English language were more likely to undergo screening (Wald Chi-Square =
5.36; p = 0.021; OR = 2.26; 95% C.I. = 1.13, 4.51). However, individuals who used
33
Medicaid were less likely to undergo screening (Wald Chi-Square = 17.10; p < 0.000;
OR = 0.30; 95% C.I. = 0.17, 0.53). Further, provider recommendation was positively
correlated with screening (Wald Chi-Square = 21.83; p < 0.000; O.R. = 25.83; 95% C.I. =
6.60, 101.04). Physician motivation was also associated with screening (Wald Chi-Square
= 3.86; p = 0.049; O.R. = 2.27; 95% C.I. = 1.02, 5.14). On the contrary, individuals who
showed greater fear were notably less likely to undergo screening (Wald Chi-Square =
4.28; p = 0.039; O.R. = 0.52; 95% C.I. = 0.28, 0.97).
While the researchers cited many strengths in their study, they also noted some
potential limitations. For example, they used participants who self-reported the
colonoscopy screening, which could be affected by participants’ bias. By including
medical or billing records in a future study, such anomaly could be eliminated or
minimized. Also, the researchers noted that their investigation was limited due to its
cross-sectional design, and hence, causality could not be assessed. Future study could
ameliorate this limitation by using longitudinal research. Furthermore, because the study
was done only in one community where the population was predominantly an older
sample of mainly Spanish speakers, the outcome could not be generalized to Hispanics in
various geographical locations across the United States at varying layers of acculturation.
To ensure generalizability, a large base of Hispanic immigrants representing both old and
young in different parts of the United States would need to be sampled to test their
generalizability (Jandorf et al., 2010).
Other health models have been successfully used to study CRC screening. Few of
them include HBF, SCHBM, and BMHSU. Using HBF, Tu et al. (2008) described a
34
synthesis of constructs of several key health behavior theories that has been applied in
multiple cancer control investigations. The HBF posits that independent and health care
system elements, and environmental and personal obstacles, jointly influence health
characteristics (Jones et al., 2015). According to Tu et al. (2008), interventions designed
to influence mutable individual patient-level characteristics, such as knowledge,
perceptions of disease susceptibility, cultural beliefs and lineage, among others within the
broader context of health system factors frequently function as hindrance to cancer
screening. By recognizing immutable factors that generate the context for individual
behaviors, interventionists could design health messages to limit health system
hindrances, promote knowledge, positively transform beliefs, empower social support,
and reduce obstacles to cancer screening (Lau et al., 2020). Thus, health beliefs constitute
an accumulation of traditional ideas, knowledge, past and present experiences, which
provide framework for cancer interventionists to explore measures to reduce obstacles
that suppress basic preventive health measures, such as CRC screening (Lau et al., 2020).
Because investigators have found in several studies that Latino adults are more likely to
be diagnosed with CRC at later stages compared to white adults (Castañeda et al., 2019),
HBF reinforces an appropriate conceptual model to address disparities in screening rates.
In a similar scenario, employing the SCHBM on 801 Vietnamese Americans from
community-based organizations, investigators noted through bivariate analysis that a
higher number of respondents who never screened for CRC reported LEP. The
respondents also reported fewer years of residency in the United States and reduced self-
efficacy associated with CRC screening with structural equation model identified self-
35
efficacy (coefficient = 0.092, p < .01) (Ma et., 2021). The SCHBM identifies and
describes associations and interplays between different elements that guide health
behavior (Ma et al., 2015). The SCHBM integrates six factors related to decision-making
and health-seeking characteristics that lead to health care utilization. The factors include:
(1) predisposition, such as educational attainment; (2) cultural influences like health
beliefs; (3) needs, such as levels of health care; (4) empowering influences, such as health
insurance coverage; (5) environmental/health systems, such as provisions of health care
resources; and (6) elements at family and community-levels, such as social norms and
social support. These factors indicate that SCHBM highlights the overarching importance
of socio-cultural factors on health behaviors, which influence decisions on CRC
screenings (Ma et al., 2021).
Researchers also used the BMHSU theoretical framework to predict factors that
enhance CRC screening among Latinos (Castañeda et al., 2019). In previous
investigations, the BMHSU had been used to understand factors influencing access and
utilization of hospital, dental, and medical care among diverse adults (Andersen et al.,
2000; Miller et al., 2008), obstacles related to access and utilization difficulties and
challenges experienced by Latinos (Andersen et al., 1986), and cervical, colorectal, and
breast cancer screening among Latinos (Fernandez and Morales, 2007). In using the
BMHSU framework, the outcome presumes that healthcare utilization is a function of a
person’s predisposition to use services (predisposing domain), factors that promote
healthcare utilization (enabling domain), and the need for care, or the need domain
(Castañeda et al., 2019). The predisposing factors describe the traits that contribute to the
36
possibility of healthcare utilization, and may include age, sex, income, education, and
acculturation (Castañeda et al., 2019). For instance, evidence suggests that lower levels of
education and income are related with lower CRC screening rates among Latinos
(DuHamel et al., 2020). However, the literature remains unsettled on the role of
acculturation factors, such as language-based acculturation, years in the United States,
and country of birth in predicting CRC screening among Hispanic Americans (Castañeda
et al., 2019). Researchers argue that enabling factors, identified as elements that promote
access to services include health insurance, availability of a regular health care source,
and utilization of services tend to motivate individuals to seek health care. Similarly,
investigators suggest that adherence to other preventive services improves adherence to
CRC screening in Latinos (Gonzalez et al., 2012). The BMHSU framework indicates that
a person must recognize the need, often measured by health-related quality of life
(HRQOL), for the likelihood of illness in order for health care utilization to occur. For
example, Latinas with a family history of cancer demonstrate enhanced breast cancer
screening use compared to others without family history, which is often credited to the
heightened awareness and perceived risk of breast cancer (Castañeda et al., 2019).
Literature Review Related to Key Variables and/or Concepts
CRC Screening
Investigators credit screening as a primary mode of cancer prevention and early
detection that leads to enhanced prognostic outcomes (Loomans-Kropp & Umar, 2019).
Clinical and experimental evidence derived from cancer screenings reveal the molecular
level development of benign adenomatous polyps from which the majority of CRCs
37
develop into adenoma-carcinoma stage (Hviding et al., 2008). Because CRC may have no
detectable symptoms while it is developing and metastasizing into nearby lymph nodes,
early detection through screening and accompanied treatment minimize tumor stages,
which is strongly associated with survival (Hviding et al., 2008; Shaukat et al., 2021).
Currently, several applications of screening are available, and the recommended
tests employed by most health care providers in screening for CRC include colonoscopy,
FIT, multitarget stool DNA testing (MT-sDNA), and computed tomography
colonography (CTC) (Redwood et al., 2021). Other methods of testing for average risk
patients include sigmoidoscopy combined with FIT (or sensitive gFOBT), sigmoidoscopy
alone, guaiac-based fecal occult blood test, and capsule colonoscopy (US Preventive
Services Task Force, Davidson et al., 2021). The USPSTF groups individuals at risk for
CRC into A, B, and C categories. The A category comprises individuals who are 50 to 75
years old, B category as individuals aged 45 to 49 years old, and C category comprises
individuals aged 76 to 85 years old. Using cancer intervention and surveillance modeling
network on CRC, the USPSTF recommends that all adults within category A (50 to 75
years) must undergo CRC screening and individuals within category B are highly
recommended to undergo CRC screening. The USPSTF also recommends that adults
aged 76 to 85 years old (category C) should receive CRC screening based on selective
clinician recommendations because the net benefit of screening is limited in this group
based on several studies (US Preventive Services Task Force, Davidson et al., 2021).
Clinicians recommend that colonoscopy should be done every 10 years for most
patients at average risk for CRC because the procedure is known to be associated with
38
reduced incidence and mortality due to its high sensitivity for CRC and adenomatous
polyps. Moreover, colonoscopy permits simultaneous lesion removal anywhere within
the colon during the procedure with the potential to detect and prevent cancer by
removing adenomatous polyps prior to malignant transformation (Stauffer & Pfeifer,
2021). For patients unwilling or unable to undergo colonoscopy, clinicians recommend
FIT for occult blood annually on a single sample as initial screening. Similarly, where
colonoscopy is limited, patients could be screened with FIT and when the result is
positive, then colonoscopy could be used promptly (Rex et al., 2017). Studies indicate
that colonoscopy and FIT have similar detection rates for CRC, but the latter has lower
rates for advanced adenomas (Cross et al., 2019). MT-sDNA, also called FIT-DNA or
multitarget fecal DNA, is tested every 3 years and it combines fecal makers for
hemoglobin and DNA mutation and methylation, and it uses one stool collection sample
(US Preventive Services Task Force, Davidson et al., 2021). Clinicians also recommend
CTC, formerly referred to as virtual colonoscopy, particularly to older patients with
comorbidities, such as cardiopulmonary disease, diabetes mellitus, or history of stroke
because the risks of colonoscopy rise with increased age (Ladabaum et al., 2021). CTC is
performed every 5 years, and it is more sensitive than all the methods except colonoscopy
(US Preventive Services Task Force, Davidson et al., 2021).
Another procedure is sensitive gFOBT, which is also a diagnostic test to look for
occult blood in the stool. Theoretically, the combination of sigmoidoscopy with FIT or
guaiac-based FOBT (gFOBT) improves lesion detection by promoting direct
visualization up to 60 cm and to also detecting colon lesion better than sigmoidoscope by
39
testing for occult blood. FIT is preferred over sensitive gFOBT (Meklin et al., 2020).
Clinicians may employ sigmoidoscopy alone every 5 to 10 years for patients where
adding stool-based test is not available or practical. While this procedure may be
conducted with minimal patient preparation and does not require sedation, it could only
identify lesions within the distal 60 cm of the bowel. The deficiency presents a challenge
for women and older patients because clinicians suggest that such patients often present
with higher frequency of more proximal lesions (Kuipers et al., 2015). Another test is
gFOBT, which is performed at home by patients and done annually on three samples as a
take-home test that patients mail back to the clinician. The gFOBT test has low
sensitivity for polyps and relatively low specificity for a clinically important disease,
making it less attractive to health care providers. Because of its poor sensitivity and
specificity, clinicians recommend that gFOBT is repeated annually if the test result is
negative (Elsafi et al., 2015).
Different CRC screening tests present with varying levels of invasiveness, patient
time investment, sensitivity for neoplasia, risks, and required supporting infrastructure
and costs, which tend to complicate efforts to regulate population effectiveness (Gupta et
al., 2014). These attributes cause uncertainty on the eventual community effectiveness of
any single modality, specifically for underserved communities. Colonoscopy is the
preferred CRC screening tool by most gastroenterologists because of its superior one-
time sensitivity for detecting polyps and cancer (Issa & Noureddine, 2017). However,
colonoscopy remains the most invasive of all CRC screening modalities, and it requires
full bowel preparation, sedation, an adult escort, and absence from work, often leading to
40
decreased wages on the procedure day for the patient. Besides the complications of
invasiveness, it is also expensive, often affordable by only few people and individuals
with health insurance (Gupta et al., 2014). Moreover, logistical and psychological
complexities due to a colonoscopy procedure have the potential to develop mistrust of the
medical system. This may lead to challenges in implementing colonoscopy screening
within racial/ethnic minorities and the socioeconomically disadvantaged because
empirical data demonstrate significant variations in accepting colonoscopy test among
underserved groups (Adams et al., 2017). For example, data suggest that providing an
informed choice between gFOBT and colonoscopy tends to promote screening uptake
compared with offering only a single modality because certain population subgroups may
choose colonoscopy due to its sensitivity and ability to detect and remove lesions
simultaneously. Other population subgroups may, however, prefer gFOBT/FIT and
sigmoidoscopy for comfort, convenience, and less invasiveness over colonoscopy. These
variations need to be considered by practitioners and public health entities to promote
CRC screening participation among individuals, particularly in underserved communities,
such as newly arrived immigrant populations (Gupta et al., 2014). However, informed
choices of types of CRC screening could be made based on patients’ ability to understand
available options, which is often impacted by their ability to understand and
communicate in English.
CRC screening may be developed into phases, including patient identification,
screening or rescreening, diagnostic follow-up, and treatment. This setting reflects on
patients who have access to care and physicians who provide care for them. Clinicians
41
need to keep track of such patients and remind them through letters and during office
visits, particularly those within the vulnerable ages from 50 to 75 years old. By
monitoring and measuring performance and outcome measures, clinicians would be able
to monitor CRC screenings for these individuals who have access to healthcare. Patient
identification within this group would not be difficult to track since they see such
physicians for their primary healthcare needs. Because identifying patients due for CRC
screening could be effectively tracked in clinical settings, physicians and their staff need
to constantly remind individuals who are due or overdue for CRC screening
(Subramanian et al., 2018). For example, using logistic regression to examine 15,866
average-risk patients aged 65 years and older, non-Hispanic white with preferred English
language and had health insurance, researchers showed that those individuals were more
likely to be up-to-date on CRC screening. In the same study, the researchers found out
that patients with no access to a gastroenterologist, who experience extreme poverty
rates, and inadequate insurance coverage or underinsured were less likely to be up-to-date
on CRC screening (Wang et al., 2018). The findings indicate that a variety of patient,
provider, and community characteristics seemed to have an impact on CRC screening. To
improve on CRC screening in a community with a variety of people, effective strategies
are needed to address multilevel factors, including focusing on patients with identified
individual barriers, modifying physician and clinical practices, and targeting populations
with low SES or inadequate levels of medical resources (Wang et al., 2018).
On the other hand, there are occasions where low participation of CRC screening,
like other preventive measures, such as mammography may be delayed due to factors
42
beyond the control of patients or limitations, such as LEP. For example, the emergence of
the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which causes
COVID-19 compelled healthcare providers to postpone preventive care appointments
since the pandemic erupted in 2019. The postponement allows healthcare resources to be
shunted away from cancer screening services leading to a steep reduction in CRC
screening and a backlog of patients awaiting screening tests (Kadakuntla et al., 2021).
Such delays in CRC screening and other preventive measures potentially result in abrupt
rise in preventable mortalities from many chronic diseases, including cancer, diabetes,
and cardiovascular disease (Schoenborn et al., 2022).
Also, most individuals of Hispanic origin confronted with LEP challenges fall
under the category of populations with low SES or inadequate levels of medical resources
(Steinberg et al., 2016). To improve screening in this underserved population, public
health campaigns need to address multiple challenges, starting with identifying
unscreened individuals (Gupta et al., 2014). In general, the absence of statewide or
national efforts to identify unscreened individuals is a key barrier to enhancing cancer
screening in target groups needed to be screened (Ellis et al., 2018). The problem is
compounded for individuals, such as foreign nationals with the burden of LEP, a key
segment of the underserved populations (Goebel et al., 2015). Systems frequently
employed to identify unscreened people depend on insurance or a regular place for health
care, such as primary care physician’s office, making data on risk factors and prevalence
of CRC screening in underserved populations limited (Song et al., 2021).
43
While the problems of identifying unscreened populations may be an obstacle to
LEP communities in Texas, another challenging factor could be the distance from where
residents live and centers where they may seek health care. Being the largest state by land
size in continental United States with 254 counties, including a significant number of
rural communities, Texas residents often must travel long distances to seek care. Thus,
transport availability may be another logistical challenge to Hispanic LEP families,
thereby suppressing efforts for CRC screening (Ioannou et al., 2021). In a survey on
transportation to the specialist’s office for colonoscopy, researchers found out that
respondents cited lack of transportation as a key impediment multiple times, “from not
having transportation, to not being able to drive the distance to the procedure, and not
having someone to go with them” (Muthukrishnan et al., 2019).
Increasingly, effects of social determinants of health (SDH), such as income level,
educational opportunities, occupation, employment status and workplace safety, gender
inequity, racial segregation, food insecurity and inaccessibility of nutritious food choices,
and access to housing and utility services tend to precipitate health issues and could
potentially suppress health preventive measures, such as CRC screening (Carethers &
Doubeni, 2020). Investigators argue that SDH effects, particularly unmet social needs
contribute adversely to poor health of an individual than either insurance status or access
to care (Fischer et al., 2021). Health analysts contend that LEP is associated with SDH,
and individuals who suffer from both SDH and LEP tend to experience increased risk for
poorer access to care, decreased healthcare utilization, and adverse health outcomes,
which could also potentially depress health preventive measures like CRC screening
44
(Sentell & Braun, 2012). These observations were found in a similar study conducted in
France where researchers examining cancer-related knowledge, awareness, self-efficacy,
and perceptions of screening barriers among low-income, illiterate immigrant women,
demonstrated that cancer screening inequalities is partly blamed on lower levels of SES,
inadequate health literacy skills, and low education (De Jesus et al., 2021). Most
individuals who are confronted with SDH deficits tend to have limited expertise, capital,
and infrastructure, all of which are factors that exacerbate poor health preventive
measures, such as CRC screening (Unger-Saldaña et al., 2020).
Furthermore, LEP individuals, including most of Hispanic Americans living in
Texas may face the calamities of SDH deficits, particularly lack of health insurance.
According to the United States Census Bureau, Texas is one of the states with the highest
number of people without health insurance (Berchick et al., 2019). For example, in 2018,
17.7% of Texas residents (nearly 5 million people) did not have health coverage, which
had ballooned from 17.3% in 2017. In 2017 and 2018, respectively, the United States had
a national average of uninsured people of 8.7% and 8.9%, indicating that Texas was more
than double the number of uninsured residents compared to the United States (Berchick
et al., 2019). The increased lack of health insurance happened in Texas because state
leaders refused to expand Medicaid, a joint state-federal program, which is the main
source of health insurance for underserved communities across the United States
(Sommers et al., 2016). Based on the provisions of the Affordable Care Act of 2010,
states receive a huge infusion of financial cushion to support Medicaid expansion for
individuals with low income (Manchikanti et al., 2017).
45
Analysis conducted by the Assistant Secretary for Planning and Evaluation
(ASPE) of the Office of Health Policy of the US Department of Health and Human
Services in 2019 showed that among the southern states, Texas accounted for a
disproportionate share of the uninsured, with a total uninsured population over 4.5
million and an uninsured rate of 19% (Bosworth et al., 2021). Even though Texas has
only 9% of the total nonelderly U.S. population, its portion of the uninsured within that
population segment exceeds 17% of the uninsured population within that age bracket in
the country, according to ASPE analysis of 2019 American Community Survey Public
Use Microdata Sample (ACS PUMS) (Bosworth et al., 2021). The ACS PUMS analysis
also noted that among the total uninsured population in the country, almost 9% reside in
households whose adults have LEP. The investigators documented that in the United
States, the highest percentage of uninsured in the LEP population occurred in the
households of North Houston, Texas, where 69% of the adults in those households
identified Spanish as their primary language. This grim situation is not limited only to
Houston area, but cuts across the whole state, where 29% of Hispanics remained
uninsured, compared to 12% non-Hispanic whites. Nationwide, among racial and ethnic
populations, Hispanic Americans have the second leading uninsured rate at nearly 15%,
surpassed only by American Indian and Alaska Native people, whose uninsured
population sits at 22% (Bosworth et al., 2021).
Like countless number of poor, uninsured and underinsured individuals, many
Hispanic populations in the United States, particularly those without legal status in the
country, often fail to gain access to health care services provided across the United States,
46
and therefore, could not participate in preventive health measures, including CRC
screening (Sohn, 2017). This finding indicates that there is a disconnect between the
healthcare system and the vulnerable Hispanic populations who may not have legal
status. Most of these individuals are poor, uninsured, and/or underinsured, which is an
indication that disparities occur when beneficial medical interventions are not shared by
all. Similarly, health disparities could emerge from a complex interaction of economic,
social, and cultural factors (Brown et al., 2019). Investigators have demonstrated the
overlapping factors of poverty, culture, and social injustice act as fundamental culprits of
health disparities, which adversely influence all aspects of the healthcare continuum from
prevention, detection, diagnosis, treatment, and survival to the end of life (Freeman &
Rodriguez, 2011). While LEP is principally connected to communication, culture, and
linguistics, most individuals who experience LEP deficiencies also have similar
resemblance of insufficient resources, risk-promoting lifestyle and behavior, and social
inequities, which are all core factors that exacerbate high rates of low CRC screening
(Floríndez et al., 2020).
In promoting CRC screening in the Hispanic community, particularly among the
low income, the uninsured, and recent immigrants who tend to underutilize CRC
screening, Shokar et al. (2016) used the Against Colorectal Cancer in Our Neighborhoods
(ACCION) to understand mechanisms in improving CRC screening. ACCION is an
evidence-based cancer control program (EBCCP) designed by the National Cancer
Institute (NCI) of the National Institutes of Health (NIH). It was designed to enhance
CRC screening among uninsured Hispanic adults. Being a community-based
47
intervention, ACCION is delivered by promotors, which is made up of an educational
program that uses presentations and videos, access to no-cost screening and navigation
services to create awareness among the Hispanic populations in medically underserved
communities (Kim et al., 2017). Using logistic regression with covariate adjustment,
researchers recruited 784 subjects (467 in intervention group, 317 controls) with a mean
age of 56.8years; 78.4% were female, 98.7% were Hispanic and 90.0% were born in
Mexico. The researchers employed ACCION protocols, where out of 784 participants,
screening uptake was 80.5% in the intervention group and 17.0% in the control group
[relative risk 4.73, 95% CI: 3.69-6.05, P<0.001]. The researchers noted that no
educational group differences were observed, and covariate adjustment did not
remarkably change the outcome, an indication that ACCION mechanism of promoting
CRC screening could be effective (Shokar et al., 2016).
Similarly, to understand mediators of CRC screening intervention among
Hispanics, investigators used HBM to create a comprehensive conceptual framework to
assess potential effects on CRC screening in the ACCION model. The investigators
integrated multiple cofactors and knowledge to explain and predict screening behavior
and to guide the development of screening interventions (Shokar et al., 2022). By
employing structural equation modelling approach, the researchers identified factors that
influence screening test completion in a successful CRC screening program, which they
designed for an uninsured Hispanic population. The researchers used generalized
structural equation models and surveys to collect information from participants who were
randomly assigned CRC screening interventions. The researchers ensured that direct and
48
indirect pathways through which cofactors, CRC knowledge and individual HBM
constructs, including perceived benefits, barriers, susceptibility, fatalism and self-
efficacy, and a latent psychosocial health construct mediated the screening efforts
(Shokar et al., 2022). The researchers recruited 723 eligible participants with a mean age
of 56 years, 79.7% were female, and 98.9% were Hispanic. The researchers identified
that the total intervention effect was comparable in both structural equation models,
Model 1 and Model 2, with both having direct and indirect effects on screening
completion (n = 715, Model 1: RC = 2.46 [95% CI: 2.20, 2.71, p < 0.001]; N = 699,
Model 2 RC =2.45, [95% CI: 2.18, 2.72, p < 0.001]. Shokar et al. (2022) realized that in
Model 1, 32% of the total impact was influenced primarily by the latent psychosocial
health construct (RC = 0.79, p < 0.001), which had its effect due primarily to self-
efficacy, perceived benefits, and fatalism. In Model 2, the authors noted that the primary
mediators were self-efficacy (RC = 0.24, p = 0.013), and fatalism (RC = 0.07, p = 0.033).
According to the researchers, the outcomes highlighted the importance of mediators in
understanding ways to improve on CRC screening, suggesting that a focus on self-
efficacy, perceived benefits and fatalism could promote the effectiveness of CRC
screening interventions particularly in Hispanic populations (Shokar et al., 2022).
Investigating CRC screening among a group of South Asians in the Metropolitan
New York and New Jersey region, Manne et al. (2015) noted that the rates are
consistently lower than the United States general population estimates of 65%. A
comparable study on Chinese, Japanese, and Vietnamese showed higher CRC rates than
most of the South Asian populations, such as Bangladeshi ethnicity who are far less
49
fluent in English. Also, individuals from the South Asian populations who have lived
fewer years in the United States compared to those who have lived in the country much
longer indicated low rates of CRC screening (Lee et al., 2011). According to Manne et al.
(2015), there is a correlation between low CRC screening among individuals of South
Asian origin with less fluency in English. Typically, such individuals also show
perspective about the United States healthcare system that may be suggestive of cultural
beliefs like those seen in the Hispanic American communities, including the extent of
trust in the healthcare system. Also, the inability to communicate freely with health care
providers has been reported as a hindrance to CRC screening among South Asians, which
is reminiscent of LEP challenges confronting individuals whose primary line of
communication is not in English like the Hispanic Americans. Thus, LEP deficiencies
transcend through different cultures and bear similar sociocultural resemblances (Berdahl
& Kirby, 2019).
While perceptions around healthcare systems bear resemblances, it is inevitable
that certain cultural and health concepts differ across cultures. For example, fundamental
values like Latino familism, collectivism, and moderation are likely to be varied among
Hispanics of different countries of origin, which may be even more widely varied
between people of Hispanic origin and the general American society, most of whom take
their roots and fundamental beliefs from the sociocultural lineage of Caucasian roots
(Valdivieso-Mora et al., 2016). These primary variations in culture between Hispanics
and traditional American approach to healthcare tend to influence the way Hispanic
populations would seek CRC screening. The appreciation of culture indicates that many
50
variables affect the ability to adopt to healthy behaviors. However, the integrative
behavioral model (IBM) shows that factors such as normative and subjective beliefs,
attitudes towards behaviors, perceived control and self-efficacy, and knowledge about
replacement behaviors have the potential to raise awareness about preventive health
measures, such as CRC screening. The IBM demonstrates that environmental conditions
could influence the intention and ability of individuals to adopt desired behaviors that
promote primary health screenings (Robb, 2021; Smith-McLallen, & Fishbein, 2008).
To further appreciate CRC screening from a cross-cultural perspective, a study on
Korean Americans (KAs), who also report suboptimal CRC screening adherence like
Hispanic Americans, was considered. In a study which applied cross-sectional survey
through self-report measurements, investigators evaluated factors that motivate KAs to
comply with CRC screening guidelines using Andersen’s Behavioral Model of Health
Services Utilization, which postulates that individuals use health care under the condition
that their predisposing characteristics, enabling resources, and need factors all operate
fully together (Jin et al., 2019). Investigators recruited 433 KAs aged 50–75 from the
Atlanta metropolitan area. Investigators measured factors linked with CRC screening,
such as predisposing factors, including gender, age, marital status, educational
attainment; enabling factors, such as income, health insurance, regular annual check-ups,
doctor’s recommendation, English proficiency, CRC knowledge, self-efficacy for CRC
screening, decisional balance in CRC screening; and need, which dealt with family
cancer history and self-reported health status (Jin et al., 2019). Using a multiple logistic
regression model including all 14 predictor variables, the authors found out that most
51
enabling factors, particularly income, regular annual health check-ups, doctor’s
recommendation, self-efficacy, and decisional balance independently predicted enhanced
CRC screening adherence in KAs. However, no predisposing or need factors could be
used independently to show improved CRC screening (Jin et al., 2019). The findings
echoed existing literature that high income is related to enhanced utilization of cancer
screening (Kelly et al., 2017). Similarly, adequate coverage of health insurance
approvingly provides access to cancer screening, and availability to primary care
physicians promotes regular check-ups or physician’s recommendation for cancer
screening, which is associated with improved CRC screening (Jin et al., 2019). Like the
Hispanic communities, KAs are also influenced by cultural and psychological factors,
such as language barriers limiting people’s ability to navigate complications associated
with the United States healthcare system and use of CRC screenings, such as
colonoscopy preparations. For example, KAs with LEP are less likely to comply with
CRC screening (Jin et al., 2019).
In a related study, researchers used nonexperimental, online survey research
design at the Minnesota State Fair to investigate whether male role norms (MRN)
(avoidance of femininity, dominance, importance of sex, negativity toward sexual
minorities, restrictive emotionality, self-reliance through mechanical skills, and
toughness), knowledge, attitudes, and perceptions could affect the intention to screen for
CRC among 297 African American men (Rogers et al., 2018). The investigators
hypothesized that the target population (Minnesota men aged 18 to 65) did not have
adequate CRC knowledge. They found out that only 33% of the sample obtained a
52
“passing” knowledge score (85% or better). Using a logistic regression model, the
investigators noted that three factors remarkably linked to an increased probability of
receiving CRC screening, including age, perceived barriers, and perceived subjective
norms. While the study focused only on African American men, the authors suggested
that their findings led to a solid basis for informing stakeholders on health policy and
promotion that early-intervention for CRC prevention programs could be responsive to
the needs of African American men (Rogers et al., 2018).
Health-related LEP
From the time it was created in 2000, the concept of LEP has affected many
policy decisions in a wide spectrum of social and public services, particularly in the areas
of healthcare for immigrant communities (Ortega et al., 2021). Using an executive order,
President Bill Clinton enshrined LEP in the lexicon of the United States federal
government policy and has since become inextricably embedded in the panoply of civil
rights protections codified in the Title VI of the Civil Rights Act of 1964 (Office for Civil
Rights (OCR), Office of the Secretary, HHS, 2016). Within the arena of public health and
health policy, the federal government defines LEP as “Individuals who do not speak
English as their primary language and who have a limited ability to read, speak, write, or
understand English”. Policymakers consider such individuals to be entitled to language
assistance with respect to a particular type of service, benefit, or encounter, and this
consideration has become operationalized in many federal agencies, such as the
Department of Health and Human Services (Foiles Sifuentes, et al., 2020). Despite its
provisions to support individuals with limited English language understanding, LEP
53
presents significant challenges within the healthcare industry. For example, owing to the
diverse and growing multilingual communities in the United States, LEP could be
problematic in three areas: the ethnocentric notion of a “primary language,” the
ambiguous idea of “limited ability,” and the deficit-oriented construct of “language
assistance” (Ortega et al., 2021).
The concept of primary language presumably demonstrates that lives of
individuals are shaped by the language they speak. However, linguistic repertoires in
more than one language show that multilingual individuals distribute their languages
along a continuum of domains, preferring to use languages they are more fluent in than
the others that they encounter vocabulary deficits (Heller, 2007). The definition of LEP
examines limited ability to speak, read, write, or understand English as a fundamental
hallmark of the LEP individual, even though principally speaking, reading, writing, and
understanding English differ significantly from context to context. Also, an individual
may have limited ability to speak, read, write, and understand English, but they may have
effective means of collaboratively use family members, such as their English-speaking
children to support them in the literacy practices in the English language (Ortega et al.,
2021). Language assistance in the LEP definition may be construed to mean handicap,
which needs to be remedied so that individuals seeking language assistance would not
consider themselves as being deficit of some sort, a condition that may have the potential
to adversely affect public health campaigns, such as initiatives for cancer screening
(Francis & Silvers, 2016).
54
Thus, linguistically, LEP has implications for healthcare. Language analysts in
healthcare indicate that linguistic minorities who are target population for effective health
delivery must be supported with language-appropriate services (Schiaffino, Al-Amin, &
Schumacher, 2014). Researchers suggest that two main approaches to providing suitable
language in communication are language-concordant care, where services are provided
by a clinician who speaks the same language as the patient, and interpreter-mediated care,
which includes a medical interpreter participating as a linguistic conduit between the
patient and clinician (Ortega et al., 2021). Language-concordant care has been shown to
enhance successful treatment outcomes, decrease healthcare costs, improve patient
satisfaction, minimize medical errors, and to promote health prevention activities, such as
cancer screening (Diamond et al., 2019). Similarly, professional medical interpretation in
language-discordant health experiences indicate remarkable benefits in the care of
linguistic minorities, although it is often underutilized (Schulson & Anderson, 2022).
Notwithstanding the practical understanding of LEP associated with language-
concordant care and interpreter-mediated care, it often assumes complexities for
healthcare providers and patients particularly in medical visits involving
multigenerational family members. For example, in cases where simultaneous presence
of both young and older generations of a family visit a clinic together, the younger ones
may prefer to communicate in English while the older generations with limited English
preferably seek medical interpreters or language-concordant health care providers.
Therefore, the involvement of linguistic practices such as Spanglish, among other ways
of translanguaging, are part of everyday lives of multilingual individuals, where a single
55
language category may not sufficiently meet their language preferences during a medical
visit (Ortega & Prada, 2020).
In some cases, direct translation from one’s dominant language to English for a
medical symptom may be difficult. Therefore, it may not be the patient whose
proficiency in English is necessarily limited but the English itself is limited in expressing
the concept that the patient would like to communicate to the clinician. Researchers
indicate that this complexity in divergent health terminology seems universal to both non-
English and English languages, which often present a challenge in health communication.
Such handicaps are not related to LEP (Ortega & Prada, 2020). It is also important to
appreciate that even people who are competently fluent in English may likely encounter
handicaps in communicating complex health concepts, especially under conditions of
illness, stress, or emergency. Such difficulties do not fit the current classification of LEP
(Ortega et al., 2021).
Patients who are confronted with challenges of LEP may seek physicians who
speak a language they understand and could use as a medium of communication, which
makes patient-provider language concordance important to CRC screening among
individuals who suffer from LEP (Kim et al., 2018). Investigators suggest that although
patient-provider language concordance is unable to fully explain all language-based
health disparities, they have identified that it predicts both access to health care and
health status (Sentell et al., 2013). However, they also find that language concordance
and CRC screening are specifically mixed because in some studies, people with LEP in
patient-provider language-concordant relationships showed increased rates of CRC
56
screening compared with individuals in language-discordant relationships (Hsueh et al.,
2021). Yet, in other studies, LEP people with language concordant providers showed
lower screening prevalence compared with individuals with a language-discordant
providers, indicating that LEP alone may not explain the complexities of lower screening
(Sentell et al., 2013). Notwithstanding such discrepancies, LEP is seen to be a dominant
factor in low rates of CRC screening among foreign-born individuals in the United States
(Manne et al., 2015).
To enhance CRC screening for LEP patients, researchers argue that patient
navigation, defined as “patient-centered healthcare service delivery model”, could guide
patients through the complex and disconnected healthcare system to remove obstacles to
timely care (Sentell et al., 2013). Because patient navigation has shown the potential to
promote effective cancer screening in vulnerable populations such as individuals who
suffer from LEP, the Affordable Care Act (ACA) of 2010 employs it as an enhancement
tool for healthcare promotion, particularly among individuals with low SES. Since its
enactment, the ACA has promoted accessibility standards that necessitate all information
to be in simple language that is culturally and linguistically receptive to LEP individuals
(Adepoju et al., 2015). Investigators have found that patient navigators significantly
augment complete screenings for breast, cervical and/or CRC. Patient navigation could
be essentially important for CRC screening among individuals like the Hispanic
Americans in Texas with LEP deficiencies since CRC screening methodologies, such as
colonoscopy preparation is significantly complex (Cotter et al., 2019).
57
Gender and CRC Screening
Researchers suggest that sex variations from both biological and sociocultural
gender differences contribute to CRC in men and women. Therefore, acknowledging the
impact of sex and gender on CRC may generate greater understanding on enhancements
to early detection and diagnosis with greater treatment outcomes and survival (White et
al., 2018). In the United States, investigators have reported higher CRC age-adjusted
incidence among men than among women persistently for more than 30 years, even
though investigators reveal that gender differences have diminished in individuals ≥60
years. The gap in the incidence of CRC narrows as people advance in age. The highest
rate reduction in incidence over a period of time is seen among individuals who are over
80 years of age (p<0.001) followed individuals in the 70-79 and 60-69 age brackets
(Abotchie et al., 2012). The findings documented by the researchers reflected findings in
recent studies conducted in the United States. For example, an investigation conducted by
Cook et al. (2009) delineated an incidence rate ratio of 1.37 between men and women in
the original 9 SEER areas in 1975–2004. In another study that broke down participants
into racial segments, Murphy et al. (2011) used the 13 SEER areas in 1992–2006 to
delineate an incidence rate ratio of 1.37 between men and women for non-Hispanic
whites, 1.48 for Hispanics, 1.30 for blacks, and 1.43 for Asians, all indicating higher
incidence for males than females. In a related study conducted in the United Kingdom,
White et al. (2018) reported that more men develop CRC, with age-standardized rates
(ASRs) of 86.1 per 100,000 men compared to 56.9 per 100,000 women in the United
Kingdom in 2014, which translated into 22,844 men and 18,421 women new cases
58
annually. Globally, the incidence rate is higher for men than women (746,298 vs 614,304
[20.6 vs14.3 ASR]) and mortality (373,639 vs 320,294 [10 vs 6.9 ASR]) for CRC
(Douaiher et al., 2017).
For CRC screening between men and women, studies indicate that both sexes
report similar rates, although men preferred a significantly higher rate of colonoscopy use
than women based on self-reporting data. However, when investigators examined test-
specific screening compliance with medical record data, there was no significance
between men and women, which may be an indication of reporting bias (Griffin et al.,
2009). Another study also showed that men and women participants reported similar
choices for CRC screenings mode, however, the researchers identified remarkable
variations in the barriers and facilitators to screening between both sexes (Friedemann-
Sánchez et al., 2007). Analysis of data indicated that women consider the preparation for
endoscopic procedures as a hindrance to screening while men have no concerns regarding
endoscopic procedure preparation. However, women and men showed varying degrees of
worry and expressed different information choices about endoscopic procedures. Also,
women considered CRC as a male disease, and therefore may be less vulnerable to CRC,
presumably due to the impact of reproductive health over women’s lifetime (Friedemann-
Sánchez et al., 2007). Another study that featured only on African Americans, researchers
reported that no stark differences were observed between men and women in their
decision to screen for CRC or in their concerns about cancer, even though the researchers
found that men and women had significantly varied understanding of CRC knowledge
(McKinney & Palmer, 2014).
59
Cultural Influences on CRC Screening
A study by Diaz et al. (2013) cited various potential socio-cultural elements that
may describe disparities in CRC screening rates between non-Hispanic Whites and
Hispanics. The researchers noted that the relationship between LEP and low CRC
screening rates among Hispanics is complicated and may be due to more than a patient-
provider communication barrier. Their argument is supported by other studies that
indicate that LEP could be a proxy for lower levels of acculturation, which is known to
have an inverse relationship with cancer screening behaviors among Hispanics (Mantwill
& Schulz, 2017). Assessing a 2005 California Health Interview Survey, Johnson-Kozlow
et al. (2009) noted that Mexican-Americans who were highly acculturated were nearly 4
times likely to participate in CRC screening compared with less acculturated Mexican-
Americans. Similar findings have been reported in other studies that featured some Asian
Americans, including Indians and Filipinos, where CRC screening rates were abysmally
low; however, the same study noted that Japanese, Chinese, Korean, and Vietnamese had
higher CRC screening rates comparable to non-Hispanic whites in the United States. The
researchers noted that higher education, acculturation, and high income was observed
among individuals with increased CRC screening, while those with less education and
have not been in the United States for a considerable period of time tend to have low
CRC screening (Burnett-Hartman et al., 2016; Ghai et al., 2018).
To minimize the impact of cultural effects on CRC screening, researchers indicate
that it is crucial to appreciate cultural implications of factors that suppress screening. For
example, previous studies showed that a significant proportion of Hispanics erroneously
60
embrace certain misconceptions about cancer that tend to decrease their participation in
cancer screening (Tejeda et al., 2017). These misconceptions may be related to perceptual
and behavioral differences associated with cultural and ethnic backgrounds of the
community, which may influence their beliefs about cancer prevention and views about
etiology of the disease. Consequently, there is likelihood of prolonging the time to seek
preventive care, such as CRC screening, leading to late detection and poor prognosis
(Diaz et al., 2013). A classic misconception within the Hispanic community highlights
the belief that rectal sex is linked with CRC, and this assertion presents as potential socio-
cultural barrier to screening among LEP Latino men (Villar-Loubet et al., 2016).
Methodology
Ratnapradipa et al. (2021) conducted cross-sectional study in a Latino-serving
federally qualified health center (FQHC) to understand obstacles to CRC screening and
related factors in a Midwest Latino community in Omaha, Nebraska. The researchers
recruited 68 Latinos at a FQHC from June to October 2017 for their investigation, and
explored factors related to scheduling, psychological, and financial barriers using t-test,
ANOVA, and multiple linear regression analyses. The subjects for the investigation
identified themselves with low education, low income, and reduced access to health
insurance or a primary care physician. Upon examination, the researchers identified
scheduling barriers as the leading obstacle compared with psychological and financial
barriers. Ratnapradipa et al. (2021) noted that being married or coupled was identified as
the only predictor of higher scheduling barriers (p < .05), and that was related to higher
psychological barriers in both univariate and multivariate analyses (p < .05). The
61
researchers also found out that higher education level positively correlate with higher
psychological barriers in univariate (p < .05) only and not in multivariate analysis. In
comparison with individual subjects with lower English proficiency, those with higher
proficiency obtained a higher financial barrier score in univariate (p < .05) only and not
in multivariate analysis. The researchers concluded that although interventions focusing
on CRC screening barriers, such as the availability of free at-home testing were
effectively provided, perceived barriers remained in place. To address the problem, the
researchers recommended bilingual patient navigators to support individuals with LEP
who would assist them with scheduling without fees for colonoscopy scheduling services.
Similarly, individuals who are well educated but are confronted with increased risk of
psychological barriers should be given more education on the relevance of CRC
screening (Ratnapradipa et al., 2021).
In a randomized clinical trial, Oyalowo et al., (2022) assessed the effectiveness of
an intervention via telephone conversation prior to fixing a date for screening or
surveillance colonoscopy and its impact on CRC screening completion rates. The
researchers argued that CRC screening is underused in the United States. They collected
data from July 2017 through August 2018 at the University of Pennsylvania Health
System, an urban academic medical center, where subjects recruited included
asymptomatic individuals, whose ages ranged from 50 to 75 years. The participants
whose primary care physicians had referred them for colonoscopy and were eligible for
CRC screening, had no scheduled colonoscopy appointment. The interventions included
individuals undergoing block randomization in a 1:1:1 ratio to 1 of 3 study groups,
62
namely “usual care group”, “generic message group”, and “tailored group”. In the usual
care group, subjects received a mailed letter and directed to make a schedule for
colonoscopy. In the generic message group, subjects were reached by telephone,
completed an evaluation, and were given a uniform, nontailored message that inspired
them to set up a colonoscopy appointment. Subjects in the tailored message group were
reached by the researchers via telephone, conducted an evaluation and were given a
tailored message that inspired them to schedule for colonoscopy appointment based on
their identified assessment cohort (Oyalowo et al., 2022).
The researchers performed data analysis from January to September 2019. The
researchers considered colonoscopy completion rate as their primary outcome, which
occurred within 120 days of enrollment, while colonoscopy scheduling appointment rate
was their secondary outcome, which occurred within 120 days of enrollment. The
researchers recruited 600 participants (median [IQR] age, 56 [51-63] years; 373 women
[62.2%]) were enrolled. The total number was divided evenly into 200 participants each,
and were randomized to usual care, generic message, and to the tailored message. Of the
total sample, 12 were Asians (2.0%), 324 were Blacks (54.0%), and 227 were Whites
(37.8%), while 9 participants (1.5%) were of Latino or Hispanic ethnicity. The analysis
indicated that colonoscopy completion was remarkable enhanced for both the
individualized message group (69 participants [34.5%]) and the generic message group
(64 participants [32.0%]) in comparison with the usual care group (37 participants
[18.5%]) (p < .001 and p = .002, respectively) (Oyalowo et al., 2022). In addition,
Oyalowo et al. (2022) noted that scheduling rates were remarkably increased in both
63
groups, with 106 participants (53.0%) in the individualized message group and 105
participants (52.5%) in the generic message group scheduling appointments in
comparison with 54 participants (27.0%) in the usual care group (p < .001 for both).
Based on the results, the researchers concluded that among people who do not have
current CRC screening, either giving them a specific message on intervention or a generic
message on intervention was effective in enhancing colonoscopy scheduling and
completion rates compared with usual care. The findings indicate that tailoring health
communications could improve personal motivation to seek CRC screening.
In assessing equity and practice issues in CRC screening, Buchman et al. (2016)
used mixed-method approach to study overall CRC screening rates, patterns in the
application of types of CRC screening, and sociodemographic features related to CRC
screening. Their effort led them to understand physicians’ perceptions regarding the use
of FOBT and colonoscopy for individuals at average risk of CRC. The researchers
applied and received research ethics board approval for the study, which was permitted
by Sunnybrook Health Sciences Centre and St Michael’s Hospital in Toronto, Canada,
where the study was conducted. The researchers employed cross-sectional administrative
data on individual sociodemographic features and semi-structured telephone interviews
with physicians. Participants were patients aged 50 to 74 years, and they were recruited
from April 1, 2009, to March 31, 2011. The long duration of recruitment allowed for
understanding the patterns of CRC screening, and all the important physician-ordered and
diagnostic tests were covered in whole without copayments or deductibles. The study
used physicians in family health teams in the Toronto Central Local Health Integration
64
Network. For quantitative analysis, descriptive measurements assessed proportions
stratified by income quintile, age, sex, and recent registration with Ontario Health
Insurance Plan (OHIP) as well as rate ratios. The researchers measured rates of CRC
screening by type; sociodemographic features related to CRC screening and thematic
evaluation that assessed constant comparative approach for semi-structured interviews.
For the physicians interviewed, their ages ranged from 27 to 62 years and had between 1
and 27 years in independent practice. The interview had an average length of 32 minutes
and was conducted between July and November 2012. The interview focused on methods
to perform CRC screening, physician preferences with respect to screening varieties, the
impact of patient preferences and beliefs, the impact of gastroenterology on screening
practices in family medicine, availability of health resources and influence of equity
matters, administrative setups and infrastructure, and influence of preventive care
financial incentive. All interviews were digitally audiorecorded to ensure exact
transcription (Buchman et al., 2016).
The results indicated that Ontario administrative data on CRC screening
demonstrated reduced total screening rates for younger individuals, male patients,
individuals identified with lower income, and people who had just immigrated into the
community. Specifically, individuals with low income and recent immigrants recorded
low rates for colonoscopy. Analysis from the semi-structured interviews indicated that
physician had divided opinions about CRC screening for average-risk patients, with one
batch of physicians noting that the evidence and recommendations for FOBT were
appropriate and another batch of physicians considering colonoscopy as the best option
65
for these patients, citing the inferiority of the FOBT method. The investigators concluded
that a clear variation of opinions exist, and physicians depend on specialist
recommendations, health care system, and patient expectations to make their choices of
CRC screening type (Buchman et al., 2016). The results of the study echoed other studies
elsewhere. For example, in the United States, investigators have documented that low
SES is an important determinant associated with low screening rates, and often
individuals in that category present with advanced CRC (Salem et al., 2021). The
findings of the study also showed that providing an informed choice of screening
applications to patients would likely promote increased screening rates and reduce
disparities, and these assertions might be realized when changes are made to policy and
physicians alter their attitudes to allow such changes (Buchman et al., 2016).
The study approach has strengths and limitations. For example, the administrative
data used by the investigators broadened the study to cover population-wide perspective
on rates of screening and the association between screening and sociodemographic
elements. Disparities in screening by area-level income, age, sex, and recent immigration
were unfolded in a way that could be generalized for the Toronto Central Local Health
Integration Network and in Ontario, although these disparities frequently varied for
colonoscopy and FOBT. Administrative data are usually gathered for multiple reasons
and not necessarily for investigations (Buchman et al., 2016). Thus, it was possible that
not all variables were covered in the data collection. For example, individual-level
income or immigration status could be missing from the data, making proxy measures
like area-level income obtained from zip codes and past registration with OHIP as
66
alternative options used in their place. Also, administrative data could not differentiate
CRC testing conducted solely for screening from diagnostic testing performed when
patients were symptomatic. The researchers also noted that the quantitative analysis
showed relevant patterns, even though the data failed to reveal purposes ascribed to
varying applications of colonoscopy and FOBT. The strength of the qualitative
assessment relies on its tendency to explain the patterns emerged from the screenings
obtained. Thorough interviews granted by the physicians about screening preferences and
the purposes behind their decision making and office practices allowed a complete
evaluation of the methods employed by the physicians (Buchman et al., 2016).
In a study where researchers sought to understand the impact of language barriers
on cancer screening for LEP patients, Genoff et al. (2016) reviewed several articles. They
set up eligibility criteria and measured the quality of the articles using the Downs and
Black Scale. The eligibility criteria focused on articles that had: (1) a study target group
of patients with LEP deficiencies suitable for breast, cervical or colorectal cancer
screenings, (2) a patient navigator intervention to give services before or at the time of
cancer screening, (3) a contrast between patient navigator intervention and either a
control group or a different intervention, and (4) language-specific results associated with
the patient navigator intervention. Their eligibility criteria were satisfied by fifteen
studies that met the inclusion criteria. The researchers measured the screening rates for
breast, colorectal, and cervical cancer in 15 language populations, out of which 14 studies
had outcomes that enhanced screening rates for LEP patients between 7 and 60%. The
researchers found great variability in the patient navigation interventions they measured.
67
Admittedly, the researchers noted that their study had limitations due to the variability in
study designs and limited reporting on patient navigator interventions, which potentially
decreased the ability to make conclusions on the complete impact of patient navigators.
However, the researchers noted that overall, there was evidence that navigators promote
screening rates for breast, cervical, and colorectal cancer screening for LEP patients.
They advocated that future investigation must systematically gather data on the training
curricula for navigators and evaluate their English and non-English language skills in an
attempt to understand means to minimize disparities for LEP patients.
In another investigation, comparing Latino community members’ and clinical
staff’s perspectives on barriers and facilitators to CRC screening, Alpert et al. (2021)
used qualitative study to contrast the views of clinical staff (CS) and Latino community
members (LCMs) in an urban Southern California community. Analyzing with purposive
sampling, 39 LCMs (mean age: 59.4 years, 79.5% female) subjects were selected to
participate in one of five focus groups. Also, 17 CS (mean age: 38.8 years, 64.7% female)
were chosen to participate in semi-structured comprehensive interviews, together with a
demographic survey. The researchers documented the interviews and focus group
recordings by transcribing them verbatim, which were then translated, and assessed with
direct content analysis. They also documented the demographic data using descriptive
statistics, and the themes considered include perspectives about CRC screening, CRC
knowledge, access to resources, commitments and responsibilities, social support,
vicarious learning, patient-provider communication, trust, and social relationships. The
results reveal that both CS and LCMs perceive barriers and facilitators of CRC screening,
68
which the investigators argue that such findings could be used to guide interventions and
policies to enhance access to CRC screening among LCMs (Alpert et al., 2021).
In a similar study, researchers investigated barriers and facilitators for CRC
screening in a low-income urban community in Mexico City, Mexico. The researchers
argued that owing to integration of changing lifestyles and improved healthcare
infrastructure to facilitate diagnosis, clinicians have an opportunity to diagnose CRC
early and improve on treatment outcomes (Unger-Saldaña et al., 2020). The researchers
sought to determine possible obstacles and facilitators for future application of FIT-based
CRC screening in a public healthcare system using qualitative study with semi-structured
individual and focus group interviews with different CRC screening stakeholders. The
stakeholders included 30 common people at average risk for CRC, 13 health care
personnel from a local public clinic, and 7 endoscopy personnel from a cancer referral
hospital. The researchers transcribed verbatim all interviews and they evaluated the data
with constant comparison method based on theoretical perspectives of the social
ecological model (SEM), the PRECEDE-PROCEED model, and the HBM. The results of
the analysis at several levels of the SEM identified both obstacles and facilitators.
Primarily, the barriers in each of the SEM levels included (1) at the social context level;
(2) at the health services organization level; and (3) at the individual level. The social
context level comprises poverty, health literacy and lay beliefs linked to gender, cancer,
allopathic medicine, and religion. The health services organization level comprises a
deficiency of CRC knowledge among health care personnel and the community
understanding of low quality of health care. The individual level deals with inadequate
69
CRC awareness, which presents with absent risk perception in association with fear of
participating in screening activities likely to expose a serious disease. Unger-Saldaña et
al. (2020) concluded that their findings postulate that multi-level CRC screening
initiatives in middle-income countries like Mexico must establish supportive plans of
action to find solutions to barriers and facilitators, such as (1) provision of free screening
tests, (2) education of primary healthcare personnel, and (3) promotion of benefits of
CRC screening information that focuses on the target population, individualized to ease
common lay beliefs of fear and uncertainty.
Also, Brand Bateman et al. (2020) conducted a qualitative study on perceptions of
Egyptian primary care physicians and specialists to explore their perspectives on CRC
screening. According to the investigators, over a third of CRC cases occur in people aged
40 years and younger, which often result in late diagnosis and poor treatment outcomes.
The researchers employed the PRECEDE-PROCEED model, which depends on
predisposing (intrapersonal), reinforcing (interpersonal), and enabling (structural) factors.
These factors were inherent in health behaviors, which they used as their theoretical
framework. Individuals who took part in the study as participants were primary health
care physicians, oncologists, and gastroenterologists practicing in Alexandria, Egypt. The
physicians participated in 1-hour semi-structured interviews, which were audiorecorded,
transcribed, translated into English, and assessed using thematic analysis. Based on the
results with 17 physician participants (n = 8 specialists and n = 9 primary care
physicians), the researchers noted that barriers to CRC screening were identified as SES,
inadequate education on prevention, fear, and cost (predisposing). Other barriers
70
mentioned included people’s assumption that only high-risk patients were eligible for
screening; disbelief that providers were not qualified to conduct and interpret screening
tests properly (reinforcing); astronomical cost of the procedure; unavailability of the
tests; and inadequate training for laboratory technicians and providers (enabling). The
researchers also identified potential facilitators as establishing a media campaign that
educate populations about early detection, curability, and prevention (predisposing);
training physicians and inducing physician engagement (reinforcing); and minimizing
costs, ensuring that screening tests are adequately provided, and supporting well-trained
providers (enabling) (Brand Bateman et al., 2020). The study suggested that Egypt would
need a CRC screening program, and for it to be successful, it must address barriers at
multiple levels to improve on participation by eligible individuals and vulnerable
populations.
Embracing Protection Motivation Theory (PMT), Wei et al. (2022) assessed
motivating elements on CRC screening among urban Chinese population in five
communities in Wuhan, China. The investigation was a qualitative study where the
researchers used cross-sectional survey, and all eligible urban Chinese were recruited and
interviewed using paper-and-pencil questionnaires. The intention of CRC screening was
assessed on six PMT subconstructs, namely perceived risk, perceived severity, fear
arousal, response efficacy, response cost, and self-efficacy. The investigators also
gathered data on sociodemographic variables and knowledge of CRC, and they employed
structural equation modeling application to conduct data analysis. The researchers had
569 respondents, of whom 83.66% agreed to take part in the CRC screening, which
71
represented adequate data that fit the proposed structural equation modeling well (Chi-
square/df = 2.04, GFI = 0.93, AGFI = 0.91, CFI = 0.91, IFI = 0.91, RMSEA = 0.04).
The researchers found out that 2 subconstructs of PMT (response efficacy and
self-efficacy) and CRC knowledge linearly associated and positively related to the
screening intention. On the contrary, through at least one of the two PMT subconstructs
(response efficacy and self-efficacy), the researchers noted that age, social status, medical
history, physical activity, and CRC knowledge were inversely associated with the
screening intention. The researchers concluded that their findings demonstrate the
significance of enhancing response efficacy and self-efficacy in promoting urban Chinese
adults to undertake CRC screening. They also acknowledged that knowledge of CRC is
remarkably linked with screening intention (Wei et al., 2022).
Employing the 2016 BRFSS survey, Viramontes et al. (2020) explored screening
modalities, predictors, and regional disparities among Hispanics and non-Hispanic whites
(NHWs) in the United States using a cross-sectional analysis of Hispanic subjects aged
50 to 75. The researchers depended on participants’ self-reported CRC screening status,
and they used the Rao-Scott Chi-square test to analyze competing screening rates and
modalities in NHWs and Hispanics. The researchers explored regional screening
disparities by using univariable and multivariable logistic regression to measure
predictors of screening among Hispanics and evaluated Hispanic-NHW screening rate
variations for each demographic area/territory. Their assessment revealed a screening rate
of 53.4% for Hispanics (N = 12,395), and 70.4% for NHWs (N = 186,331) (p < 0.001).
The findings also showed that 75.9% of Hispanics preferred colonoscopy to other forms
72
of CRC screening. Also, uninsured status (aOR = 0.51; 95% CI = 0.38-0.70) and limited
access to medical care (aOR = 0.38; 95% CI = 0.29-0.49) was associated with lack of
screening. The researchers concluded that comparatively, Hispanics have lower CRC
screening rates than NHWs across most United States, but disparities were more
significant in some states than other highest screening variations occurring in North
Carolina (33.9%), Texas (28.3%), California (25.1%), and Nebraska (25.6%), while New
York (2.6%), Indiana (3.1%), and Delaware (4.0%) had smallest disparities (Viramontes
et al., 2020).
Hill et al. (2022) undertook a retrospective cohort study to compare mt-sDNA test
among participants with LEP and English proficient participants, from 2015 to 2018. The
researchers matched participants with LEP to English proficient participants by age at a
3:1 ratio. Results obtained indicated that among participants with LEP, 53% had mt-
sDNA tests without useful results compared to 29% among English proficient
participants (p < 0.0001). Also, individuals with LEP had 62.5 median number of days
from order placement to test completion compared to 33 for English proficient
individuals (p = 0.003). The researchers concluded that disparity was remarkable in CRC
screening completion with mt-sDNA test among subjects with LEP, which may be partly
a reason for increased disparities in CRC mortality among individuals with LEP
challenges (Hill et al., 2022).
Summary and Conclusions
Studies that use factors of key behavioral theories, such as HBM, TTM, PMT,
IBM, BMHSU, SCHBM, and HBF tend to independently identify important elements
73
that shape the interest of individuals in receiving CRC screening. Understanding the
factors that promote participation is critical to raise participation and reduce attendant
CRC morbidities and mortalities (Jung, 2020). Several investigations have highlighted
screening behavior factors associated with potential participants, providers, or healthcare
system. These influencing factors could be described as non-modifiable, which may
include demographic factors (e.g., age, sex, race), education, health insurance, or income,
and modifiable factors, such as knowledge about CRC and screening, patient and
provider attitudes or structural barriers for screening (Wang et al., 2019). Health care
providers could not do much to alter the trajectory of nonmodifiable factors. However,
modifiable determinants are suitable points of alteration, and they are considered as the
plausible targets for intervention, leading to substantial decline in CRC risk. Knowledge,
perceived susceptibility, perceived severity, perceived benefits, perceived self-efficacy,
behavioral intention, and preventive behaviors tend to promote CRC screening, which
low SDH, such as low SES, and recent migration and LEP tend to depress CRC screening
participation. Researchers indicate that these factors are influential in public health
campaigns that enhance CRC screening (Rakhshanderou et al., 2020). Among Hispanic
Americans, particularly those whose primary language is not English, their participation
in CRC screening is abysmally low (Heintzman et al., 2022).
Numerous studies confirm that LEP is a hindrance to effective CRC screening due
to its limitations on communication between health care providers and their patients who
face shortcomings in patient-provider language concordance (Diamond et al., 2019).
Researchers extrapolate LEP challenges to potential socio-cultural impediments that
74
exacerbate disparities in CRC screening because they argue that LEP could be a proxy for
lower levels of acculturation, which is known to have an inverse relationship with cancer
screening behaviors among Hispanics (Mantwill & Schulz, 2017). To understand the
impact of LEP on CRC screening, confounding factors, such as age, income, occupation,
health care access, and educational levels of participants could be isolated to appreciate
the effects of LEP on low CRC screening among Hispanic Americans in Texas.
75
Chapter 3: Research Method
Overview
The purpose of this study was to explore the perceptions of LEP and its
repercussions on Hispanic Americans in Texas about CRC screening rates. In many
respects, LEP as a language barrier leads to unmet challenges for non-English speaking
communities in the United States. Investigators suggest that poor and marginal levels of
English proficiency may be blamed as key culprits in suboptimal participation of CRC
screening among individuals whose primary language is not English (Mojica et al.,
2015). Because of their low participation in CRC screening, people with LEP tend to
receive diagnosis for CRC at late stages of disease, which often results in poor prognosis.
According to one study, recent immigrants and LEP populations tend to have greater
odds of late-stage diagnosis of CRC (Batai et al., 2019; Stern et al., 2016).
To ensure an effective preparation for a CRC screening procedure, such as
colonoscopy, patients need to follow instructions that would lead to proper visualization
of the mucosa of the colon and rectum. Studies indicate that colonoscopy preparation
involves adequate bowel cleansing, which is critical to ensure good quality of the
procedure (Kastenberg et al., 2018). Unfortunately, bowel cleansing has been noted as
inadequate in nearly one-third of procedures (Tontini et al., 2021). A key factor noted by
health care providers involves drawbacks in patient-dependent factors, mainly their
inability to follow instructions needed for bowel preparation. Cleansing quality relies on
the understanding of solution preparation, volume, taste, and timing of consumption
76
(Tontini et al., 2021). This study would bring to light the extent of impact of LEP on
CRC screening rates among Hispanic Americans in Texas.
This study would also examine if gender differences existed for CRC knowledge,
intention to screen, perceived risk, and cancer worry among Hispanic Americans in Texas
when potential confounding variables, such as age, income, occupation, health care
access, and educational levels of participants are controlled. The study would highlight
differences in gender among individuals who suffer from LEP, which would examine
whether sex differences of the participants influenced their CRC screening rates. An
investigation conducted by Friedemann-Sánchez et al. (2007) on gender differences in
CRC screening barriers and information needs revealed that female and male subjects had
similar preferences for CRC screening mode; however, both males and females were
noted to have variations in the barriers and facilitators to screening. This study sheds light
on variations in the participation of both Hispanic men and women who undertake CRC
screening in Texas. The study assesses whether gender alone and/or gender with LEP
effects contribute to any differences between men and women of Hispanic origin for
CRC screening. Understanding the influence of gender differences on CRC screening
would lead policymakers to address shortfalls and improve on the overall CRC screening
intervention for both men and women.
Instrument to Measure Limited English Proficiency
Hospitals that tend to operate without reliable language screening tools often fail
to support their LEP patients with proper interpretation of care they provide. For effective
delivery of health, providers need good appraisal of the number of individuals with LEP
77
in a community to evaluate language need and language service provision (Taira et al.,
2020). Boscolo-Hightower et al. (2014) used a capture-recapture approach, which is an
audit tool, to identify families with LEP by conducting a cross-sectional assessment of a
retrospective cohort of patients on admission at a large pediatric hospital from July 1 to
December 31, 2009. The capture-recapture method they employed was used to evaluate
language screening, assess the rate of language interpretation, and determine the number
of LEP individuals receiving service. The researchers depended on two independent
sources to assess the captures, namely identification of language during registration and
patients seeking assistance for interpretation when on admission. The researchers
determined the number of LEP individuals missed by both captures on an assumption of a
closed population, which included 6887 patients on admission. Out of the 6887 patients,
the researchers found 948 LEP families during registration, and 847 families who sought
interpretation assistance at least once while on admission. The researchers evaluated the
size of the “ascertainment corrected” to be 1031 (95% confidence interval: 1022-1040)
and the size of patients missing by both approaches was 15 (95% confidence interval: 7-
24), leading to only 76% of LEP patients identified in both data sources. Boscolo-
Hightower et al. (2014) concluded that a simple language audit tool could be employed to
evaluate language need, interpretation rates, and inadequate language services needed by
individuals who need interpretation.
In another study, to identify LEP patients in clinical care, Karliner et al. (2008)
used the United States Census English proficiency question (Census-LEP) to assess
patients’ preparedness to communicate effectively in English. The researchers recruited
78
302 patients older than 18 years of age from a cardiology clinic and an inpatient general
medical-surgical ward from 2004-2005. The participants spoke Spanish and/or English.
The study focused on the sensitivity and specificity of the Census-LEP alone, or in
integration with, a question on a chosen language for medical care. That would be used to
estimate patient-reported capacity to understand provider recommendation on how to
address symptomatic conditions associated with their health. The researchers determined
that 198 (66%) participants reported speaking English less than “very well,” while 166
(55%) were documented as less than “well.” Also, 157 (52%) opted to have their medical
care in Spanish, and in total, 135 (45%) had the ability to discuss symptoms; further, 43
(48%) indicated that they understood physician recommendations in English. Karliner et
al. (2008) set up the Census-LEP as high-threshold (less than “very well”), which had the
highest sensitivity for estimating effective communication (100% Discuss; 98.7%
Understand).
The researchers also determined the lowest specificity (72.6% Discuss; 67.1%
Understand), noting that the composite assessment of Census-LEP and chosen language
for medical care led to a remarkable rise in specificity (91.9% Discuss; 83.9%
Understand), with only a marginal reduction sensitivity (99.4% Discuss; 96.7%
Understand). The analysis led to the conclusion that it was recommended to opt for a
suitable language for medical care questions and the Census-LEP provision with a high-
threshold of less than “very well” and a screening question. The researchers stated that
such an approach would lead to inclusive and precise identification of patients most
likely to gain from language support (Karliner et al., 2008).
79
Research Design and Rationale
In this study, I used quantitative methodology and cross-sectional design to
examine the association between LEP and CRC screening among Hispanic Americans in
Texas at the time of the study. The investigation explores the relationships in the RQs and
hypothesis between and among the variables using secondary data from the Texas
BRFSS (2020) on the general population to deduce general health information of
noninstitutionalized civilian residents. The Texas BRFSS provides information on all
races and ethnic groups for analyses, including cancer incidence to support health care
assessment, evaluation, and planning, identifying populations at increased risk of cancer,
improving research associated with cancer etiology, prevention, and control with
appropriate interventions. Using the BRFSS, the LEP was divided into 8 categories: Non-
Hispanic White Men, Non- Hispanic White Women, Non- Hispanic Black Men, Non-
Hispanic Black women, Hispanic Men Responding in English, Hispanic Women
Responding in English, Hispanic Men Responding in Spanish, and Hispanic Women
Responding in Spanish. The dependent variable (DV) was based on CRC screening tests
and described as reporting FOBT within the past year, and/or sigmoidoscopy within the
past 5 years, and/or colonoscopy within the past 10 years.
The IV in this study was dichotomous (LEP as a barrier to screening - yes/no).
Thus, responses to the LEP were dichotomized, where responses representing LEP
individuals were coded “1-Yes,” and categorical responses showing English proficiency
or no LEP were coded “2-No.” Given that English is the main language spoken in Texas,
“Spanish only” respondents were categorized as having LEP, whereas respondents who
80
spoke “only English” or “both English and Spanish” were categorized as having no LEP.
CRC screening was the dependent variable (DV), or outcome variable of this study. The
DVs (screening rate differences between male and female gender) were categorical
variables. The DVs may be described as “high screening rate, average screening rate, and
low screening rate,” indicating that these categorical variables were ordinal variables.
The secondary data contained sociodemographic information on participants in
the study. To compute for LEP levels, I employed sociodemographic features such as
age, gender, Hispanic origin (place of birth), Spanish as first language, highest
educational attainment, income level, and English proficiency. To find variables that
could promote screening for CRC, I evaluated cancer screening knowledge, accessibility
and utilization of health care services, health literacy, and environmental barriers such as
legal status and preparation for and fear of colonoscopy procedure. The Texas BRFSS
database provided all sociodemographic characteristics needed to analyze LEP as an
unmet challenge to screen for CRC and improve upon early detection of CRC for good
prognosis and improved quality of life after diagnosis.
To ensure that only IVs of interest would influence the statistical outcome, the
covariates of age, income, and education were assessed on a continuous dependent-
response Likert-type scale (Russell & Bobko, 1992). Using SPSS, I recoded some of the
continuous variables such as LEP level as categorical variables, for analytic purposes. I
also used the data to examine the relationship between gender and CRC screening rates,
where age and SES (income, education, and employment) were controlled. The IV of
81
gender and LEP (based on sociodemographic features) were measured against the DV of
CRC screening.
The BFRSS is publicly available secondary data from the CDC enabled this study
into the relationship between the variables and provided answers to the RQs. The
application of secondary data enhanced the research to be completed in a timely fashion
and assisted in facilitating scientific knowledge and comprehensiveness of the
investigation. Moreover, the application of primary data may be difficult to obtain
because of the high cost involved in gathering of data and extended follow-up time that
could be futile if there was a loss to follow up (Trinh, 2018).
Population
In the United States, the Hispanic population is identified by individuals of
Mexican, Puerto Rican, Cuban, Dominican, and additional Central/South American as
well as other Spanish ancestry based on self-identification (Jackson et al., 2016).
According to the United States Census Bureau (2020), the Hispanic American population
was 18.5% (60.6 million) of the total population in the country in 2020. The United
States Census Bureau projects that the Hispanic population would be 28% (111 million)
of the entire United States population in 2060 (United States Census Bureau, 2018). In
2020, the population of Hispanics in Texas was 11.4 million or more than 39% of the
total state population of 29.1 million (United States Census Bureau, 2020).
The study focused on the target population with a minimal age of 50 years old up
to individuals aged 79 years old. The age limits were chosen based on the
recommendations from the USPSTF. The USPSTF recommendation for CRC screening
82
begins at age 50 and up to 75 years. For individuals 85 years of age or older, health care
providers consider benefits to be minimal, but may be accompanied by increased risk,
especially procedures, such as colonoscopy which is invasive. I also used the data to
evaluate if gender differences existed for CRC screening among men and women of
Hispanic origin in Texas when potential confounding variables, such as age, income,
occupation, health care access, and educational levels of participants are controlled. Thus,
the study would highlight gender differences among individuals who suffer from LEP,
which would examine whether sex differences of the participants impacted their CRC
screening rates.
Sampling Design
The BFRSS (2020) cross-sectional study involved random-digit-dialed telephone
survey of noninstitutionalized adults who were at least 18 years old residing in the state
of Texas. In general, the BRFSS survey is an ongoing, state-based program that is
conducted across the United States and the Commonwealth of Puerto Rico, Guam, and
other territories. The BRFSS data gathering dwells on health risk behaviors, chronic
diseases and conditions, access to health care, and use of preventive health services and
practices associated with the main causes of mortality and morbidity in the United States
and participating territories (Gamble et al., 2017). This indicates that BRFSS survey
helps understand underlying challenges that people experience, such as excessive alcohol
intake, tobacco use, unhealthy diet, recurrent mental instability, and insomnia, which are
key culprits associated with pathophysiology of chronic diseases and conditions,
including heart disease, cancer, stroke, arthritis, and diabetes, which are often the leading
83
causes of morbidity and mortality in the United States (Blackwelder et al., 2021; Liu et
al., 2013; Williams et al., 2016). Using BRFSS survey, researchers could appreciate
positive health behaviors like maintaining physically active, avoiding smoking use,
undertaking routine physical checkups, such as CRC screening, monitoring blood
pressure and cholesterol levels, which are fundamental public health applications to
decrease preventable deaths and disability (Gamble et al., 2017).
Based on the target population in this study, I excluded individuals less than 50
years of age, non-Texas residents, individuals older than 79 years of age, active-duty
military personnel, and individuals incarcerated. Like the BRFSS in other states, the
Texas BRFSS initiated in 1987, is a federally supported landline and cellular telephone
survey. The Texas BRFSS gathers data on Texas residents about risk characteristics
associated with health, chronic health conditions, and application of preventive services,
which make it a useful tool for health policymakers in the state’s Department of State
Health Services (DSHS). Texas BRFSS is used by both public and private health
personnel at the federal, state, and local levels to identify public health challenges,
develop informed priorities and goals, develop policies and interventions, and assess their
impacts over a definite period (Texas Department of State Health Services, 2022).
In a previous study using a BRFSS survey, participants were asked demographic
and health-related questions to know their insurance status, doctors’ office visits, types of
preventive measures they received, impacts of uninsured status and out-of-pocket
payments for provider services. The researchers noted that among participants with cost
constraints, the rates of visiting a doctor’s office for CRC screening was low (Perisetti et
84
al., 2018). In another study, Sauer et al., (2018) estimated cancer screening by comparing
BRFSS and National Health Interview Survey (NHIS). The BRFSS was designed to give
state-level estimates, while the NHIS was used to give national estimates. Both surveys
were used to estimate progress control in cancer prevalence. By deducting NHIS
estimates from BRFSS estimates, the researchers calculated the direct differences (DD),
and by dividing the DD by NHIS estimates, the researchers obtained relative differences
(RR). Sauer et al. (2018) used two-sample t-test (2-tails) to test for statistically significant
differences, BRFSS screening estimates showed higher than those from NHIS for breast
(78.4% vs. 72.5%; DD = 5.9%, p<0.0001); colorectal (65.5% versus 57.6%; DD=7.9%,
p<0.0001); and cervical (83.4% versus 81.8%; DD=1.6%, p<0.0001) cancers. Based on
their computations, the researchers noted that DDs were higher in racial/ethnic minorities
than whites. Similarly, individuals with limited education and those without health
insurance obtained higher DDs compared with most educated individuals and fully
insured individuals, respectively (Sauer et al., 2018).
Data Collection
The Institutional Review Board (IRB) of Walden University approved the
dissertation topic and provided an approval number 07-29-22-0638019 that permitted me
to begin data collection. I used data collected through the BRFSS survey. The BRFSS is a
collaborative project initiated in 1984 by the CDC in partnership with all the 50 states in
the United States and participating US territories. This statewide project is conducted and
assisted by the National Center for Chronic Disease Prevention and Health Promotion, an
umbrella unit of the CDC. The BRFSS is a system of ongoing health-related telephone
85
surveys conducted monthly by state departments over landline and cellular telephones.
For the landline telephone survey, data collection is conducted randomly with selected
adults in a household. For the cellular telephone survey, data collection is conducted on
participants residing in private homes or college housing. The project is designed to
collect data using a standardized questionnaire with technical and methodologic support
from CDC on health-related risk behaviors and chronic health conditions. It also collects
data on preventive services, such as cancer screening from the noninstitutionalized adult
residents in the United States who are at least 18 years old (Centers for Disease Control
and Prevention, 2021). In 2020, the core factors assessed by the BRFSS health status and
healthy days, exercise, inadequate sleep, chronic health conditions, oral health, tobacco
use, cancer screenings, and health-care access, while its optional module topics covered
prediabetes and diabetes, cognitive decline, electronic cigarettes, cancer survivorship
(type, treatment, pain management) and sexual orientation/gender identity (SOGI).
Many questions are extracted from national surveys that have been developed,
such as the National Health Interview Survey or the National Health and Nutrition
Examination Survey. In addition to both core and optional questionnaire generated by the
CDC, states are allowed to add their own questionnaire to the survey. All new questions
that are proposed by states, federal agencies, or other organizations are subjected to
cognitive evaluation and field assessment as well as voting for approval from state
representatives on the BRFSS project prior to be accepted into BRFSS questionnaire.
States with significant populations of individuals who speak another language other than
English must have the BRFSS translated into such languages. Currently, the CDC has a
86
Spanish version of BRFSS questionnaire, which states may adapt if they need it for their
Hispanic communities (Centers for Disease Control and Prevention, 2021).
The BRFSS telephone survey follows a sample description, where a sample
record indicates a telephone number among a group of all telephone numbers the system
randomly chooses for dialing. The BRFSS employs a disproportionate stratified sample
(DSS) design for landline samples, where the BRFSS divides telephone numbers into two
groups, or strata, which are sampled separately. These groups are referred to as the high-
density and medium-density strata, which have telephone numbers expected to be owned
by households. The determination of high-density or medium-density stratum depends on
the number of listed household numbers in its block, or a set of 100 telephone numbers
having the same area code, prefix, and first 2 digits of the suffix in addition to all possible
combinations of the last 2 digits (Centers for Disease Control and Prevention, 2021).
Variables
The main IV was LEP. I grouped LEP into 8 categories in the BRFSS in Spanish,
including Non-Hispanic White Men, Non- Hispanic White Women, Non- Hispanic Black
Men, Non- Hispanic Black women, Hispanic Men Responding in English, Hispanic
Women Responding in English, Hispanic Men Responding in Spanish, and Hispanic
Women Responding in Spanish. The DV was based on CRC screening tests, and it was
described as reporting FOBT within the past year, and/or sigmoidoscopy within the past 5
years, and/or colonoscopy within the past 10 years. The DVs were described as “high
screening rate, average screening rate, and low screening rate”. Entries excluded from the
87
analysis included those missing a language variable, gender, race, or Hispanic/Latino
background.
Sociodemographic variables investigated in the analyses were categorized as: age
(50–54, 55–59, 60–64, 65–74, 75-79); educational status (less than high school, some
high school, high school graduate, some college, college graduate); gender of participants
(male or female). Availability of health insurance was an important determinant for
effective CRC screening, making access to care variables important to be investigated:
insurance status (private, public, uninsured) and having a reliable source of health care
(yes/no). The models constructed to examine the association between LEP and CRC
screening among Hispanic Americans were conducted with logistic regression analyses.
CRC screening was DV and LEP was IV for Module I, adjusting for age, income,
occupation, health care access, educational levels of participants and gender status and
Module II involved LEP and gender status (male or female), adjusting for age, income,
occupation, health care access, and educational levels of participants.
Threats to Validity
The validity of a study is important for other researchers and readers to accept the
outcome of a study, and it is affected by both internal and external threats. Researchers
stress that a catalog of threats to validity may be blamed for why empirical studies would
fail to deliver causal effects (Matthay & Glymour, 2020). There are four key types of a
thread to validity, including statistical conclusion validity, internal validity, construct
validity, and external validity (Campbell & Stanley, 2015). The statistical construct
validity refers to the proper application of statistical methods to evaluate the association
88
among variables of interest. In this study, I used the 2020 BFRSS data surveyed on
noninstitutionalized adults who were at least 18 years old and residing in the state of
Texas. This geographical limitation ensured that participants were only residents of
Texas.
The threats to internal validity, which are often a central concern of most causal
analyses, are blamed on violations corresponding to factors, such as history (extraneous
effects), group composition effects, and regression. All these factors are associated with
confounding, which could adversely affect the estimated association between the target
population and the causal effect from exposure to outcome (Matthay & Glymour, 2020).
Even though my efforts to minimize internal threats were limited because my study
depended on secondary data, the BRFSS data are generally collected by random
telephone calls. The randomization of the selection of participants minimized selection
bias, a key factor in confounding as an internal threat (Lesko et al., 2020). Also, to
improve on internal validity, researchers need to take necessary precautions for study
planning and sufficient quality control and implementation mechanisms, such as
sufficient recruitment plans, data gathering, data evaluation, and sample size (Patino &
Ferreira, 2018).
The construct validity focuses on the quality of choices of the independent and
dependent variables. The threats of construct validity may arise from the
operationalization of the IV and the choice of outcome measure, which refers to the
operationalization of the DV (Petursdottir & Carr, 2018). For example, lack of reliability
(when IV varies significantly from one occasion to the other, thereby potentially
89
increasing variability, which may obscure an association intended to be studied) and lack
of representativeness of the IV that deviates significantly from the theoretical construct of
interest (Deaton & Cartwright, 2018). For situations where the relationship between IV
and DV fail to capture all aspects of the variables that are operating to establish the
relationship, there becomes a possibility of threats conceptualized as measurement error,
confounding or a violation of consistency (Rothman et al., 2008). My study focused on
LEP and CRC screening among Hispanic Americans resident in Texas, with the main IV
as LEP and DV as CRC screening RQ1. For RQ2, the study focused on gender and CRC
screening, with the main IV as gender and CRC screening as DV. The operationalization
of the IVs were related to the operationalization of the DVs, in that the IVs would likely
have direct impacts on the DVs.
Also, the threats to statistical conclusion validity could adversely affect the
outcome of a study. Significant threats to statistical conclusion validity may include
conditions that affect measurement error or modifications, which are measured variables
that tend to diminish statistical power. Low statistical power, violated assumptions of
statistical tests, fishing, and the error rate problem refer to null hypothesis significance
testing, which contribute significantly to problematic outcomes of studies, thereby
presenting as threats. Such threats are important to estimation as they tend to negatively
alter estimates, leading to imprecise and potentially uninformative outcomes, which
promote wrong establishments of policies that depend on the study (Matthay & Glymour,
2020).
90
Similar to threats of internal validity, threats to external validity have a profound
impact on the outcome of research. Concerns about external validity tend to be associated
with populations and communities where research outcomes may be generalized. Under
such conditions, causal associations of choice may demonstrate interactions with
participant traits, surroundings, and the types of outcomes evaluated. Often, threats of
external validity are controlled in the explanation of results, where researchers seek to
depict population of the participants referred to in the results about their
sociodemographics or geography. To overcome the threats of external validity,
researchers address concerns with design or analytic characteristics, such as
oversampling of underrepresented groups or modeling casual interactions. External
validity may also be maximized through application of broad inclusion strategies, which
may result in the target population that more actually resembles real-life participants
(Patino & Ferreira, 2018). Because my study used Texas BRFSS, which is a statewide
data, the sample population was adequate to overcome underrepresentation of the target
population. Thus, the outcome would be safely generalized to extrapolate the results to
address the Hispanic American population in Texas, the target group in the study.
Descriptive Statistical Analysis
Investigators use descriptive statistics to describe the fundamental characteristics
of data in a study, and they provide simple summaries about the sample and the estimates
(Kaliyadan & Kulkarni, 2019). There are three main types of descriptive statistics,
including estimations of frequency (frequency and percent), estimations of central
tendency (mean, median, and mode), and estimations of dispersion or variation (standard
91
deviation, standard error, quartile, interquartile range, percentile, range, and coefficient of
variation [CV]), which are used to describe quantitative data. While researchers consider
an estimation of frequency for the categorical data, they use other forms of descriptive
statistical estimates, such as measures of central tendency and dispersion to describe
quantitative data (Mishra et al., 2019). In assessing continuous data, researchers use test
of the normality to evaluate the estimates of central tendency and statistical methods for
data analysis. The mean or median measurements, skewness, and kurtosis were
documented on the continuous variables of age, income, and the highest level of
educational achievements. On the other hand, for the categorical and nominal variables
(CRC screening rate, ethnicity, health insurance status, and health care use), the
descriptive statistics were considered as the total counts and their related percentages
(Kaliyadan & Kulkarni, 2019).
Inferential Statistics
The research questions (RQs) and hypotheses were as follows.
RQ1: Are CRC screening rates different between residents in Texas with and
without proficiency in English, when potential confounding variables including age,
income, occupation, health care access, and educational levels of participants are
controlled?
H01: There is no significant relationship between language barrier and CRC
screening rates among residents with and without proficiency in English in Texas, when
potential confounding variables including age, income, occupation, health care access,
and educational levels of participants are controlled.
92
Ha1: There is a significant relationship between language barrier and CRC
screening rates among residents with and without proficiency in English in Texas, when
potential confounding variables including age, income, occupation, health care access,
and educational levels of participants are controlled.
RQ2: Are CRC screening rates different between male and female residents in
Texas with and without proficiency in English, when potential confounding variables
including age, income, occupation, health care access, and educational levels of
participants are controlled?
H02: There is no significant relationship between gender and CRC screening
among residents with and without proficiency in English in Texas, when potential
confounding variables including age, income, occupation, health care access, and
educational levels of participants are controlled.
Ha2: There is a significant relationship between gender and CRC screening
among residents with and without proficiency in English in Texas, when potential
confounding variables including age, income, occupation, health care access, and
educational levels of participants are controlled.
Test of Assumption
Prior to testing the hypotheses, I analyzed the normality of the data as a
prerequisite for the statistical test by assessing the sampling distribution means and the
appropriate parameter estimates. I explored statistical tests of normality to the data to test
normality, which is considered as an underlying assumption in parametric testing (Kim &
Park, 2019). To assess for a nonsignificant p value for the Levene’s test to measure for
93
the presence of homoscedasticity or homogeneity of variance, I performed ANOVA. I
considered a p value of .05 or higher as not significant, making it possible to reject the
null hypothesis, taken as equal variance among varying groups.
Continuous and Categorical Variables
I used t test to detect any significant differences between respondents with LEP
deficiency and others without LEP deficiency by evaluating all continuous variables (age,
income, and highest education achieved: some high school, High school, some college,
college). The categorical variables of health care access (yes/no) and language barrier
were assessed with Chi-square test to examine the presence of significant differences
between respondents with LEP and respondents without LEP. Similarly, the t test was
used to examine the differences of CRC screening with respect to the gender status by
evaluating all continuous variables (age, income, and highest education achieved).
Inferential Statistical Analysis
The IV of LEP was dichotomized into low LEP (fluent in English/both English
and Spanish) and high LEP (fluent in Spanish only), while the DV of CRC screening was
assessed on nominal scales where a Chi-square test for the association was conducted to
measure whether there was a significant relationship between the variables of interest. I
used the Chi-square goodness-of-fit test to measure the normality of distribution by
assessing the magnitude of the Chi-square statistic and the p-value. In addition, counts
with related percentages and 95% CIs for the CRC screening were evaluated by
sociodemographic features. For bivariate comparisons, I performed bivariate logistic
regression and Chi-square tests. Bivariate analysis permits an evaluation of how the value
94
of the CRC screening (outcome variable) relies upon the extent of LEP (explanatory
variable). Also, in comparing gender (explanatory variable) and CRC screening (outcome
variable). While CRC screening may be influenced by gender, gender may not depend on
the CRC screening (Bertani et al., 2018). Additionally, I employed forced multivariate
logistic regression analysis to assess the relationship between the variables of interest,
such as LEP and CRC screening among the target population. I used bivariate logistic
regression to establish the association between LEP and CRC screening variables. I
conducted a multivariate logistic regression for odds ratio (ORs) of CRC screening
between respondents with low LEP (fluent in English/both English and Spanish) and high
LEP (fluent in Spanish only). The model was modified for the impact of clustering
design. All statistical analyses were performed with the application of SPSS version 28.
Ethical Procedures
I used secondary data from 2020 Texas BRFSS database. Secondary data deals
with the use of existing research data to answer a RQ, which most often differs from the
intended question answered with the data in its original work (Tripathy, 2013). Ethical
concerns raised about the use of secondary data often relates to potential harm to
individual participants and matters of return consent. Secondary data may differ based on
identifying information it may contain. The data may be de-identified or appropriately
coded so that investigators do not have access to the codes for identification. De-
identification refers to a mechanism of removing identifiers, such as names of individuals
and social security numbers that directly or indirectly relate to participants (or entity) and
removing such identifiers from the data (Kayaalp, 2018). Under such circumstances,
95
ethical procedures may not necessarily be required for full review by the IRB. Prior to
such a waiver, the IRB needs to confirm that the data are devoid of information needed
for identification. On the contrary, if the data contain associated information on
participating subjects, or information that could link the subjects to the data, IRB would
need researchers to meet the full confidentiality, security, and informed consent
requirements. The researchers would have to demonstrate how the subjects’ privacy and
confidentiality would be protected (Tripathy, 2013).
Because the BRFSS is de-identified publicly available data, it is exempted from
IRB approval (Crews et al., 2014). Even though the BRFSS data are publicly available,
they are managed and regulated by the CDC (federal), states, and participating territories
in charge of their respective BRFSS data. For my study I requested permission from the
Texas DSHS who are responsible for managing and regulating the use of the BRFSS
data. In my submission I explained that I needed to use their BRFSS data for dissertation
as part of requirements to complete my study at the Walden University. Prior to the
application, I applied for IRB approval from Walden University with the same request,
where I was cleared to seek permission to use the Texas BRFSS data for my study. For
both applications, I explained that the data would be kept safe from unauthorized access,
accidental loss, or destruction. To ensure that the data were kept safe and secured,
hardcopies shall be kept in locked cabinets and softcopies in encrypted files on my
computer, which was password protected. Since secondary data could be used for various
studies, I evaluated it to determine whether important criteria, such as the methodology of
data collection, accuracy, period of data collection, purpose for which it was collected,
96
and the content of the data were suitable to answer my RQs. The data would not be kept
for no longer than was necessary for my investigation. I would delete it completely to
prevent it from getting into unintended hands.
Summary
The Texas BRFSS was used to extract sociodemographic and CRC screening and
language data for analysis. The data were cleaned to remove cases with missing
variables. All sociodemographic characteristics necessary to measure LEP levels,
motivation variables, such as the availability of health insurance, and the CRC screening
rates were obtained from the 2020 Texas BRFSS data. The data were selected carefully to
ensure that only noninstitutionalized adults with a minimum age of 50 years old and not
more than 79 years old were chosen for the study.
The selection was based on recommendations by both USPSTF and ACS.
Comparing the invasive nature of some of the CRC screening procedures, such as
colonoscopy, clinicians suggest that elderly patients older than 85 years old may receive
minimal benefits compared with the risk involved. The ACS recommended to begin CRC
screening from 45 years of age due to cases found in individuals who fall within that age
category, however, this study used data from BFRSS, which is created by the CDC, a
federal government agency that uses recommendations from the USPSTF, not from ACS.
Only residents of the state of Texas were included in the study to support the descriptive
statistics analyzed. The two RQs were answered using the inferential statistics such as the
t test, multivariate logistical regression, and chi-square, which were all extracted from the
BRFSS (2020) data on Hispanic Americans residing in the state of Texas.
97
In the data analysis section, I performed a series of statistical analyses to compare the
associations between the variables. As part of the statistical analyses, I assessed the two
null hypotheses in this study and the accompanied statistical assumptions. I employed t
test to detect any significant differences between respondents with LEP deficiency and
others without LEP deficiency by evaluating all continuous variables. I used bivariate
logistic regression to establish the association between LEP and CRC screening variables.
I conducted a multivariate logistic regression for odds ratio (ORs) of CRC screening
between respondents with low LEP (fluent in English/both English and Spanish) and high
LEP (fluent in Spanish only). I conducted Chi-square goodness-of-fit test to measure the
normality of distribution by assessing the magnitude of the Chi-square statistic and the p-
value.
98
Chapter 4: Results
Introduction
In this study, I conducted analysis to assess whether there was a relationship
between LEP and CRC among Hispanic Americans living in Texas. The study focused on
variations on CRC screening rates among Hispanic, non-Hispanic White, and non-
Hispanic Black Americans, with the sample population aged 50 to 79 years old. I used
the 2020 Texas BRFSS on the overall population to deduce general health information of
noninstitutionalized civilian residents of Texas for the study. Respondents’ CRC
screening rates were analyzed by comparing sociodemographic variables, including age,
income, LEP, employment status, health care access, and educational levels, employing a
validated model by Diaz et al (2013) based on CDC’s BRFSS 2008 survey.
The second purpose of the research was to assess whether CRC screening rates
varied between male and female residents in Texas with and without LEP. LEP and
gender were the IVs, and CRC screening was the DV. The IV was divided into 8
categories: non-Hispanic White Men, non- Hispanic White Women, non- Hispanic Black
Men, non- Hispanic Black women, Hispanic Men Responding in English, Hispanic
Women Responding in English, Hispanic Men Responding in Spanish, and Hispanic
Women Responding in Spanish. The DV was defined as CRC screening as described by
the USPSTF for individuals aged 50 to 79 years old, and/or reporting FOBT within the
past year, and/or sigmoidoscopy within the past 5 years, and/or colonoscopy within the
past 10 years. Age, income, employment status, health care access, and educational levels
of participants were considered as other explanatory variables.
99
I measured the relationships between the IVs of LEP and gender and the DV of
CRC screening using the Texas BRFSS (2020). All statistical analyses on the data were
conducted using the application of SPSS version 28. I conducted a t-test for continuous
variables (age, income, and highest education achieved: some high school, high school,
some college, college) and Chi-square tests for categorical variables to detect any
significant differences between respondents with LEP deficiency and others without LEP
deficiency. Similarly, the t-test was used to examine the differences of CRC screening
with respect to the gender status by evaluating all continuous variables (age, income, and
highest education achieved). I used the Chi-square goodness-of-fit test to measure the
normality of distribution by assessing the magnitude of the Chi-square statistic and the p-
value. In addition, counts with related percentages and 95% CIs for the CRC screening
were evaluated by sociodemographic features. I used bivariate logistic regression to
establish the association between LEP and CRC screening variables. Additionally, I
conducted a multivariate logistic regression for ORs of CRC screening between
respondents with low LEP (fluent in English/both English and Spanish) and high LEP
(fluent in Spanish only). To control endogeneity, I employed a two-stage, predicted,
residual inclusion technique with the application of the instrumental variables and
confounders.
Research Questions and Hypotheses
The RQs and hypotheses were as follows.
RQ1: Are CRC screening rates different between Hispanic and non-Hispanic
residents in Texas with and without LEP, when potential confounding variables including
100
age, income, occupation, health care access, and educational levels of participants are
controlled?
H01: There is no significant relationship between LEP and CRC screening rates
among Hispanic and non-Hispanic residents in Texas, when potential confounding
variables including age, income, occupation, health care access, and educational levels of
participants are controlled.
Ha1: There is a significant relationship between LEP and CRC screening rates
among Hispanic and non-Hispanic residents in Texas, when potential confounding
variables including age, income, occupation, health care access, and educational levels of
participants are controlled.
RQ2: Are CRC screening rates different between male and female residents in
Texas with and without LEP, when potential confounding variables including age,
income, occupation, health care access, and educational levels of participants are
controlled?
H02: There is no significant relationship between gender and CRC screening of
Hispanic and non-Hispanic residents in Texas, when potential confounding variables
including age, income, occupation, health care access, and educational levels of
participants are controlled.
Ha2: There is a significant relationship between gender and CRC screening of
Hispanic and non-Hispanic residents in Texas, when potential confounding variables
including age, income, occupation, health care access, and educational levels of
participants are controlled.
101
Data Analysis
The data analysis focused on descriptive and inferential statistics. The predictor
variables in the study were LEP and gender; the outcome variable was CRC screening.
The descriptive statistics provided information on the sociodemographic variables that
supported the computation of the LEP and gender, the predictor variables in the study.
Descriptive Statistics
Sociodemographic Factors
The study had a total of 766 participants. The descriptive statics showed the
sociodemographic factors of participants grouped into 5-year age groups, sex, race,
language of participants, income of participants, employment status, healthcare access,
and education status, which were applied to the validated model by Diaz et al (2013) to
measure the LEP of participants. The statistics indicated that there were some participants
who were missing from certain categories; for example, 162 participants failed to show
income status, 12 participants did not show employment status, one participant did not
indicate healthcare status, and three participants did not have education status. Besides
those missing participants, other categories of participants under sex, race, and language
of participants all had all the total respondents of 766 (Table 1).
Table 1
Sample Size per Instrumental Variables
5-year
age
groups
Sex of
participant
s
Race of
participan
ts
Language
of
participant
s
Income of
participan
ts
Employme
nt status
Healthca
re access
Educatio
n status
N Valid 766 766 766 766 604 754 765 763
Missing 0 0 0 0 162 12 1 3
102
Sex and Age
The study participants selected randomly from the 2020 Texas BRFSS included
450 females or 58.7% and 316 males or 41.3% (Table 2). The minimum age of the
subjects was 50 years, and the maximum age was 79 years, with a mean age of 65.81
years. Age categories were not grouped based on sex. The ages of the participants
showed a standard deviation of 8.432, indicating the age distribution was widespread
across the range of the participants’ age. Also, the age distribution was approximately
symmetric (skewness = -.281) and platykurtic (kurtosis = -1.006).
Table 2
Sex of Participants
Frequency Percent Valid Percent
Cumulative
Percent
Valid Male 316 41.3 41.3 41.3
Female 450 58.7 58.7 100.0
Total 766 100.0 100.0
Participants Grouped by Age
Study participants were categorized into 5-year age groups from 50 to 79 years of
age. All the age groups had significant representations, with the age group 70-74 having
the largest group of 163 (21.3%) out of the overall subjects of 766 in the study. The age
group 55-59 had the least representation with 91 participants, or 11.9%, with the other
age groups falling in between (Table 3). The CRC screening characteristics varied
between age groups. For individuals within the age brackets 50-54, only 52 or 55.3%
screened for cancer, with 44.7% unscreened. Individuals within the 75-79 years age
group had the lowest sample of 22 out of 766 participants; however, they had 20 of them
103
or 90.9% screened, with only 2 or 9.1% unscreened. There were more screened
individuals among participants within the 70-74 age group with 125 participants
screened, representing 83.3% with 25 or 16.7% unscreened. The analysis had p <.001
(Table 5). The study sample indicated that there were more females than males with 486
or 58.7% female subjects compared with 316 or 41.3% males (Table 3). With a p >.5,
female participants had a higher CRC screening rate with 251 or 74.3% screened
compared with males 189 or 72.9% who screened (Table 5).
Race and Language
The study focused on three major races in Texas, including White only, non-
Hispanic, Black only, non-Hispanic, and Hispanic residents. The White only, non-
Hispanic variable was represented by 525 or 68.5% of the participants, Black only, non-
Hispanic was represented by 77 subjects, or 10.1%, and Hispanic was represented by 164
subjects or 21.4% (Table 3). When languages spoken by participants were considered, an
overwhelming 696 out 766 or 90.9% chose English, including a significant number of
individuals who identified themselves as Hispanic, non-White, non-Black. Out of 164
individuals who identified themselves as Hispanic, only 70 or 42.7% indicated that they
speak Spanish (Table 3). The race assessment indicated a p <.05. White only, non-
Hispanic had more participants of 313 or 76.7% screened and 95 or 23.3% unscreened
compared with Black only, non-Hispanic participants of 40, or 71.4% screened, and 16 or
28.6% unscreened, while Hispanic had 87 or 64.9% screened and 47 or 35.1%
unscreened (Table 5).
104
Income (Annual Household)
Many study participants had incomes of $75,000 or more, representing 200 out of
the 766 subjects, or 26.1%. However, an overwhelming 52.4% had an annual household
income of less than $50,000. Also, 162 or 21.1% of the participants did not report an
income at all in 2020, or their income data were missing (Table 4). Among the
participants, 180 had annual household income of $75,000 or more, 136 or 75.6% were
screened, and 44 or 24.4% were not screened. For individuals within the lowest bracket
of annual household income of less than $20,000, 62 or 72.9% underwent screening for
CRC while 23 or 27.1% did not receive screening for CRC. Participants with annual
household income from $50,000 to under $75,000, 56 or 80.0% received CRC screening
compared with 14 or 20.0% who did not receive screening for CRC, and all had a p <.001
(Table 5).
Employment Status
Out of 766 participants, 516 or 67.4% indicated that they were unemployed, 238
or 31.1% identified themselves as employed while 12 or 1.6 were identified as missing.
The huge number of unemployed participants may be partly due to retirement since all
participants were at least 50 years of age (Table 4). For individuals employed, 142 or
66.4% received screening for CRC while 72 or 33.6% did not. For the unemployed, 294
or 77.6% received screening while only 85 or 22.4% did not screen for CRC. The p <.05
(Table 5).
105
Healthcare Access
Because many of the study participants were Medicare eligible, 710 or 92.8% had
healthcare access while 55 individuals did not have healthcare access with one person’s
healthcare access information missing (Table 4). Health care access indicated that
individuals with access, 420 or 76.5% of the study participants received CRC screening
compared to 129 or 23.5 who had health insurance but did not screen for CRC.
Conversely, for individuals without healthcare access only 19 or 39.6% received CRC
screening, while 29 or 60.4% did not undertake CRC screening, p <.001 (Table 5).
Education Status
A significant number, 311 of the participants or 40.6% had college education
compared to 90 individuals or 11.7% who had less than high school education status.
Also, 209 or 27.3% had some college education while 153 or 20% of the participants had
completed high school. There were three individuals in the study whose educational
status was missing (Table 4). Considering the impact of education on CRC screening, the
trend of undertaking screening increased from less educated participants to more
educated participants. For individuals with less than high school education, 41 or 64.1%
received CRC screening with 23 or 35.9% not receiving screening. For high school
graduates/GED, 80 or 67.8% received screening and 38 or 32.2 stayed away from
screening. Participants who attained some college education, 119 or 74.4% received CRC
screening and 41 or 25.6% did not receive screening. The participants with college
degree or higher, 198 or 78.3% received CRC screening, while 55 or 21.7% did not
receive it, p <.05 (Table 5).
106
Table 3
Frequency Table 1: Five-year Groups, Sex, Race, Language
Demographic Category Frequency % Valid % Cumulative %
Five-year Groups
Valid
50-54 102 13.3 13.3 13.3
55-59 91 11.9 11.9 25.2
60-64 118 15.4 15.4 40.6
65-69 154 20.1 20.1 60.7
70-74 163 21.3 21.3 82.0
75-79 138 18.0 18.0 100.0
Total 766 100.0 100.0
Sex
Valid
Male 316 41.3 41.3
Female 450 58.7 58.7
Total 766 100.0 100.0
Participant’s Race
Valid
White only, non-
Hispanic 525 68.5 68.5 68.5
Black only, non-
Hispanic 77 10.1 10.1 78.6
Hispanic 164 21.4 21.4 100.0
Total 766 100.0 100.0
Participant’s Language
Valid
English 696 90.9 90.9 90.9
Spanish 70 9.1 9.1 100.0
Total 766 100.0 100.0
107
Table 4
Frequency Table 2: Participant Annual Income, Employment Status, Healthcare Access,
Education Status
Demographic Category Frequency % Valid
% Cumulative %
Participant Annual Income
Valid
$0 — < $20,000 109 14.2 18.0 18.0
$20,000 — < $50,000 208 27.2 34.4 52.4
$50,000 — < $75,000 87 11.4 14.4 66.8
$75,000 or more 200 26.1 33.1 100.0
Total 604 78.9 100.0
Missing System 162 21.1
Total 766 100.0
Employment Status
Valid
Yes, Employed 238 31.1 31.6 31.6
No, not employed 516 67.4 68.4 100.0
Total 754 98.4 100.0
Missing System 12 1.6
Total 766 100.0
Healthcare Access
Valid
Yes, Healthcare access 710 92.7 92.8 Yes, Healthcare access
No healthcare access 55 7.2 7.2 No healthcare access
Total 765 99.9 100.0
Missing System 1 .1 100.0
Total 766 100.0
Education Status
Less than high school 90 11.7 11.8 11.8
High school graduate 153 20.0 20.1 31.8
Some college 209 27.3 27.4 59.2
College graduate 311 40.6 40.8 100.0
Total 763 99.6 100.0
Missing System 3 .4
Total 766 100.0
108
Table 5
Descriptive Statistics of Study Population Based on Screened and Unscreened for
Colorectal Cancer (N=766).
Characteristics Screened
N (%)
Unscreened
N (%)
Total
N p-value
Age
50-54 52 (55.3) 42 (44.7) 94 <.001
55-59 54 (63.5) 31 (36.5) 85
60-64 79 (74.5) 27 (25.5) 106
65-69 110 (78.0) 31 (22.0) 141
70-74 125 (83.3) 25 (16.7) 150
75-79 20 (90.9) 2 (9.1) 22
Sex
Male 189 (72.9) 71 (27.3) 260 >.5
Female 251 (74.3) 87 (25.7) 338 >.5
Participant’s Race
White only, non-Hispanic 313 (76.6) 95 (23.2) 408 <.05
Black only, non-Hispanic 40 (71.4) 16 (28.6) 56 <.05
Hispanic 87 (64.9) 47 (35.1) 143 <.05
Language
English 410 (75.4) 134 (24.6) 544 <.05
Spanish 30 (55.6) 24 (44.4) 54 <.05
Education Status
Less than high school 41 (64.1) 23 (35.9) 64 <.05
High school graduate 80 (67.8) 38 (32.2) 118 <.05
Some college 119 (74.4) 41 (25.6) 160 <.05
College degree or higher 198 (78.3) 55 (21.7) 253 <.05
Employment Status
Employed 142 (66.4) 72 (33.6) 214 <.05
Unemployed 294 (77.6) 86 (22.4) 379 <.05
Annual Household Income ($)
$0 — < $20,000 62 (72.9) 23 (27.1) 85 <.001
$20,000 — < $50,000 111 (68.1) 52 (31.9) 163 <.001
$50,000 — < $75,000 56 (80.0) 14 (20.0) 70 <.001
$75,000 or more 136 (75.6) 44 (24.4) 180 <.001
Health Insurance Coverage 763 99.6 100.0
Yes 420 (76.5) 129 (23.5) 549 <.001
No 19 (39.6) 29 (60.4) 48 <.001
109
Using standard means for continuous variables and proportions/frequencies for
categorical variables, I calculated respondent characteristics. Chi-square tests were
employed to assess the association between the dependent and independent variables as
well as each potential confounder. The study participants were grouped into 5-year age
groups from 50 to 79 years of age with their screening rates increasing from the youngest
age group to oldest age group. However, the number of participants screened was higher
within groups 60-64, 65-69, and 70-74 compared to age groups 50-54, 55-59, and 75-79.
Also, groups 60-64, 65-69, and 70-74 had higher screening rates compared to age groups
50-54 and 55-59. While individuals in 75-79 age group were only 22 study participants,
20 or 90.1% received screening. Overall, 440 out of 766 or 73.6% study participants
received CRC screening (Table 6).
Table 6
Five-year Age Groups and Colorectal Screening Status
Yes, screened for
colorectal cancer
No, not screened for
colorectal cancer Total
50-54
Count 52 42 94
% within 5-year age groups 55.3% 44.7% 100.0%
55-59 Count 54 31 85
% within 5-year age groups 63.5% 36.5% 100.0%
60-64 Count 79 27 106
% within 5-year age groups 74.5% 25.5% 100.0%
65-69 Count 110 31 141
% within 5-year age groups 78.0% 22.0% 100.0%
70-74 Count 125 25 150
% within 5-year age groups 83.3% 16.7% 100.0%
75-79 Count 20 2 22
% within 5-year age groups 90.9% 9.1% 100.0%
Total Count 440 158 598
% within 5-year age groups 73.6% 26.4% 100.0%
110
The Chi-square tests indicated that Pearson Chi-square χ(1) = 32.754, p <.001,
showing that there was a statistically significant association between different age groups
(Table 7).
Table 7
Chi-Square Tests on Five-Year Age Groups
Value df Asymptotic Significance (2-sided)
Pearson Chi-Square 32.754a 5 <.001
Likelihood Ratio 32.403 5 <.001
Linear-by-Linear Association 31.668 1 <.001
N of Valid Cases 598
Note: a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is
5.81. Considering symmetric measures of Phi and Cramer’s V of 0.234, each at p <.001
indicated a relatively stronger association between age and CRC screening (Table 8).
Table 8
Symmetric Measures on Five-year Age Groups
Value Approximate Significance
Nominal by Nominal Phi .234 <.001
Cramer’s V .234 <.001
N of Valid Cases 598
Among 260 male study participants, 189 underwent CRC screening while 71 did
not. Male study participants represented 43.0% of 43.5 of the total study participants. The
number of females in the study was 338, out of which 251 underwent CRC screening
while 87 stayed away. In all, females represented 56.5% of the study participants with
74.3% receiving screening compared to 72.7% of men who received the CRC screening.
Together, 73.6% of both sexes received CRC screening while 26.4% did not (Table 9).
111
Table 9
Sex of Participants and Colorectal Screening Status
Yes, screened for
colorectal cancer
No, not screened
for colorectal
cancer
Total
Sex of
participants
Male
Count 189 71 260
% within Sex of
participants 72.7 27.3 100.0
Female
Count 251 87 338
% within Sex of
participants 74.3 25.7 100.0
Total
Count 440 158 598
% within Sex of
participants 73.6 26.4 100.0
The Chi-square tests indicated that Pearson Chi-square χ(1) = 0.186, p = .666, showing
that there was no statistically significant association between different sexes of
participants (Table 10).
Table 10
Chi-Square Tests on Sex of Participants and Colorectal Screening Status
Value df
Asymptotic Significance
(2-sided)
Exact Sig.
(2-sided)
Exact Sig.
(1-sided)
Pearson Chi-Square .186 a 1 .666
Continuity Correction b .114 1 .736
Likelihood Ratio .186 1 .667
Fisher’s Exact Test .708 .367
Linear-by-Linear
Association
.186 1 .667
N of Valid Cases 598
Note: a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is
68.70; b. Computed only for a 2x2 table.
112
Symmetric measures of Phi and Cramer’s V were -0.018 and 0.018, respectively,
each at p = 0.666. The Phi and Cramer’s V values being less than 0.10 indicated no
association between gender differences and CRC screening (Table 11).
Table 11
Symmetric Measures on Sex of Participants and Colorectal Screening Status
Value
Approximate
Significance
Nominal by Nominal Phi -.018 .666
Cramer’s V .018 .666
N of Valid Cases 598
Out of the 766 study participants, 440 received CRC screening, of which 313 or
71.1% were White only, non-Hispanic, 40 or 9.1% were Black only, non-Hispanic, and
87 or 19.8% were Hispanic. Among White only, non-Hispanic 76.7% underwent
screening, Black only, non-Hispanic 71.4% received screening, while 64.9% of Hispanic
also received screening. The data also indicated that 95 or 23.3% White only, non-
Hispanic, 16 or 10.1% Black only, non-Hispanic, and 47 or 29.7% Hispanic did not
screen for CRC. In all, 440 or 73.6% of respondents screened for CRC, another 158 or
26.4% of respondents did not screen, while 168 or 21.9% were missing from screening
among study participants (Table 12).
113
Table 12
Participant Race and Colorectal Screening Status
Yes, screened
for colorectal
cancer
No, not
screened for
colorectal
cancer
Race of
participants
White only,
non-Hispanic
Count 313 95
% within Race of participants 76.7% 23.3%
Black only,
non-Hispanic
Count 40 16
% within Race of participants 71.4% 28.6%
Hispanic Count 87 47
% within Race of participants 64.9% 35.1%
Total Count 440 158
% within Race of participants 73.6% 26.4%
The Chi-square tests indicated that Pearson Chi-square χ(1) = 7.360, p = <.025,
showing that there was a statistically significant association among White only, non-
Hispanic, Black only, non-Hispanic, and Hispanic samples of participants (Table 13).
Table 13
Chi-Square Tests on Race of Participants and Colorectal Screening Status
Value df Asymptotic Significance (2-sided)
Pearson Chi-Square 7.360a 2 .025
Likelihood Ratio 7.113 2 .029
Linear-by-Linear Association 7.011 1 .008
N of Valid Cases 598
Note: a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 14.80.
Symmetric measures of Phi and Cramer’s V were the same, 0.111, each at p =
0.025. The Phi and Cramer’s V values being more than 0.10 indicated there were
differential influences among White only, non-Hispanic, Black only, non-Hispanic, and
Hispanic samples of participants and CRC screening (Table 14).
114
Table 14
Symmetric Measures on Participant Race and Colorectal Status
Value Approximate Significance
Nominal by Nominal Phi .111 .025
Cramer’s V .111 .025
N of Valid Cases 598
This study focused only on two languages used by participants, English and
Spanish. While 544 or 90% of respondents identified themselves as English speakers, 54
or 9.0% identified themselves as Hispanic. Out of the 544 English speakers, 410 or
75.4% received CRC screening, while 134 or 24.6% did not. Among the Spanish-
speaking respondents, 30 or 55.6% received CRC screening, while 24 or 44.4% did not
receive CRC screening (Table 15). The data indicated that some participants who
identified themselves as Hispanic respondents in the BRFSS survey, chose to respond to
the survey questionnaire in English instead of their native Spanish language, thereby
decreasing the Spanish respondents from 134 Hispanic only participants to 54 as Spanish-
speaking respondents (Table 12).
Table 15
Language of Participants and Colorectal Screening Status
Yes, screened for
colorectal cancer
No, not screened for
colorectal cancer
Language of
participants
English
Count 410 134 544
% within Language of
participants 75.4% 24.6% 100.0%
Spanish
Count 30 24 54
% within Language of
participants 55.6% 44.4% 100.0%
Total
Count 440 158 598
% within Language of
participants 73.6% 26.4% 100.0%
115
The Chi-square tests indicated that Pearson Chi-square χ(1) = 9.918, p = <.002, showing
that there was a statistically significant association between English- and Spanish-
speaking respondents (Table 16).
Table 16
Chi-Square Tests on Participants’ Language and Colorectal Screening Status
Value df
Asymptotic
Significance (2-sided)
Exact Sig.
(2-sided)
Exact Sig.
(1-sided)
Pearson Chi-Square 9.918a 1 .002
Continuity
Correctionb
8.926 1 .003
Likelihood Ratio 9.014 1 .003
Fisher’s Exact Test .003 .002
Linear-by-Linear
Association
9.902 1 .002
N of Valid Cases 598
Note: a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 14.27; b.
Computed only for a 2x2 table.
Symmetric measures of Phi and Cramer’s V were the same, 0.129, each at p = 0.002. The
Phi and Cramer’s V values being more than 0.10 indicated there were differential
influences among English- and Spanish-speaking respondents (Table 17).
Table 17
Symmetric Measures on Language of Participants and Colorectal Screening Status
Value
Approximate
Significance
Nominal by Nominal Phi .129 .002
Cramer’s V .129 .002
N of Valid Cases 598
The study indicated that out of 214 participants in the study who were employed, 142 or
66.4% screened and 72 or 33.6% did not screen for CRC. Also, 379 participants in the
study were unemployed but 294 or 77.6% received screening while 85 or 22.4% did not
116
receive screening for CRC. The increased higher number of unemployment would likely
be due to retirement. Because individuals who are at least 65 years of age are Medicare-
eligible, most of those unemployed but retired would have Medicare for health
screening, including CRC screening (Table 18).
Table 18
Employment Status and Colorectal Cancer Screening
Yes, screened
for colorectal
cancer
No, not screened
for colorectal
cancer
Employme
nt status
Yes,
Employed
Count 142 72 214
% within
Employmen
t status
66.4% 33.6% 100.0%
No, not
employed
Count 294 85 379
% within
Employmen
t status
77.6% 22.4% 100.0%
Total
Count 436 157 593
% within
Employmen
t status
73.5% 26.5% 100.0%
The Chi-square tests indicated that Pearson Chi-square χ(1) = 8.841, p = .003,
showing that there was a statistically significant association between employment and
CRC screening (Table 19).
117
Table 19
Chi-Square Tests on Employment of Participants and Colorectal Screening Status
Value df Asymptotic Significance
(2-sided)
Exact Sig.
(2-sided)
Exact
Sig.
(1-
sided)
Pearson Chi-Square 8.841 a 1 .003
Continuity Correction b 8.274 1 .004
Likelihood Ratio 8.676 1 .003
Fisher’s Exact Test .004 .002
Linear-by-Linear
Association 8.826 1 .003
N of Valid Cases 593
Note: a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is
56.66. b. Computed only for a 2x2 table.
Symmetric measures of Phi and Cramer’s V were -0.122 and 0.122, respectively, each at
p = 0.003. The Phi and Cramer’s V values being more than 0.10 indicated there was an
association between employment status and CRC screening (Table 20).
Table 20
Symmetric Measures on Employment of Participants and Colorectal Screening Status
Value
Approximate
Significance
Nominal by Nominal Phi -.122 .003
Cramer’s V .122 .003
N of Valid Cases 593
Out of 549 respondents who had healthcare access, 420 or 76.5% screened for
CRC while 129 or 23.5% did not. Among the uninsured, only 19 or 39.6% underwent
CRC screening with 29 or 60.4% not undertaking CRC screening, an indication that
healthcare access plays a role in CRC screening (Table 21).
118
Table 21
Healthcare Access and Colorectal Cancer Screening
Yes, screened
for colorectal
cancer
No, not screened
for colorectal
cancer
Total
Healthcare
access
Yes,
Healthcare
access
Count 420 129 549
% within
Healthcare access
76.5 23.5 100.0
No healthcare
access
Count 19 29 48
% within
Healthcare access
39.6 60.4 100.0
Total
Count 439 158 597
% within
Healthcare access
73.5 26.5 100.0
The Chi-square tests indicated that Pearson Chi-square χ(1) = 30.915, p <.001, showing
that there was a statistically significant association between healthcare access and CRC
screening (Table 22).
Table 22
Chi-Square Tests on Healthcare Access of Participants and Colorectal Screening Status
Value df
Asymptotic
Significance (2-
sided)
Exact Sig.
(2-sided)
Exact Sig.
(1-sided)
Pearson Chi-Square 30.915 a 1 <.001
Continuity Correction b 29.047 1 <.001
Likelihood Ratio 26.889 1 <.001
Fisher’s Exact Test <.001 <.001
Linear-by-Linear
Association
30.864 1 <.001
N of Valid Cases 597
Note: a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 12.70. b.
Computed only for a 2x2 table.
119
Symmetric measures of Phi and Cramer’s V were both 0.228, each at p <.001. The Phi
and Cramer’s V values being more than 0.10 indicated there was an association between
healthcare access and CRC screening (Table 23).
Table 23
Symmetric Measures on Health Care Access of Participants and Colorectal Screening
Status
Value
Approximate
Significance
Nominal by Nominal Phi .228 <.001
Cramer’s V .228 <.001
N of Valid Cases 597
I performed a one-sample t test to assess the age variations within the English and
Spanish study participants for CRC screening. Table 24 shows the result of the t test. The
English-speakers had a sample mean age of 65.86 (SD = ± 8.43) and the Spanish-
speakers had a sample mean age of 65.34 (SD = ± 8.55), t(596)=-3.170, p =.002 (two-
tailed), 95% CI (-.321, -.075) (Table 26). The result in Table 24 shows that while the
English-speaking participants largely outnumbered the Spanish-speaking participants,
their mean ages were similar, with the English-speaking group at 65.86 years and
Spanish-speaking group at 65.34 years of age (Table 24). The Levene’s test for equality
of variance, F = 16.803, p <.001, an indication that the variances are not equal (Table
25).
120
Table 24
Group Statistics for English Speaking and Spanish Speaking Participants in CRC
Screening
Language of
participants N Mean
Std.
Deviation Std. Error Mean
Colorectal cancer
screening status
English 544 1.25 .431 .018
Spanish 54 1.44 .502 .068
Age of
participants
English 696 65.86 8.426 .319
Spanish 70 65.34 8.545 1.021
Table 25
One Sample Statistics 1: Independent Samples Test on Colorectal Cancer Screening and
Age of Participants
Levene’s Test for
Equality of Variances
(EV)
F Sig.
Colorectal cancer
screening status
Equality of variances
assumed 16.80 <.001
Equality of variances not
assumed 3
Age of participant
Equality of variances
assumed .013 .909
Equality of variances not
assumed
121
Table 26
One Sample Statistics 2: Independent Samples Test on Colorectal Cancer Screening and
Age of Participants
t-test for Equality of Means
t df
Significance
Mean
Difference
Std. Error
Difference
95%
Confidence
Interval of the
Difference
One-
Sided p
Two-
Sided p Lower Upper
Colorectal
cancer
screening
status
EV
assumed
3.17
0 596 <.001 .002 -.198 .062 -.321 -.075
EV not
assumed
2.80
2
61.03
2 .003 .007 -.198 .071 -.340 -.057
Age of
participant
EV
assumed .485 764 .314 .628 .513 1.058 -1.563 2.590
EV not
assumed .480
83.07
6 .316 .633 .513 1.070 -1.615 2.642
Also, conducted a one-sample t test for continuous variables (income and highest
education achieved: some high school, high school, some college, college) to assess the
implications of income and higher education on CRC screening. Table 25 shows the
result of the t test. The income of participants had a sample mean age of 5.83 (SD = ±
2.202) and the education status of participants had a sample mean age of 2.97 (SD = ±
1.039), t(597) = 70.057, p <.001(two-tailed), 95% CI (1.23, 1.30) (Table 28). The result
in Table 27 shows that participants with income and education, N was 604 and 763,
respectively.
122
Table 27
Number of Participants (N), Mean, Std Deviation, and Std Error Mean and CRC
Screening and Participants with Income and Education
N Mean Std. Deviation Std. Error Mean
Colorectal cancer screening status 598 1.26 .441 .018
Income of participants 604 5.83 2.202 .090
Education status 763 2.97 1.039 .038
Table 28
One-Sample T Test for CRC Screening with Income and Education of Participants
Test Value = 0
t df
Significance
Mean
Difference
95% Confidence
Interval of the
Difference
One-
Sided
p
Two-
Sided
p
Lower Upper
Colorectal cancer
screening status 70.057 597 <.001 <.001 1.264 1.23 1.30
Income of
participants 65.041 603 <.001 <.001 5.828 5.65 6.00
Education status 78.953 762 .000 .000 2.971 2.90 3.05
Colonoscopy screening was more prominent among groups 60-64, 65-69, 70-74, and 75-
79 compared to age groups 50-54 and 55-59. Groups 60-64 (82 or 15.6%), 65-69 (109 or
20.8%), 70-74 (126 or 24.0%), and 75-79 (110 or 21.0%) had colonoscopy compared to
groups 50-54 (45 or 8.6%) and 55-59 (52 or 9.9%) had colonoscopy (Table 29).
123
Table 29
Crosstabulation of Five-year Age Groups and Colonoscopy Screening Status
5-year age groups Total
50-54 55-59 60-64 65-69 70-74 75-79
Have you
ever had a
colonoscopy?
Yes
Count 45 52 82 109 126 110 524
% within Have
you ever had a
colonoscopy?
8.6 9.9 15.6 20.8 24.0 21.0 100.0
No
Count 52 33 26 36 27 22 196
% within Have
you ever had a
colonoscopy?
26.5 16.8 13.3 18.4 13.8 11.2 100.0
Total
Count 97 85 108 145 153 132 720
% within Have
you ever had a
colonoscopy?
13.5 11.8 15.0 20.1 21.3 18.3 100.0
The Chi-square tests indicated that Pearson Chi-square χ(1) = 55.326, p <.001, showing
that there was a statistically significant association among five-year age groups 50-54,
55-59, 60-64, 65-69, 70-74, and 75-79 (Table 30).
Table 30
Chi-Square Tests on Five-year Age Groups and Colonoscopy Screening Status
Value df Asymptotic Significance (2-sided)
Pearson Chi-Square 55.326 a 5 <.001
Likelihood Ratio 52.254 5 <.001
Linear-by-Linear
Association 47.064 1 <.001
N of Valid Cases 720
Note: a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 23.14.
Symmetric measures of Phi and Cramer’s V were the same, 0.277, each at p
<.001. The Phi and Cramer’s V values being more than 0.10 indicated there were
124
differential influences among five-year age groups and colonoscopy screening status
(Table 31).
Table 31
Symmetric Measures on Five-year Age Groups and Colorectal Screening Status
Value
Approximate
Significance
Nominal by Nominal Phi .277 <.001
Cramer’s V .277 <.001
N of Valid Cases 720
There was a significant variation between males and female for colonoscopy screening,
with 315 or 60.1% of females undergoing colonoscopy screening compared to 209 or
39.9% of males who received the same screening (Table 32).
Table 32
Sex of Participants and Colonoscopy Screening Status
Sex of participants
Total Male Female
Have you
ever had a
colonoscopy?
Yes
Count 209 315 524
% within Have
you ever had a
colonoscopy?
39.9 60.1 100.0
No
Count 92 104 196
% within Have
you ever had a
colonoscopy?
46.9 53.1 100.0
Total
Count 301 419 720
% within Have
you ever had a
colonoscopy?
41.8 58.2 100.0
125
The Chi-square tests indicated that Pearson Chi-square χ(1) = 2.917, p <.088, showing
that there was no statistically significant association between male and female
participants in the study (Table 33).
Table 33
Chi-Square Tests on Males and Females and Colonoscopy Screening Status
Value df
Asymptotic
Significance
(2-sided)
Exact Sig.
(2-sided)
Exact Sig.
(1-sided)
Pearson Chi-Square 2.917 a 1 .088
Continuity Correction b 2.634 1 .105
Likelihood Ratio 2.900 1 .089
Fisher’s Exact Test .090 .053
Linear-by-Linear
Association 2.913 1 .088
N of Valid Cases 720
Note: a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 81.94. b.
Computed only for a 2x2 table.
Symmetric measures of Phi and Cramer’s V were - 0.064 and 0.064, respectively, each at
p = .088. The Phi and Cramer’s V values being less than 0.10 indicated there were no
differential influences between males and females and colonoscopy screening status
(Table 34).
Table 34
Symmetric Measures on Sex of Participants and Colorectal Screening Status
Value
Approximate
Significance
Nominal by Nominal Phi -.064 .088
Cramer’s V .064 .088
N of Valid Cases 720
126
The choice of colonoscopy among racial groups in the study varied significantly. Among
White only, non-Hispanic, 387 or 73.9% had colonoscopy compared with 47 or 9.0% of
Black only, non-Hispanic, while 90 or 17.2% of Hispanics had colonoscopy (Table 35).
Table 35
Race of Participants and Colonoscopy Screening Status
Race of participants
Total
White only,
non-Hispanic
Black only,
non-Hispanic Hispanic
Have you
ever had a
colonoscopy?
Yes
Count 387 47 90 524
% within Have
you ever had a
colonoscopy?
73.9 9.0 17.2 100.0
No
Count 108 22 66 196
% within Have
you ever had a
colonoscopy?
55.1 11.2 33.7 100.0
Total
Count 495 69 156 720
% within Have
you ever had a
colonoscopy?
68.8 9.6 21.7 100.0
The Chi-square tests indicated that Pearson Chi-square χ(1) = 25.973, p <.001,
showing that there was a statistically significant association among White, non-Hispanic,
Black, non-Hispanic, and Hispanic participants in the study (Table 36).
127
Table 36
Chi-Square Tests on Race of Participants and Colonoscopy Screening Status
Value df
Asymptotic Significance
(2-sided)
Pearson Chi-Square 25.973a 2 <.001
Likelihood Ratio 24.765 2 <.001
Linear-by-Linear Association 24.342 1 <.001
N of Valid Cases 720
Note: a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 18.78.
Symmetric measures of Phi and Cramer’s V were the same, 0.190, each at p
<.001. The Phi and Cramer’s V values being more than 0.10 indicated there were
differential influences among White, non-Hispanic, Black, non-Hispanic, and Hispanic
and colonoscopy screening status (Table 37).
Table 37
Symmetric Measures on Race of Participants and Colonoscopy Screening Status
Value
Approximate
Significance
Nominal by Nominal Phi .190 <.001
Cramer’s V .190 <.001
N of Valid Cases 720
The study indicated that 524 participants received colonoscopy screening. However, 492
or 93.9% were English-speaking while only 32 or 6.1% were Spanish-speaking
participants in the study (Table 38).
128
Table 38
Language of Participants and Colonoscopy Screening Status
English Spanish Total
Have you ever
had a
colonoscopy?
Yes
Count 492 32 524
% within Have you ever had a
colonoscopy? 93.9 6.1 100.0
No
Count 163 33 196
% within Have you ever had a
colonoscopy? 83.2 16.8 100.0
Total
Count 655 65 720
% within Have you ever had a
colonoscopy? 91.0 9.0 100.0
The Chi-square tests indicated that Pearson Chi-square χ(1) = 19.996, p <.001, showing
that there was a statistically significant association between English-speaking and
Spanish-speaking participants in the study (Table 39).
Table 39
Chi-square Tests on Language of Participants and Colonoscopy Screening Status
Value df
Asymptotic
Significance
(2-sided)
Exact Sig.
(2-sided)
Exact Sig.
(1-sided)
Pearson Chi-Square 19.996a 1 <.001
Continuity
Correctionb
18.711 1 <.001
Likelihood Ratio 17.957 1 <.001
Fisher’s Exact Test <.001 <.001
Linear-by-Linear
Association
19.969 1 <.001
N of Valid Cases 720
Note: 0 cells (0.0%) have expected count less than 5. The minimum expected count is 17.69. b.
Computed only for a 2x2 table.
Symmetric measures of Phi and Cramer’s V were the same, 0.167, each at p <.001. The
Phi and Cramer’s V values being less than 0.10 indicated there were differential
129
influences between English-speaking and Spanish-speaking study participants and
colonoscopy screening status (Table 40).
Table 40
Symmetric Measures on Language of Participants and Colonoscopy Screening Status
Value
Approximate
Significance
Nominal by Nominal Phi .167 <.001
Cramer’s V .167 <.001
N of Valid Cases 720
The study indicated that within the past 10 years, 381 participants received
colonoscopy screening, with the highest screening taking place among individuals in 70-
74 age group (112 or 29.4%), followed by individuals in 65-69 age group (92 or 24.1%).
Groups 50-54 (41 or 10.8%), 55-59 (50 or 13.1%), 60-64 (68 or 17.8%), and 75-79 (18 or
4.7%) (Table 41).
Table 41
Five-year Age Groups and Colonoscopy Screening Status Within the Past 10 Years
5-year age groups Total
50-54 55-59 60-64 65-69 70-74 75-79
Colonoscopy
within the
past 10 years
Yes
Count 41 50 68 92 112 18 381
% within
Colonoscopy within
the past 10 years
10.8 13.1 17.8 24.1 29.4 4.7 100.0
No
Count 55 35 39 49 39 4 221
% within
Colonoscopy within
the past 10 years,
50-75
24.9 15.8 17.6 22.2 17.6 1.8 100.0
Total
Count 96 85 107 141 151 22 602
% within
Colonoscopy within
the past 10 years,
50-75
15.9 14.1 17.8 23.4 25.1 3.7 100.0
130
The Chi-square tests indicated that Pearson Chi-square χ(1) = 29.415, p <.001, showing
that there was a statistically significant association among five-year age groups in the
study (Table 42).
Table 42
Chi-square Tests on Five-year Age Groups and Colonoscopy Screening Status Within the
Past 10 Years
Value df Asymptotic Significance (2-sided)
Pearson Chi-Square 29.415 a 5 <.001
Likelihood Ratio 29.403 5 <.001
Linear-by-Linear Association 27.004 1 <.001
N of Valid Cases 602
Note: a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 8.08.
Symmetric measures of Phi and Cramer’s V were the same, 0.221, each at p <.001. The
Phi and Cramer’s V values being more than 0.10 indicated there were differential
influences among five-year age groups and colonoscopy screening status (Table 43).
Table 43
Symmetric Measures on Five-year Age Groups and Colonoscopy Screening Status Within
the Past 10 Years
Value
Approximate
Significance
Nominal by Nominal Phi .221 <.001
Cramer’s V .221 <.001
N of Valid Cases 602
The results indicated that within the past 10 years, 162 males (42.5%) and 219 females
(57.5%) received colonoscopy screening (Table 44).
131
Table 44
Sex of Participants and Colonoscopy Screening Status Within the Past 10 Years
Sex of participants
Total Male Female
Colonoscopy
within the
past 10 years
Yes
Count 162 219 381
% within
Colonoscopy within
the past 10 years
42.5 57.5 100.0
No
Count 100 121 221
% within
Colonoscopy within
the past 10 years
45.2 54.8 100.0
Total
Count 262 340 602
% within
Colonoscopy within
the past 10 years
43.5 56.5 100.0
The Chi-square tests indicated that Pearson Chi-square χ(1) = 0.424, p =.515, showing
that there was no statistically significant association between male and female
participants in the study (Table 45).
Table 45
Chi-Square Tests on Sex of Participants and Colonoscopy Screening Status Within the
Past 10 Years
Value df
Asymptotic
Significance
(2-sided)
Exact Sig.
(2-sided)
Exact Sig.
(1-sided)
Pearson Chi-Square .424 a 1 .515
Continuity
Correction b
.320 1 .572
Likelihood Ratio .423 1 .515
Fisher’s Exact Test .551 .286
Linear-by-Linear
Association
.423 1 .515
N of Valid Cases 602
Note: a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 96.18.
b. Computed only for a 2x2 table.
132
Symmetric measures of Phi and Cramer’s V were -.027 and .027, respectively
each at p = .515. The Phi and Cramer’s V values being less than 0.10 indicated there
were no differential influences between male and female study participants and
colonoscopy screening status (Table 46).
Table 46
Symmetric Measures on Sex of Participants and Colonoscopy Screening Status Within
the Past 10 Years
Value
Approximate
Significance
Nominal by Nominal Phi -.027 .515
Cramer’s V .027 .515
N of Valid Cases 602
The results indicated that within the past 10 years, 274 (71.9%) White, non-Hispanic, 34
(8.9%) Black only, non-Hispanic, and 73 (19.2%) Hispanic participants received
colonoscopy screening (Table 47).
133
Table 47
Race of Participants and Colonoscopy Screening Status Within the Past 10 Years
Race of participants Total
White only,
non-
Hispanic
Black only,
non-Hispanic Hispanic
Colonoscopy
within the
past 10 years
Yes
Count 274 34 73 381
% within
Colonoscopy
within the past
10 years
71.9 8.9 19.2 100.0
No
Count 133 24 64 221
% within
Colonoscopy
within the past
10 years
60.2 10.9 29.0 100.0
Total
Count 407 58 137 602
% within
Colonoscopy
within the past
10 years
67.6 9.6 22.8 100.0
The Chi-square tests indicated that Pearson Chi-square χ(1) = 9.295, p =.010,
showing that there was statistically significant association between male and female
participants in the study (Table 48).
Table 48
Chi-Square Tests on Race of Participants and Colonoscopy Screening Status Within the
Past 10 Years
Value df Asymptotic Significance (2-sided)
Pearson Chi-Square 9.295a 2 .010
Likelihood Ratio 9.159 2 .010
Linear-by-Linear Association 8.276 1 .004
N of Valid Cases 602
Note: a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 21.29.
134
Symmetric measures of Phi and Cramer’s V were the same, .124, each at p = .010. The
Phi and Cramer’s V values being more than 0.10 indicated there were differential
influences among White, non-Hispanic, Black, non-Hispanic, and Hispanic study
participants and colonoscopy screening status (Table 49).
Table 49
Symmetric Measures on Race of Participants and Colonoscopy Screening Status Within
the Past 10 Years
Value
Approximate
Significance
Nominal by Nominal Phi .124 .010
Cramer’s V .124 .010
N of Valid Cases 602
The results indicated that within the past 10 years, 356 (93.4%) English-speaking and 25
(6.6%) of Spanish-speaking study subjects received colonoscopy screening (Table 50).
Table 50
Language of Participants and Colonoscopy Screening Status Within the Past 10 Years
Language of
participants
Total English Spanish
Colonosco
py within
the past 10
years
Yes
Count 356 25 381
% within
Colonoscopy within
the past 10 years
93.4 6.6 100.0
No
Count 191 30 221
% within
Colonoscopy within
the past 10 years
86.4 13.6 100.0
Total
Count 547 55 602
% within
Colonoscopy within
the past 10 years
90.9 9.1 100.0
135
The Chi-square tests indicated that Pearson Chi-square χ(1) = 8.286, p =.004, showing
that there was statistically significant association between English-speaking and Spanish-
speaking participants in the study (Table 51).
Table 51
Chi-Square Tests on Language of Participants and Colonoscopy Screening Status Within
the Past 10 Years
Value df
Asymptotic
Significance
(2-sided)
Exact Sig.
(2-sided)
Exact Sig.
(1-sided)
Pearson Chi-Square 8.286 a 1 .004
Continuity Correction b 7.463 1 .006
Likelihood Ratio 7.970 1 .005
Fisher’s Exact Test .005 .004
Linear-by-Linear
Association 8.273 1 .004
N of Valid Cases 602
Note: a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 20.19. b.
Computed only for a 2x2 table.
Symmetric measures of Phi and Cramer’s V were the same, .117, each at p =
.004. The Phi and Cramer’s V values being more than 0.10 indicated there were
differential influences between English-speaking and Spanish-speaking study participants
and colonoscopy screening status (Table 52).
Table 52
Symmetric Measures on Language of Participants and Colonoscopy Screening Status
Within the Past 10 Years
Value
Approximate
Significance
Nominal by Nominal Phi .117 .004
Cramer’s V .117 .004
N of Valid Cases 602
136
Sigmoidoscopy screening was more prominent among groups 60-64, 65-69 and
70-74 compared to age groups 50-54, 55-59, and 74-79 age groups. However, in
comparison to colonoscopy, fewer participants in the study used sigmoidoscopy
screening. Groups 60-64 (4 or 12.9%), 65-69 (8 or 25.8%), 70-74 (11 or 35.5%) had
sigmoidoscopy compared to groups 50-54 (3 or 9.7%), 55-59 (2 or 6.5%), and 75-79 (3
or 9.7%) had sigmoidoscopy. In all, only 31 participants had sigmoidoscopy while 555
respondents stated that they had not received sigmoidoscopy (Table 53).
Table 53
Five-year Age Groups and Sigmoidoscopy Screening Status Within the Past Five Years
5-year age groups Total
50-54 55-59 60-64 65-69 70-74 75-79
Sigmoidoscopy
within the past
5 years
Yes
Count 3 2 4 8 11 3 31
% within
Sigmoidoscopy
within the past
5 years
9.7 6.5 12.9 25.8 35.5 9.7 100.0
No
Count 92 80 99 130 135 19 555
% within
Sigmoidoscopy
within the past
5 years
16.6 14.4 17.8 23.4 24.3 3.4 100.0
Total
Count 95 82 103 138 146 22 586
% within
Sigmoidoscopy
within the past
5 years
16.2 14.0 17.6 23.5 24.9 3.8 100.0
The Chi-square tests indicated that Pearson Chi-square χ(1) = 7.196, p =.206, showing
that there was no statistically significant association between male and female
participants in the study (Table 54).
137
Table 54
Chi-Square Tests on Five-year Age Groups and Sigmoidoscopy Screening Status Within
the Past Five Years
Value df
Asymptotic
Significance (2-sided)
Pearson Chi-Square 7.196a 5 .206
Likelihood Ratio 6.645 5 .248
Linear-by-Linear Association 5.637 1 .018
N of Valid Cases 586
Note: a. 2 cells (16.7%) have expected count less than 5. The minimum expected count is 1.16.
Symmetric measures of Phi and Cramer’s V were the same, 0.111, each at p =.206. The
Phi and Cramer’s V values being more than 0.10 indicated there were differential
influences among five-year age groups and sigmoidoscopy screening status (Table 55).
Table 55
Symmetric Measures on Five-year Age Groups and Sigmoidoscopy Screening Status
Within the Past Five Years
Value
Approximate
Significance
Nominal by Nominal Phi .111 .206
Cramer’s V .111 .206
N of Valid Cases 586
Among respondents only 31 had taken sigmoidoscopy screening within the past 5 years,
out of which 19 or 61.3% were females and 12 or 38.7% were females (Table 56).
138
Table 56
Sex of Participants and Sigmoidoscopy Screening Status Within the Past Five Years
Sex of participants
Total Male Female
Sigmoidoscop
y within the
past 5 years
Yes
Count 12 19 31
% within Sigmoidoscopy
within the past 5 years 38.7 61.3 100.0%
No
Count 240 315 555
% within Sigmoidoscopy
within the past 5 years 43.2 56.8 100.0%
Total
Count 252 334 586
% within Sigmoidoscopy
within the past 5 years 43.0 57.0 100.0%
The Chi-square tests indicated that Pearson Chi-square χ(1) = 0.246, p =.620, showing
that there was no statistically significant association between male and female
participants in the study (Table 57).
Table 57
Chi-Square Tests on Sex of Participants and Sigmoidoscopy Screening Status Within the
Past Five Years
Value df
Asymptotic
Significance
(2-sided)
Exact Sig.
(2-sided)
Exact Sig.
(1-sided)
Pearson Chi-Square .246 a 1 .620
Continuity Correction b .096 1 .757
Likelihood Ratio .248 1 .618
Fisher’s Exact Test .711 .381
Linear-by-Linear Association .246 1 .620
N of Valid Cases 586
Note: a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 13.33. b.
Computed only for a 2x2 table.
139
Symmetric measures of Phi and Cramer’s V were -.020 and .020, respectively,
each at p = .620. The Phi and Cramer’s V values being less than 0.10 indicated there
were no differential influences between male and female study participants and
sigmoidoscopy screening status (Table 58).
Table 58
Symmetric Measures on Sex of Participants and Sigmoidoscopy Screening Status Within
the Past Five Years
Value Approximate Significance
Nominal by Nominal Phi -.020 .620
Cramer’s V .020 .620
N of Valid Cases 586
Sigmoidoscopy screening within the past 5 years was highest among White only,
non-Hispanic (16 or 51.6%) and lowest among Black only non-Hispanic (5 or 16.1%)
with Hispanics (10 or 32.3%) fallen in between (Table 59).
140
Table 59
Race of Participants and Sigmoidoscopy Screening Status Within the Past Five Years
Race of participants
Total
White only,
non-Hispanic
Black only,
non-Hispanic Hispanic
Sigmoidoscopy
within the past
5 years
Yes
Count 16 5 10 31
% within
Sigmoidoscopy
within the past 5
years
51.6 16.1 32.3 100.0
No
Count 382 53 120 555
% within
Sigmoidoscopy
within the past 5
years
68.8 9.5 21.6 100.0
Total
Count 398 58 130 586
% within
Sigmoidoscopy
within the past 5
years
67.9 9.9 22.2 100.0
The Chi-square tests indicated that Pearson Chi-square χ(1) = 4.063, p =.131,
showing that there was no statistically significant association among White only, non-
Hispanic, Black only non-Hispanic, and Hispanics participants in the study (Table 60).
Table 60
Chi-Square Tests on Race of Participants and Sigmoidoscopy Screening Status Within
the Past Five Years
Value df
Asymptotic Significance
(2-sided)
Pearson Chi-Square 4.063a 2 .131
Likelihood Ratio 3.802 2 .149
Linear-by-Linear Association 2.336 1 .126
N of Valid Cases 586
Note: a. 1 cells (16.7%) have expected count less than 5. The minimum expected count is 3.07.
141
Symmetric measures of Phi and Cramer’s V were the same, 0.083, each at p = .131. The
Phi and Cramer’s V values being less than 0.10 indicated there were no differential
influences among White, non-Hispanic, Black, non-Hispanic, and Hispanic study
participants and sigmoidoscopy screening status (Table 61).
Table 61
Symmetric Measures on Race of Participants and Sigmoidoscopy Screening Status Within
the Past Five Years
Value Approximate Significance
Nominal by Nominal Phi .083 .131
Cramer’s V .083 .131
N of Valid Cases 586
Out of 31 respondents who indicated that they had taken sigmoidoscopy screening within
the past 5 years, 25 or 80.6% responded in English and 6 or 19.4% responded in Spanish
to the survey questions (Table 62).
Table 62
Language of Participants and Sigmoidoscopy Screening Status Within the Past Five
Years
Language of participants
Total English Spanish
Sigmoidoscopy
within the past
5 years
Yes
Count 25 6 31
% within Sigmoidoscopy
within the past 5 years 80.6 19.4 100.0
No
Count 510 45 555
% within Sigmoidoscopy
within the past 5 years 91.9 8.1 100.0
Total
Count 535 51 586
% within Sigmoidoscopy
within the past 5 years 91.3 8.7 100.0
142
The Chi-square tests indicated that Pearson Chi-square χ(1) = 4.674, p =.031, showing
that there was statistically significant association between English-speaking and Spanish-
speaking participants in the study (Table 63).
Table 63
Chi-Square Tests on Language of Participants and Sigmoidoscopy Screening Status
Within the Past Five Years
Value df
Asymptotic
Significance
(2-sided)
Exact Sig.
(2-sided)
Exact Sig.
(1-sided)
Pearson Chi-Square 4.674 a 1 .031
Continuity
Correction b 3.366 1 .067
Likelihood Ratio 3.641 1 .056
Fisher’s Exact Test .044 .044
Linear-by-Linear
Association 4.666 1 .031
N of Valid Cases 586
Note: a. 1 cells (25.0%) have expected count less than 5. The minimum expected count is 2.70.
b. Computed only for a 2x2 table.
Symmetric measures of Phi and Cramer’s V were -0.089 and 0.089, respectively,
each at p = .031. The Phi and Cramer’s V values being less than 0.10 indicated there
were no differential influences between English-speaking and Spanish-speaking study
participants and sigmoidoscopy screening status (Table 64).
Table 64
Symmetric Measures on Language of Participants and Sigmoidoscopy Screening Status
Within the Past Five Years
Value Approximate Significance
Nominal by Nominal Phi -.089 .031
Cramer’s V .089 .031
N of Valid Cases 586
143
Among respondents 105 had taken blood stool test screening for CRC within the past
year. Groups 50-54 (13 or 12.4%), 60-64 (18 or 17.1%), 65-69 (25 or 23.8%), 70-74 (40
or 38.1%) had blood stool test screening within the past year compared to groups 55-59
(5 or 4.8%) and 75-79 (4 or 3.8%) had blood stool test screening (Table 65).
Table 65
Five-year Age Groups and Blood Stool Test Within the Past Year
5-year age groups Total
50-54 55-59 60-64 65-69 70-74 75-79
Blood
stool
test
within
the past
year
Yes
Count 13 5 18 25 40 4 105
% within Blood
stool test
within the past year
12.4 4.8 17.1 23.8 38.1 3.8 100.0
No
Count 81 80 91 117 113 17 499
% within Blood
stool test
within the past year
16.2 16.0 18.2 23.4 22.6 3.4 100.0
Total
Count 94 85 109 142 153 21 604
% within Blood
stool test
within the past year
15.6 14.1 18.0 23.5 25.3 3.5 100.0
The Chi-square tests indicated that Pearson Chi-square χ(1) = 16.933, p =.005,
showing that there was a statistically significant association among five-year age group
participants in the study (Table 66).
144
Table 66
Chi-Square Tests for Five-year Age Groups and Blood Stool Test Within the Past Year
Value df
Asymptotic
Significance
(2-sided)
Pearson Chi-Square 16.933a 5 .005
Likelihood Ratio 18.312 5 .003
Linear-by-Linear Association 10.566 1 .001
N of Valid Cases 604
Note: a. 1 cells (8.3%) have expected count less than 5. The minimum expected count is 3.65
Symmetric measures of Phi and Cramer’s V were the same, 0.167, each at p =.005. The
Phi and Cramer’s V values being more than 0.10 indicated there were differential
influences among five-year age groups and blood stool test screening status (Table 67).
Table 67
Symmetric Measures on Five-year Age Groups and Blood Stool Test Within the Past Year
Value Approximate Significance
Nominal by Nominal Phi .167 .005
Cramer’s V .167 .005
N of Valid Cases 604
Among respondents 105 had taken blood stool test screening for CRC within the past
year, with 52 or 49.5% males and 53 or 50.5% females involved. This indicated that
blood stool test was nearly even between males and females (Table 68).
145
Table 68
Sex of Participants and Blood Stool Test Within the Past Year
Sex of participants Total
Male Female
Blood
stool test
within the
past year
Yes Count 52 53 105
% within Blood stool test
within the past year 49.5 50.5 100.0
No Count 206 293 499
% within Blood stool test
within the past year 41.3 58.7 100.0
Total Count 258 346 604
% within Blood stool test
within the past year 42.7 57.3 100.0
The Chi-square tests indicated that Pearson Chi-square χ(1) = 2.408, p =.121, showing
that there was no statistically significant association between male and female
participants in the study (Table 69).
Table 69
Chi-Square Tests on Sex of Participants and Blood Stool Test Within the Past Year
Value df
Asymptotic
Significance
(2-sided)
Exact Sig.
(2-sided)
Exact Sig.
(1-sided)
Pearson Chi-Square 2.408 a 1 .121
Continuity
Correction b 2.083 1 .149
Likelihood Ratio 2.389 1 .122
Fisher’s Exact Test .129 .075
Linear-by-Linear
Association 2.404 1 .121
N of Valid Cases 604
Note: a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 44.85. b.
Computed only for a 2x2 table.
Symmetric measures of Phi and Cramer’s V were the same, 0.063, each at p =.121. The
Phi and Cramer’s V values being less than 0.10 indicated there were no differential
146
influences between male and female participants and blood stool test screening status
(Table 70).
Table 70
Symmetric Measures on Sex and Blood Stool Test Within the Past Year
Value Approximate Significance
Nominal by Nominal Phi .063 .121
Cramer’s V .063 .121
N of Valid Cases 604
The results indicated that 68 or 64.8% of White only, non-Hispanic, 15 or 14.3%
of Black, non-Hispanic, and 22 or 21.0% Hispanic had taken blood stool test screening
for CRC within the past year (Table 71).
Table 71
Race of Participants and Blood Stool Test Within the Past Year
Race of participants Total
White only,
non-Hispanic
Black only,
non-Hispanic Hispanic
Blood stool
test within
the past
year
Yes
Count 68 15 22 105
% within
Blood stool
test within the
past year
64.8 14.3 21.0 100.0
No
Count 341 44 114 499
% within
Blood stool
test within the
past year
68.3 8.8 22.8 100.0
Total
Count 409 59 136 604
% within
Blood stool
test within the
past year
67.7 9.8 22.5 100.0
147
The Chi-square tests indicated that Pearson Chi-square χ(1) = 2.957, p =.228, showing
that there was no statistically significant association among White, non-Hispanic, Black,
non-Hispanic, and Hispanic study participants in the study (Table 72).
Table 72
Chi-Square Tests on Race of Participants and Blood Stool Test Within the Past Year
Symmetric measures of Phi and Cramer’s V were the same, 0.063, each at p
=.121. The Phi and Cramer’s V values being less than 0.10 indicated there were no
differential influences among White, non-Hispanic, Black, non-Hispanic, and Hispanic
study participants and blood stool test screening status (Table 73).
Table 73
Symmetric Measures on Race and Blood Stool Test Within the Past Year
Value Approximate Significance
Nominal by Nominal Phi .070 .228
Cramer’s V .070 .228
N of Valid Cases 604
The results indicated that 97 or 92.4% of English-speaking and 8 or 7.6% of
Spanish-speaking participants in the study had taken blood stool test screening for CRC
within the past year (Table 74).
Value df
Asymptotic
Significance
(2-sided)
Pearson Chi-Square 2.957a 2 .228
Likelihood Ratio 2.701 2 .259
Linear-by-Linear Association .063 1 .802
N of Valid Cases 604
148
Table 74
Language of Participants and Blood Stool Test Within the Past Year
Language of participants
Total English Spanish
Blood
stool test
within the
past year
Yes
Count 97 8 105
% within Blood stool test
within the past year 92.4 7.6 100.0
No
Count 451 48 499
% within Blood stool test
within the past year 90.4 9.6 100.0
Total
Count 548 56 604
% within Blood stool test
within the past year 90.7 9.3 100.0
The Chi-square tests indicated that Pearson Chi-square χ(1) = 0.413, p =.521,
showing that there was no statistically significant association between English-speaking
and Spanish-speaking participants in the study (Table 75).
Table 75
Chi-Square Tests on Language of Participants and Blood Stool Test Within the Past Year
Value df
Asymptotic
Significance
(2-sided)
Exact Sig.
(2-sided)
Exact Sig.
(1-sided)
Pearson Chi-Square .413 a 1 .521
Continuity
Correction b .209 1 .648
Likelihood Ratio .432 1 .511
Fisher’s Exact Test .711 .334
Linear-by-Linear
Association .412 1 .521
N of Valid Cases 604
Note: a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 9.74. b.
Computed only for a 2x2 table.
Symmetric measures of Phi and Cramer’s V were the same, 0.026, each at p
=.521. The Phi and Cramer’s V values being less than 0.10 indicated there were no
149
differential influences between English-speaking and Spanish-speaking study participants
and FOBT screening status (Table 76).
Table 76
Symmetric Measures on Language of Participants and Blood Stool Test Within the Past
Year
Value Approximate Significance
Nominal by Nominal Phi .026 .521
Cramer’s V .026 .521
N of Valid Cases 604
Inferential Statistical Analysis
I performed bivariate logistic regression and Chi-square tests to evaluate whether
there were significant associations between the DV and the motivating variables of CRC
screening including 5-year age groups, race, language, annual household income,
employment status, healthcare access, and education status. Bivariate analysis permits an
evaluation of how the value of the CRC screening (outcome variable) relies upon the
extent of LEP (explanatory variable). Also, I used bivariate analyses to compare gender
(explanatory variable) and CRC screening (outcome variable). Additionally, I employed
forced multivariate logistic regression analysis to assess the relationship between the
variables of interest, such as LEP and CRC screening among the target population. I
conducted a multivariate logistic regression for odds ratio (ORs) of CRC screening
between respondents with low LEP (fluent in English/both English and Spanish) and high
LEP (fluent in Spanish only).
150
Model with Interaction Effects
In a forced logistic regression that included 362 respondents representing 73.4%
of the total sample, Table 25 shows how the DV of CRC screening status was recorded.
For the BRFSS question, “Have you ever screened for CRC?” individuals who answered
Yes were coded 0, and all “No” respondents were coded 1 (Table 77). Table 78 shows
that the model itself explained 73.4% of the variance.
Table 77
DV Encoding
Original Value Internal Value
Yes, screened for
colorectal cancer
0
No, not screened
for colorectal
cancer
1
Table 78
Classification Table: Colorectal Cancer Status
Observed
Predicted
Colorectal cancer screening status
Percentage
Correct
Yes, screened
for colorectal
cancer
No, not screened
for colorectal
cancer
Step 0 Colorectal
cancer
screening status
Yes, screened for
colorectal cancer
362 0 100.0
No, not screened for
colorectal cancer
131 0 .0
Overall Percentage 73.4
Note: a. Constant is included in the model. b. The cut value is .500.
The Omnibus Tests of Model Coefficients (Table 79) indicated that the overall fit of the
model was significant: χ2 (7) = 47.080, p < .001. The Hosmer and Lemeshow Test, χ2 (8)
151
= 16.886, p = .031, was significant, indicating that the model was not predictable (Table
80).
Table 79
Omnibus Tests of Model Coefficients
Chi-square df Sig.
Step 1 Step 47.080 7 <.001
Block 47.080 7 <.001
Model 47.080 7 <.001
Table 80
Hosmer and Lemeshow Test
Step Chi-square df Sig.
1 16.886 8 .031
Table 81 indicates that the Nagelkerke R2 value of 13.3%, which is adjusted version of
Cox and Snell R2, and it shows that the predictor variables explained 13.3% of the
variance. The predictor variables improved the overall predictability of the model from
73.4% to 75.3% (Table 82).
Table 81
Model Summary
Step -2 Log likelihood Cox & Snell R Square Nagelkerke R Square
1 523.769 a .091 .133
Note: a. Estimation terminated at iteration number 5 because parameter estimates changed by less
than .001.
152
Table 82
Improved Predictor Variables
Observed
Predicted
Colorectal cancer screening status Percentage
Correct Yes, screened for
colorectal cancer
No, not screened for
colorectal cancer
Step 1 Colorectal
cancer
screening
status
Yes, screened for
colorectal cancer 347 15 95.9
No, not screened
for colorectal
cancer
107 24 18.3
Overall Percentage 75.3
Note: a. The cut value is .500.
I performed a series of bivariate analyses to evaluate whether there were
significant associations between the DV and the motivating variables of CRC screening
including 5-year age groups, race, language, annual household income, employment
status, healthcare access, and education status. The Wald test is employed to assess
statistical significance for the IVs. From Table 83, race with a Wald value of 0.087, p
=.768; language Wald value of 0.574, p =.449; income of participants Wald value of
0.175, p =.675; employment status Wald value of 1.168, p =.280; healthcare access Wald
value of 4.857, p <.028; education status Wald value of 5.070, p <.024; and 5-year age
groups with a Wald value of 16.163 had a p <.001. From these results, variables with p
<.05 were significant, while the variables with p >.05 were not significant. Thus, 5-year
age groups, healthcare access, and education status were significant, while race of
participants, language of participants, income of participants, and employment status
were not significant (Table 83).
153
Table 83 presents the unadjusted bivariate ORs for sociodemographic variables.
The sociodemographic variables and their associations with the unadjusted OR vary from
one another. For race of participants, OR = 0.987; 95% CI 0.902, 1.079, p = .768;
language, OR = 1.415; 95% CI 0.576, 3.476, p = .449; income, OR = 0.973; 95% CI
0.854, 1.108, p = .675; employment status, OR = 0.753; 95% CI 0.451, 1.259, p = .280;
healthcare access, OR = 2.448; 95% CI, 1.104, 5.429, p = .028; education status, OR =
0.746; 95% CI 0.578, 0.963, p = .024; sex, OR = 0.912; 95% CI 0.595, 1.398, p = .673,
and 5-year age groups, OR = 0.717; CI 95% 0.610, 0.843, p <.001. From the analyses,
race, language, income, employment status, and sex were not significant predictors of
CRC screening since their respective p >.05. On the other hand, healthcare access,
education status, and 5-year groups, all had p < .05, indicating that they were significant
predictors of CRC screening among participants in the study.
Table 83
Variables in the Equation
B S.E. Wald df Sig. Exp(B)
95% C.I. for
EXP(B)
Lower Upper
Step 1a Race of participants -.013 .046 .087 1 .768 .987 .902 1.079
Language of
participants (1)
.347 .458 .574 1 .449 1.415 .576 3.476
Income of
participants
-.028 .066 .175 1 .675 .973 .854 1.108
Employment status -.283 .262 1.168 1 .280 .753 .451 1.259
Healthcare access .895 .406 4.857 1 .028 2.448 1.104 5.429
Education status -.293 .130 5.070 1 .024 .746 .578 .963
Sex of participants -.092 .218 .178 1 .673 .912 .595 1.398
5-year age groups -.332 .083 16.163 1 <.001 .717 .610 .843
Constant 2.709 1.219 4.939 1 .026 15.022
Note: a. Variable(s) entered on Step 1: race, language, income, employment status, healthcare
access, education status, sex, 5-year age groups
154
Interaction of the Independent Variable
In accordance with the 2021 USPSTF CRC screening guidelines, I considered
those who had a FOBT within the past year, flexible sigmoidoscopy within the past five
years, and colonoscopy within the past ten years for the analysis, leaving other forms of
CRC screening out of the study (Davidson et al., 2021).
Tables 84 - 88 showed the parameter estimates, which described the coefficients
of the model and adjusted ORs that explained various forms of CRC screenings based on
various IVs. I conducted multivariate logistic regressions to evaluate the interaction effect
of the variables based on p-values. For there to be a significant interaction effect, p-value
must be <.05 and no significant interaction effect would show a p > .05. I found an
interaction effect between CRC screening and the following IVs: healthcare access,
education status, and 5-year age groups. The adjusted OR for healthcare access = 0.400;
95% CI 0.180, 0.889, p = .025; education status = 1.358; 95% CI 1.015, 1.755, p = .019,
and 5-year age groups = 1.390; 95% CI 1.182, 1.635, p <.001 (Table 84). There were no
interaction effects between CRC screening and language spoken by participants, annual
household incomes, employment status, sex, and race of participants. For colonoscopy
and IVs (Table 85), I found interactive effects with annual household incomes,
employment status, sex, 5-year age groups. The adjusted OR for annual household =
1.238; 95% CI 1.095, 1.400, p <.001; employment status = 1.713; 95% CI 1.027, 2.857, p
= .039; and 5-year age groups = 1.449; 95% CI 1.256, 1.672, p <.001. There was no
interaction with healthcare access, education status, race, and language spoken by
participants (Table 85). For participants who had colonoscopy within the past 10 years
155
(Table 86), no interaction effect was observed in annual household incomes,
employment, healthcare access, education status, sex, race, and language. For participants
who had sigmoidoscopy within the past 5 years (Table 87), only annual household
income had a positive interaction effect with an adjusted OR = 0.767; 95% CI 0.621,
0.949, p = .014. In these participants, employment status, education status, healthcare
access, sex, race, and language showed no significant interaction effect. With FOBT
within the past year (Table 88), only annual household income showed a positive
interaction effect with adjusted OR = 0.868; 95% CI 0.758, 0.993, p = .040, while
employment status, education status, healthcare access, sex, race, and language all
showing no significant interaction effect (Table 88).
Table 84
Parameter Estimates of Variables on Colorectal Cancer: Yes - Screened for Colorectal
Status
B Std. Error Wald df Sig. Exp(B)
95% Confidence Interval for
Exp(B)
Lower Bound Upper Bound
Intercept -2.702 1.137 5.651 1 .017
Income of
participants .018 .067 .074 1 .786 1.018 .893 1.161
Employment
status .284 .262 1.177 1 .278 1.329 .795 2.222
Healthcare
access -.917 .408 5.050 1 .025 .400 .180 .889
Education
status .306 .131 5.466 1 .019 1.358 1.051 1.755
5-year age
groups .329 .083 15.860 1 <.001 1.390 1.182 1.635
Sex of
participants=1 -.103 .218 .224 1 .636 .902 .588 1.383
Sex of
participants=2 0b . . 0 . . . .
Race of
participants=1 -.080 .321 .062 1 .803 .923 .492 1.732
Race of
participants=2 -.454 .455 .998 1 .318 .635 .261 1.548
156
Race of
participants=8 0b . . 0 . . . .
Language of
participants=1 .382 .461 .686 1 .407 1.465 .593 3.617
Table 85
Parameter Estimates of Variables on Colorectal Cancer: Yes - Had a Colonoscopy
B Std. Error Wald df Sig. Exp(B)
95% Confidence Interval for
Exp(B)
Lower Bound Upper Bound
Intercept -4.720 1.076 19.228 1 <.001
Income of
participants .213 .063 11.583 1 <.001 1.238 1.095 1.400
Employment
status .538 .261 4.257 1 .039 1.713 1.027 2.857
Healthcare
access -.710 .407 3.046 1 .081 .492 .222 1.091
Education
status .174 .121 2.073 1 .150 1.190 .939 1.509
5-year age
groups .371 .073 25.731 1 <.001 1.449 1.256 1.672
Sex of
participants=1 -.468 .209 5.018 1 .025 .626 .416 .943
Sex of
participants=2 0b . . 0 . . . .
Race of
participants=1 .255 .295 .750 1 .387 1.291 .724 2.302
Race of
participants=2 -.031 .423 .005 1 .942 .969 .423 2.223
Race of
participants=8 0b . . 0 . . . .
Language of
participants=1 .418 .422 .979 1 .322 1.519 .664 3.477
157
Table 86
Parameter Estimates of Variables on Colorectal Cancer: Yes - Colonoscopy Within Past
10 Years
B Std. Error Wald df Sig. Exp(B)
95% Confidence Interval for
Exp(B)
Lower Bound Upper Bound
Intercept -3.463 1.073 10.416 1 .001
Income of
participants .099 .060 2.745 1 .098 1.104 .982 1.241
Employment
status .241 .241 .994 1 .319 1.272 .793 2.041
Healthcare
access -.717 .413 3.014 1 .083 .488 .217 1.097
Education
status .222 .120 3.430 1 .064 1.248 .987 1.578
5-year age
groups .315 .076 17.207 1 <.001 1.370 1.181 1.590
Sex of
participants=1 -.204 .199 1.046 1 .306 .816 .552 1.205
Sex of
participants=2 0b . . 0 . . . .
Race of
participants=1 .001 .288 .000 1 .998 1.001 .569 1.760
Race of
participants=2 -.385 .412 .873 1 .350 .680 .304 1.526
Race of
participants=8 0b . . 0 . . . .
Language of
participants=1 .357 .437 .666 1 .415 1.429 .606 3.365
158
Table 87
Parameter Estimates of Variables on Colorectal Cancer: Yes - Sigmoidoscopy Within
Past 5 Years
B Std. Error Wald df Sig. Exp(B)
95% Confidence Interval for
Exp(B)
Lower Bound Upper Bound
Intercept -3.737 2.245 2.770 1 .096
Income of
participants -.265 .108 5.981 1 .014 .767 .621 .949
Employment
status -.214 .542 .156 1 .693 .807 .279 2.336
Healthcare
access -.105 .823 .016 1 .898 .900 .179 4.512
Education
status .119 .228 .272 1 .602 1.126 .720 1.761
5-year age
groups .284 .163 3.036 1 .081 1.329 .965 1.830
Sex of
participants=1 -.134 .414 .105 1 .746 .875 .389 1.969
Sex of
participants=2 0b . . 0 . . . .
Race of
participants=1 -.054 .608 .008 1 .929 .947 .287 3.120
Race of
participants=2 .665 .740 .809 1 .368 1.945 .456 8.292
Race of
participants=8 -.286 .767 .138 1 .710 .752 .167 3.382
Language of
participants=1 -.265 .108 5.981 1 .014 .767 .621 .949
159
Table 88
Parameter Estimates of Variables on Colorectal Cancer: Yes – Blood Stool Test Within
the Past Year
B Std. Error Wald df Sig. Exp(B)
95% Confidence Interval for
Exp(B)
Lower Bound Upper Bound
Intercept -2.566 1.378 3.466 1 .063
Income of
participants -.142 .069 4.225 1 .040 .868 .758 .993
Employment
status .318 .319 .990 1 .320 1.374 .735 2.568
Healthcare
access -.616 .588 1.100 1 .294 .540 .171 1.708
Education
status -.104 .145 .515 1 .473 .901 .678 1.198
5-year age
groups .168 .095 3.099 1 .078 1.183 .981 1.425
Sex of
participants=1 .459 .247 3.452 1 .063 1.583 .975 2.569
Sex of
participants=2 0b . . 0 . . . .
Race of
participants=1 .031 .361 .007 1 .932 1.031 .509 2.092
Race of
participants=2 .392 .487 .647 1 .421 1.479 .570 3.842
Race of
participants=8 .419 .548 .585 1 .444 1.520 .520 4.446
Language of
participants=1 -.142 .069 4.225 1 .040 .868 .758 .993
Note: a. The reference category is: No, not screened for colorectal cancer; b. This parameter is
set to the 0 because it is redundant.
Table 89 showed CRC screening based on sex, race, and language of participants.
Out of 766 sample size, the number of valid participants indicated 598 with 168 missing.
Among the valid participants, 440 or 73.6% received CRC screening. There were fewer
males than females with 260 or 43.5% representing males and 338 or 56.5% representing
females. Because all White, non-Hispanic, all Black, non-Hispanic and some Hispanic
participants chose English as their medium of communication, individual participants
who identified themselves with the English was 544 or 91.0%, while only 54 or 9.0%
identified themselves as Spanish-speaking participants in the study.
160
Table 89
Case Processing Summary
N
Marginal
Percentage
Colorectal cancer screening
status
Yes, screened for colorectal
cancer
440 73.6%
No, not screened for
colorectal cancer
158 26.4%
Sex of participants Male 260 43.5%
Female 338 56.5%
Race of participants White only, non-Hispanic 408 68.2%
Black only, non-Hispanic 56 9.4%
Hispanic 134 22.4%
Language of participants English 544 91.0%
Spanish 54 9.0%
Valid 598 100.0%
Missing 168
Total 766
Subpopulation 46 a
Note: a. The dependent variable has only one value observed in 8 (17.4%) subpopulations.
The observed and predicted frequencies indicate varied percentages for the CRC
screening participants. White only, non-Hispanic male and female and Black only, non-
Hispanic male and female participants communicate in English. For the Hispanic
participants, some Hispanic males and females speak English, while some other Hispanic
males and females speak Spanish only. From Table 90, for White only, non-Hispanic
males and females, there were 146 or 78.1% and 167 or 75% of participants who received
CRC screening, respectively. Among the Black only, non-Hispanic males and females,
12 or 60.0% and 28 or 77.8% received CRC screening in that order. Hispanics who speak
English, 22 or 66.7% of males received CRC screening while 35 or 74.5% of females
also had CRC screening. For Hispanic male and female participants who speak Spanish,
9 or 45.0% of males had CRC screening and 21 or 61.8% of females had CRC screening.
161
These variations in the percentages indicated that individuals who speak Spanish had the
lowest rates of CRC screening (Table 90).
Considering colonoscopy (Table 90), there were stark variations in the results.
White only, non-Hispanic, 167 or 76.6% of males and 220 or 79.4% of females; Black
only, non-Hispanic, 15 or 62.5% of males and 32 or 71.1% of females; English-speaking
Hispanic, 21 or 56.8% of males and 37 or 68.5% of females; and Spanish-speaking
Hispanic, 6 or 27.3% of males and 26 or 60.5% of females received colonoscopy. For
colonoscopy within the past 10 years (Table 90), White only, non-Hispanic, 128 or
68.8% of males and 146 or 66.1% of females; Black only, non-Hispanic, 11 or 52.4% of
males and 23 or 62.2% of females; English-speaking Hispanic, 18 or 51.4% of males and
30 or 63.8% of females; and Spanish-speaking Hispanic, 5 or 25.0% of males and 20 or
57.1% of females received colonoscopy. For sigmoidoscopy within the past 5 years
(Table 90), White only, non-Hispanic, 7 or 3.9% of males and 9 or 4.1% of females;
Black only, non-Hispanic, 1 or 5.0% of males and 4 or 10.5% of females; English-
speaking Hispanic, 1 or 3.0% of males and 3 or 6.5% of females; and Spanish-speaking
Hispanic, 3 or 15.0% of males and 3 or 9.7% of females received sigmoidoscopy. For
FOBT in the past year (Table 90), White only, non-Hispanic, 34 or 18.6% of males and
34 or 15.0% of females; Black only, non-Hispanic, 6 or 28.6% of males and 9 or 23.7%
of females; English-speaking Hispanic, 6 or 18.2% of males and 8 or 17.0% of females;
and Spanish-speaking Hispanic, 6 or 28.6% of males and 2 or 5.7% of females received
blood stool test in the past year (Table 90).
162
Table 90
Observations and Predicted Frequencies of Males and Females Among the Racial
Groups in the Study
Race of
participants
Language of
participants
Sex of
participants
Have you ever had
colorectal cancer
screening
Frequency Percentage
Observed Predicted Pearson Residual Observed Predicted
White only,
non-Hispanic English
Male Yes 146 140.451 .938 78.1% 75.1%
No 41 46.549 -.938 21.9% 24.9%
Female Yes 167 170.505 -.562 75.6% 77.2%
No 54 50.495 .562 24.4% 22.8%
Black only,
non-Hispanic English
Male Yes 12 14.869 -1.469 60.0% 74.3%
No 8 5.131 1.469 40.0% 25.7%
Female Yes 28 27.515 .190 77.8% 76.4%
No 8 8.485 -.190 22.2% 23.6%
Hispanic
English
Male Yes 22 22.919 -.347 66.7% 69.5%
No 11 10.081 .347 33.3% 30.5%
Female Yes 35 33.740 .408 74.5% 71.8%
No 12 13.260 -.408 25.5% 28.2%
Spanish
Male Yes 9 10.761 -.790 45.0% 53.8%
No 11 9.239 .790 55.0% 46.2%
Female Yes 21 19.239 .609 61.8% 56.6%
No 13 14.761 -.609 38.2% 43.4%
White only,
non-Hispanic
English
Male
Ever had a colonoscopy?
Yes
167
160.776
.958
76.6%
73.8%
No 51 57.224 -.958 23.4% 26.2%
Female Yes 220 221.549 -.233 79.4% 80.0%
No 57 55.451 .233 20.6% 20.0%
Black only,
non-Hispanic
English Male Yes 15 17.217 -1.005 62.5% 71.7%
No 9 6.783 1.005 37.5% 28.3%
Female Yes 32 35.237 -1.171 71.1% 78.3%
No 13 9.763 1.171 28.9% 21.7%
Hispanic
English Male Yes 21 21.452 -.150 56.8% 58.0%
No 16 15.548 .150 43.2% 42.0%
Female Yes 37 35.769 .354 68.5% 66.2%
No 17 18.231 -.354 31.5% 33.8%
Spanish Male Yes 6 9.555 -1.529 27.3% 43.4%
No 16 12.445 1.529 72.7% 56.6%
Female Yes 26 22.445 1.085 60.5% 52.2%
No 17 20.555 -1.085 39.5% 47.8%
(table continues)
163
Table 90
Observations and Predicted Frequencies of Males and Females Among the Racial
Groups in the Study cont.
Race of
participants
Language of
participants
Sex of
participants
Sigmoidoscopy within the past 5 years
Frequency Percentage
Observed Predicted
Pearson
Residual Observed Predicted
White only,
non-Hispanic
English
Male
Yes
7
7.367
-.138
3.9%
4.1%
No 172 171.633 .138 96.1% 95.9%
Female Yes 9 10.577 -.497 4.1% 4.8%
No 210 208.423 .497 95.9% 95.2%
Black only,
non-Hispanic
English Male Yes 1 .846 .171 5.0% 4.2%
No 19 19.154 -.171 95.0% 95.8%
Female Yes 4 1.886 1.579 10.5% 5.0%
No 34 36.114 -1.579 89.5% 95.0%
Hispanic
English Male Yes 1 1.642 -.514 3.0% 5.0%
No 32 31.358 .514 97.0% 95.0%
Female Yes 3 2.682 .200 6.5% 5.8%
No 43 43.318 -.200 93.5% 94.2%
Spanish Male Yes 3 2.145 .618 15.0% 10.7%
No 17 17.855 -.618 85.0% 89.3%
Female Yes 3 3.855 -.465 9.7% 12.4%
No 28 27.145 .465 90.3% 87.6%
White only,
non-Hispanic
English
Male
Blood stool test within the past year
Yes
34
37.137
-.577
18.6%
20.3%
No 149 145.863 .577 81.4% 79.7%
Female Yes 34 34.966 -.178 15.0% 15.5%
No 192 191.034 .178 85.0% 84.5%
Black only,
non-Hispanic
English Male Yes 6 4.291 .925 28.6% 20.4%
No 15 16.709 -.925 71.4% 79.6%
Female Yes 9 5.923 1.376 23.7% 15.6%
No 29 32.077 -1.376 76.3% 84.4%
Hispanic
English Male Yes 6 7.029 -.437 18.2% 21.3%
No 27 25.971 .437 81.8% 78.7%
Female Yes 8 7.655 .136 17.0% 16.3%
No 39 39.345 -.136 83.0% 83.7%
Spanish Male Yes 6 3.543 1.432 28.6% 16.9%
No 15 17.457 -1.432 71.4% 83.1%
Female Yes 2 4.457 -1.246 5.7% 12.7%
No 33 30.543 1.246 94.3% 87.3%
(table continues)
164
Table 90
Observations and Predicted Frequencies of Males and Females Among the Racial
Groups in the Study cont.
White only,
non-Hispanic
English
Male
Colonoscopy within the past 10 years
Yes
128
120.459
1.157
68.8%
64.8%
No 58 65.541 -1.157 31.2% 35.2%
Female Yes 146 150.082 -.588 66.1% 67.9%
No 75 70.918 .588 33.9% 32.1%
Black only,
non-Hispanic
English Male Yes 11 13.343 -1.062 52.4% 63.5%
No 10 7.657 1.062 47.6% 36.5%
Female Yes 23 24.693 -.591 62.2% 66.7%
No 14 12.307 .591 37.8% 33.3%
Hispanic
English Male Yes 18 19.551 -.528 51.4% 55.9%
No 17 15.449 .528 48.6% 44.1%
Female Yes 30 27.872 .632 63.8% 59.3%
No 17 19.128 -.632 36.2% 40.7%
Spanish Male Yes 5 8.647 -1.646 25.0% 43.2%
No 15 11.353 1.646 75.0% 56.8%
Female Yes 20 16.353 1.236 57.1% 46.7%
No 15 18.647 -1.236 42.9% 53.3%
Summary
Statistical analysis of the data led to the rejection of the null hypothesis (H01)
which stated that there is no significant relationship between LEP and CRC screening
rates among Hispanic and non-Hispanic residents in Texas, when potential confounding
variables including age, income, occupation, health care access, and educational levels of
participants are controlled. The results showed significant associations between LEP and
CRC screening uptake. However, the analysis of the data did not indicate that the null
hypothesis (H02) could not be rejected as gender did not independently show a significant
relationship between males and females with CRC screening of Hispanic and non-
Hispanic residents in Texas, when potential confounding variables including age, income,
occupation, health care access, and educational levels of participants are controlled.
165
Nonetheless, Hispanic males are likely to be exposed to increased risk for CRC due to
their low screening rates among all participants in the study.
To understand any discrepancies in this study, it was important to compare its
findings with extant knowledge with possible explanation. The next chapter would focus
on discussion of the results and implications of the findings. It would also present
recommendations and discuss the social change associated with the study.
166
Chapter 5: Discussion, Conclusions, and Recommendations
Introduction
Numerous studies have focused on the impact of language barrier as a
fundamental cause of low screening of CRC among non-English-speaking ethnic groups
in the United States, including Spanish-speaking Americans (DuHamel et al., 2020).
However, no study has been done to confirm this finding in the state of Texas where
nearly 11.3 million or 40.2% of the population were Hispanic in 2020 (United States
Census Bureau, 2021). I conducted this study using descriptive statistics, such as
frequencies and percentages, to assess the association between each of the two main IVs,
LEP and sex, and the DV, which was the CRC screening. To understand the preferences
of CRC screening methods among participants, I ran statistical analyses on several types
of screening for CRC (ever taken colonoscopy, colonoscopy in the past 10 years,
sigmoidoscopy in the past 5 years, and blood stool test or fecal occult blood test (FOBT)
within the past year), using crosstabulation descriptive analysis, bivariate, and
multivariate models. These models demonstrated the associations between LEP and sex
on CRC screening. Whereas English language preference correlated with increased
screening uptake in bivariate comparisons, in multivariate logistic regression models,
LEP did not independently predict CRC screening. I assessed the crude relationship
between each covariate and the binary outcome by using univariate logistic regression
models, followed by a series of multivariable logistic regression models. Such models
include various sets of independent variables, namely demographic variables (age, sex,
167
and language) and socioeconomic and health-related variables (household income,
employment status, education attainment, and health care access).
Sample Description Summary
The study sample of 766 noninstitutionalized civilian residents of Texas was
chosen randomly from the 2020 Texas BRFSS. The sample population was made up of
non-Hispanic White, non-Hispanic Black, and Hispanic adults aged from 50-79 years old.
Out of the 766 participants, 525 or 68.5% described themselves as non-Hispanic White,
77 or 10.1% as non-Hispanic Black, and 164 or 21.4% identified themselves as Hispanic.
However, not all the self-described Hispanic sample population identified themselves
with the Spanish language in the BRFSS sample when they responded to the
questionnaire on CRC screening. Only 70 participants or 9.1% identified themselves as
Spanish-speaking individuals (or limited English proficient participants), whereas 696
individuals or 90.9% identified themselves as English-speaking (English proficient)
participants in the survey. With the CRC screening, 544 participants in the study who
underwent CRC screening did the BRFSS survey in English, while 54 individuals who
did the CRC screening were Spanish speakers.
Also, in terms of age, 696 participants considered themselves as English speakers,
whereas 70 participants identified themselves as Spanish speakers. The study sample had
450 female, or 58.7%, and 316 male, or 41.3% participants. Out of the total sample of
766 participants, 604 reported earning income while 162 did not indicate their incomes;
754 indicated that they were employed with 12 showing no employment status; 765 had
168
healthcare access with one participant not providing a healthcare access status; and 763
showed education status while 3 did not indicate their education status.
Interpretation of Findings
Participants Grouped by Age
Analysis from the study participants categorized into 5-year age groups from 50
to 79 years of age indicated that 598 out of 766 had CRC screening. Within the groups,
CRC screening rates increased as the group age advanced from 50-54, 55-59, 60-64, 65-
69, 70-74, and 75-79; the 50-54 group reported 55.3% CRC screening, and 75-79
reported 90.9% screened for CRC, with the rest in between the percent ranges. Similarly,
the unscreened rates decreased as the group age advanced. The Chi-square tests indicated
that Pearson Chi-square χ(1) = 32.754, p < .001, showing that there was a statistically
significant association between different age groups. Considering symmetric measures of
Phi and Cramer’s V of 0.234, each at p < .001, indicated a relatively stronger association
between age and CRC screening. The bivariate logistic analysis of the 5-year age groups
showed a Wald value of 16.163 and a p < .001. From the multivariate analysis, the
adjusted OR for 5-year age groups = 1.449; 95% CI 1.256, 1.672, p < .001.
The results from this statewide representative sample were consistent with the
literature that CRC screening increases with age from 50 years age (US Preventive
Services Task Force, Davidson et al., 2021); however, recent studies have revealed
increased incidence rates of CRC in individuals younger than 50 years of age, even if
there were no genetic connections associated with their family lines, despite the vast
majority of CRC diagnoses still occurring in individuals aged 65 to 74 years (Rawla et
169
al., 2019; US Preventive Services Task Force, Davidson et al., 2021). For example, the
US Preventive Services Task Force (2021) noted in a publication that 10.5% of new CRC
cases occur in individuals younger than 50 years old, citing particularly adenocarcinoma
with a significant incidence rate of nearly 15% from 2000-2002 to 2014-2016 in persons
aged 40 to 49 years.
Sex
Analysis of the results showed that female participants had a relatively higher
CRC screening rate, with 74.3% vs. 72.9% males screened for CRC, p >.5, indicating that
statistically, the difference was not significant. Similarly, the Chi-square tests showed
that Pearson Chi-square χ(1) = 0.186, p = .666. Thus, there was no statistically significant
association between males and females. Also, symmetric measures of Phi and Cramer’s
V were -0.018 and 0.018, respectively, each at p = 0.666, with the Phi and Cramer’s V
values being less than 0.10, which indicated no association between gender differences
and CRC screening. Bivariate analysis on sex indicated OR = 0.912; 95% CI 0.595,
1.398, p = .673. For both Hispanic men and women, a personal history of cancer survival
has been found to be correlated to increased CRC screening (Shah et al., 2022). For
women, acculturation, which involves learning new language, adopting new customs, and
changing religious beliefs was seen as embraced by Hispanic females. Such individuals
prefer learning English language, leading them to become nearly assimilated into the
American society, which improves health screening rates. Also, people who have been in
the American society for at least 10 years are associated with higher odds of receiving
CRC screening (Castañeda et al., 2019).
170
Race
This study analyzed three major races in Texas, including White only, non-
Hispanic; Black only, non-Hispanic; and Hispanic residents. The results showed that
within the White only, non-Hispanic, 76.7% underwent CRC screening; Black only, non-
Hispanic, 71.4% received screening, while 64.9% of Hispanic also received screening,
with 23.2%, 28.6%, and 35.1% of White only, non-Hispanic; Black only, non-Hispanic;
and Hispanic participants not receiving CRC screening, in that order. The Chi-square
tests indicated that Pearson Chi-square χ(1) = 7.360, p < .025, showing that there was a
statistically significant association among White only, non-Hispanic; Black only, non-
Hispanic; and Hispanic samples of participants. Although the results of bivariate analysis
of race indicated OR = 0.987; 95% CI 0.902, 1.079, p = .768, symmetric measures of Phi
and Cramer’s V were the same, 0.111, each at p = 0.025. The Phi and Cramer’s V values
being more than 0.10 indicated there were differential influences among White only, non-
Hispanic; Black only, non-Hispanic; and Hispanic samples of participants and CRC
screening.
The results support a consistent disparity in the literature that while there have
been increasing CRC screening rates among Hispanics in recent years, there remains
persistent gaps in comparison with White only, non-Hispanic groups (Wittich et al.,
2019). In another study, May et al. (2020) found that CRC screening was 17% lower in
Hispanic population compared to Whites in the United States. In the same study, the
authors showed that CRC screening rates among Blacks were 4% lower than Whites (p <
.001).
171
Language
Even though disparities exist in CRC screening among White only, non-Hispanic
and Hispanic populations, this disparity widened more significantly when language was
considered. The results revealed that 44.4% of Hispanics who speak Spanish did not
receive CRC screening, while 35.1% of the general Hispanic participants did not receive
CRC screening, indicating that language limitation contributes to low rates of CRC
screening. For the individuals who speak English, CRC screening rate was 75.4%
compared to 55.6% for Spanish speakers. The results showed that the Pearson Chi-square
χ(1) = 9.918, p < .002, indicating that there was a statistically significant association
between English- and Spanish-speaking respondents. Similarly, the symmetric measures
of Phi and Cramer’s V were the same, 0.129, each at p = 0.002, showing there were
differential influences among English- and Spanish-speaking respondents for screening.
Although English language preference correlated with increased screening uptake in
bivariate comparisons, language did not seem to indicate a remarkable predictor of CRC
screening uptake in multivariate logistic regression models in the study (OR = 1.415, p =
.449). This was likely resulted from inter-relations with other notable predictors like
education and income, which have the propensity to promote increased awareness of
screening, a necessary drive to achieving greater screening compliance (Juon et al.,
2018).
Income (Annual Household)
Analysis of the results with respect to income indicators showed that among
participants with an income of $75,000 or higher, 75.6% had screening, individuals
172
whose incomes ranged from $50,000 to under $75,000, 80% had screening, while 72.9%
compared to 68.1% for those whose incomes ($0 — < $20,000 vs. $20,000 — < $50,000)
received screening, p < .001. For individuals at least 65 years of age, income status could
not predict a particular trend regarding health care access or CRC screening because
people who are 65 years or older have health care access through Medicare with most of
them having primary care physicians. Because CRC incidence is higher within their age
brackets, screening rates are often higher. Thus, income may not be an influential factor
affecting screening among individuals who are eligible for Medicare (Shapiro et al.,
2021). The multivariate analyses showed that the adjusted OR for annual household =
1.238; 95% CI 1.095, 1.400, p < .001. On the other hand, those with income of $75,000
or higher are likely to be those in the 50-54 and 55-59 age groups who are likely actively
employed and relatively stronger than their older counterparts. Analysis of the study
indicated that those aged 50-59 had lower CRC screening rates.
Employment Status
Considering the employment status and CRC screening of participants in the
study, 77.6% of individuals who were unemployed compared with 66.4% of those
employed received CRC screening, p < .05. The discrepancy arose from the age of the
study participants (50-79 years of age) with the majority of individuals aged 65 years or
older most likely retired from the active workforce. In the United States, people who are
65 years or older have Medicare, a federal health insurance program that guarantees
health care access, thereby addressing obstacles of uninsurance for individuals who want
173
to undertake CRC screening (Zhao et al., 2018). In the multivariate analyses, the adjusted
OR for employment status = 1.713; 95% CI 1.027, 2.857, p = .039.
Healthcare Access
The results showed that 76.5% of those with health insurance compared to 39.6%
of those without it screened for CRC. For individuals who did not screen, comparing
people with access to health insurance to those without health insurance access (23.5%
vs. 60.4%), the uninsured individuals were nearly threefold those insured, indicating that
health care access contributes significantly to screening, p < .001. This shows that lack of
health care access is a major obstacle to screening. Health analysts blame Texas as the
state with the least health care access because the state fails to provide Medicaid to most
eligible residents even though the Affordable Care Act (ACA) has been successful in
expanding Medicaid to all eligible individuals across the United States (Allen et., 2017).
Texas is one of the few states where there is a Medicaid gap, defined as people whose
annual household income is less than $35,000 per family with only one person, or
$47,000 with two-person without any health insurance (Zhang & Wu, 2021). The
multivariate analyses showed that the adjusted OR for healthcare access = 0.400; 95% CI
0.180, 0.889, p = .025.
Education Status
The results indicated that individuals with higher educational attainments, such as
college graduates, had increased screening rates compared to those whose educational
attainments were lower than high school. For individuals with college degree or higher,
78.3% received screening, 74.4% of individuals who had some college education
174
received screening, 67.8% of participants with high school/GED had screening, while
only 64.1% of individuals with less than high school education had screening of CRC, all
these comparisons had a p < .05. Studies have shown that in the general United States
population, age, health care access, nativity, level of education, language, socioeconomic
elements, overall health, and provider practices are seen as some influential factors
associated with increased CRC screening (Mayhand et al., 2021). In the bivariate
analyses, the adjusted OR for education status = 1.358; 95% CI 1.015, 1.755, p = .019.
Colorectal Screening Types
Colonoscopy
The crosstabulation analysis showed that colonoscopy screening was more
prominent among groups 60-64, 65-69, 70-74, and 75-79 compared to age groups 50-54
and 55-59. For individuals in age groups 70-74, 24%, 75-79, 21%, 65-69, 20.8%, and 60-
64, 15.6%, had colonoscopy compared to groups 50-54 and 55-59 where only 8.6% and
9.9% had colonoscopy. The Chi-square tests indicated that Pearson Chi-square χ(1) =
55.326, p < .001, showing that there was a statistically significant association among
five-year age groups 50-54, 55-59, 60-64, 65-69, 70-74, and 75-79 participants in the
study. Symmetric measures of Phi and Cramer’s V were the same, 0.277, each at p <
.001. The Phi and Cramer’s V values being more than 0.10 indicated there were
differential influences among five-year age groups and colonoscopy screening status. The
results also showed that although the USPSTF recommends CRC screening from age 50-
75 years, people older than 75 participate in CRC screening to some extent, despite the
175
benefits of CRC screening diminishes as individuals get much older (US Preventive
Services Task Force, Davidson et al., 2021).
Studies demonstrate that because of shorter life expectancy and more frequent
comorbidities commonly correlated with old age, certain adverse reactions associated
with colonoscopy, such as bleeding and perforation of the colon, which exacerbate
morbidity factors in elderly people (Kim et al., 2019), the USPSTF recommends
colonoscopy up to 75 years of age. Healthcare providers may encourage patients older
than 75 years old to undertake colonoscopy only after careful consideration of potential
benefits, risks, and patient preferences due to other pathophysiological factors that may
be clinically visible (Lin, 2014). The adjusted OR from covariates, such as household
annual income 1.238; 95% CI 1.095, 1.400, p < .001; employment status = 1.713; 95%
CI 1.027, 2.857, p = .039; and 5-year age groups = 1.449; 95% CI 1.256, 1.672, p < .001,
indicated interactive effects in the multivariate analysis for colonoscopy.
However, multivariate logistic regression analysis showed that there was no
interaction with healthcare access, education status, race, and language spoken by
participants, p > .05. Researchers indicate that SES, such as healthcare access and
education status, represent one of the strongest and most consistent predictors of health
(Carethers & Doubeni, 2020). Despite the profound significance of SES on predictors of
health, other factors may also alter their expected impacts. For example, effective
preventive healthcare may be positively influenced by health literacy, which has a
fundamental impact on how individuals consider the importance of health (Coughlin et
al. 2020). Thus, low health literacy could potentially limit the positive impact of health
176
care access and education status, although education may somehow (but not absolutely)
be associated with improved health literacy (Bayati et al., 2018; Wigfall & Tanner,
2018).
In this study, the descriptive analysis from crosstabulation showed that the choice
of colonoscopy among racial groups in the study varied significantly. Among White only,
non-Hispanic, 73.9% had colonoscopy compared with 9.0% of Black only, non-Hispanic
and 17.2% of Hispanics who had colonoscopy. The Chi-square tests revealed that
Pearson Chi-square χ(1) = 25.973, p < .001, showing that there was a statistically
significant association among White, non-Hispanic, Black, non-Hispanic, and Hispanic
participants in the study. Symmetric measures of Phi and Cramer’s V were the same,
0.190, each at p < .001 with Phi and Cramer’s V values being more than 0.10, indicating
there were differential influences among White, non-Hispanic, Black, non-Hispanic, and
Hispanic and colonoscopy screening status. When language used for the survey was
considered, the analysis of the results showed that 93.9% of 524 of study participants
took the survey in English and only 6.1% identified themselves as Spanish-speaking
participants with Pearson Chi-square χ(1) = 19.996, p < .001, showing that there was a
statistically significant association between English-speaking and Spanish-speaking
participants in the study. Phi and Cramer’s V symmetric measures were the same, 0.167,
each at p < .001, with the Phi and Cramer’s V values being more than 0.10, an indication
that there were differential influences between English-speaking and Spanish-speaking
study participants and colonoscopy screening status. The multivariate logistic regression
analysis, however, showed that like healthcare access and education status, there was no
177
interaction with race and language spoken by participants, p > .05. Thus, it was likely
other factors could have potentially influenced the interactions between race and
language and colonoscopy screening. For example, acculturation and improved health
literacy could reduce the adverse impact of race and language and decrease their
perceived differences in colonoscopy screening (Rogers et al., 2021).
Crosstabulation analysis showed a variation between males and female for
colonoscopy screening, more than 60% of females underwent colonoscopy screening
compared to nearly 40% of males who received the same screening. However, the
Pearson Chi-square χ(1) = 2.917, p < .088, showing that there was no statistically
significant association between male and female participants in the study. Also,
symmetric measures of Phi and Cramer’s V were - 0.064 and 0.064, respectively, each at
p = .088, with the Phi and Cramer’s V values being less than 0.10 indicating there were
no differential influences between males and females and colonoscopy screening status.
This discrepancy based on previous studies may be due to other factors, such as
acculturation, which tends to minimize some cultural effects on preventive healthcare,
including colonoscopy screening (Ma et al., 2020; Savas et al., 2015).
The clinical summary released by the USPSTF (2021) recommended that
colonoscopy screening should be done once every 10 years. Analysis from
crosstabulation on 5-year age groups showed that 381 participants received colonoscopy
screening within the past 10 years, with the highest screening taking place among
individuals in 70-74 age group (29.4%), followed by individuals in 65-69 age group
(24.1%). Groups 50-54 (10.8%), 55-59 (13.1%), 60-64 (17.8%), and 75-79 (4.7%) had
178
rates that were less, with individuals in 75-79 age group having the least rate. The Chi-
square tests indicated that Pearson Chi-square χ(1) = 29.415, p < .001, showing that there
was a statistically significant association between male and female participants in the
study, while symmetric measures of Phi and Cramer’s V were the same, 0.221, each at p
< .001. The Phi and Cramer’s V values indicated there were differential influences
among five-year age groups and colonoscopy screening status. The lowest rate shown in
the age group 75-79 reflects recommendations by the USPSTF that benefits for
colonoscopy diminishes, but is also associated with increased risk, as individuals get
older (US Preventive Services Task Force, Davidson et al, 2021).
Analysis of the results of males and females who received colonoscopy screening
within the past 10 years showed that 42% of males and 57.5% of females had
colonoscopy screening. The p =.515, showing that there was no statistically significant
association between male and female participants, Pearson Chi-square χ(1) = 0.424 and
symmetric measures of Phi and Cramer’s V were -.027 and .027, which also indicates
that there were no differential influences between male and female study participants and
colonoscopy screening status within the past 10 years. Considering race, colonoscopy
uptake within the past 10 years shows that 71.9 % of White, non-Hispanic, 8.9% of Black
only, non-Hispanic, and 19.2% Hispanic participants received colonoscopy screening.
The Chi-square tests indicated that Pearson Chi-square χ(1) = 9.295, p = .010, showing
that there was statistically significant association the racial groups. Symmetric measures
of Phi and Cramer’s V were the same, .124, each at p = .010, an indication that there
were differential influences among White, non-Hispanic, Black, non-Hispanic, and
179
Hispanic study participants and colonoscopy screening status. While the results showed
that Black only, non-Hispanic had the least rate of colonoscopy screening among the
races in the study, another study conducted in New York City indicated that Black only
non-Hispanic had the highest colonoscopy rate of 72.2%, Latinos 71.1%, Whites 67.2%,
and Asians, 60.9% (Brown et al., 2021). The discrepancy between the outcomes from
New York City and the state of Texas may be due to other factors, such as health care
access. In New York State, 27% of New York population was low-income (<200 FPL),
but 28% of New York population or 2 in 9 adults aged 19-64 years old was covered by
Medicaid/CHIP compared to 17% of Texas population or 1 in 14 adults aged 19-64 years
old was covered by Medicaid/CHIP where 33% of the population was low-income (<200
FPL), according to a study by the Kaiser Permanente in October 2022 (Kaiser Family
Foundation, 2022).
Assessing the colonoscopy rates between English-speaking and Spanish-speaking
participants in the study, the results showed that within the past 10 years, 93.4% English-
speaking and 6.6% of Spanish-speaking study subjects received colonoscopy screening.
The Chi-square tests showed that Pearson Chi-square χ(1) = 8.286, p = .004, indicating
that there was statistically significant association between English-speaking and Spanish-
speaking participants in the study, with symmetric measures of Phi and Cramer’s V were
the same, .117, each at p = .004, showing there were differential influences between
English-speaking and Spanish-speaking study participants and colonoscopy screening
status. Thus, within the study, while Hispanic participants had 19.2% of colonoscopy
screening within the past 10 years, individual Hispanics who responded to the survey
180
questions in Spanish, only 6.6% received colonoscopy screening within the same period.
This was an indication that LEP may have contributed to the low rate of colonoscopy
screening among the limited English proficient participants in the study.
Investigations into social, behavioral, and medical sciences show that there are
remarkable perceptual and behavioral variations associated with patients’ cultural and
ethnic backgrounds that could potentially impact cancer prevention behaviors, such as
delay in seeking preventive care, understanding the concerns of adverse impact about
etiology of disease, and expectations about treatment outcomes and prognosis (Diaz et
al., 2013; Hall et al., 2022; Yedjou et al., 2019). For the covariates in the study, such as
annual household incomes, employment, healthcare access, and education status, the
multivariate logistic regression showed no significance for colonoscopy within the past
10 years as the p > .05.
Sigmoidoscopy
Procedural preparations for both colonoscopy and sigmoidoscopy require patient
preparation to clear out the colon, and most sigmoidoscopy and colonoscopy preparations
involve large intake of cleansing solutions, such as polyethylene glycol (MiraLAX). In
addition, patients take laxatives, enemas, and possibly several days of a clear liquid diet
prior to the procedure. Despite their similarities, many people consider colonoscopy as
more painful than sigmoidoscopy, and they tend to choose the latter for screening. To
understand these options as a choice, I analyzed sigmoidoscopy screening alongside
colonoscopy and FOBT. In assessing sigmoidoscopy screening among 5-year age groups,
181
the procedure was found to be more prominent among age groups 60-64, 65-69 and 70-
74 compared to age groups 50-54, 55-59, and 74-79.
However, in comparison to colonoscopy, fewer participants in the study used
sigmoidoscopy screening. Groups 60-64 (12.9%), 65-69 (25.8%), 70-74 (35.5%) had
sigmoidoscopy compared to groups 50-54 (9.7%), 55-59 (6.5%), and 75-79 (9.7%) who
also had sigmoidoscopy. In all, only 31 participants had sigmoidoscopy while 555
respondents stated that they had not received sigmoidoscopy within the past 5 years vs
381 who had colonoscopy within the past 10 years, and 221 who did not have the
procedure within the past 10 years. The Chi-square tests for the sigmoidoscopy indicated
that Pearson Chi-square χ(1) = 7.196, p =.206, which showed no statistically significant
association among the age groups. Symmetric measures of Phi and Cramer’s V were the
same, 0.111, each at p = .206.
Considering gender, there were more women (61.3%) than men (38.7%) who took
sigmoidoscopy screening. In terms of race, sigmoidoscopy screening within the past 5
years was highest among White only, non-Hispanic (51.6%) and lowest among Black
only non-Hispanic (16.1%) with Hispanics (32.3%) fallen in between White only, non-
Hispanic and Black only, non-Hispanic. For English-speaking (80.6%) and Spanish-
speaking (19.4%) of 31 participants received sigmoidoscopy screening within the past 5
years. The trend was the same as in colonoscopy, with English-speaking individuals
receiving more sigmoidoscopy than Spanish-speaking individuals, most likely due to
LEP as a hindrance. The Chi-square tests indicated that Pearson Chi-square χ(1) = 4.674,
p = .031, showing that there was a statistically significant association between English-
182
speaking and Spanish-speaking participants in the study. In evaluating the results from
the multivariate logistic regression, only annual household income had a positive
interaction effect with an adjusted OR = 0.767; 95% CI 0.621, 0.949, p = .014. for
participants who had sigmoidoscopy within the past 5 years, while employment status,
education status, healthcare access, sex, race, and language showed no significant
interaction effect among participants.
Fecal Occult Blood Test
Because several choices are available as CRC screening procedures, I chose to
consider both invasive and noninvasive approaches to understand which options are
chosen often for screening among participants in the study. Study results showed that 105
respondents had taken FOBT screening for CRC within the past year. The age groups
with higher rates of participants included 50-54 (12.4%), 60-64 (17.1%), 65-69 (23.8%),
70-74 (38.1%) compared to age groups 55-59 (4.8%) and 75-79 (3.8%) whose FOBT
screening rates were much lower, with Chi-square tests indicating Pearson Chi-square
χ(1) = 16.933, p = .005, showing that there was a statistically significant association
among five-year age group participants in the study. Symmetric measures of Phi and
Cramer’s V were the same, 0.167, each at p = .005, and the Phi and Cramer’s V values
being more than 0.10, indicates differential influences among five-year age groups and
FOBT screening status. Out of the 105 participants who took FOBT screening for CRC
within the past year, 49.5% were males and 50.5% were females, demonstrating that
FOBT was nearly even between males and females. The Chi-square tests indicated that
Pearson Chi-square χ(1) = 2.408, p =.121, showing that there was no statistically
183
significant difference between male and female participants with respect to FOBT within
the past year. The FOBT rates among the racial groups varied with 64.8% White only,
non-Hispanic, 14.3% Black, non-Hispanic, and 21.0% Hispanic undertaken it within the
past year. When FOBT rates were assessed using the language respondents used for the
survey, 92.4% of the respondents were identified with English and 7.6% of respondents
were identified with Spanish, demonstrating that LEP was likely a reason for lower rates
of FOBT screening for CRC. The multivariate logistic analysis of FOBT within the past
year showed that annual household income had a positive interaction effect with adjusted
OR = 0.868; 95% CI 0.758, 0.993, p = .040. However, employment status, education
status, healthcare access, sex, race, and language all show no significant interaction
effect, p >.05.
Sensitivity, Compliance, and Access of CRC Screening Types
Researchers suggest that effective CRC screening emanates from the combination
of 3 critical factors, namely sensitivity, compliance, and access. Screening
methodologies, such as colonoscopy, sigmoidoscopy, and FOBT, vary in their sensitivity
or accuracy. While sensitivities for early-stage CRC are essentially equivalent between
colonoscopy and MT-sDNA, they are notably higher than with FOBT whose sensitivity
for detecting CRC is about 70%-75% the sensitivity of colonoscopy (Ahlquist, 2019).
Also, investigators argue that sensitivity is poor for precursor lesions, nearly 20% to 25%
for advanced adenomas and less than 5% for advanced sessile serrated polyps, thereby
making it difficult for tests like FOBT that inherently have low sensitivities (Chang et al.,
2017). Because of their low sensitivities, such tests are recommended to be performed
184
annually, a situation considered as a disadvantage by some patients due to increased
inconvenience of performing it annually, resulting in an increased possibility of low
patient compliance (Ahlquist, 2019).
Although test sensitivity is important, program logistics to maintain an effective
intervention needs to address challenges associated with compliance and access. Among
other things, individuals eligible to undertake CRC screening may be less compliant if
they have to do the test every year, find it difficult to follow instructions for the test, do
not have access to health care, and do not have a primary care physician (Barthold et al.,
2022). Other researchers also note that although stool collection may be offputting to
some patients and affects FOBT compliance, a significant number of surveys has
demonstrated that patients prefer noninvasive over invasive tests, which was seen in this
study where FOBT rates were higher than sigmoidoscopy (Ahlquist, 2019; Ferrari et al.,
2021). Thus, notwithstanding its outcome as being the gold standard for CRC screening,
because of its labor-intensiveness, invasiveness, and potential for injury among elderly
persons, colonoscopy is not universally preferred over other types of CRC screening,
such as FOBT, which is used by a significant number of individuals as an alternative
screening test despite its associated detection inaccuracies and the need to do it annually
(Tepus & Yau, 2020).
Gender Influences on CRC Screening
The multivariate logistic regression analysis that featured observed and predicted
frequencies for CRC screening using colonoscopy, sigmoidoscopy, and FOBT showed
varied percentages for CRC screening among English-speaking White only, non-Hispanic
185
male and female; Black only, non-Hispanic male and female; English-speaking Hispanic
male and female; and Spanish-speaking Hispanic male and female participants. The
Spanish-speaking Hispanic participants are considered as individuals who speak Spanish
only and have LEP. The results indicated that overall, English-speaking participants in
the state of Texas received significantly higher rates of CRC screening, with slight
variations from one race or gender to the other. However, both male and female LEP
individuals had low rates of CRC screening, particularly when rates were compared
males to males and females to females within the Hispanic English-speaking and
Hispanic LEP groups. Taking all forms of CRC screening together, the results indicated
White only, non-Hispanic males (78.1%) and females (75.0%); Black only, non-Hispanic
males (60.0%) and females (77.8); and Hispanics who did the survey in English, males
(66.7%) and females (74.5%), while LEP Spanish males (45.0%) and females (61.8%).
The results are consistent with previous studies where Hispanic men with LEP were
found with lower CRC screening, suggesting that limited-English proficient Hispanic
males are at the greatest risk of not being screened for CRC compared to non-Hispanic
Whites and Blacks, English proficient Hispanics, and even Hispanic females with LEP
(Diaz et al., 2013; Hall et al., 2022).
For colonoscopy, females in each racial group had higher screening rates than
their male counterparts, with White, non-Hispanic females (79.4%) vs males (76.6%);
non-Hispanic Black females (71.1%) vs males (62.5%); English proficient Hispanic
females (68.5%) vs males (56.8%); and limited-English proficient Hispanic females
(60.5%) vs males (27.3%). The USPSTF (2022) recommends that colonoscopy screening
186
should be taken every 10 years for individuals aged 50-75 years old. From the results,
however, colonoscopy taken within the past 10 years was lower for each gender when
compared to colonoscopy taken overall. For example, colonoscopy within the past 10
years had a lower rate for White only, non-Hispanic females (66.1%) compared with
overall colonoscopy taken by White only, non-Hispanic females (79.4%). For limited-
English proficient Hispanic females (60.5%) vs males (27.3%) were the overall rates of
colonoscopy compared with colonoscopy taken within the past 10 years by limited-
English proficient Hispanic females (57.1%) vs males (25.0%). This indicates that
effective CRC screening requires an adequate public health campaign to educate
communities about health literacy and the importance of adhering to colonoscopy
schedules as recommended by the USPSTF (Edwardson et al., 2023).
Among the 3 choices of CRC screening in this study, sigmoidoscopy was used
least by the participants. For White only, non-Hispanic, 3.9% of males and 4.1% of
females; Black only, non-Hispanic, 5.0% of males and 10.5% of females; English-
speaking Hispanics, 3.0% of males and 6.5% of females; and Spanish-speaking
Hispanics, 15.0% of males and 9.7% of females received sigmoidoscopy. Out of 766
participants in the sample, 586 responded to the questionnaire on sigmoidoscopy, of
which 31 responded to have taken sigmoidoscopy screening and 555 stated that they did
not. In comparison, 720 surveyed participants responded to the questionnaire on
colonoscopy, and 524 stated to have received colonoscopy while 196 did not.
Health care providers indicate that flexible sigmoidoscopy is quicker, safer, less
complicated, and more affordable than colonoscopy, and usually it does not require
187
intravenous sedation while enemas employed in its preparation have fewer side effects
and greater acceptability than oral solution employed in the preparation for colonoscopy.
Also, because its procedural approach is relatively less complicated than colonoscopy,
sigmoidoscopy could be performed competently by non-physician endoscopists (Cross et
al., 2019; Maslekar et al., 2010; Riegert et al., 2020).
Several reasons account for colonoscopy as a preferred choice over
sigmoidoscopy. While sigmoidoscopy has high sensitivity for CRC screening in the distal
colon and rectum, it could only reach splenic flexure at best, making it ineffective in
accessing abnormalities in the proximal colon (Cross et al., 2019). On the contrary,
colonoscopy could examine the whole large bowel up to the ileocecal valve, which is the
end of the small intestine. Colonoscopy is also both diagnostic and therapeutic, capable
of detecting and removing adenomas, which are precancerous polyps (Safiyeva &
Bayramov, 2019). Sigmoidoscopy is only diagnostic and does not have the capability of
removing adenomas (Issa & Noureddine, 2017). Another possible reason why
colonoscopy is preferred over sigmoidoscopy is that the former is recommended to be
performed every 10 years as against the latter, which is recommended to be done every 5
years (Bénard et al., 2018).
In comparing FOBT and colonoscopy, multivariate logistic regression showed
that for FOBT in the past year, White only, non-Hispanic, 18.6% of males and 15.0% of
females; Black only, non-Hispanic, 28.6% of males and 23.7% of females; English-
speaking Hispanics, 18.2% of males and 17.0% of females; and Spanish-speaking
Hispanics, 28.6% of males and 5.7% of females received FOBT in the past year. While
188
participants engaged with higher rates of FOBT than sigmoidoscopy, their overall use of
FOBT was much lower than colonoscopy. Also, similar to both colonoscopy and
sigmoidoscopy, Hispanic with LEP had lower rates of screening with FOBT compared
with other participants who took the 2020 Texas BRFSS survey in English.
Research Questions and Hypotheses
This study explored the impact of LEP on CRC screening among Hispanic
Americans living in the state of Texas. The data was obtained from the 2020 Texas
BRFSS survey on CRC. Language that participants used to respond to the BRFSS survey
questionnaires was considered as the language they speak. Thus, Hispanics who
responded to the survey in Spanish were individuals who did not speak English, or they
were limited English proficient individuals. The main IV was LEP. I grouped the IV into
8 categories, including White only, non-Hispanic males; White only, non-Hispanic
females; Black only, non-Hispanic males; Black only, non-Hispanic females; Hispanic
males responding in in English; Hispanic females responding in English; Hispanic males
responding in Spanish; and Hispanic females responding in Spanish. The CRC screening
rates define the DV. The variations in the IV caused changes in the rates of the CRC
screening test, which was the DV. The study focused on three forms of the DV, namely
FOBT (blood stool test) within the past year, and/or sigmoidoscopy within the past 5
years, and/or colonoscopy within the past 10 years. Entries excluded from the analysis
included those missing a language variable, gender, race, or Hispanic/Latino origin.
The RQs and hypotheses were addressed using the analyses from the results. The
RQ 1 stated: Are CRC screening rates different between Hispanic and non-Hispanic
189
residents in Texas with and without LEP, when potential confounding variables including
age, income, occupation, health care access, and educational levels of participants are
controlled? The results showed that White only, non-Hispanic, both male and female
participants in the study, had higher rates of CRC screening, using colonoscopy,
sigmoidoscopy, and FOBT compared to Black only, non-Hispanic and Hispanic
participants, irrespective of the language they speak. However, when language was
considered, English-speaking participants in the study had far higher rates of screening
compared to Hispanics with LEP whether they used colonoscopy, sigmoidoscopy, or
FOBT. For example, participants who were English proficient had 93.9% colonoscopy
screening compared with 6.1% of Hispanics with LEP, p <.001. For English-speaking
(80.6%) and Hispanics with LEP (19.4%) of 31 participants received sigmoidoscopy
screening within the past 5 years, p = .031. Similarly, when FOBT rates were assessed
using the language respondents used for the survey, 92.4% of the respondents were
identified with English and 7.6% of respondents were identified with Spanish,
demonstrating that LEP was likely a reason for lower rates of FOBT screening for CRC,
p = .040. This analysis showed that, the null hypothesis, H01, which states that there is no
significant relationship between LEP and CRC screening rates among Hispanic and non-
Hispanic residents in Texas, when potential confounding variables including age, income,
occupation, health care access, and educational levels of participants are controlled
should be rejected, since the CRC screening demonstrated that p ˂ .05. Thus, the
alternative hypothesis, Ha1, which states that there is a significant relationship between
LEP and CRC screening rates among Hispanic and non-Hispanic residents in Texas,
190
when potential confounding variables including age, income, occupation, health care
access, and educational levels of participants are controlled could be accepted.
The RQ 2 stated: Are CRC screening rates different between male and female
residents in Texas with and without LEP, when potential confounding variables including
age, income, occupation, health care access, and educational levels of participants are
controlled? Grouping the study participants into 8 categories, including White only, non-
Hispanic males; White only, non-Hispanic females; Black only, non-Hispanic males;
Black only, non-Hispanic females; Hispanic males proficient in English; Hispanic
females proficient in English; Hispanic males with LEP; and Hispanic females with LEP,
I used multivariate regression analysis to assess the observed and predicted frequencies
and percentages for CRC screening using colonoscopy, sigmoidoscopy, and FOBT. The
results showed varied percentages and frequencies for CRC screening among English-
speaking White only, non-Hispanic male and female; Black only, non-Hispanic male and
female; Hispanic male and female; and Hispanic with LEP male and female participants.
The limited English proficient Hispanic participants are considered as individuals who
speak Spanish only. Analysis of the results indicated that overall, English-speaking
participants in the state of Texas received significantly higher rates of CRC screening,
with variations between males and females. However, both male and female LEP
individuals had significantly low rates of CRC screening, particularly when rates were
compared males to males and females to females within the Hispanic English-speaking
and Hispanic LEP groups. In comparing all forms of CRC screening together, the
following results were obtained: White only, non-Hispanic males (78.1%) and females
191
(75.0%); Black only, non-Hispanic males (60.0%) and females (77.8); and Hispanics who
did the survey in English, males (66.7%) and females (74.5%), while LEP Spanish males
(45.0%) and females (61.8%). The multivariate logistic regression analysis showed that
OR for sex = 0.894; 95% CI 0.599, 1.333, p = .581. The OR ˂ 1.0 indicates that the odds
of difference between male and female CRC screening is smaller or less significant. This
assessment showed that, the null hypothesis, H02, which states that there is no significant
relationship between gender and CRC screening of Hispanic and non-Hispanic residents
in Texas, when potential confounding variables including age, income, occupation, health
care access, and educational levels of participants are controlled could not be rejected.
Also, the p ˃ .05 indicates that there is no significant relationship between gender and
CRC screening among the study participants. Thus, the alternative hypothesis, Ha2,
which states that there is a significant relationship between gender and CRC screening of
Hispanic and non-Hispanic residents in Texas, when potential confounding variables
including age, income, occupation, health care access, and educational levels of
participants are controlled, could not be accepted.
Although generally, there was no clear significance between males and females
for CRC screening, in considering Hispanic groups alone, the differences in the CRC
screening showed remarkable differences between males and females. For Hispanics who
did the survey in English, males (66.7%) and females (74.5%), while LEP Spanish males
(45.0%) and females (61.8%) had CRC screening, showing an uptake of screening among
the female groups vs their male counterparts. These results are consistent with previous
studies where Hispanic men with LEP were found with lower CRC screening, suggesting
192
that limited-English proficient Hispanic males are at the greatest risk of not being
screened for CRC compared to non-Hispanics Whites and Blacks, English proficient
Hispanics, and even Hispanic females with LEP (Diaz et al., 2013; Hall et al., 2022).
Analysis: The Health Belief Model Conceptual Framework
The HBM has assumed a central position among theoretical frameworks that seek
to account for the broad failure of individuals to participate in events to avert or detect
asymptomatic disease (Hochbaum, 1958; Rosenstock, 1966, 1974). In addition, HBM
seeks to explain individuals’ responses based on their experienced symptoms (Kirscht,
1974) and to their behavior in response to clinically diagnosed illnesses, especially in
compliance with medical regimens (Becker, 1974). The HBM integrates six constructs,
including risk susceptibility, risk severity, benefits to action, barriers to action, perceived
self-efficacy, and cues to action predict health behavior (Becker, 1974). Historically, the
HBM was developed to screen for disease and to immunize against viral diseases like
poliomyelitis and influenza, and of attempts to improve on compliance with medical
advice regarding diabetes, hypertension, cancer, obesity, exercise, seat-belt use, and HIV-
risk behavior (Janz & Becker, 1984; Stretcher & Rosenstock, 1997).
The constructs of HBM form a logical connection that explain the nature of this
study. These constructs do not only highlight the impact of language barrier in identifying
asymptomatic CRC patients through screening for early detection, which leads to positive
disease prognosis, but also associate clinical approach of managing CRC symptoms and
improve on core determinants of disease prognosis (Hochbaum, 1958; Rosenstock, 1966,
1974). The risk susceptibility describes an individual’s subjective awareness of the risk of
193
acquiring an illness or disease. The risk of severity creates awareness on individual’s
assumptions of contracting CRC, and the possibility of not receiving an appropriate
treatment, which may result in a wide variation of apprehensiveness (Karl et al., 2022).
Such apprehensiveness may include medical consequences like death or disability and
social consequences that could negatively affect their families with economic hardships.
The benefits to action explain an individual’s appreciation of the effectiveness of
available actions to reduce the adverse impacts of CRC (or to cure CRC).
Individuals who recognize these dangers take actions to avoid or cure the cancer
based on their understanding and evaluation of both perceived susceptibility and
perceived benefit as a motivation to accept the recommended health action they consider
beneficial (Sharifikia et al., 2019). Barriers to action refers to an individual’s perception
of barriers to take a recommended health action on CRC, which may include a wide
variation of obstacles, such as costs, time-consumption, and inconvenience related to
health actions they accept in dealing with CRC (Rakhshanderou et al., 2020).
Consequently, individuals perform cost and benefit analysis and weigh the effectiveness
of decisions they accept. Perceived self-efficacy highlights on individuals’ confidence in
their ability to appropriately perform a behavior, which was a construct that was added to
the model in mid-1980s, and it is a construct in many behavioral theories as it is directly
associated with whether an individual takes up the desired behavior (Rosenstock,
Strecher, & Becker, 1988). The cues to action, which could be internal, such as fatigue
due to anemia, lack of appetite and weight loss, which are often associated with CRC
residual symptoms, or external, such as advice from closed relatives and friends, illness
194
of family member, and newspaper articles, serve as a stimulus that influence decision-
making process to embrace a recommended health action (Tsai et al., 2021).
Therefore, the fundamental importance of HBM constructs makes it a suitable
public health theory in appreciating the unmet challenges, such as language barrier for
individuals in identifying asymptomatic CRC patients and the ability to associate it with
clinical instructions of managing CRC residual symptoms (Rakhshanderou et al., 2020).
This demonstrates that HBM serves as an example of a value-expectancy theory, where
behavior assumes the function of the subjective value of an outcome and the subjective
probability, or expectation, where a specific action would lead to the expected outcome
(Lewin et al., 1944). In the context of health-related behavior, the value-expectancy
theory shows the willingness to prevent illness or to get well (value), which also signifies
the belief that a particular health action available to individuals would avoid or
ameliorate illness (expectancy), where individual’s estimate of personal susceptibility to
and the severity of an illness could be linked to the likelihood of being able to reduce that
threat through personal action (Ban & Kim, 2020).
Mediating Effects of Limited English Proficiency
To understand the mediating effect of LEP on each motivating variable included
in the model, a series of bivariate logistic regression on the full sample that included LEP
and other motivating variables as regressors indicated that age was remarkably significant
in the presence of LEP. Healthcare access and education status were also significant,
while race of participants, income of participants, and employment status were not
significant. Besides these factors that were assessed using bivariate logistic regression
195
model, several other factors, including prevailing environmental conditions, such as
Coronavirus disease 2019 (COVID-19) caused by SARS-CoV-2, which erupted in China
and spread around the globe from the latter part of 2019 throughout 2020 and beyond,
had serious effects on health screening on almost all chronic diseases, including CRC.
Researchers found out that the COVID-19 pandemic decreased physician’s office
visits significantly from the early months of 2020. For example, in the United States, one
study found out that the COVID-19 pandemic decreased the total number of outpatient
visits per provider by 70% during the week of April 5-11, 2020, when the effects of the
pandemic became acutely significant relative to the same week in prior years (Chatterji &
Li, 2021). Because of acute reduction in primary care visitations, several chronic
diseases, such as hypertension, diabetes, asthma, chronic obstructive pulmonary disease
(COPD), and cancer conditions exacerbated and became unstoppably top comorbidities
with COVID-19, which led to increased mortality and morbidity among such patients
(Fekadu et al., 2021). Among cancers, due to the precipitous reduction in lung cancer
screening, health care providers reported unprecedented proportions of nodules
suspicious for malignancy when COVID-19 spread was abated and screening resumed
(Van Haren et al., 2021).
Several other factors potentially mediated low CRC screening rates, such as lack
of transportation, poor health literacy, physician-patient communications in situations
like physician-patient concordant and discordant relationships. Transportation is an
important social determinant of health, particularly in rural and suburban communities
where public transportation services are often unavailable. Researchers identify that
196
transportation barriers disproportionately affect the most vulnerable groups of society
who carry the highest burden of chronic diseases, such as cancer, cardiovascular disease,
diabetes, and COPD (Lin & Cui, 2021).
For individuals to undertake CRC screening in a physician’s office, they need to
travel to the office. Transportation becomes an obstacle to screening when individuals do
not have the means to visit the physician’s office. Such an impediment could discourage
vulnerable people from scheduling an appointment for screening. Thus, it is critical to
identify interventions that improve access to transportation for individuals to be able to
visit their primary care physicians (Starbird et al., 2019). Importantly, individuals who
have high knowledge about health could seek preventive care through regular visits to
their primary care physicians and often request screening for any pathophysiological
variations in their health systems. Studies show that higher health literacy supports
guideline-concordant screening which could identify a disease (Rutan et al., 2021), such
as cancer in its rudimentary stage before it becomes cancerous (Cartwright et al., 2022).
Furthermore, stage 1 cancers are easier to treat than stage 3 or 4, which has already
metastasized to distant organs (Welch & Hurst, 2019). Lack of health care access is an
impediment that makes cancer screening rates lower even when individuals have
concerns of potential cancer like individuals aware of the disease in their families
(Aleshire et al., 2021). While some studies indicate that physician-patient concordance do
not play any significant role in primary care effectiveness (Saha & Beach, 2020), other
researchers also argue that physician-patient concordant relationships contribute
immensely to effective health care delivery and screening for chronic diseases, such as
197
CRC (Shen et al., 2018). Physician-patient concordant relationships often tend to trace
both physicians and patients to similar cultures, where relatedness in culture enhance dual
understanding and appreciation of values (Moore et al., 2022).
Summary
Low utilization of screening is a major barrier to minimizing CRC morbidity and
mortality. Barriers, such as limited language proficiency, cultural values that may obscure
the importance of screening, new migration status, low health literacy, lack of health care
access, prevailing environmental conditions, such as acute pandemics, and lack of
transportation to visit doctor’s office are some of the principal constraints that depress
health screening to identify and treat chronic diseases in the early stages. Among
Hispanic communities in the United States, LEP has been particularly a challenge, and
the impact of LEP is more prominent in situations such as CRC screening because of
complex process involved in preparation for procedures, new immigration status,
ineffective communication, and low health literacy. For example, colonoscopy
preparation entails a complex web of processes in ensuring that solutions that patients
must drink to clear their colon must be of certain concentration to achieve the desired
goals. Other methods, such as FOBT, involve the embarrassing process of collecting
one’s stool for laboratory analysis, an offputting process that tends to diminish people’s
interest in undertaking FOBT.
This study revealed that LEP is a central barrier to improved CRC screening. For
example, within the past 10 years Hispanic participants in the study had 19.2% of
colonoscopy screenings, however separation of English proficient Hispanics from
198
Hispanics with LEP revealed that only 6.6% of individuals with LEP received
colonoscopy screening within the same period. Considering sigmoidoscopy, 32.3% of all
Hispanics in the study received screening, but when only Spanish-speaking group was
assessed, the rate decreased to 19.4% for Hispanics who received sigmoidoscopy within
the past 5 years. Similarly, 21.0% of Hispanic respondents participated in FOBT for CRC
within the past year, but when only LEP Hispanics were considered, only 7.6% of used
FOBT within the same period. Thus, the trend was the same for all the three forms of
CRC screening with LEP Hispanics receiving the least rates of screening. The study did
not indicate that gender influences were prominent in undertaking CRC screening,
although in most cases, females had higher rates of CRC screening than their male
counterparts.
Strengths and Limitations
This study focused on three forms of CRC screening to assess the impact of LEP
on screening among Hispanic and non-Hispanic residents of Texas using the 2020
BRFSS. Because the study assessed three forms of screening, it created a clear picture of
the impact of LEP on screening. The study was also the first to be conducted in the state
of Texas, although it has been conducted elsewhere in the United States as well as
nationally. The results were consistent with the outcome of the past studies. Although the
study did not include all eligible participants in the BRFSS data, the selection was
conducted randomly thereby minimizing any potential selection bias. Also, because the
study focused only on White only, Black only, and Hispanic only populations for sample
collection, the data for the study excluded all individuals who did not identify with the
199
definitions set up for the study. The study sample’s population was from 50 to 79 years
old. All other individuals outside the defined age range were excluded from the study.
The number of study participants was 766, which was large enough to ensure the
power of the study, which is the probability of finding a difference that exists in a
population. The power relies on the chosen level of significance (p <.05 for this study),
effect size, variability of the measured variables, and sample size (Serdar et al., 2021).
However, out of the 766 participants, 525 or 68.5% were White only, non-Hispanic, 77 or
10.1% were Black only, non-Hispanic, and 164 or 21.4% were Hispanic. Although the
Black only, non-Hispanic participants were far fewer than either White or Hispanic
participants, their representation reflected their population in the state of Texas, where
Black only, non-Hispanic were 13.2% in 2020. However, the sample size of Hispanics in
the study could have been higher than 21.4% since in 2020, the Hispanic population in
the state of Texas was more than 39% (nearly 11.4 million) compared to White only,
non-Hispanic population of nearly 40.0% (over 11.6 million). Thus, while the size of the
Hispanic representation of 164 participants was an adequate sample size, the difference
of 361 or 47.1% in representation between White only, non-Hispanic and Hispanic
sample sizes was likely to potentially influence the generalizability of the findings. Thus,
the results should be interpreted with caution and generalized only within the state of
Texas and not in the United States. In 2020, Black only, non-Hispanic population was
13.2% of the state of Texas population, making their representation of 10.1% in the study
reasonably adequate.
200
The BRFSS data depends on respondents’ responses to questionnaire in the
survey. Any incorrect information the respondents provided was likely to find their way
into the data, which could be influenced by social desirability bias or recall bias. This
means that the accuracy of the data depended partly on the correct information provided
by the survey respondents. This study could not establish validity tests for the
participants’ responses to the survey, and therefore, did not verify any potential
misinformation bias or misclassification. Researchers also cite an inherent challenge in
the BRFSS as the effectiveness in managing an increasingly complex surveillance system
that serves the needs of several programs in the ever-evolving environment of
telecommunication technology and the higher demand for more local-level data (Mokdad
et al., 2003). Any inherent disadvantages of BRFSS data collection could potentially
affect the quality of the results.
Recommendations
The study indicated that CRC screening was lower among individuals younger
than 59 years of age. The USPSTF recommends that CRC screening should start from
age 50 up to 75 years old in both males and females. However, studies demonstrate that a
significant number of individuals between ages 45 and 50 are being diagnosed with CRC.
Because the USPSTF has not lowered the initial age, BRFSS data are available only for
individuals 50 years and older. Due to increased incidence of CRC in individuals younger
than 50 years old, it would be beneficial to the general population for the CDC to
recommend 45 years as the initial screening age for CRC. By so doing health care
providers may be motivated to educate their patients to receive CRC screening at a
201
younger age than current initial screening age of 50 years old. That would also ensure
that all health insurance providers would be obliged to pay for such screenings. Even at
the current initial screening age of 50 years old where individuals aged 50 -59 are not
undertaking screening as effectively as recommended, public health campaign needs to
be used to create awareness in the general population to motivate individuals within the
age bracket of 50 - 59 to ensure that they receive screening. Such a campaign could
involve health insurance organizations and employers to remind their employees to take
CRC screening when they attain 50 years of age.
Individuals of Hispanic origin, particularly Hispanic men, have wrong perceptions
of cancer, which affect their understanding and preparedness to undergo screening for
early detection of cancer. Several investigations in the social, behavioral, and medical
sciences suggest that there are notable variations in their perception and behavior toward
patients’ cultural and ethnic backgrounds (Wittich et al., 2019). Such characteristics may
influence cancer prevention behaviors, including delay in seeking preventive care, views
about etiology of disease, and beliefs about treatment and prognosis (Gast et al., 017). To
address cultural misgivings and misperceptions, public health practitioners and
policymakers may need to use social and community centers, such as religious centers
like the church to educate the Hispanic communities, especially new immigrants to
understand the importance of health screening for early detection of cancer and other
chronic diseases, such as diabetes and hypertension. Also, Hispanic nonprofit
organizations that specialize on cultural integration into the American society could
provide health care education materials in both English and Spanish to support them.
202
Future studies may investigate the impact of such cultural tendencies on Hispanic interest
in CRC screening, and how social organizations and nonprofits that engage with the
Hispanic community may promote CRC screening.
Implications
Several studies on Hispanic Americans and CRC implicate them as one of the key
racial groups that are often diagnosed with CRC at advanced states compared to their
White counterparts; the other racial group that are also diagnosed with CRC in later
stages is African Americans. Hispanics and Blacks are diagnosed at advanced stages
because they fail to seek CRC screening as recommended by the USPSTF (Muller et al.,
2021). In this study, LEP Hispanic men results highlight the particularly low odds of
CRC screening relative to their female counterparts, and other racial/ethnic groups. The
poor screening rates were consistent with multiple studies conducted on Hispanics across
the United States (Diaz et al., 2013; Garcia et al., 2018; Wittich et al., 2019). Qualitative
studies suggest that maintenance of masculinity, embarrassment about invasive
procedures, such as colonoscopy, and fear are some of the reasons why low screening
rates are seen among Hispanic males (Christy et al., 2014; Rogers et al., 2022).
Understanding the challenges confronting Hispanic men about the implications of their
low participation in CRC screening and the possible health burdens, including increased
cost of treatment to themselves and their families as well as morbidity and potential death
may influence them to consider screening with some degree of seriousness. In this regard,
health care providers and public health practitioners could use the available resources to
educate them through public health campaigns using community centers, religious
203
groups, employers, and health insurance organizations to engage them with preventive
health promotion initiatives.
Epidemiological evidence on CRC survival possibilities could be compelling facts
that public health practitioners and other stakeholders need to use as part of the education
for individuals eligible for CRC screening. For example, the overall 5-year survival rate
for CRC is approximately 65.2%, however, according to stages defined by the American
Joint Committee on Cancer (AJCC) fifth edition system, 5-year stage-specific survivals
were 93.2% for stage I, 82.5% for stage II, 59.5% for stage III, and 8.1% for stage IV,
with stages III and IV already spread to distant organs (Joachim et al., 2019). This
indicates that survival rates with good prognosis are associated with CRC detection at its
rudimentary stage up to stage II before it metastasizes to distant organs.
Although in recent years, CRC incidence rates declined among older (age ≥50
years) Whites, non-Hispanic and Hispanic populations, the former underwent a steeper
decline (31% vs. 27%) relative decline among Hispanics. However, CRC incidence
among young adults (age 20-49) have experienced increased incidence rates from 2001 to
2014 (Garcia et al., 2018). The authors noted that the largest relative rise in incidence was
seen among Hispanics from 20–29 years (90% vs. 50% relative increase among Whites,
non-Hispanic). These changes in disease onset must influence the USPSTF to make
changes from the recommended initial screening age of 50 years to age 45, which is
currently recommended by the ACS.
204
Social Change
The HBM conceptual framework employed to highlight its constructs and their
relevance to CRC screening lays the foundation to motivate eligible people to undertake
screening for CRC. All the six HBM constructs, namely risk susceptibility, risk severity,
benefits to action, barriers to action, perceived self-efficacy, and cues to action predict
health behavior seek to promote preventive health care (Zewdie et al., 2022). The
constructs lay the roadmap for health care providers and public health campaigners to
engage society and communities to participate in health screening to avoid the glaring
consequences of poor health (Karimy et al., 2021). Using the HBM constructs as
motivating factors, public health campaigners, health care advocates, and health care
providers could promote the importance of health screening as a basis to empower
eligible individuals to undergo CRC screening (Rakhshanderou et al., 2020). Social
groups may be used as a conduit to embark on health screening drive and spur
communities to accept screening as a basic method to minimize increased incidence of
chronic diseases, including CRC.
In addition to serving an impetus to push public health practitioners and other
stakeholders to promote CRC screening, it is also a wealth of knowledge regarding
inherent advantages in screening to detect CRC early enough for good prognosis and
treatment. Similarly, it is a rich source of academic information that could drive health
policies in the right direction where policymakers would be able to utilize it in
formulating workable policies to enhance public health initiatives for the target
populations in the study. For example, some policies may be tailored to meet the needs of
205
Hispanic males who are historically the least likely to seek CRC screening. The study
also identified important variables, particularly types of screening employed to identify
CRC, such as colonoscopy, sigmoidoscopy, and FOBT. Health care providers need to
educate the eligible public about these types of screening and their advantages and
disadvantages so that individuals could make informed choices when they need to
undertake screening for CRC detection.
Conclusion
Results of the study confirmed the low CRC screening rates among LEP Hispanic
Americans in the state of Texas. The results are consistent with other studies done
elsewhere in the United States, which concluded that limited English proficient Hispanics
are at an increased risk for CRC due to late diagnosis of the disease blamed principally on
lack of screening for early detection. The results are also consistent with other studies
that Hispanic men, particularly those with LEP, do not participate in CRC screening
compared with their female counterparts and men of White only, non-Hispanic and Black
only, non-Hispanic origins. It also shows that Blacks are at an increased risk for CRC due
to their low participation in CRC screening, possibly due to lack of health insurance or
reasons I did not investigate. Despite the availability of Medicaid through the ACA, the
state of Texas remains one of the few states in the United States where Medicaid-eligible
citizens remain uninsured in huge numbers. The study found out that while women
undertake CRC screening more than men, the difference is not significant.
206
References
Abotchie, P. N., Vernon, S. W., & Du, X. L. (2012). Gender differences in colorectal
cancer incidence in the United States, 1975-2006. Journal of Women’s Health
(2002), 21(4), 393–400. https://doi.org/10.1089/jwh.2011.2992
Abualkhair, W. H., Zhou, M., Ahnen, D., Yu, Q., Wu, X. C., & Karlitz, J. J. (2020).
Trends in incidence of early-onset colorectal cancer in the United States among
those approaching screening age. JAMA Network Open, 3(1), e1920407.
https://doi.org/10.1001/jamanetworkopen.2019.20407
Adams, L. B., Richmond, J., Corbie-Smith, G., & Powell, W. (2017). Medical mistrust
and colorectal cancer screening among African Americans. Journal of Community
Health, 42(5), 1044–1061. https://doi.org/10.1007/s10900-017-0339-2
Adepoju, O. E., Preston, M. A., & Gonzales, G. (2015). Health care disparities in the
post-Affordable Care Act era. American Journal of Public Health, 105 Suppl
5(Suppl 5), S665–S667. https://doi.org/10.2105/AJPH.2015.302611
Ahlquist D. A. (2019). Stool-based tests vs screening colonoscopy for the detection of
colorectal cancer. Gastroenterology & hepatology, 15(8), 437–440.
Aleshire, M. E., Adegboyega, A., Escontrías, O. A., Edward, J., & Hatcher, J. (2021).
Access to care as a barrier to mammography for Black Women. Policy, Politics &
Nursing Practice, 22(1), 28–40. https://doi.org/10.1177/1527154420965537
Allen, E. M., Call, K. T., Beebe, T. J., McAlpine, D. D., & Johnson, P. J. (2017). Barriers
to care and health care utilization among the publicly insured. Medical
Care, 55(3), 207–214. https://doi.org/10.1097/MLR.0000000000000644
207
Alpert, E. N., Clark, T., Garcia-Alcaraz, C., Eddington, S. N., Carrizosa, C., Haughton,
J., de la Torre, C. L., Garcia-Bigley, F., Arredondo, E. M., Ramers, C., Nodora, J.,
& Wells, K. J. (2022). Comparing Latino community members’ and clinical
staff’s perspectives on barriers and facilitators to colorectal cancer
screening. Journal of Cancer Education, 37(6), 1645–1653.
https://doi.org/10.1007/s13187-021-02007-0
Al Shamsi, H., Almutairi, A. G., Al Mashrafi, S., & Al Kalbani, T. (2020). Implications
of language barriers for healthcare: A systematic review. Oman Medical
Journal, 35(2), e122. https://doi.org/10.5001/omj.2020.40
Andersen, R., Bozzette, S., Shapiro, M., St Clair, P., Morton, S., Crystal, S., Goldman,
D., Wenger, N., Gifford, A., Leibowitz, A., Asch, S., Berry, S., Nakazono, T.,
Heslin, K., & Cunningham, W. (2000). Access of vulnerable groups to
antiretroviral therapy among persons in care for HIV disease in the United States.
HCSUS Consortium. HIV Cost and Services Utilization Study. Health Services
Research, 35(2), 389–416. Retrieved from
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1089126/
Andersen, R. M., Giachello, A. L., & Aday, L. A. (1986). Access of Hispanics to health
care and cuts in services: A state-of-the-art overview. Public Health Reports
(Washington, D.C.: 1974), 101(3), 238–252.
Andrew, A. S., Parker, S., Anderson, J. C., Rees, J. R., Robinson, C., Riddle, B., &
Butterly, L. F. (2018). Risk factors for diagnosis of colorectal cancer at a late
stage: A population-based study. Journal of General Internal Medicine, 33(12),
208
2100–2105. https://doi.org/10.1007/s11606-018-4648-7
Arnold, C. L., Rademaker, A., Wolf, M. S., Liu, D., Lucas, G., Hancock, J., & Davis, T.
C. (2016). Final results of a 3-year literacy-informed intervention to promote
annual fecal occult blood test screening. Journal of Community Health, 41(4),
724–731. https://doi.org/10.1007/s10900-015-0146-6
Ban, H. J., & Kim, H. S. (2020). Applying the modified Health Belief Model (HBM) to
Korean medical tourism. International Journal of Environmental Research and
Public Health, 17(10), 3646. https://doi.org/10.3390/ijerph17103646
Bandura, A. (1965). Influence of models’ reinforcement contingencies on the acquisition
of imitative responses. Journal of Personality and Social Psychology, 1, 589-595.
https://doi.org/10.1037/h0022070
Barthold, D., Yeung, K., Lieberman, D., & Fendrick, A. M. (2022). Comparison of
screening colonoscopy rates after positive noninvasive testing for colorectal
cancer in states with and without cost-sharing. JAMA Network Open, 5(6),
e2216910. https://doi.org/10.1001/jamanetworkopen.2022.16910
Bastos, J. L., Duquia, R. P., González-Chica, D. A., Mesa, J. M., & Bonamigo, R. R.
(2014). Field work I: selecting the instrument for data collection. Anais
Brasileiros de Dermatologia, 89(6), 918–923. https://doi.org/10.1590/abd1806-
4841.20143884
Batai, K., Gachupin, F. C., Estrada, A. L., Garcia, D. O., Gomez, J., & Kittles, R. A.
(2019). Patterns of cancer related health disparities in Arizona. Cancer Health
Disparities, 3, e1–e20. Retrieved from
209
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6959217/
Bayati, T., Dehghan, A., Bonyadi, F., & Bazrafkan, L. (2018). Investigating the effect of
education on health literacy and its relation to health-promoting behaviors in
health center. Journal of Education and Health Promotion, 7, 127.
https://doi.org/10.4103/jehp.jehp_65_18
Becker, M. H., Drachman, R. H., & Kirscht, J. P. (1974). A new approach to explaining
sick-role behavior in low-income populations. American Journal of Public
Health, 64(3), 205–216. https://doi.org/10.2105/ajph.64.3.205
Bénard, F., Barkun, A. N., Martel, M., & von Renteln, D. (2018). Systematic review of
colorectal cancer screening guidelines for average-risk adults: Summarizing the
current global recommendations. World Journal of Gastroenterology, 24(1), 124–
138. https://doi.org/10.3748/wjg.v24.i1.124
Benson, M., Johannes, A., Weiss, J. M., Lucey, M., Pier, J., & Pfau, P. (2021). Colorectal
Cancer Screening After Changes in US Preventive Services Task Force
Guidelines with Increased Screening Options. Wisconsin Medical
Journal, 120(2), 127–130. Retrieved from https://wmjonline.org/wp-
content/uploads/2021/120/2/127.pdf
Berchick, E. R., Barnett, J. C., & Upton, R. D. (2019). Health insurance coverage in the
United States: 2018. Report Number P60-267 (RV). Retrieved from:
https://www.census.gov/
Berdahl, T. A., & Kirby, J. B. (2019). Patient-provider communication disparities by
limited English proficiency (LEP): Trends from the US Medical Expenditure
210
Panel Survey, 2006-2015. Journal of General Internal Medicine, 34(8), 1434–
1440. https://doi.org/10.1007/s11606-018-4757-3
Bertani, A., Di Paola, G., Russo, E., & Tuzzolino, F. (2018). How to describe bivariate
data. Journal of Thoracic Disease, 10(2), 1133–1137.
https://doi.org/10.21037/jtd.2018.01.134
Blackwelder, A., Hoskins, M., & Huber, L. (2021). Effect of Inadequate Sleep on
Frequent Mental Distress. Preventing Chronic Disease, 18, E61.
https://doi.org/10.5888/pcd18.200573
Boscolo-Hightower, A., Rafton, S. A., Tolman, M., Zhou, C., & Ebel, B. E. (2014).
Identifying families with limited English proficiency using a capture-recapture
approach. Hospital Pediatrics, 4(1), 16–22. https://doi.org/10.1542/hpeds.2013-
0040
Bosworth, A., Finegold, K., & Ruhter, J. (2021). The remaining uninsured: Geographic
and demographic variation. Assistant Secretary for Planning and Evaluation
(ASPE) of the Office of Health Policy of the US Department of Health and
Human Services. Issue Brief Report Number HP-2021-06. Retrieved from
https://aspe.hhs.gov/
Bramer, W. M., de Jonge, G. B., Rethlefsen, M. L., Mast, F., & Kleijnen, J. (2018). A
systematic approach to searching: an efficient and complete method to develop
literature searches. Journal of the Medical Library Association, 106(4), 531–541.
https://doi.org/10.5195/jmla.2018.283
Brand Bateman, L., Khamess, S., Abdelmoneim, S. E., Arafat, W., Fouad, M. N.,
211
Khamis, Y., Omar, A., Abdelmoneim, R. S., & Scarinci, I. (2020). Designing an
effective colorectal cancer screening program in Egypt: A qualitative study of
perceptions of Egyptian primary care physicians and specialists. The
Oncologist, 25(10), e1525–e1531. https://doi.org/10.1634/theoncologist.2019-
0687
Brown, J. J., Asumeng, C. K., Greenwald, D., Weissman, M., Zauber, A., Striplin, J.,
Weng, O., List, J. M., Farley, S. M., & Winawer, S. J. (2021). Decreased
colorectal cancer incidence and mortality in a diverse urban population with
increased colonoscopy screening. BMC Public Health, 21(1), 1280.
https://doi.org/10.1186/s12889-021-11330-6
Brown, M. T., & Bussell, J. K. (2011). Medication adherence: WHO cares? Mayo Clinic
Proceedings, 86(4), 304–314. https://doi.org/10.4065/mcp.2010.0575
Brown, A. F., Ma, G. X., Miranda, J., Eng, E., Castille, D., Brockie, T., Jones, P.,
Airhihenbuwa, C. O., Farhat, T., Zhu, L., & Trinh-Shevrin, C. (2019). Structural
interventions to reduce and eliminate health disparities. American Journal of
Public Health, 109(S1), S72–S78. https://doi.org/10.2105/AJPH.2018.304844
Buchman, S., Rozmovits, L., & Glazier, R. H. (2016). Equity and practice issues in
colorectal cancer screening: Mixed-methods study. Canadian Family physician
Medecin de Famille Canadien, 62(4), e186–e193. Retrieved from
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4830674/
Burnett-Hartman, A. N., Mehta, S. J., Zheng, Y., Ghai, N. R., McLerran, D. F., Chubak,
J., Quinn, V. P., Skinner, C. S., Corley, D. A., Inadomi, J. M., Doubeni, C. A., &
212
PROSPR Consortium (2016). Racial/ethnic disparities in colorectal cancer
screening across healthcare systems. American Journal of Preventive
Medicine, 51(4), e107–e115. https://doi.org/10.1016/j.amepre.2016.02.025
Busch, E. L., Martin, C., DeWalt, D. A., & Sandler, R. S. (2015). Functional health
literacy, chemotherapy decisions, and outcomes among a colorectal cancer
cohort. Cancer Control, 22(1), 95–101.
https://doi.org/10.1177/107327481502200112
Butterly L. F. (2020). Proven strategies for increasing adherence to colorectal cancer
screening. Gastrointestinal Endoscopy Clinics of North America, 30(3), 377–392.
https://doi.org/10.1016/j.giec.2020.02.003
Campbell, D. T., & Stanley, J. C. (2015). Experimental and quasi-experimental designs
for research. Ravenio books. Retrieved from https://www.scirp.org/
Cano-Ibáñez, N., Zolfaghari, Y., Amezcua-Prieto, C., & Khan, K. S. (2021). Physician-
patient language discordance and poor health outcomes: A systematic scoping
review. Frontiers in Public Health, 9, 629041.
https://doi.org/10.3389/fpubh.2021.629041
Carethers, J. M., & Doubeni, C. A. (2020). Causes of socioeconomic disparities in
colorectal cancer and intervention framework and strategies. Gastroenterology,
158(2), 354–367. https://doi.org/10.1053/j.gastro.2019.10.029
Cartwright, K., Leekity, S., Sheche, J., Kanda, D., Kosich, M., Rodman, J., Gonya, M.,
Kelly, K., Edwardson, N., Pankratz, V. S., & Mishra, S. I. (2022). Health literacy,
health numeracy, and cancer screening patterns in the Zuni Pueblo: Insights from
213
and limitations of “standard” questions. Journal of Cancer Education, 1–11.
Advance online publication. https://doi.org/10.1007/s13187-022-02227-y
Castañeda, S. F., Gallo, L. C., Nodora, J., Talavera, G. A., Penedo, F. J., Evenson, K. R.,
Lopez-Gurrola, M., Smoller-Wassertheil, S., Sanchez-Johnsen, L., Gonzalez, P.,
Perreira, K. M., & Gupta, S. (2019). Colorectal cancer screening among
Hispanics/Latinos in the HCHS/SOL sociocultural ancillary study. Preventive
Medicine Reports, 15, 100947. https://doi.org/10.1016/j.pmedr.2019.100947
Centers for Disease Control and Prevention (2021). Behavioral risk factor surveillance
system overview: BRFSS 2020. Retrieved from https://www.cdc.gov/brfss/.
Chang, L. C., Shun, C. T., Hsu, W. F., Tu, C. H., Tsai, P. Y., Lin, B. R., Liang, J. T., Wu,
M. S., & Chiu, H. M. (2017). Fecal immunochemical test detects sessile serrated
adenomas and polyps with a low level of sensitivity. Clinical Gastroenterology
and Hepatology, 15(6), 872–879.e1. https://doi.org/10.1016/j.cgh.2016.07.029
Charrois T. L. (2015). Systematic reviews: what do you need to know to get started? The
Canadian Journal of Hospital Pharmacy, 68(2), 144–148.
https://doi.org/10.4212/cjhp.v68i2.1440
Chatterji, P., & Li, Y. (2021). Effects of the COVID-19 pandemic on outpatient providers
in the United States. Medical Care, 59(1), 58–61.
https://doi.org/10.1097/MLR.0000000000001448
Christy, S. M., Mosher, C. E., & Rawl, S. M. (2014). Integrating men’s health and
masculinity theories to explain colorectal cancer screening behavior. American Journal of
Men’s Health, 8(1), 54–65. https://doi.org/10.1177/1557988313492171
214
Cook, M. B., Dawsey, S. M., Freedman, N. D., Inskip, P. D., Wichner, S. M., Quraishi, S.
M., Devesa, S. S., & McGlynn, K. A. (2009). Sex disparities in cancer incidence
by period and age. Cancer Epidemiology, Biomarkers & Prevention, 18(4), 1174–
1182. https://doi.org/10.1158/1055-9965.EPI-08-1118
Costas-Muñiz, R., Jandorf, L., Philip, E., Cohen, N., Villagra, C., Sriphanlop, P.,
Schofield, E., & DuHamel, K. (2016). Examining the Impact of Latino Nativity,
Migration, and Acculturation Factors on Colonoscopy Screening. Journal of
Community Health, 41(5), 903–909. https://doi.org/10.1007/s10900-016-0168-8
Cotter, G., DuHamel, K., Schofield, E., & Jandorf, L. (2019). Processes of change for
colonoscopy: Limited support for use among navigated Latinos. Journal of Racial
and Ethnic Health Disparities, 6(2), 327–334. https://doi.org/10.1007/s40615-
018-0528-4
Coughlin, S. S., Vernon, M., Hatzigeorgiou, C., & George, V. (2020). Health literacy,
social determinants of health, and disease prevention and control. Journal of
Environment and Health Sciences, 6(1), 3061. Retrieved from
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7889072/
Crews, J. E., Chou, C. F., Zhang, X., Zack, M. M., & Saaddine, J. B. (2014). Health-
related quality of life among people aged ≥65 years with self-reported visual
impairment: findings from the 2006-2010 behavioral risk factor surveillance
system. Ophthalmic Epidemiology, 21(5), 287–296.
https://doi.org/10.3109/09286586.2014.926556
Cross, A. J., Wooldrage, K., Robbins, E. C., Kralj-Hans, I., MacRae, E., Piggott, C.,
215
Stenson, I., Prendergast, A., Patel, B., Pack, K., Howe, R., Swart, N., Snowball,
J., Duffy, S. W., Morris, S., von Wagner, C., Halloran, S. P., & Atkin, W. S.
(2019). Faecal immunochemical tests (FIT) versus colonoscopy for surveillance
after screening and polypectomy: a diagnostic accuracy and cost-effectiveness
study. Gut, 68(9), 1642–1652. https://doi.org/10.1136/gutjnl-2018-317297
Cross, A. J., Wooldrage, K., Robbins, E. C., Pack, K., Brown, J. P., Hamilton, W.,
Thompson, M. R., Flashman, K. G., Halligan, S., Thomas-Gibson, S., Vance, M.,
Saunders, B. P., & Atkin, W. (2019). Whole-colon investigation vs. flexible
sigmoidoscopy for suspected colorectal cancer based on presenting symptoms and
signs: a multicentre cohort study. British Journal of Cancer, 120(2), 154–164.
https://doi.org/10.1038/s41416-018-0335-z
De Jesus, M., Rodrigue, C. M., Rahmani, S., & Balamou, C. (2021). Addressing cancer
screening inequities by promoting cancer prevention knowledge, awareness, self-
efficacy, and screening uptake among low-income and illiterate immigrant
women in France. International Journal of Public Health, 66, 1604055.
https://doi.org/10.3389/ijph.2021.1604055
Diamond, L., Izquierdo, K., Canfield, D., Matsoukas, K., & Gany, F. (2019). A
systematic review of the impact of patient-physician non-English language
concordance on quality of care and outcomes. Journal of General Internal
Medicine, 34(8), 1591–1606. https://doi.org/10.1007/s11606-019-04847-5
Diaz, J. A., Roberts, M. B., Clarke, J. G., Simmons, E. M., Goldman, R. E., & Rakowski,
W. (2013). Colorectal cancer screening: language is a greater barrier for Latino
216
men than Latino women. Journal of Immigrant and Minority Health, 15(3), 472–
475. https://doi.org/10.1007/s10903-012-9667-6
Douaiher, J., Ravipati, A., Grams, B., Chowdhury, S., Alatise, O., & Are, C. (2017).
Colorectal cancer-global burden, trends, and geographical variations. Journal of
Surgical Oncology, 115(5), 619–630. https://doi.org/10.1002/jso.24578
DuHamel, K. N., Schofield, E. A., Villagra, C., Sriphanlop, P., Itzkowitz, S. H., Cotter,
G., Cohen, N., Erwin, D. O., Winkel, G., Thompson, H. S., Zauber, A. G., &
Jandorf, L. H. (2020). Promoting colonoscopy screening among low-income
Latinos at average risk of colorectal cancer: A randomized clinical
trial. Cancer, 126(4), 782–791. https://doi.org/10.1002/cncr.32541
Edwardson, N., Cartwright, K., Sheche, J., Pankratz, V. S., Kosich, M., Kanda, D.,
Leekity, S., & Mishra, S. I. (2023). Colorectal cancer screening among adults in
Zuni Pueblo: Factors associated with FOBT and colonoscopy utilization. Journal
of Community Health, 1–11. https://doi.org/10.1007/s10900-023-01196-7
Ellis, L., Abrahão, R., McKinley, M., Yang, J., Somsouk, M., Marchand, L. L., Cheng, I.,
Gomez, S. L., & Shariff-Marco, S. (2018). Colorectal cancer incidence trends by
age, stage, and racial/ethnic group in California, 1990-2014. Cancer
Epidemiology, Biomarkers & Prevention, 27(9), 1011–1018.
https://doi.org/10.1158/1055-9965.EPI-18-0030
Ellis, E. M., Erwin, D. O., Jandorf, L., Saad-Harfouche, F., Sriphanlop, P., Clark, N.,
Dauphin, C., Johnson, D., Klasko-Foster, L. B., Martinez, C., Sly, J., White, D.,
Winkel, G., & Kiviniemi, M. T. (2018). Designing a randomized controlled trial
217
to evaluate a community-based narrative intervention for improving colorectal
cancer screening for African Americans. Contemporary Clinical Trials, 65, 8–18.
https://doi.org/10.1016/j.cct.2017.11.019
Elsafi, S. H., Alqahtani, N. I., Zakary, N. Y., & Al Zahrani, E. M. (2015). The sensitivity,
specificity, predictive values, and likelihood ratios of fecal occult blood test for
the detection of colorectal cancer in hospital settings. Clinical and Experimental
Gastroenterology, 8, 279–284. https://doi.org/10.2147/CEG.S86419
Fang, C. Y., & Ragin, C. C. (2020). Addressing disparities in cancer screening among
U.S. immigrants: Progress and opportunities. Cancer Prevention Research
(Philadelphia, Pa.), 13(3), 253–260. https://doi.org/10.1158/1940-6207.CAPR-
19-0249
Fekadu, G., Bekele, F., Tolossa, T., Fetensa, G., Turi, E., Getachew, M., Abdisa, E.,
Assefa, L., Afeta, M., Demisew, W., Dugassa, D., Diriba, D. C., & Labata, B. G.
(2021). Impact of COVID-19 pandemic on chronic diseases care follow-up and
current perspectives in low resource settings: a narrative review. International
Journal of Physiology, Pathophysiology and Pharmacology, 13(3), 86–93.
Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8310882/
Ferrari, A., Neefs, I., Hoeck, S., Peeters, M., & Van Hal, G. (2021). Towards novel non-
invasive colorectal cancer screening methods: A comprehensive
review. Cancers, 13(8), 1820. https://doi.org/10.3390/cancers13081820
Feng, G. C., Lin, Z., Ou, W., Su, X., & Yan, Q. (2021). A model-based meta-analysis of
willingness to participate in cancer screening. International Journal of
218
Environmental Research and Public Health, 18(5), 2580.
https://doi.org/10.3390/ijerph18052580
Fernández, L. E., & Morales, A. (2007). Language and use of cancer screening services
among border and non-border Hispanic Texas women. Ethnicity & Health, 12(3),
245–263. https://doi.org/10.1080/13557850701235150
Fernandez, A., Schillinger, D., Warton, E. M., Adler, N., Moffet, H. H., Schenker, Y.,
Salgado, M. V., Ahmed, A., & Karter, A. J. (2011). Language barriers, physician-
patient language concordance, and glycemic control among insured Latinos with
diabetes: The Diabetes Study of Northern California (DISTANCE). Journal of
General Internal Medicine, 26(2), 170–176. https://doi.org/10.1007/s11606-010-
1507-6
Firouzbakht, M., Hajian-Tilaki, K., & Bakhtiari, A. (2021). Comparison of competitive
cognitive models in explanation of women breast cancer screening behaviours
using structural equation modelling: Health belief model and theory of reasoned
action. European Journal of Cancer Care, 30(1), e13328.
https://doi.org/10.1111/ecc.13328
Fischer, A., Conigliaro, J., Allicock, S., & Kim, E. J. (2021). Examination of social
determinants of health among patients with limited English proficiency. BMC
Research Notes, 14(1), 299. https://doi.org/10.1186/s13104-021-05720-7
Floríndez, L. I., Floríndez, D. C., Como, D. H., Secola, R., & Duker, L. (2020). Differing
interpretations of health care encounters: A qualitative study of non-Latinx health
care providers’ perceptions of Latinx patient behaviors. PloS One, 15(8),
219
e0236706. https://doi.org/10.1371/journal.pone.0236706
Foiles Sifuentes, A. M., Robledo Cornejo, M., Li, N. C., Castaneda-Avila, M. A., Tjia, J.,
& Lapane, K. L. (2020). The role of limited English proficiency and access to
health insurance and health care in the Affordable Care Act Era. Health
Equity, 4(1), 509–517. https://doi.org/10.1089/heq.2020.0057
Francis, L., & Silvers, A. (2016). Perspectives on the meaning of “disability”. AMA
Journal of Ethics, 18(10), 1025–1033.
https://doi.org/10.1001/journalofethics.2016.18.10.pfor2-1610
Freeman, H. P., & Rodriguez, R. L. (2011). History and principles of patient
navigation. Cancer, 117(15 Suppl), 3539–3542.
https://doi.org/10.1002/cncr.26262
Friedemann-Sánchez, G., Griffin, J. M., & Partin, M. R. (2007). Gender differences in
colorectal cancer screening barriers and information needs. Health
Expectations, 10(2), 148–160. https://doi.org/10.1111/j.1369-7625.2006.00430.x
Gamble, S., Mawokomatanda, T., Xu, F., Chowdhury, P. P., Pierannunzi, C., Flegel, D.,
Garvin, W., & Town, M. (2017). Surveillance for certain health behaviors and
conditions among states and selected local areas - Behavioral Risk Factor
Surveillance System, United States, 2013 and 2014. Morbidity and Mortality
Weekly Report, 66(16), 1–144. https://doi.org/10.15585/mmwr.ss6616a1
Garcia, S., Pruitt, S. L., Singal, A. G., & Murphy, C. C. (2018). Colorectal cancer
incidence among Hispanics and non-Hispanic Whites in the United States. Cancer
Causes & Control, 29(11), 1039–1046. https://doi.org/10.1007/s10552-018-1077-
220
1
Gast, J., Peak, T., & Hunt, A. (2017). Latino health behavior: An exploratory analysis of
health risk and health protective factors in a community sample. American
Journal of Lifestyle Medicine, 14(1), 97–106.
https://doi.org/10.1177/1559827617716613
Genoff, M. C., Zaballa, A., Gany, F., Gonzalez, J., Ramirez, J., Jewell, S. T., & Diamond,
L. C. (2016). Navigating language barriers: A systematic review of patient
navigators’ impact on cancer screening for limited English proficient
patients. Journal of General Internal Medicine, 31(4), 426–434.
https://doi.org/10.1007/s11606-015-3572-3
Ghai, N. R., Jensen, C. D., Corley, D. A., Doubeni, C. A., Schottinger, J. E., Zauber, A.
G., Lee, A. T., Contreras, R., Levin, T. R., Lee, J. K., & Quinn, V. P. (2018).
Colorectal cancer screening participation among Asian Americans overall and
subgroups in an integrated health care setting with organized screening. Clinical
and Translational Gastroenterology, 9(9), 186. https://doi.org/10.1038/s41424-
018-0051-2
Goebel, M., Singal, A. G., Nodora, J., Castañeda, S. F., Martinez, E., Doubeni, C.,
Laiyemo, A., & Gupta, S. (2015). How can we boost colorectal and hepatocellular
cancer screening among underserved populations? Current Gastroenterology
Reports, 17(6), 22. https://doi.org/10.1007/s11894-015-0445-1
Gonzalez, P., Castaneda, S. F., Mills, P. J., Talavera, G. A., Elder, J. P., & Gallo, L. C.
(2012). Determinants of breast, cervical and colorectal cancer screening
221
adherence in Mexican-American women. Journal of Community Health, 37(2),
421–433. https://doi.org/10.1007/s10900-011-9459-2
Griffey, R. T., Shin, N., Jones, S., Aginam, N., Gross, M., Kinsella, Y., Williams, J. A.,
Carpenter, C. R., Goodman, M., & Kaphingst, K. A. (2015). The impact of teach-
back on comprehension of discharge instructions and satisfaction among
emergency patients with limited health literacy: A randomized, controlled
study. Journal of Communication in Healthcare, 8(1), 10–21.
https://doi.org/10.1179/1753807615Y.0000000001
Griffin, J. M., Burgess, D., Vernon, S. W., Friedemann-Sanchez, G., Powell, A., van
Ryn, M., Halek, K., Noorbaloochi, S., Grill, J., Bloomfield, H., & Partin, M.
(2009). Are gender differences in colorectal cancer screening rates due to
differences in self-reporting? Preventive Medicine, 49(5), 436–441.
https://doi.org/10.1016/j.ypmed.2009.09.013
Gunn, C. M., Paasche-Orlow, M. K., Bak, S., Wang, N., Pamphile, J., Nelson, K.,
Morton, S., & Battaglia, T. A. (2020). Health literacy, language, and cancer-
related needs in the first 6 months after a breast cancer diagnosis. JCO Oncology
Practice, 16(8), e741–e750. https://doi.org/10.1200/JOP.19.00526
Gupta, S., Sussman, D. A., Doubeni, C. A., Anderson, D. S., Day, L., Deshpande, A. R.,
Elmunzer, B. J., Laiyemo, A. O., Mendez, J., Somsouk, M., Allison, J., Bhuket,
T., Geng, Z., Green, B. B., Itzkowitz, S. H., & Martinez, M. E. (2014). Challenges
and possible solutions to colorectal cancer screening for the underserved. Journal
of the National Cancer Institute, 106(4), dju032.
222
https://doi.org/10.1093/jnci/dju032
Hall, I. J., Rim, S. H., & Dasari, S. (2022). Preventive care use among Hispanic adults
with limited comfort speaking English: An analysis of the Medical Expenditure
Panel Survey data. Preventive Medicine, 159, 107042.
https://doi.org/10.1016/j.ypmed.2022.107042
Heintzman, J. D., Ezekiel-Herrera, D. N., Quiñones, A. R., Lucas, J. A., Carroll, J. E.,
Gielbultowicz, S. H., Cottrell, E. C., & Marino, M. (2022). Disparities in
colorectal cancer screening in Latinos and non-Hispanic Whites. American
Journal of Preventive Medicine, 62(2), 203–210.
https://doi.org/10.1016/j.amepre.2021.07.009
Heller, M. (Ed.). (2007). Bilingualism: A social approach. Springer. Retrieved from
https://link.springer.com/chapter/10.1057/9780230596047_1
Hill, J. C., Njeru, J. W., Wieland, M. L., Rutten, L. J., Jacobson, D. J., Jenkins, G. D., &
Fan, C. (2021). Disparities among limited English proficient patients in colon
cancer screening with multi-target DNA Stool Test. Journal of Immigrant and
Minority Health, https://doi.org/10.1007/s10903-021-01214-7.
Hochbaum, G. M. (1958). Public participation in medical screening programs: A socio-
psychological study (No. 572). US Department of Health, Education, and
Welfare, Public Health Service, Bureau of State Services, Division of Special
Health Services, Tuberculosis Program.
Horshauge, P. M., Gabel, P., Larsen, M. B., Kirkegaard, P., Edwards, A., & Andersen, B.
(2020). The association between health literacy and colorectal cancer screening
223
uptake in a publicly funded screening program in Denmark: Cross-sectional
study. Preventive Medicine Reports, 19, 101132.
https://doi.org/10.1016/j.pmedr.2020.101132
Hsueh, L., Hirsh, A. T., Maupomé, G., & Stewart, J. C. (2021). Patient–provider
language concordance and health outcomes: A systematic review, evidence map,
and research agenda. Medical Care Research and Review, 78(1), 3–23.
https://doi.org/10.1177/1077558719860708
Huang, J., Chen, G., Liu, H., Zhang, Y., Tang, R., Huang, Q., Fu, K., Peng, X., & Xiao,
S. (2020). Surgery improves the prognosis of colon mucinous adenocarcinoma
with liver metastases: a SEER-based study. BMC Cancer, 20(1), 908.
https://doi.org/10.1186/s12885-020-07400-4
Hviding, K., Juvet, L. K., Vines, D., & Fretheim, A. (2008). Colorectal cancer screening
effect on mortality and incidence rate of colorectal cancer: Overview of
documentation and international recommendations. Knowledge Centre for the
Health Services at The Norwegian Institute of Public Health (NIPH). Retrieved
from https://www.ncbi.nlm.nih.gov/books/NBK464909/
Iachan, R., Pierannunzi, C., Healey, K., Greenlund, K. J., & Town, M. (2016). National
weighting of data from the Behavioral Risk Factor Surveillance System
(BRFSS). BMC Medical Research Methodology, 16(1), 155.
https://doi.org/10.1186/s12874-016-0255-7
Ioannou, S., Sutherland, K., Sussman, D. A., & Deshpande, A. R. (2021). Increasing
uptake of colon cancer screening in a medically underserved population with the
224
addition of blood-based testing. BMC Cancer, 21(1), 966.
https://doi.org/10.1186/s12885-021-08678-8
Issa, I. A., & Noureddine, M. (2017). Colorectal cancer screening: An updated review of
the available options. World Journal of Gastroenterology, 23(28), 5086–5096.
https://doi.org/10.3748/wjg.v23.i28.5086
Jackson, C. S., Oman, M., Patel, A. M., & Vega, K. J. (2016). Health disparities in
colorectal cancer among racial and ethnic minorities in the United States. Journal
of Gastrointestinal Oncology, 7(Suppl 1), S32–S43.
https://doi.org/10.3978/j.issn.2078-6891.2015.039
Jacobson, H. E., Hund, L., & Soto Mas, F. (2016). Predictors of English health literacy
among U.S. Hispanic immigrants: The importance of language, bilingualism and
sociolinguistic environment. Literacy & Numeracy Studies, 24(1), 43–64.
https://doi.org/10.5130/lns.v24i1.4900
Jandorf, L., Ellison, J., Villagra, C., Winkel, G., Varela, A., Quintero-Canetti, Z.,
Castillo, A., Thélémaque, L., King, S., & Duhamel, K. (2010). Understanding the
barriers and facilitators of colorectal cancer screening among low-income
immigrant hispanics. Journal of Immigrant and Minority Health, 12(4), 462–469.
https://doi.org/10.1007/s10903-009-9274-3
Janz, N. K., & Becker, M. H. (1984). The Health Belief Model: A decade later. Health
Education Quarterly, 11(1), 1–47. https://doi.org/10.1177/109019818401100101
Jin, S. W., Yun Lee, H., & Lee, J. (2019). Analyzing factors enabling colorectal cancer
screening adherence in Korean Americans using the Andersen’s Behavioral
225
Model of Health Services Utilization. Journal of Psychosocial Oncology, 37(6),
729–745. https://doi.org/10.1080/07347332.2019.1608347
Johnson-Kozlow, M., Roussos, S., Rovniak, L., & Hovell, M. (2009). Colorectal cancer
test use among Californians of Mexican origin: influence of language
barriers. Ethnicity & Disease, 19(3), 315–322. Retrieved from
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2757758/pdf/nihms124281.pdf
Jones, C. L., Jensen, J. D., Scherr, C. L., Brown, N. R., Christy, K., & Weaver, J. (2015).
The Health Belief Model as an explanatory framework in communication
research: exploring parallel, serial, and moderated mediation. Health
Communication, 30(6), 566–576. https://doi.org/10.1080/10410236.2013.873363
Joachim, C., Macni, J., Drame, M., Pomier, A., Escarmant, P., Veronique-Baudin, J., &
Vinh-Hung, V. (2019). Overall survival of colorectal cancer by stage at diagnosis:
Data from the Martinique Cancer Registry. Medicine, 98(35), e16941.
https://doi.org/10.1097/MD.0000000000016941
Jung Y. S. (2020). Strategies to increase the participation rate of colorectal cancer
screening. Gut and Liver, 14(3), 277–278. https://doi.org/10.5009/gnl20110
Juon, H. S., Guo, J., Kim, J., & Lee, S. (2018). Predictors of colorectal cancer knowledge
and screening among Asian Americans aged 50-75 years old. Journal of Racial
and Ethnic Health Disparities, 5(3), 545–552. https://doi.org/10.1007/s40615-
017-0398-1
Kadakuntla, A., Wang, T., Medgyesy, K., Rrapi, E., Litynski, J., Adynski, G., & Tadros,
M. (2021). Colorectal cancer screening in the COVID-19 era. World Journal of
226
Gastrointestinal Oncology, 13(4), 238–251.
https://doi.org/10.4251/wjgo.v13.i4.238
Kaiser Family Foundation (2022). Medicaid State Fact Sheets. Percent of people covered
by Medicaid/CHIP, 2022. Retrieved from https://www.kff.org/. Date accessed:
March 10, 2023
Kaliyadan, F., & Kulkarni, V. (2019). Types of variables, descriptive statistics, and
sample size. Indian Dermatology Online Journal, 10(1), 82–86.
https://doi.org/10.4103/idoj.IDOJ_468_18
Karimy, M., Bastami, F., Sharifat, R., Heydarabadi, A. B., Hatamzadeh, N., Pakpour, A.
H., Cheraghian, B., Zamani-Alavijeh, F., Jasemzadeh, M., & Araban, M. (2021).
Factors related to preventive COVID-19 behaviors using health belief model
among general population: a cross-sectional study in Iran. BMC public
health, 21(1), 1934. https://doi.org/10.1186/s12889-021-11983-3
Karl, J. A., Fischer, R., Druică, E., Musso, F., & Stan, A. (2022). Testing the
effectiveness of the Health Belief Model in predicting preventive behavior during
the COVID-19 pandemic: The case of Romania and Italy. Frontiers in
Psychology, 12, 627575. https://doi.org/10.3389/fpsyg.2021.627575
Karliner, L. S., Napoles-Springer, A. M., Schillinger, D., Bibbins-Domingo, K., & Pérez-
Stable, E. J. (2008). Identification of limited English proficient patients in clinical
care. Journal of General Internal Medicine, 23(10), 1555–1560.
https://doi.org/10.1007/s11606-008-0693-y
Kastenberg, D., Bertiger, G., & Brogadir, S. (2018). Bowel preparation quality scales for
227
colonoscopy. World Journal of Gastroenterology, 24(26), 2833–2843.
https://doi.org/10.3748/wjg.v24.i26.2833
Kayaalp M. (2018). Modes of De-identification. AMIA ... Annual Symposium
proceedings. AMIA Symposium, 2017, 1044–1050. Retrieved from
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5977668/pdf/2730548.pdf
Kelly, D. M., Estaquio, C., Léon, C., Arwidson, P., & Nabi, H. (2017). Temporal trend in
socioeconomic inequalities in the uptake of cancer screening programmes in
France between 2005 and 2010: results from the Cancer Barometer surveys. BMJ
Open, 7(12), e016941. https://doi.org/10.1136/bmjopen-2017-016941
Kim, B., Lairson, D. R., Chung, T. H., Kim, J., & Shokar, N. K. (2017). Budget compact
analysis of Against Colorectal Cancer In Our Neighborhoods (ACCION): A
successful community-based Colorectal Cancer Screening Program for a
Medically Underserved Minority Population. Value in Health, 20(6), 809–818.
https://doi.org/10.1016/j.jval.2016.11.025
Kim, T. K., & Park, J. H. (2019). More about the basic assumptions of t-test: normality
and sample size. Korean Journal of Anesthesiology, 72(4), 331–335.
https://doi.org/10.4097/kja.d.18.00292
Kim, K., Quinn, M., & Lam, H. (2018). Promoting colorectal cancer screening in foreign-
born Chinese-American women: Does racial/ethnic and language concordance
matter? Journal of Racial and Ethnic Health Disparities, 5(6), 1346–1353.
https://doi.org/10.1007/s40615-018-0484-z
Kim, S. Y., Kim, H. S., & Park, H. J. (2019). Adverse events related to colonoscopy:
228
Global trends and future challenges. World Journal of Gastroenterology, 25(2),
190–204. https://doi.org/10.3748/wjg.v25.i2.190
Kirscht, J. P. (1974). The health belief model and illness behavior. Health Education
Monographs, 2(4), 387–408. https://doi.org/10.1177/109019817400200406
Kuipers, E. J., Grady, W. M., Lieberman, D., Seufferlein, T., Sung, J. J., Boelens, P. G.,
van de Velde, C. J., & Watanabe, T. (2015). Colorectal cancer. Nature Reviews.
Disease Primers, 1, 15065. https://doi.org/10.1038/nrdp.2015.65
Ladabaum, U., Mannalithara, A., Desai, M., Sehgal, M., & Singh, G. (2021). Age-
specific rates and time-courses of gastrointestinal and nongastrointestinal
complications associated with screening/surveillance colonoscopy. The American
Journal of Gastroenterology, 116(12), 2430–2445.
https://doi.org/10.14309/ajg.0000000000001531
Lau, J., Lim, T. Z., Jianlin Wong, G., & Tan, K. K. (2020). The health belief model and
colorectal cancer screening in the general population: A systematic
review. Preventive Medicine Reports, 20, 101223.
https://doi.org/10.1016/j.pmedr.2020.101223
Lee, H. Y., Lundquist, M., Ju, E., Luo, X., & Townsend, A. (2011). Colorectal cancer
screening disparities in Asian Americans and Pacific Islanders: which groups are
most vulnerable? Ethnicity & Health, 16(6), 501–518.
https://doi.org/10.1080/13557858.2011.575219
Lesko, C. R., Ackerman, B., Webster-Clark, M., & Edwards, J. K. (2020). Target
validity: Bringing treatment of external validity in line with internal
229
validity. Current Epidemiology Reports, 7(3), 117–124.
https://doi.org/10.1007/s40471-020-00239-0
Lewin, K., Dembo, T., Festinger, L., & Sears, P. S. (1944). Level of aspiration. In J. Hunt
(Ed.), Personality and the Behavior Disorders (pp. 333–378). Somerset, NJ:
Ronald Press.
Li, D. (2018). Recent advances in colorectal cancer screening. Chronic Diseases and
Translational Medicine, 4(3), 139–147.
https://doi.org/10.1016/j.cdtm.2018.08.004
Lin, D., & Cui, J. (2021). Transport and Mobility Needs for an Ageing Society from a
Policy Perspective: Review and Implications. International Journal of
Environmental Research and Public Health, 18(22), 11802.
https://doi.org/10.3390/ijerph182211802
Lin, O. S. (2014). Performing colonoscopy in elderly and very elderly patients: Risks,
costs, and benefits. World Journal of Gastrointestinal Endoscopy, 6(6), 220–226.
https://doi.org/10.4253/wjge.v6.i6.220
Lin, T. Y., Chuang, S. T., Huang, S. F., Hsu, H. P., Lu, L. T., & Guo, J. L. (2019).
Likelihood of a fecal occult blood test uptake among older adults: comparisons
between health professionals and healthcare volunteers based on the health belief
model. BMC Geriatrics, 19(1), 51. https://doi.org/10.1186/s12877-019-1067-5
Lin, J. S., Perdue, L. A., Henrikson, N. B., Bean, S. I., & Blasi, P. R. (2021). Screening
for Colorectal Cancer: An Evidence Update for the U.S. Preventive Services Task
Force. Agency for Healthcare Research and Quality (US).
230
doi:10.1001/jama.2021.4417.
Liu, Y., Croft, J. B., Wheaton, A. G., Perry, G. S., Chapman, D. P., Strine, T. W.,
McKnight-Eily, L. R., & Presley-Cantrell, L. (2013). Association between
perceived insufficient sleep, frequent mental distress, obesity, and chronic
diseases among US adults, 2009 Behavioral Risk Factor Surveillance
System. BMC Public Health, 13, 84. https://doi.org/10.1186/1471-2458-13-84
Loomans-Kropp, H. A., & Umar, A. (2019). Cancer prevention and screening: the next
step in the era of precision medicine. NPJ Precision Oncology, 3, 3.
https://doi.org/10.1038/s41698-018-0075-9
Ma, G. X., Lee, M., Zhu, L., Tan, Y., Do, P., Ma, X., Tran, T., & Johnson, C. K. (2020).
Colorectal cancer-related knowledge, acculturation, and healthy lifestyle
behaviors among low-income Vietnamese Americans in the Greater Philadelphia
Metropolitan Area. Journal of Community Health, 45(6), 1178–1186.
https://doi.org/10.1007/s10900-020-00931-8
Ma, G. X., Zhu, L., Lin, T. R., Tan, Y., & Do, P. (2021). Multilevel pathways of
colorectal cancer screening among low-income Vietnamese Americans: A
structural equation modeling analysis. Cancer Control, 28, 10732748211011077.
https://doi.org/10.1177/10732748211011077
Ma, G. X., Zhang, G. Y., Zhai, S., Ma, X., Tan, Y., Shive, S. E., & Wang, M. Q. (2015).
Hepatitis B screening among Chinese Americans: A structural equation modeling
analysis. BMC Infectious Diseases, 15, 120. https://doi.org/10.1186/s12879-015-
0854-7
231
Manchikanti, L., Helm Ii, S., Benyamin, R. M., & Hirsch, J. A. (2017). A critical analysis
of Obamacare: Affordable Care or insurance for many and coverage for
few? Pain Physician, 20(3), 111–138. Retrieved from
https://www.painphysicianjournal.com/
Manne, S., Steinberg, M. B., Delnevo, C., Ulpe, R., & Sorice, K. (2015). Colorectal
cancer screening among foreign-born South Asians in the Metropolitan New
York/New Jersey Region. Journal of Community Health, 40(6), 1075–1083.
https://doi.org/10.1007/s10900-015-0053-x
Mantwill, S., & Schulz, P. J. (2017). Does acculturation narrow the health literacy gap
between immigrants and non-immigrants-An explorative study. Patient Education
and Counseling, 100(4), 760–767. https://doi.org/10.1016/j.pec.2016.10.021
Matthay, E. C., & Glymour, M. M. (2020). A graphical catalog of threats to validity:
Linking social science with epidemiology. Epidemiology (Cambridge,
Mass.), 31(3), 376–384. https://doi.org/10.1097/EDE.0000000000001161
Maslekar, S., Hughes, M., Gardiner, A., Monson, J. R., & Duthie, G. S. (2010). Patient
satisfaction with lower gastrointestinal endoscopy: doctors, nurse and nonmedical
endoscopist. Colorectal Disease, 12(10), 1033–1038.
https://doi.org/10.1111/j.1463-1318.2009.01989.x
May, F. P., Yang, L., Corona, E., Glenn, B. A., & Bastani, R. (2020). Disparities in
colorectal cancer screening in the United States before and after implementation
of the Affordable Care Act. Clinical Gastroenterology and Hepatology, 18(8),
1796–1804.e2. https://doi.org/10.1016/j.cgh.2019.09.008
232
Mayhand, K. N., Handorf, E. A., Ortiz, A. G., Gonzalez, E. T., Devlin, A., Sorice, K. A.,
Esnaola, N., Fisher, S., & Lynch, S. M. (2021). Effect of neighborhood and
individual-level socioeconomic factors on colorectal cancer screening
adherence. International Journal of Environmental Research and Public
Health, 18(9), 4398. https://doi.org/10.3390/ijerph18094398
McKinney, S. Y., & Palmer, R. C. (2014). The influence of gender on colorectal cancer
knowledge, screening intention, perceived risk and worry among African
Americans in South Florida. Journal of Community Health, 39(2), 230–238.
https://doi.org/10.1007/s10900-013-9812-8
Meklin, J., SyrjÄnen, K., & Eskelinen, M. (2020). Fecal occult blood tests in colorectal
cancer screening: Systematic review and meta-analysis of traditional and new-
generation fecal immunochemical tests. Anticancer Research, 40(7), 3591–3604.
https://doi.org/10.21873/anticanres.14349
Miller, D. C., Gelberg, L., Kwan, L., Stepanian, S., Fink, A., Andersen, R. M., & Litwin,
M. S. (2008). Racial disparities in access to care for men in a public assistance
program for prostate cancer. Journal of Community Health, 33(5), 318–335.
https://doi.org/10.1007/s10900-008-9105-9
Mishra, P., Pandey, C. M., Singh, U., Gupta, A., Sahu, C., & Keshri, A. (2019).
Descriptive statistics and normality tests for statistical data. Annals of Cardiac
Anaesthesia, 22(1), 67–72. https://doi.org/10.4103/aca.ACA_157_18
Mojica, C. M., Glenn, B. A., Chang, C., & Bastani, R. (2015). The relationship between
neighborhood immigrant composition, limited English proficiency, and late-stage
233
colorectal cancer diagnosis in California. BioMed Research International, 2015,
460181. https://doi.org/10.1155/2015/460181
Mokdad, A. H., Stroup, D. F., Giles, W. H., & Behavioral Risk Factor Surveillance Team
(2003). Public health surveillance for behavioral risk factors in a changing
environment. Recommendations from the Behavioral Risk Factor Surveillance
Team. MMWR. Recommendations and reports, 52(RR-9), 1–12. Retrieved from
https://www.cdc.gov/mmwr/preview/mmwrhtml/rr5209a1.htm
Moore, C., Coates, E., Watson, A., de Heer, R., McLeod, A., & Prudhomme, A. (2022).
“It’s Important to Work with People that Look Like Me”: Black Patients’
Preferences for Patient-Provider Race Concordance. Journal of Racial and Ethnic
Health Disparities, 1–13. https://doi.org/10.1007/s40615-022-01435-y
Muller, C., Ihionkhan, E., Stoffel, E. M., & Kupfer, S. S. (2021). Disparities in Early-
Onset Colorectal Cancer. Cells, 10(5), 1018.
https://doi.org/10.3390/cells10051018
Murphy, G., Devesa, S. S., Cross, A. J., Inskip, P. D., McGlynn, K. A., & Cook, M. B.
(2011). Sex disparities in colorectal cancer incidence by anatomic subsite, race
and age. International Journal of Cancer, 128(7), 1668–1675.
https://doi.org/10.1002/ijc.25481
Muthukrishnan, M., Arnold, L. D., & James, A. S. (2019). Patients’ self-reported barriers
to colon cancer screening in federally qualified health center settings. Preventive
Medicine Reports, 15, 100896. https://doi.org/10.1016/j.pmedr.2019.100896
Office for Civil Rights (OCR), Office of the Secretary, HHS (2016). Nondiscrimination
234
in health programs and activities. Fine rule. Federal Register,81(96),31375-
31473. Retrieved from https://www.federalregister.gov/
Office of Disease Prevention and Health Promotion (2021). Health Literacy in Healthy
People 2030. United States Department of Health and Human Services. Retrieved
from: https://www.cdc.gov/healthliteracy/learn/index.html
Office of Minority Health (OMH) at the United States Department of Health and Human
Services (HHS) (2021). Profile: Hispanic/Latino Americans. Retrieved from
https://minorityhealth.hhs.gov/
Ortega, P., & Prada, J. (2020). Words matter: Translanguaging in medical
communication skills training. Perspectives on Medical Education, 9(4), 251–
255. https://doi.org/10.1007/s40037-020-00595-z
Ortega, P., Shin, T. M., & Martínez, G. A. (2021). Rethinking the term “Limited English
Proficiency” to improve language-appropriate healthcare for all. Journal of
Immigrant and Minority Health, 1–7. Advance online publication.
https://doi.org/10.1007/s10903-021-01257-w
Ou, J. Y., Warner, E. L., Nam, G. E., Martel, L., Carbajal-Salisbury, S., Fuentes, V.,
Wetter, D. W., Kirchhoff, A. C., & Kepka, D. (2019). Colorectal cancer
knowledge and screening adherence among low-income Hispanic
employees. Health Education Research, 34(4), 400–414.
https://doi.org/10.1093/her/cyz013
Oyalowo, A., Forde, K. A., Lamanna, A., & Kochman, M. L. (2022). Effect of Patient-
Directed Messaging on Colorectal Cancer Screening: A Randomized Clinical
235
Trial. JAMA Network Open, 5(3), e224529.
https://doi.org/10.1001/jamanetworkopen.2022.4529
Patino, C. M., & Ferreira, J. C. (2018). Internal and external validity: can you apply
research study results to your patients? Jornal Brasileiro de Pneumologia, 44(3),
183. https://doi.org/10.1590/S1806-37562018000000164
Perisetti, A., Khan, H., George, N. E., Yendala, R., Rafiq, A., Blakely, S., Rasmussen, D.,
Villalpando, N., & Goyal, H. (2018). Colorectal cancer screening use among
insured adults: Is out-of-pocket cost a barrier to routine screening? World Journal
of Gastrointestinal Pharmacology and Therapeutics, 9(4), 31–38.
https://doi.org/10.4292/wjgpt.v9.i4.31
Petursdottir, A. I., & Carr, J. E. (2018). Applying the taxonomy of validity threats from
mainstream research design to single-case experiments in applied behavior
analysis. Behavior Analysis in Practice, 11(3), 228–240.
https://doi.org/10.1007/s40617-018-00294-6
Rahman, R., Schmaltz, C., Jackson, C. S., Simoes, E. J., Jackson-Thompson, J., & Ibdah,
J. A. (2015). Increased risk for colorectal cancer under age 50 in racial and ethnic
minorities living in the United States. Cancer Medicine, 4(12), 1863–1870.
https://doi.org/10.1002/cam4.560
Rakhshanderou, S., Maghsoudloo, M., Safari-Moradabadi, A., & Ghaffari, M. (2020).
Theoretically designed interventions for colorectal cancer prevention: a case of
the health belief model. BMC Medical Education, 20(1), 270.
https://doi.org/10.1186/s12909-020-02192-4
236
Ratnapradipa, K. L., Chen, K., Watanabe-Galloway, S., & Farazi, P. A. (2021). Cross-
sectional Study of Colorectal Cancer Screening Barriers in a Latino-Serving
Federally Qualified Health Center. Journal of Cancer Education,
10.1007/s13187-021-02097-w. Advance online publication.
https://doi.org/10.1007/s13187-021-02097-w
Rawla, P., Sunkara, T., & Barsouk, A. (2019). Epidemiology of colorectal cancer:
incidence, mortality, survival, and risk factors. Przeglad
Gastroenterologiczny, 14(2), 89–103. https://doi.org/10.5114/pg.2018.81072.
Redwood, D. G., Dinh, T. A., Kisiel, J. B., Borah, B. J., Moriarty, J. P., Provost, E. M.,
Sacco, F. D., Tiesinga, J. J., & Ahlquist, D. A. (2021). Cost-Effectiveness of
Multitarget Stool DNA Testing vs Colonoscopy or Fecal Immunochemical
Testing for Colorectal Cancer Screening in Alaska Native People. Mayo Clinic
Proceedings, 96(5), 1203–1217. https://doi.org/10.1016/j.mayocp.2020.07.035
Rex, D. K., Boland, C. R., Dominitz, J. A., Giardiello, F. M., Johnson, D. A., Kaltenbach,
T., Levin, T. R., Lieberman, D., & Robertson, D. J. (2017). Colorectal cancer
screening: recommendations for physicians and patients from the U.S. multi-
society task force on colorectal cancer. Gastroenterology, 153(1), 307–323.
https://doi.org/10.1053/j.gastro.2017.05.013
Riegert, M., Nandwani, M., Thul, B., Chiu, A. C., Mathews, S. C., Khashab, M. A., &
Kalloo, A. N. (2020). Experience of nurse practitioners performing colonoscopy
after endoscopic training in more than 1,000 patients. Endoscopy International
Open, 8(10), E1423–E1428. https://doi.org/10.1055/a-1221-4546
237
Robb K. A. (2021). The integrated screening action model (I-SAM): A theory-based
approach to inform intervention development. Preventive Medicine Reports, 23,
101427. https://doi.org/10.1016/j.pmedr.2021.101427
Rogers, C. R., Goodson, P., Dietz, L. R., & Okuyemi, K. S. (2018). Predictors of
intention to obtain colorectal cancer screening among African American men in a
state fair setting. American Journal of Men’s Health, 851–862.
https://doi.org/10.1177/1557988316647942
Rogers, C. R., Obidike, O. J., Wallington, S. F., Hussein, M., Mahamed, Z. A., &
Sampson, J. (2021). A qualitative study of barriers and enablers associated with
colorectal cancer screening among Somali men in Minnesota. Ethnicity &
Health, 26(2), 168–185. https://doi.org/10.1080/13557858.2018.1493438
Rogers, C. R., Perdue, D. G., Boucher, K., Korous, K. M., Brooks, E., Petersen, E.,
Inadomi, J. M., Tuuhetaufa, F., Levant, R. F., & Paskett, E. D. (2022).
Masculinity barriers to ever completing colorectal cancer screening among
American Indian/Alaska Native, Black, and White Men (Ages 45-
75). International Journal of Environmental Research and Public Health, 19(5),
3071. https://doi.org/10.3390/ijerph19053071
Rosenstock, I. M. (1974). Historical origins of the health belief model. Health Education
Monographs, 2(4), 328–335. https://doi.org/10.1177/109019817400200403
Rosenstock, I. M. (1966). Why people use health services. The Milbank Memorial Fund
Quarterly, 44, 94-127.
Rosenstock, I. M., Strecher, V. J., & Becker, M. H. (1988). Social learning theory and the
238
Health Belief Model. Health Education Quarterly, 15(2), 175–183.
https://doi.org/10.1177/109019818801500203
Rothman, K. J., Greenland, S., & Lash, T. L. (2008). Modern epidemiology (Vol. 3).
Philadelphia: Wolters Kluwer Health/Lippincott Williams & Wilkins.
Russell, C. J., & Bobko, P. (1992). Moderated regression analysis and Likert scales: too
coarse for comfort. The Journal of Applied Psychology, 77(3), 336–342.
https://doi.org/10.1037/0021-9010.77.3.336
Rutan, M. C., Sammon, J. D., Nguyen, D. D., Kilbridge, K. L., Herzog, P., & Trinh, Q.
D. (2021). The relationship between health literacy and nonrecommended cancer
screening. American Journal of Preventive Medicine, 60(2), e69–e72.
https://doi.org/10.1016/j.amepre.2020.08.018
Safiyeva, A. K., & Bayramov, N. Y. (2019). The importance of colonoscopy in the
treatment of colorectal polyps and colorectal cancer screening. Annali Italiani di
Chirurgia, 90, 311–317. Retrieved from
https://pubmed.ncbi.nlm.nih.gov/30946027/
Saha, S., & Beach, M. C. (2020). Impact of physician race on patient decision-making
and ratings of physicians: A randomized experiment using video
vignettes. Journal of General Internal Medicine, 35(4), 1084–1091.
https://doi.org/10.1007/s11606-020-05646-z
Salem, M. E., Puccini, A., Trufan, S. J., Sha, W., Kadakia, K. C., Hartley, M. L.,
Musselwhite, L. W., Symanowski, J. T., Hwang, J. J., & Raghavan, D. (2021).
Impact of sociodemographic disparities and insurance status on survival of
239
patients with early-onset colorectal cancer. The Oncologist, 26(10), e1730–e1741.
https://doi.org/10.1002/onco.13908
Savas, L. S., Vernon, S. W., Atkinson, J. S., & Fernández, M. E. (2015). Effect of
acculturation and access to care on colorectal cancer screening in low-income
Latinos. Journal of Immigrant and Minority Health, 17(3), 696–703.
https://doi.org/10.1007/s10903-014-0061-4
Schiaffino, M. K., Al-Amin, M., & Schumacher, J. R. (2014). Predictors of language
service availability in U.S. hospitals. International Journal of Health Policy and
Management, 3(5), 259–268. https://doi.org/10.15171/ijhpm.2014.95
Schillinger D. (2020). The intersections between social determinants of health, health
literacy, and health disparities. Studies in Health Technology and
Informatics, 269, 22–41. https://doi.org/10.3233/SHTI200020
Schoenborn, N. L., Boyd, C. M., & Pollack, C. E. (2022). Impact of the COVID-19
pandemic on cancer screening attitudes, intentions, and behaviors in older
adults. Journal of the American Geriatrics Society, 70(1), 67–69.
https://doi.org/10.1111/jgs.17449
Schulson, L. B., & Anderson, T. S. (2022). National estimates of professional interpreter
use in the ambulatory setting. Journal of General Internal Medicine, 37(2), 472–
474. https://doi.org/10.1007/s11606-020-06336-6
Sentell, T., & Braun, K. L. (2012). Low health literacy, limited English proficiency, and
health status in Asians, Latinos, and other racial/ethnic groups in
California. Journal of Health Communication, 17 Suppl 3(Suppl 3), 82–99.
240
https://doi.org/10.1080/10810730.2012.712621
Sentell, T., Braun, K. L., Davis, J., & Davis, T. (2013). Colorectal cancer screening: low
health literacy and limited English proficiency among Asians and Whites in
California. Journal of Health Communication, 18 Suppl 1(Suppl 1), 242–255.
https://doi.org/10.1080/10810730.2013.825669
Sentell, T. L., Tsoh, J. Y., Davis, T., Davis, J., & Braun, K. L. (2015). Low health
literacy and cancer screening among Chinese Americans in California: a cross-
sectional analysis. BMJ Open, 5(1), e006104. https://doi.org/10.1136/bmjopen-
2014-006104
Serdar, C. C., Cihan, M., Yücel, D., & Serdar, M. A. (2021). Sample size, power and
effect size revisited: simplified and practical approaches in pre-clinical, clinical
and laboratory studies. Biochemia Medica, 31(1), 010502.
https://doi.org/10.11613/BM.2021.010502
Shah, S. K., Narcisse, M. R., Hallgren, E., Felix, H. C., & McElfish, P. A. (2022).
Assessment of colorectal cancer screening disparities in U.S. men and women
using a demographically representative sample. Cancer Research
Communications, 2(6), 561–569. https://doi.org/10.1158/2767-9764.crc-22-0079
Shapiro, J. A., Soman, A. V., Berkowitz, Z., Fedewa, S. A., Sabatino, S. A., de Moor, J.
S., Clarke, T. C., Doria-Rose, V. P., Breslau, E. S., Jemal, A., & Nadel, M. R.
(2021). Screening for colorectal cancer in the United States: Correlates and time
trends by type of test. Cancer Epidemiology, Biomarkers & Prevention, 30(8),
1554–1565. https://doi.org/10.1158/1055-9965.EPI-20-1809
241
Sharifikia, I., Rohani, C., Estebsari, F., Matbouei, M., Salmani, F., & Hossein-Nejad, A.
(2019). Health Belief Model-based intervention on women’s knowledge and
perceived beliefs about warning signs of cancer. Asia-Pacific Journal of
Oncology Nursing, 6(4), 431–439. https://doi.org/10.4103/apjon.apjon_32_19
Shaukat, A., Kaalby, L., Baatrup, G., Kronborg, O., Duval, S., Shyne, M., Mandel, J. S.,
& Church, T. R. (2021). Effects of screening compliance on long-term reductions
in all-cause and colorectal cancer mortality. Clinical Gastroenterology and
Hepatology, 19(5), 967–975.e2. https://doi.org/10.1016/j.cgh.2020.06.019
Shen, M. J., Peterson, E. B., Costas-Muñiz, R., Hernandez, M. H., Jewell, S. T.,
Matsoukas, K., & Bylund, C. L. (2018). The effects of race and racial
concordance on patient-physician communication: A systematic review of the
literature. Journal of Racial and Ethnic Health Disparities, 5(1), 117–140.
https://doi.org/10.1007/s40615-017-0350-4
Shokar, N. K., Salinas, J., & Dwivedi, A. (2022). Mediators of screening uptake in a
colorectal cancer screening intervention among Hispanics. BMC Cancer, 22(1),
37. https://doi.org/10.1186/s12885-021-09092-w
Shokar, N. K., Byrd, T., Salaiz, R., Flores, S., Chaparro, M., Calderon-Mora, J.,
Reininger, B., & Dwivedi, A. (2016). Against colorectal cancer in our
neighborhoods (ACCION): A comprehensive community-wide colorectal cancer
screening intervention for the uninsured in a predominantly Hispanic
community. Preventive Medicine, 91, 273–280.
https://doi.org/10.1016/j.ypmed.2016.08.039
242
Siegel, R. L., Fedewa, S. A., Anderson, W. F., Miller, K. D., Ma, J., Rosenberg, P. S., &
Jemal, A. (2017). Colorectal cancer incidence patterns in the United States, 1974-
2013. Journal of the National Cancer Institute, 109(8), djw322.
https://doi.org/10.1093/jnci/djw322
Siegel, R. L., Jemal, A., & Ward, E. M. (2009). Increase in incidence of colorectal cancer
among young men and women in the United States. Cancer Epidemiology,
Biomarkers & Prevention, 18(6), 1695–1698. https://doi.org/10.1158/1055-
9965.EPI-09-0186
Siegel, R. L., Miller, K. D., Fuchs, H. E., & Jemal, A. (2021). Cancer statistics,
2021. CA, 71(1), 7–33. https://doi.org/10.3322/caac.21654
Siegel, R. L., Miller, K. D., Goding Sauer, A., Fedewa, S. A., Butterly, L. F., Anderson,
J. C., Cercek, A., Smith, R. A., & Jemal, A. (2020). Colorectal cancer statistics,
2020. CA, 70(3), 145–164. https://doi.org/10.3322/caac.21601
Sineshaw, H. M., Robbins, A. S., & Jemal, A. (2014). Disparities in survival
improvement for metastatic colorectal cancer by race/ethnicity and age in the
United States. Cancer Causes & Control, 25(4), 419–423.
https://doi.org/10.1007/s10552-014-0344-z
Skinner, B. F. (1938). The Behavior of Organisms. New York: Appleton-Century-
Crofts. Smith-McLallen, A., & Fishbein, M. (2008). Predictors of intentions to
perform six cancer-related behaviours: roles for injunctive and descriptive
norms. Psychology, Health & Medicine, 13(4), 389–401.
https://doi.org/10.1080/13548500701842933
243
Sohn H. (2017). Racial and Ethnic Disparities in health insurance coverage: Dynamics of
gaining and losing coverage over the life-course. Population Research and Policy
Review, 36(2), 181–201. https://doi.org/10.1007/s11113-016-9416-y
Sommers, B. D., Blendon, R. J., Orav, E. J., & Epstein, A. M. (2016). Changes in
utilization and health among low-income adults after Medicaid expansion or
expanded private insurance. JAMA Internal Medicine, 176(10), 1501–1509.
https://doi.org/10.1001/jamainternmed.2016.4419
Song, E. Y., Swanson, J., Patel, A., MacDonald, M., Aponte, A., Ayoubi, N., Guerra, L.,
Gonzalez, E., Mhaskar, R., Mirza, A-S. (2021). Colorectal cancer risk factors and
screening among the uninsured of Tampa Bay: A free clinic study. Prev Chronic
Dis;18:200496. DOI: http://dx.doi.org/10.5888/pcd18.200496
Starbird, L. E., DiMaina, C., Sun, C. A., & Han, H. R. (2019). A systematic review of
interventions to minimize transportation barriers among people with chronic
diseases. Journal of Community Health, 44(2), 400–411.
https://doi.org/10.1007/s10900-018-0572-3
Stauffer, C. M., & Pfeifer, C. (2021). Colonoscopy. In: StatPearls [Internet]. Treasure
Island (FL): StatPearls Publishing; 2022 Jan-. Retrieved from:
https://www.ncbi.nlm.nih.gov/books/NBK559274/.
Stephen, J. M., & Zoucha, R. (2020). Spanish Speaking, Limited English Proficient
Parents whose Children are Hospitalized: An Integrative Review. Journal of
Pediatric Nursing, 52, 30–40. https://doi.org/10.1016/j.pedn.2020.02.033
Stretcher, V. and Rosenstock, I.M. (1997). The Health Belief Model. In Glanz, K., Lewis,
244
F.M. and Rimer, B.K., (Eds.). Health Behavior and Health Education: Theory,
Research and Practice. San Francisco: Jossey-Bass.
Subramanian, S., Hoover, S., Tangka, F., DeGroff, A., Soloe, C. S., Arena, L. C.,
Schlueter, D. F., Joseph, D. A., & Wong, F. L. (2018). A conceptual framework
and metrics for evaluating multicomponent interventions to increase colorectal
cancer screening within an organized screening program. Cancer, 124(21), 4154–
4162. https://doi.org/10.1002/cncr.31686
Taira, B. R., Torres, J., Nguyen, A., Guo, R., & Samra, S. (2020). Language assistance
for the care of limited English proficiency (LEP) patients in the emergency
department: A survey of providers and staff. Journal of Immigrant and Minority
Health, 22(3), 439–447. https://doi.org/10.1007/s10903-019-00964-9
Tatari, C. R., Andersen, B., Brogaard, T., Badre-Esfahani, S. K., Jaafar, N., &
Kirkegaard, P. (2020). Perceptions about cancer and barriers towards cancer
screening among ethnic minority women in a deprived area in Denmark - a
qualitative study. BMC Public Health, 20(1), 921. https://doi.org/10.1186/s12889-
020-09037-1
Tejeda, S., Gallardo, R. I., Ferrans, C. E., & Rauscher, G. H. (2017). Breast cancer delay
in Latinas: the role of cultural beliefs and acculturation. Journal of Behavioral
Medicine, 40(2), 343–351. https://doi.org/10.1007/s10865-016-9789-8
Tepus, M., & Yau, T. O. (2020). Non-Invasive Colorectal Cancer Screening: An
Overview. Gastrointestinal Tumors, 7(3), 62–73.
https://doi.org/10.1159/000507701
245
Texas Department of State Health Services (2022). Texas Behavioral Risk Factor
Surveillance System. Retrieved from https://dshs.texas.gov/
Tontini, G. E., Prada, A., Sferrazza, S., Ciprandi, G., & Vecchi, M. (2021). The unmet
needs for identifying the ideal bowel preparation. Journal of Gastroenterology
and Hepatology Open, 5(10), 1135–1141. https://doi.org/10.1002/jgh3.12653
Topaloğlu, M. M. & Aydoğdu, N. G. (2021). The Role of Nurses in Health Belief Model-
Based Interventions to Increase Participation of Poor Individuals in Colorectal
Cancer Screening. Journal of Education and Research in Nursing;18(2):280-283.
DOI: 10.5152/jern.2021.26428
Trinh Q. D. (2018). Understanding the impact and challenges of secondary data
analysis. Urologic Oncology, 36(4), 163–164.
https://doi.org/10.1016/j.urolonc.2017.11.003
Tripathy J. P. (2013). Secondary data analysis: Ethical issues and challenges. Iranian
Journal of Public Health, 42(12), 1478–1479.
Tsai, F. J., Hu, Y. J., Chen, C. Y., Tseng, C. C., Yeh, G. L., & Cheng, J. F. (2021). Using
the Health Belief Model to explore nursing students’ relationships between
COVID-19 knowledge, health beliefs, cues to action, self-efficacy, and behavioral
intention: A cross-sectional survey study. Medicine, 100(11), e25210.
https://doi.org/10.1097/MD.0000000000025210
Tu, S. P., Yip, M. P., Chun, A., Choe, J., Bastani, R., & Taylor, V. (2008). Development
of intervention materials for individuals with limited English proficiency: lessons
learned from “Colorectal Cancer Screening in Chinese Americans”. Medical
246
Care, 46(9 Suppl 1), S51–S61. https://doi.org/10.1097/MLR.0b013e31817f0cde
Tseng, C. W., Tierney, E. F., Gerzoff, R. B., Dudley, R. A., Waitzfelder, B., Ackermann,
R. T., Karter, A. J., Piette, J., Crosson, J. C., Ngo-Metzger, Q., Chung, R., &
Mangione, C. M. (2008). Race/ethnicity and economic differences in cost-related
medication underuse among insured adults with diabetes. Diabetes Care, 31(2),
261–266. https://doi.org/10.2337/dc07-1341
Unger-Saldaña, K., Saldaña-Tellez, M., Potter, M. B., Van Loon, K., Allen-Leigh, B., &
Lajous, M. (2020). Barriers and facilitators for colorectal cancer screening in a
low-income urban community in Mexico City. Implementation Science
Communications, 1, 64. https://doi.org/10.1186/s43058-020-00055-z
United States Census Bureau (2011). 2010 Census Shows America’s Diversity. Retrieved
from https://www.census.gov/
United States Census Bureau (2020). Hispanic Heritage Month 2020. Retrieved from
https://www.census.gov/
United States Census Bureau (2021). About the Hispanic Population and its Origin.
Retrieved from https://www.census.gov/
United States Census Bureau (2018). Hispanic population to reach 111 million by 2060.
Retrieved from https://www.census.gov/
US Preventive Services Task Force (2021). Clinician summary of USPSTF
recommendation screening for colorectal cancer. Retrieved from
https://www.uspreventiveservicestaskforce.org/uspstf.
United States Preventive Services Task Force (2021). Final recommendation statement:
247
colorectal cancer: screening. Retrieved from
https://www.uspreventiveservicestaskforce.org
United States Preventive Services Task Force, Davidson, K. W., Barry, M. J., Mangione,
C. M., Cabana, M., Caughey, A. B., Davis, E. M., Donahue, K. E., Doubeni, C.
A., Krist, A. H., Kubik, M., Li, L., Ogedegbe, G., Owens, D. K., Pbert, L.,
Silverstein, M., Stevermer, J., Tseng, C. W., & Wong, J. B. (2021). Screening for
Colorectal Cancer: US Preventive Services Task Force Recommendation
Statement. JAMA, 325(19), 1965–1977. https://doi.org/10.1001/jama.2021.6238
Valdivieso-Mora, E., Peet, C. L., Garnier-Villarreal, M., Salazar-Villanea, M., &
Johnson, D. K. (2016). A systematic review of the relationship between familism
and mental health outcomes in Latino population. Frontiers in Psychology, 7,
1632. https://doi.org/10.3389/fpsyg.2016.01632
Van Haren, R. M., Delman, A. M., Turner, K. M., Waits, B., Hemingway, M., Shah, S.
A., & Starnes, S. L. (2021). Impact of the COVID-19 Pandemic on Lung Cancer
Screening Program and Subsequent Lung Cancer. Journal of the American
College of Surgeons, 232(4), 600–605.
https://doi.org/10.1016/j.jamcollsurg.2020.12.002
Velasco-Mondragon, E., Jimenez, A., Palladino-Davis, A. G., Davis, D., & Escamilla-
Cejudo, J. A. (2016). Hispanic health in the USA: a scoping review of the
literature. Public Health Reviews, 37, 31. https://doi.org/10.1186/s40985-016-
0043-2
Villar-Loubet, O., Weiss, S. M., Marks, G., O’Daniels, C., Jones, D., Metsch, L. R., &
248
McLellan-Lemal, E. (2016). Social and psychological correlates of unprotected
anal intercourse among Hispanic-American women: implications for STI/HIV
prevention. Culture, Health & Sexuality, 18(11), 1221–1237.
https://doi.org/10.1080/13691058.2016.1182217
Viramontes, O., Bastani, R., Yang, L., Glenn, B. A., Herrmann, A. K., & May, F. P.
(2020). Colorectal cancer screening among Hispanics in the United States:
Disparities, modalities, predictors, and regional variation. Preventive
Medicine, 138, 106146. https://doi.org/10.1016/j.ypmed.2020.106146
Wang, X., O’Connell, K., Jeon, J., Song, M., Hunter, D., Hoffmeister, M., Lin, Y.,
Berndt, S., Brenner, H., Chan, A. T., Chang-Claude, J., Gong, J., Gunter, M. J.,
Harrison, T. A., Hayes, R. B., Joshi, A., Newcomb, P., Schoen, R., Slattery, M.
L., Vargas, A., … Du, M. (2019). Combined effect of modifiable and non-
modifiable risk factors for colorectal cancer risk in a pooled analysis of 11
population-based studies. BMJ Open Gastroenterology, 6(1), e000339.
https://doi.org/10.1136/bmjgast-2019-000339
Wang, D. Y., Thrift, A. P., Zarrin-Khameh, N., Wichmann, A., Armstrong, G. N.,
Thompson, P. A., Bondy, M. L., & Musher, B. L. (2017). Rising incidence of
colorectal cancer among young Hispanics in Texas. Journal of Clinical
Gastroenterology, 51(1), 34–42.
https://doi.org/10.1097/MCG.0000000000000563
Wang, J., Moehring, J., Stuhr, S., & Krug, M. (2013). Barriers to colorectal cancer
screening in Hispanics in the United States: an integrative review. Applied
249
Nursing Research: ANR, 26(4), 218–224.
https://doi.org/10.1016/j.apnr.2013.08.005
Wang, H., Qiu, F., Gregg, A., Chen, B., Kim, J., Young, L., Wan, N., & Chen, L. W.
(2018). Barriers and facilitators of colorectal cancer screening for patients of rural
accountable care organization clinics: A multilevel analysis. The Journal of Rural
Health, 34(2), 202–212. https://doi.org/10.1111/jrh.12248
Wei, W., Zhang, M., Zuo, D., Li, Q., Zhang, M., Chen, X., Yu, B., & Liu, Q. (2022).
Screening intention prediction of colorectal cancer among urban Chinese based on
the Protection Motivation Theory. International Journal of Environmental
Research and Public Health, 19(7), 4203. https://doi.org/10.3390/ijerph19074203
Welch, D. R., & Hurst, D. R. (2019). Defining the Hallmarks of Metastasis. Cancer
Research, 79(12), 3011–3027. https://doi.org/10.1158/0008-5472.CAN-19-0458
White, A., Ironmonger, L., Steele, R., Ormiston-Smith, N., Crawford, C., & Seims, A.
(2018). A review of sex-related differences in colorectal cancer incidence,
screening uptake, routes to diagnosis, cancer stage and survival in the UK. BMC
Cancer, 18(1), 906. https://doi.org/10.1186/s12885-018-4786-7
Wigfall, L. T., & Tanner, A. H. (2018). Health Literacy and Health-Care Engagement as
Predictors of Shared Decision-Making Among Adult Information Seekers in the
USA: a Secondary Data Analysis of the Health Information National Trends
Survey. Journal of Cancer Education, 33(1), 67–73.
https://doi.org/10.1007/s13187-016-1052-z
Williams, N. J., Grandner, M. A., Wallace, D. M., Cuffee, Y., Airhihenbuwa, C.,
250
Okuyemi, K., Ogedegbe, G., & Jean-Louis, G. (2016). Social and behavioral
predictors of insufficient sleep among African Americans and Caucasians. Sleep
Medicine, 18, 103–107. https://doi.org/10.1016/j.sleep.2015.02.533
Williams, J. S., Walker, R. J., & Egede, L. E. (2016). Achieving equity in an evolving
healthcare system: Opportunities and challenges. The American Journal of the
Medical Sciences, 351(1), 33–43. https://doi.org/10.1016/j.amjms.2015.10.012
Williams, R., White, P., Nieto, J., Vieira, D., Francois, F., & Hamilton, F. (2016).
Colorectal cancer in African Americans: An Update. Clinical and Translational
Gastroenterology, 7(7), e185. https://doi.org/10.1038/ctg.2016.36
Wittich, A. R., Shay, L. A., Flores, B., De La Rosa, E. M., Mackay, T., & Valerio, M. A.
(2019). Colorectal cancer screening: Understanding the health literacy needs of
Hispanic rural residents. AIMS Public Health, 6(2), 107–120.
https://doi.org/10.3934/publichealth.2019.2.107
Woudstra, A. J., Smets, E., Verdam, M., & Fransen, M. P. (2019). The role of health
literacy in explaining the relation between educational level and decision making
about colorectal cancer screening. International Journal of Environmental
Research and Public Health, 16(23), 4644.
https://doi.org/10.3390/ijerph16234644
Xie, Y. H., Chen, Y. X., & Fang, J. Y. (2020). Comprehensive review of targeted therapy
for colorectal cancer. Signal Transduction and Targeted Therapy, 5(1), 22.
https://doi.org/10.1038/s41392-020-0116-z
Yedjou, C. G., Sims, J. N., Miele, L., Noubissi, F., Lowe, L., Fonseca, D. D., Alo, R. A.,
251
Payton, M., & Tchounwou, P. B. (2019). Health and Racial Disparity in Breast
Cancer. Advances in Experimental Medicine and Biology, 1152, 31–49.
https://doi.org/10.1007/978-3-030-20301-6_3
Zare, M., Ghodsbin, F., Jahanbin, I., Ariafar, A., Keshavarzi, S., & Izadi, T. (2016). The
effect of Health Belief Model-based education on knowledge and prostate cancer
screening behaviors: A randomized controlled trial. International Journal of
Community-based Nursing and Midwifery, 4(1), 57–68.
Zewdie, A., Mose, A., Sahle, T., Bedewi, J., Gashu, M., Kebede, N., & Yimer, A. (2022).
The health belief model’s ability to predict COVID-19 preventive behavior: A
systematic review. SAGE open medicine, 10, 20503121221113668.
https://doi.org/10.1177/20503121221113668
Zhang, J., & Wu, X. (2021). Predict Health Care Accessibility for Texas Medicaid
Gap. Healthcare (Basel, Switzerland), 9(9), 1214.
https://doi.org/10.3390/healthcare9091214
Zhang, W., Hong, S., Takeuchi, D. T., & Mossakowski, K. N. (2012). Limited English
proficiency and psychological distress among Latinos and Asian
Americans. Social Science & Medicine (1982), 75(6), 1006–1014.
https://doi.org/10.1016/j.socscimed.2012.05.012
Zhao, G., Okoro, C. A., Li, J., & Town, M. (2018). Health Insurance Status and Clinical
Cancer Screenings Among U.S. Adults. American Journal of Preventive
Medicine, 54(1), e11–e19. https://doi.org/10.1016/j.amepre.2017.08.024.
252
Appendix: Abbreviations
ACA Affordable Care Act
ACCION Against Colorectal Cancer in Our Neighborhood
ACS American Cancer Society
ACS PUMPS American Community Survey Public Use Microdata Sample
AJCC American Joint Committee on Cancer
ANCOVA Analysis of Covariance
ANOVA Analysis of Variance
ASPE Assistant Secretary for Planning and Evaluation
ASR Age-standardized Rates
BMHSU Behavioral Model of Health Services Use
BRFSS Behavioral Risk Factor Surveillance System
CDC Centers for Disease Control and Prevention
CI Confidence Interval
COVID-19 Coronavirus-2019
CRC Colorectal Cancer
CS Clinical Staff
CTC Computed Tomography and Colonography
DSHS Department of State Health Services, Texas
DSS Disproportional Stratified Sample
DV Dependent Variable
EBCCP Evidence-based Cancer Control Program
FIT-DNA Fecal Immunochemical Test-DNA
FOBT Fecal Occult Blood Test
FPL Federal Poverty Level
FQHC Federally Qualified Health Center
gFOBT Guaiac-based FOBT
Ha Alternative Hypothesis
HBF Health Behavioral Framework
HBM Health Belief Model
HHS Health and Human Services
HIV Human Immunodeficiency Virus
Ho Null Hypothesis
HRQOL Health-related Quality of Life
IBM Integrative Behavioral Model
IRB Institutional Review Board
IV Independent Variable
KA Korean American
LCM Latino Community Members
253
LEP Limited English Proficiency
LHL Low Health Literacy
MRN Male Role Norm
MT-sDNA Multitarget Fecal DNA
NCI National Cancer Institute
NHIS National Health Interview Survey
NHW non-Hispanic White
NIH National Institutes of Health
OCR Office of Civil Rights
OHIP Ontario Health Insurance Plan
OMB Office of Management and Budget
OMH Office of Minority Health
OR Odds Ratio
P Power (Statistics)
PMT Protection Motivation Theory
RQ Research Question
SARS-CoV-2 Severe Acute Respiratory Syndrome Coronavirus 2
SCHBM Sociocultural Health Behavioral Model
SDH Social Determinants of Health
SEER Surveillance, Epidemiology, and End Results
SEM Social Ecological Model
Sensitive gFOBT Sigmoidoscopy combined with FIT
SES Socioeconomic Status
SOGI Sexual Orientation Gender Identity
SPSS Statistical Package for the Social Sciences
TRA Theory of Reasoned Action
TTM Transtheoretical Model
USPSTF United States Preventive Services Task Force
- Language Barrier: An Unmet Challenge for Low Screening of Colorectal Cancer Among Hispanic Americans in Texas
- PhD Dissertation Template, APA 7