Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Investigating racial and ethnic disparities in patient experiences with care and health services use following colorectal cancer diagnosis among older adults with comorbid chronic conditions
(USC Thesis Other)
Investigating racial and ethnic disparities in patient experiences with care and health services use following colorectal cancer diagnosis among older adults with comorbid chronic conditions
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
INVESTIGATING RACIAL AND ETHNIC DISPARITIES IN PATIENT EXPERIENCES WITH CARE AND HEALTH
SERVICES USE FOLLOWING COLORECTAL CANCER DIAGNOSIS AMONG OLDER ADULTS WITH COMORBID
CHRONIC CONDITIONS
by
Stephanie Navarro
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(EPIDEMIOLOGY)
May 2023
Copyright 2023 Stephanie Navarro
ii
TABLE OF CONTENTS
List of Tables………………………………………………………………………………………………………………………………… v
List of Figures……………………………………………………………………………………………………………………………….. vii
Abstract………………………………………………………………………………………………………………………………………… viii
Introduction……………………………………………………………………………………………………………………………….… 1
Background………………………………………………………………………………………………………………………… 1
Conceptual Framework……………………………………………………………………………………………………… 3
Specific Aims and Hypotheses……………………………………………………………………………………………. 3
Data Source and Study Cohort Creation……………………………………………………………………………… 5
CAHPS Patient Experiences with Care……………………………….………………………..………….. 5
Approach to Studying CAHPS Patient Experience Survey Data……………………….……….. 7
Identification of Patient Race and Ethnicity in SEER-CAHPS Data……………………………. 9
Identification of Comorbid Chronic Conditions in SEER-CAHPS Data…………….…………. 10
Chapter 1: Will Improvements in Patient Experience With Care Impact Clinical and Quality of
Care Outcomes?: A Systematic Review…………………………………………………………………………………………. 12
Preface………………………………………………………………………………………………………………………………. 12
Introduction………………………………………………………………………………………………………………………. 12
Conceptual Framework…………………………………………………………………………………………… 13
Methods……………………………………………………………………………………………………………………………. 14
Data Sources and Searches…………………………………………………………………………………….. 14
Study Selection……………………………………………………………………………………………………….. 15
Data Extraction and Quality Assessment…………………………………………………………………. 15
Data Synthesis and Analysis……………………………………………………………………………………. 16
Results……………………………………………………………………………………………………………………………….. 16
Study Selection……………………………………………………………………………………………………….. 16
Study Characteristics………………………………………………………………………………………………. 16
Clinical Outcomes of Care……………………………………………………………………………………….. 16
Quality Outcomes of Care……………………………………………………………………………………….. 18
Study Quality………………………………………………………………………………………………………….. 19
Discussion………………………………………………………………………………………………………………………….. 20
Strengths and Limitations……………………………………………………………………………………….. 23
Conclusion………………………………………………………………………………………………………………. 24
Chapter 2: Investigating the Associations Between Patient Race and Ethnicity, Patient
Experiences with Care, and Health Care Use Following Colorectal Cancer Diagnosis Among
Medicare Beneficiaries with Comorbid Chronic Conditions………………………………………………………….. 38
Introduction………………………………………………………………………………………………………………………. 38
Methods……………………………………………………………………………………………………………………………. 40
Comorbid Chronic Condition Status……………………………………………………………………….. 41
Patient Race and Ethnicity………………………………………………………………………………………. 41
CAHPS Patient Experiences with Care……………………………………………………………………… 41
Covariates…………………………………………………………………………………………………………….… 42
Health Services Use………………………………………………………………………………………………… 42
iii
Statistical Analyses………………………………………………………………………………………………….. 44
Results………………………………………………………………………………………………………………………………. 45
Associations with the patient experience of care coordination………………………………. 47
Associations with the patient experience of getting care quickly……………………………. 47
Associations with the patient experience of getting needed care……………………………. 48
Associations with the patient experience of physician communication…………………… 49
Bivariate associations with change in use of outpatient care following
colorectal cancer diagnosis…………………………………………………………………………………….. 51
Discussion………………………………………………………………………………………………………………………….. 51
Patient experiences with access to care and health services use…………………………….. 52
Patient experience with physician communication and health services use………..….. 53
Patient experience with care coordination and health services use………………………… 54
Improving patient experiences with care among racial and ethnic minority
patients………………………………………………………………………………………………………………….. 55
Patient experiences with care coordination among racial and ethnic minority
patients………………………………………………………………………………………………………………….. 56
Changes in use of outpatient care for chronic conditions following colorectal
cancer diagnosis…………………………………………………………………………………………………….. 57
Strengths and Limitations……………………………………………………………………………………….. 58
Conclusion……………………………………………………………………………………………………………… 61
Chapter 3: Investigating Associations Between Patient Race and Ethnicity, Patient Experiences
with Care, and Receipt of Stage-Specific Guideline Concordant Treatment (GCT) for Colorectal
Cancer Among Medicare Beneficiaries with Comorbid Chronic Conditions…………………………………… 74
Introduction………………………………………………………………………………………………………………………. 74
Methods…………………………………………………………………………………………………………………………….. 75
Comorbid Chronic Conditions…………………………………………………………………………………. 76
Patient Race and Ethnicity………………………………………………………………………………………. 76
CAHPS Patient Experiences with Care…………………………………………………………………….. 77
Receipt of GCT………………………………………………………………………………………………………… 77
Covariates………………………………………………………………………………………………………………. 78
Statistical Analyses…………………………………………………………………………………………………. 78
Results……………………………………………………………………………………………………………………………….. 79
Experiences with care and receipt of GCT……………………………………………………………….. 80
Patient race and ethnicity and receipt of GCT by experiences with care......…………… 80
Discussion………………………………………………………………………………………………………………………….. 81
Experience with getting care quickly and receipt of GCT among non-Hispanic
Black patients………………………………………………………………………………………….……………… 81
Experience with getting needed care and receipt of GCT among Hispanic
patients………………………………………………………………………………………………………………….. 83
Patient experiences with care and receipt of GCT among older adults with
colorectal cancer and comorbid chronic conditions………………………………………………… 84
Strengths and Limitations……………………………………………………………………………………….. 85
Conclusion……………………………………………………………………………………………………………… 88
Conclusion……………………………………………………………………………………………………………………………………. 95
Implications for improving patient experience with care coordination………………………………. 96
iv
Implications for improving patient experiences with getting care quickly and getting
needed care………………………………………………………………………………………………………………………. 98
Implications for improving patient experience with physician communication………………….. 100
Summary of recommendations for clinical practice…………………………………………………………… 102
Next steps and future directions………………………………………………………………………………………… 103
References……………………………………………………………………………………………………………………………………. 106
Appendices…………………………………………………………………………………………………………………………………… 121
Appendix A: Systematic review search strategy…………………………………………………………………. 121
Appendix B: PRISMA 2009 checklist……………………………………………………………………………………. 129
Appendix C: Quality appraisal of included studies……………………………………………………………… 130
Appendix D: Descriptions of how diagnoses of diabetes, hyperlipidemia, and
hypertension were identified in Medicare claims files……………………………………………………….. 138
Appendix E: List of individual questions comprising each included composite
measure of CAHPS patient experience………………………………………………………………………………. 140
Appendix F: Descriptions of how measures of health care services use were
identified in Medicare claims files for the 24 months before and the 24 months after
colorectal cancer diagnosis………..………………………………………………………………………………………. 142
Appendix G: Matrix of outcome modelling decision processes………………………………………….. 146
Appendix H: Descriptions of how stage-specific guideline concordant treatment
was identified in Medicare claims files…………………………………………….………………………………… 148
v
LIST OF TABLES
Table 1: Chronic condition diagnoses from the Centers for Medicare & Medicaid Services list of
chronic conditions and reasons for exclusion from this research……………………………………………………… 11
Table 2: Description of included studies…………………………………………………………………………………………… 27
Table 3a: Summary of significant findings involving clinical outcomes of care (1-5)……….………………… 36
Table 3b: Summary of significant findings involving clinical outcomes of care (6-10).......………………… 36
Table 4a: Summary of significant findings involving quality outcomes of care (1-5)....……………………… 37
Table 4b: Summary of significant findings involving quality outcomes of care (6-10).……………………… 37
Table 5: Characteristics of SEER-CAHPS participants with at least 1 comorbid chronic condition
and non-distant colorectal cancer (CRC) diagnosed from 2001-2017 at age 67+ who met
inclusion criteria.……………………………………………………………………………………………………………………………… 63
Table 6a: Distribution of self-reported CAHPS patients experiences with care coordination and
getting care quickly by patient characteristics for study cohort..……………………………………………………… 64
Table 6b: Distribution of self-reported CAHPS patients experiences with getting needed care and
physician communication by patient characteristics for study cohort……………………………………………… 66
Table 7: Results of simple survey weighted multivariable logistic regression models analyzing
the associations between excellent experience with care coordination and 1) at least three
outpatient evaluation and management visits for chronic conditions, 2) any all-cause
emergency department visits, and 3) any all-cause hospitalizations in the 24 months after
diagnosis of colorectal cancer by CAHPS patient experiences with care…………………………………………… 68
Table 8: Results of simple survey weighted multivariable
logistic regression models analyzing
the associations between excellent experience with getting care quickly and 1) at least three
outpatient evaluation and management visits for chronic conditions, 2) any all-cause
emergency department visits, and 3) any all-cause hospitalizations in the 24 months after
diagnosis of colorectal cancer by CAHPS patient experiences with care…………………………………………… 68
Table 9: Results of simple survey weighted multivariable
logistic regression models stratified
by patient experience with getting care quickly analyzing associations between patient race
and ethnicity and the odds of 1) at least three outpatient evaluation and management visits
for chronic conditions, 2) any all-cause emergency department visits, and 3) any all-cause
hospitalizations in the 24 months after diagnosis of colorectal cancer…………………………………………….. 69
Table 10: Results of simple survey weighted multivariable
logistic regression models analyzing
the associations between excellent experience with getting needed care and 1) at least three
outpatient evaluation and management visits for chronic conditions, 2) any all-cause
emergency department visits, and 3) any all-cause hospitalizations in the 24 months after
diagnosis of colorectal cancer by CAHPS patient experiences with care…………………………………………… 69
Table 11: Results of simple survey weighted multivariable
logistic regression models stratified
by patient experience with getting needed care analyzing associations between patient race
and ethnicity and the odds of 1) at least three outpatient evaluation and management visits
for chronic conditions, 2) any all-cause emergency department visits, and 3) any all-cause
hospitalizations in the 24 months after diagnosis of colorectal cancer…………………………………………….. 70
vi
Table 12: Results of simple survey weighted multivariable
logistic regression models analyzing
the associations between excellent experience with physician communication and 1) at least
three outpatient evaluation and management visits for chronic conditions, 2) any all-cause
emergency department visits, and 3) any all-cause hospitalizations in the 24 months after
diagnosis of colorectal cancer by CAHPS patient experiences with care…………………………………………… 70
Table 13: Results of simple survey weighted multivariable logistic regression models stratified
by patient experience with physician communication analyzing associations between patient
race and ethnicity and the odds of 1) at least three outpatient evaluation and management
visits for chronic conditions, 2) any all-cause emergency department visits, and 3) any
all-cause hospitalizations in the 24 months after diagnosis of colorectal cancer……………………………… 71
Table 14: Patient characteristics and experiences with care by change in use of outpatient care for
chronic conditions in the 24 months following colorectal cancer diagnosis, compared to in the 24
months prior to CRC diagnosis…………………………………………………………………………………………………………. 72
Table 15: Characteristics of SEER-CAHPS participants with pre-existing chronic conditions and
non-distant colorectal cancer diagnosed between 2001-2017 at age 67+ who completed a
CAHPS patient experience survey within three years of CRC diagnosis and had continuous
Medicare parts A and B (with no part C) coverage in the 2 years before and after CRC
diagnosis…………………………………………………………………………………………………………………………………………. 90
Table 16a: Distribution of self-reported patient experiences with care coordination and getting
care quickly by patient characteristics……………………………………………………………………………………………… 91
Table 16b: Distribution of self-reported patient experiences with getting needed care and
physician communication by patient characteristics………………………………………………………………………… 92
Table 17: Distribution of patient characteristics by receipt of stage-specific guideline concordant
treatment for colorectal cancer……………………………………………………………………………………………………….. 93
Table 18: Results of survey weighted multivariable
logistic regression models analyzing the
odds of receiving stage-specific guideline concordant treatment for colorectal cancer by
CAHPS patient experiences with care………………………………………………………………………………………………. 94
Table 19: Results of survey weighted multivariable
logistic regression models analyzing the
odds of receiving stage-specific guideline concordant treatment for colorectal cancer by
CAHPS patient experiences with care and patient race and ethnicity………………………………………………. 94
vii
LIST OF FIGURES
Figure 1: Conceptual framework describing theorized, denoted by solid lines, and hypothesized,
denoted by dashed line, associations between measures of patient experiences with care and
health services use for older adult patients with colorectal cancer and comorbid chronic
conditions…………………………………………………………………………………………………………………………….............. 4
Figure 2: Conceptual framework describing theorized, denoted by solid lines, and hypothesized,
denoted by dashed line, associations between specific and actionable measures of CAHPS patient
experiences with care and clinical and quality outcomes of care, moderated by associations with
aspects of health and health care…………………………………………………………………………………………………….. 25
Figure 3: PRISMA flow diagram………………………………………………………………………………………………………… 26
Figure 4: Summary of selection of study cohort from SEER-CAHPS participants with colorectal
cancer……………………………………………………………………………………………………………………………………………… 62
Figure 5: Selection of study cohort of SEER-CAHPS participants with colorectal cancer…………………… 89
viii
ABSTRACT
Older adults with chronic conditions who are subsequently diagnosed with colorectal cancer
experience declines in care for chronic conditions, increases in use of acute health care, and
substandard care for colorectal cancer. Given compounded disparities in care for racial and ethnic
minority patients, non-White older adults with colorectal cancer and comorbid chronic conditions
experience even more disparate care and outcomes. Yet, when patients report better experiences with
their health care, adherence to care is greater and health outcomes are improved. We sought to
determine relationships between patient race and ethnicity, patient experiences with care, and health
care use after colorectal cancer diagnosis for older adults with comorbid chronic conditions. We used
Surveillance, Epidemiology, and End Results (SEER)-Consumer Assessment of Healthcare Providers and
Systems (CAHPS) data to identify Medicare fee-for-service beneficiaries diagnosed with noninvasive
colorectal cancer between 2001-2017 at age 67 or greater who had diabetes, hyperlipidemia, and/or
hypertension, and responded to a CAHPS patient experience survey within three years of colorectal
cancer diagnosis. Individual patient experiences with care were derived from CAHPS composite
measures of experiences with care coordination, getting care quickly, getting needed care, and doctor
communication. Health services use was determined from Medicare claims files and was identified as (1)
use of outpatient visits for chronic conditions in the 24 months following colorectal cancer diagnosis, (2)
all-cause emergency department use in the 24 months following colorectal cancer diagnosis, (3) all-
cause hospitalizations in the 24 months following colorectal cancer diagnosis, and (4) stage-specific
guideline concordant treatment for colorectal cancer. Multivariable logistic regression models stratified
by experiences with care determined associations between patient race and ethnicity and measures of
health services use by patient experiences with care. Several aspects of experiences with care were
associated with care use after colorectal cancer diagnosis among Medicare fee-for-service beneficiaries
ix
with comorbid chronic conditions. Furthermore, by patient race and ethnicity, findings suggested that
targeted improvements in specific experiences with care among specific racial and ethnic patient groups
may help to alleviate racial and ethnic disparities in health services use following colorectal cancer
diagnosis. Specifically, we found that compared to non-Hispanic White patients, improvements in
patient experiences of getting care quickly may mitigate disparities in health services use among non-
Hispanic Black patients, improvements in experiences of getting needed care may mitigate disparities in
care use among non-Hispanic Black and Hispanic patients, and improvements in experiences of
physician communication may mitigate disparities among non-Hispanic Black, Hispanic, and non-
Hispanic Asian patients. To promote racial and ethnic equity in health services use among older adults
with colorectal cancer and comorbid chronic conditions, health care providers and systems should seek
to make targeted improvements to patient experiences with specific elements of healthcare informed
by unique needs of patients from different racial and ethnic backgrounds.
1
INTRODUCTION
BACKGROUND
In the United States, colorectal cancer is the third most common cancer diagnosis as well as the
third most common cause of cancer-related death. With an overall incidence of 35.9 per 100,000 and an
incidence of 35.7 per 100,000 in non-Hispanic White patients, diagnoses of colorectal cancer are, on
average, more common in non-Hispanic Black patients (incidence of 41.7 per 100,000), and less
common in Hispanic (32.5 per 100,000) and non-Hispanic Asian patients (28.6 per 100,000) compared to
both the general population and to non-Hispanic White patients. In addition, 56% of all colorectal cancer
diagnoses are among patients aged 65 and older.
1
However, despite earlier detection in this age group
compared to younger counterparts, patients with colorectal cancer who are 65 and older account for
68% of all colorectal cancer mortality.
2
Scholars from the American Cancer Society and others have
suggested that this morality age disparity can be attributed to “age-related disadvantages, such as
comorbidities”.
2,3
In fact, nearly 64% of all individuals 65 and older in the US are living with at least 2
diagnosed chronic conditions.
4
In a literature review of comorbid chronic diseases and cancer diagnoses,
Renzi et al. found that living with comorbid chronic conditions may delay cancer surveillance and limit
both patient care seeking for symptoms of cancer development as well as physician referrals for cancer
diagnostic tests following presentation of these symptoms.
5
Specifically, following a diagnosis of colorectal cancer, patients with preexisting chronic
conditions tend to experience declines in care use for these conditions, increases in use of high resource
acute care services (such as emergency department visits and hospitalizations), and are less likely than
those with no comorbid chronic conditions to receive guideline concordant treatment for their cancer.
6–
11
Among racial and ethnic minority patients, disparities in care and outcomes for patients with chronic
conditions who are subsequently diagnosed with colorectal cancer are augmented, given increased
2
prevalence of chronic conditions and decreased likelihoods of receiving appropriate care for both cancer
and comorbid chronic conditions compared to non-Hispanic White patients.
10,12–18
Given the public health challenges of caring for older adults with colorectal cancer and comorbid
chronic conditions, finding ways to improve healthcare delivery that promote equity in care and
outcomes within this vulnerable population is of paramount importance. As originally described in
Andersen’s Behavioral Model of Health Services Use, one actionable area of health care quality
improvement which may be influential in the care of older adult patients with colorectal cancer and
comorbid chronic conditions is the patient experience of care around the time of cancer diagnosis and
first course of treatment.
19
As defined by the Agency for Healthcare Research and Quality (AHRQ)
“patient experience encompasses the range of interactions that patients have with the healthcare
system, including care from their health plans, and from doctors, nurses, and staff in hospitals, physician
practices, and other healthcare facilities”.
20
Measures of patient experiences with care are related to
key processes and outcomes of care, such as health services use, individual clinical outcomes of care,
and overall patient health and well-being.
21,22
Yet, a wealth of evidence has demonstrated consistent
differences in patient experience measures for racial and ethnic minority patients, compared to non-
Hispanic White patients. Specifically, work studying racial and ethnic differences in patient experience
measures has consistently found that non-Hispanic Black patients tend to report better experiences with
care than non-Hispanic White patients, Asian patients tend to report worse experiences with care than
non-Hispanic White patients, and Hispanic patients tend to report a mix of better, worse, and similar
experiences with care compared to non-Hispanic White patients.
23–25
In addition, patients in worse
health states tend to report worse patient experiences with care and racial and ethnic differences in
patient experiences with care may change depending on patient comorbidity load.
26–28
3
CONCEPTUAL FRAMEWORK
Therefore, we sought to investigate the relationships between patient experiences with care,
patient race and ethnicity, and post-colorectal cancer diagnosis health services use among older adults
with comorbid chronic conditions. Our research questions and hypotheses were informed by the
theoretical model depicted in Figure 1, which is based on Navarro et al.’s work in a systematic review of
clinical and quality outcomes of care associated with patient experiences with care.
22
This model
describes the theoretical associations between patient experiences with care and clinical and quality of
care outcomes for older adult patients with colorectal cancer and comorbid chronic conditions and was
informed by prior theoretical models from Andersen and Anhang Price, which posit connections
between outcomes of health care use and experiences with health care.
19,21
As seen in Figure 1, patient
experiences with care indirectly influence outcomes of health services use. In addition, this conceptual
model describes a broadly hypothesized direct relationship between patient experiences with care and
health services use.
SPECIFIC AIMS AND HYPOTHESES
In brief, our specific aims were the following:
(1) to synthesize findings of the existing literature regarding associations between patient
experiences with care and clinical and quality outcomes of care (Chapter 1);
(2) to measure associations between patient experiences with care, patient race and ethnicity,
and the following outcomes after colorectal cancer diagnosis among older adult patients with
comorbid chronic conditions (Chapter 2):
a. use of outpatient care for chronic conditions,
b. all-cause emergency department use, and
c. all-cause hospitalizations; and
4
(3) to measure associations between patient experiences with care, patient race and ethnicity,
and receipt of guideline concordant treatment for colorectal cancer among older adult patients
with comorbid chronic conditions (Chapter 3).
We hypothesized that compared to those reporting less than excellent experiences with care,
those reporting excellent experiences with care would have greater likelihoods of appropriate health
services use following colorectal cancer diagnosis. In addition, we hypothesized that compared to non-
Hispanic White patients and to those reporting less than excellent experiences with care, reporting
excellent experiences with care would be associated with mitigated racial and ethnic disparities in
health services use following colorectal cancer diagnosis.
Figure 1. Conceptual framework describing theorized, denoted by solid lines, and hypothesized, denoted by
dashed line, associations between measures of patient experiences with care and health services use for
older adult patients with colorectal cancer and comorbid chronic conditions. Main associations are
moderated by associations with drivers of health and health care.
DRIVERS OF HEALTH AND
HEALTH CARE
NEED FOR CARE
Medically-Evaluated
Self-Perceived
HEALTH BEHAVIORS
Health Maintenance
Health Care Seeking
CAHPS PATIENT EXPERIENCES
WITH CARE
EXPERIENCE OF CARE
FROM PROVIDERS
Care Coordination
Patient-Provider
Communication
HEALTH SERVICES USE
Chronic Condition Care
Guideline Cancer Care
Emergency Department
Use
EXPERIENCE OF ACCESS TO
CARE
Getting Care Quickly
Thus, to seek
Getting Needed Care
Thus, to seek
Hospital Use
5
DATA SOURCE AND STUDY COHORT CREATION
To complete this research we used Surveillance, Epidemiology, and End Results (SEER)-
Consumer Assessment of Healthcare Providers and Systems (CAHPS) data. This data source is
maintained by the National Cancer Institute (NCI) and the Centers for Medicare & Medicaid Services
(CMS) and includes a linkage of information from NCI’s SEER national cancer database, CMS’s beneficiary
administrative data and claims files, and CMS’s Medicare CAHPS patient experience survey data.
Accordingly, the data set includes information on cancer diagnosis, treatment, and outcomes for a
nationally representative sample of Medicare insured cancer patients from the NCI’s SEER cancer
registries, as well as Medicare fee-for-service claims data and CAHPS patient experience survey data for
these patients.
29
Information from 16 regional SEER cancer registries is included in the linked SEER-CAHPS
dataset, covering about 28% of the US population.
29
Patients captured by SEER registries are racially and
ethnically diverse, with registries capturing approximately 42% of Whites in the US, 45% of African
Americans, 66% of Hispanics, 60% of American Indians and Alaska Natives, 71% of Asians, and 70% of
Hawaiian/Pacific Islanders.
30
However, of note, the SEER-CAHPS dataset does not include the Alaska
Native or Arizona American Indian SEER registries.
31
Compared with the total US population, there is a
slightly greater proportion of foreign-born residents living in SEER areas. However, the proportions of
residents with less than a high school degree and living below the poverty level in SEER areas are
comparable to those of the total US population.
32
CAHPS Patient Experiences with Care
The CAHPS survey was introduced in 1995 and was developed through a collaboration of AHRQ,
Harvard Medical School, RAND, and the Research Triangle Institute (RTI). These organizations
approached survey development by conducting comprehensive literature reviews, holding consumer
focus groups, and mining existing validated surveys of patient experience and care quality.
33,34
Various
6
survey elements, based on RTI qualitative research on survey design as well as cognitive testing, were
designed and implemented to limit subjectivity of responses, produce actionable and timing-specific
data, and minimize measurement error.
33,35
Several versions and iterations of the CAHPS survey have
been shown to be internally consistent, reliable, and valid measures of reporting patient experiences in
aggregate.
34,36,37
The current Medicare CAHPS survey contains seven composite measures of patient experience
and four global ratings of health care which measure patients’ experiences with health care and
Medicare health plans within the last six months. Composite measures include care coordination,
getting care quickly, getting information about prescription drugs, getting needed care, getting needed
prescription drugs, health plan customer service, and physician communication. A calculated composite
score (ranging from 0-100) for each composite measure is computed by combining individual responses
from several questions which ask about experiences with care received in each respective domain of
experience. Global ratings of experiences with care include global rating of health care, global rating of
health plan, global rating of personal doctor, and global rating of specialist doctor. Each global rating
score is derived from a single question which asks patients to rate their care with the respective domain
on a scale from 0-10. For this research, we excluded CAHPS global ratings with care as they are
inherently less actionable survey items, as patients may choose their ratings based on any number of
factors relating to the care that they received within each domain. CAHPS composite measures of
patient experience are based on specific components of care received and responses are, thus, more
actionable as they detail specific aspects of care experiences which may need improvement.
To study how experiences of medical care are related with outcomes of health services use in
this population (as opposed to experiences of Medicare health plans) we only included Medicare CAHPS
composite measures focused on directly receiving care. Therefore, we excluded the Medicare CAHPS
composite measures of getting information about prescription drugs (as this is only relevant for those
7
with Medicare Part D), getting needed prescription drugs (as this is, again, only relevant for those with
Medicare Part D), and health plan customer service (as this measure asks about experiences with
Medicare, as opposed to experiences with health care providers). Of note the Medicare CAHPS
composite measure of experience with care coordination was the most recently introduced composite
measure in 2012. Therefore, there is a substantially smaller cohort of respondents to this CAHPS
question. Despite this smaller sample, the care coordination measure provides valuable insight into an
aspect of care which is important for patients receiving medical care from multiple providers and for
multiple conditions, such as the older adult patients with multimorbidity who are the subject of our
investigation. Accordingly, it remained important to explore the role of patient experiences with care
coordination and we chose to investigate the care coordination measure with the caveat that results
may have to be carefully interpreted due to smaller sample sizes.
Medicare CAHPS survey data is collected annually from Medicare beneficiaries through a
stratified sampling method based on the number of beneficiaries in a plan or in a geographically defined
area.
38
Previous work has shown that overall response to Medicare CAHPS surveys is about 50%.
39
In
SEER-CAHPS data specifically, response rates vary by year of survey administration and range from 49%-
82%.
29
In addition, there are differences in response rates for different population groups. For example,
while those between the ages of 65-84 have response rates of about 50%, those 85 and over have a
response rate of about 43%. In addition, while White beneficiaries have a response rate of about 52%,
African American beneficiaries have a response rate of about 41%, Hispanics have a response rate of
about 33%, and Asians have a response rate of about 29%. Lastly, those dually eligible for Medicare and
Medicaid have a response rate of about 37%.
39
Approach of Studying CAHPS Patient Experience Survey Data
To deal with differential response rates to Medicare CAHPS surveys, SEER-CAHPS makes simple
survey weights available to researchers so that the data can be weighted to represent the Medicare
8
enrolled population of each state. SEER-CAHPS data also offers raked survey weights that correct for
biases which may arise due to differential response rates among patients with different characteristics;
however, raked survey weights are only available for patients diagnosed with cancer in 2011 or later. As
our research contained CAHPS participants from 1999-2019, we used simple survey weights for
applicability across our study cohort. As we were unable to use raked survey weights, there remains the
possibility that differential response rates may introduce bias into this research and affect the overall
generalizability of findings. In addition, as with any survey data, there may be individual questions which
patients chose not to respond to. However, for CAHPS survey variables included in this research, SEER-
CAHPS reports that the fraction of missing information ranges from 5% on questions concerning self-
reported general and mental health status to 8% concerning individual education level. Therefore, as
reported by SEER-CAHPS, and subsequently verified in our specific cohort data, we did not experience a
very high degree of missing data within CAHPS survey data. In addition, an overwhelming proportion of
missing CAHPS survey data in our research was intended missing data, which SEER-CAHPS guidance
suggests excluding from research. Therefore, despite only minimal missing survey data and mostly
intended missing data, we followed this guidance and took a complete case analysis approach to this
research.
40
Results of this research must be interpreted considering that CAHPS survey participants were
included in this research if they completed a CAHPS survey between three years before to three years
after colorectal cancer diagnosis. This inclusion period was selected to capture experiences with care
relatively close to the time of colorectal cancer diagnosis and first course of treatment while maintaining
a broad enough time frame that a relatively large sample size could be achieved in our cohort. While this
may introduce some risk of misclassifying experiences which may be differentially related to receipt of
care received around the time of cancer diagnosis and treatment, we included a covariate measure of
time from CAHPS survey completion to colorectal cancer diagnosis in all multivariable analyses to
9
account for any potential confounding bias that differential time of survey completion may introduce. In
addition, for each analysis, we completed several sensitivity analyses to test the robustness of our
findings under different inclusion constraints for time window of CAHPS survey completion. Findings of
these sensitivity analyses are discussed in subsequent research chapters, however, in general, we saw
that results changed little under more narrow windows of inclusion for CAHPS survey completion, while
substantially limiting cohort size, and subsequently, statistical power of our analyses.
Identification of Patient Race and Ethnicity in SEER-CAHPS Data
Within SEER-CAHPS data there are three potential sources to identify individual patient race and
ethnicity: (1) CAHPS self-reported survey data, (2) SEER cancer registrar-reported data, and (3)
Medicare-reported administrative data. Following SEER-CAHPS guidance (informed by evidence of
moderate agreement between sources of race and ethnicity information in SEER-CAHPS data), to
identify individual patient race and ethnicity within our cohort we prioritized use of self-reported race
and ethnicity from CAHPS surveys. However, if patients did not respond to race and ethnicity questions
included on CAHPS surveys, we identified race and ethnicity from SEER registry data. Subsequently, if
information on race and ethnicity was missing from SEER data, we identified individual patient race and
ethnicity from Medicare administrative data.
41
In addition, we used this method to first identify
participants who reported Hispanic ethnicity. Among all patients who were not identified as Hispanic in
any of the three sources of race and ethnicity information, we further classified participants into non-
Hispanic racial groups. This definition of Hispanic may introduce some misclassification bias into our
research if participants in SEER or Medicare are being labeled Hispanic based on last name, instead of
actual ethnicity. Nonetheless, approximately 82% of Hispanic patients included in this research were
identified through self-report of Hispanic ethnicity on CAHPS surveys, while only about 7% of all included
Hispanic participants did not report any race or ethnicity on a CAHPS survey, so we are fairly confident
that Hispanic ethnicity is being accurately represented within our research.
10
Identification of Comorbid Chronic Conditions in SEER-CAHPS Data
To identify patients with chronic condition diagnoses and to identify outcomes of health services
use for this research, we used Medicare claims data provided in the SEER-CAHPS data linkage. This data
includes Medicare Part A inpatient claims data included in the Medicare Provider Analysis and Review
(MEDPAR) file, Medicare Part B provider claims data included in the Carrier Claims file, and Medicare
Part B outpatient claims data included in the Outpatient Claims file. Accordingly, for inclusion in this
research participants needed to have continuous Medicare parts A and B coverage (with no
supplemental part C coverage) from 24 months before colorectal cancer diagnosis (to identify comorbid
chronic condition diagnoses) to 24 months following colorectal diagnosis (to identify outcomes of health
services use following colorectal cancer diagnosis).
To select patients with comorbid chronic conditions for inclusion in this research we used the
CMS list of chronic conditions to identify diagnoses that were both sensitive to ambulatory care and
which presented a relatively low risk of short-term mortality. A summary of these chronic conditions and
of our process of selecting eligible chronic conditions for inclusion in this research is provided in Table 1.
Briefly, of the non-cancer chronic conditions included in this list, we first excluded any conditions which
were not sensitive to ambulatory care or which presented a relatively high risk of short term mortality.
From the subsequent list of conditions, we selected the three most prevalent conditions among older
adults as eligible chronic conditions for inclusion in this research. These conditions included diabetes,
hyperlipidemia, and hypertension, with prevalences of 27.0%, 47.7% and 57.2%, respectively, among
Medicare fee-for-service beneficiaries.
42
Subsequent identification of our three comorbid chronic
condition diagnoses was performed in Medicare claims data using established CMS chronic condition
identification algorithms.
43
While inclusion of these three comorbid chronic conditions, provides a representative sample of
older adults with colorectal cancer and comorbid chronic conditions, choice of only the three most
11
prevalent conditions limits the extent to which we can generalize the findings of our research. In the
future, generalizability of findings may be expanded by including a complete list of ambulatory care
sensitive conditions as inclusion criteria for comorbid chronic conditions. In addition, we included a
covariate measure of relative comorbidity load in all multivariable analyses by including modified
Charlson comorbidity index score (excluding cancer and diabetes). Future work may broaden the
inclusion criteria for comorbid chronic conditions by seeking to understand relationships between
overall comorbidity load, experiences with care, and health services use following cancer diagnosis by
studying a similar measure of comorbidity load as a main independent variable (as opposed to inclusion
as a covariate).
Table 1. Chronic condition diagnoses from the Centers for Medicare & Medicaid Services list of chronic conditions and
reasons for exclusion from this research. Of those conditions which were not excluded, we chose the three most prevalent
as eligible conditions for inclusion in this research (including diabetes, hyperlipidemia, and hypertension).
Non-Cancer Chronic Condition Not Sensitive to Ambulatory Care
Relatively High Risk of Short Term
Mortality
Alcohol abuse
Alzheimer’s disease and related dementia X
Arthritis (osteoarthritis and rheumatoid) X
Asthma
Atrial fibrillation X
Autism spectrum disorders X
Chronic kidney disease
Chronic obstructive pulmonary disease
Depression
Diabetes
Drug abuse/substance abuse
Heart failure X
Hepatitis (chronic viral B & C)
HIV/AIDS
Hyperlipidemia (high cholesterol)
Hypertension (high blood pressure)
Ischemic heart disease X
Osteoporosis
Schizophrenia and other psychotic disorders
Stroke X
12
CHAPTER 1: Will Improvements in Patient Experience With Care Impact Clinical and Quality of Care
Outcomes?: A Systematic Review
PREFACE
The following research chapter features a previously published systematic review studying
clinical and quality of care outcomes associated with Consumer Assessment of Healthcare Providers and
Systems (CAHPS) patient experiences with care.
22
As there are multiple authors of this manuscript, a
summary of individual contributions is included here. Stephanie Navarro served as first author on the
manuscript and designed the research question, designed the systematic review search protocol and
research methods, completed the search of included databases for included literature, identified
literature from searches which matched inclusion and exclusion criteria through (a) abstract review and
(b) full text review, wrote the first draft of the completed manuscript, and led edits to subsequent
manuscript drafts. Dr. Carol Y Ochoa assisted with design of the systematic review search protocol,
assisted with identification of studies meeting inclusion and exclusion criteria through (a) abstract
review and (b) full text review, and assisted with drafting and edits of the completed manuscript. Emily
Chan and Senxi Du assisted with drafting and edits of the systematic review manuscript. Dr. Albert J
Farias served as senior author for the manuscript and assisted with design and conception of the
systematic review research question and search protocol as well as assisted with drafting and editing of
the completed manuscript.
INTRODUCTION
The Consumer Assessment of Healthcare Providers and Systems (CAHPS) survey is a widely used
standardized tool to assess the quality performance of health care practices and systems.
21,44,45
Since its
inception in 1995 by the Agency of Healthcare Research and Quality, the CAHPS survey tool has been
validated as a reliable measure of patient experiences in a variety of health care settings and across
racial/ethnic groups.
21,25,46
Currently, CAHPS ratings are components of national hospital rankings,
13
Medicare star ratings, and Medicare reimbursement.
47–51
In addition, health care systems, stand-alone
clinics, providers, and researchers regularly develop interventions to improve care by improving CAHPS
patient experiences.
52
Previous reviews have investigated how patient experiences impact quality, safety, and
outcomes of care.
21,53
However, these reviews have not exclusively considered patient experiences
captured by the CAHPS survey. Thus, results have compared a variety of measures of patient
experiences with care reported by different tools which were not designed for these comparisons,
presenting a major limitation to the content validity of results. In addition, while CAHPS patient
experiences may hold strong ties to outcomes of care, it remains unclear whether these experiences are
a direct driver of outcomes or whether there are additional factors at play.
Conceptual Framework
Adapted from Anhang Price’s work investigating associations between patient experiences with
care and care quality, as well as Andersen’s work modeling associations between satisfaction, health
services use, and outcomes, we developed a conceptual framework explaining how specific actionable
aspects of care described by CAHPS composite measures of patient experience relate to health and
healthcare outcomes (Figure 2).
19,21
Mediating relationships between actionable measures of CAHPS
patient experience and outcomes are aspects of health and health care; in which experiences inform
health needs and behaviors which lead to health and healthcare outcomes, and which update
perceptions of care experiences. In addition, our framework hypothesizes that the actionable aspects of
care described by CAHPS composite measures may also have direct relationships with outcomes of care.
Within our conceptual framework, CAHPS composite measures are divided into two categories:
the experience of care from providers and the experience of access to care. Drawing from Andersen’s
Behavioral Model of Health Services Use, we describe both categories as drivers of health behaviors, but
only the experience of care from providers as a driver of the need for care. As CAHPS global ratings of
14
care do not describe specific actionable aspects of care provision, these measures were included as
quality outcomes. This is consistent with previous work describing overall ratings of care as more
subjective care measures based on experiences with multiple specific aspects of care.
54,55
To investigate relationships in our conceptual framework of patient experiences with care and
outcomes, this systematic review summarizes the literature discussing the associations between patient
experiences with care, captured by specific CAHPS composite measures, and healthcare sensitive clinical
health outcomes and health care quality measures.
METHODS
Data Sources and Searches
We conducted a systematic review using PubMed, Embase, and the Cumulative Index to Nursing
and Allied Health Literature (CINAHL) databases. We performed two separate searches for quality and
clinical outcomes of care for each database through December 14, 2019 (see file, Appendix A, describing
complete search strategy). Both searches included keywords for CAHPS surveys which were connected
by the Boolean operator OR to keywords of the CAHPS composite domains. All six composite domains of
the core CAHPS survey were included in our search. These composite domains include 1) care
coordination, 2) customer service, 3) getting care quickly, 4) getting needed care, 5) getting needed
prescription drugs, and 6) patient-provider communication.
For the clinical outcomes search, ten sub-searches were completed to include variations of
keywords for conditions within each of ten domains of diagnoses amenable to healthcare interventions
and utilization. We included terms for infections; neoplasms; nutritional, metabolic, and endocrine
disorders; neurological disorders; cardiovascular diseases; diseases of the respiratory system; diseases
of the digestive system; diseases of the genitourinary system; maternal and perinatal mortality; and
external causes.
56
The keyword sets for each category of disease were connected to CAHPS keyword sets
by the Boolean operator AND. For the quality outcomes search, we used keywords for health care
15
quality, including terms like “quality of care” OR “clinical quality”, as well as terms capturing Health
Effectiveness Data and Information Set (HEDIS) quality measures. Each set of keywords was attached by
the Boolean operator AND. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses
(PRISMA) guidelines were used for the design of this literature review (see file, Appendix B,
documenting PRISMA 2009 checklist) and the review was registered with The International Prospective
Register of Systematic Reviews (PROSPERO) (CDR42020155758) before study activities were initiated.
57
Study Selection
We included studies at the title and abstract screening stage if they were US adult human
quantitative studies of CAHPS patient experience ratings and clinical or quality outcomes of care. At the
full-text screening stage, we included peer-reviewed empirical research examining CAHPS composite
measures as an independent variable and clinical or quality outcomes as the dependent variable. We
excluded studies that used CAHPS from hospice and pediatric patients, as these surveys are all
completed by proxy caregivers and CAHPS surveys completed by proxies are known to have inconsistent
responses relative to those completed by patients and depending on the degree of relationship of proxy
responders to patients.
58
We also excluded studies using the Dental CAHPS to maintain a focus on
medical care and outcomes. Title and abstract screening and full-text screening were performed in
duplicate, with conflicts resolved by group consensus.
Data Extraction and Quality Assessment
The first author performed data extraction and quality appraisal of included studies. Data
extraction focused on study characteristics (i.e. aims, study design, sampling information) and the
associations between outcomes of care and CAHPS composite domain scores. Quality appraisal
evaluated issues of selection bias, misclassification, random error, residual confounding, and external
validity. Covidence online software was used to assist with the screening, data extraction, and quality
16
appraisal processes (Covidence systematic review software, Veritas Health Innovation, Melbourne,
Australia. Available at covidence.org).
Data Synthesis and Analysis
Due to heterogeneity of outcomes and a limited number of studies that met inclusion criteria,
meta-analysis was not feasible and the results of meta-analysis may not have been valid. Therefore, we
summarized findings narratively and qualitatively synthesized broad trends in included studies.
RESULTS
Study Selection
Our search of the three databases returned 3,040 articles, of which 1,174 were unique. We
excluded 1,075 studies at the title and abstract screening phase and 80 articles at the full-text screening
phase (Figure 3). A total of 19 articles met our inclusion criteria: ten studying clinical outcomes of care
59–68
and nine studying quality outcomes of care.
69–77
Study Characteristics
Descriptions of all included studies are summarized in Table 2. Ten studies were conducted with
individual patients as the unit of analysis,
59,63,64,66,70–73,76,77
six studies were conducted at the facility or
health plan level,
60,62,68,69,74,75
and three studies used individual patient outcomes data and facility level
CAHPS ratings.
61,65,67
Among studies conducted at the individual level or which included demographic
composition information, eight reported findings largely for non-Hispanic white patients (>60%),
61–
63,66,67,71,73,77
three included a large percentage of minorities,
64,70,76
and one focused specifically on African
American adults.
72
Clinical Outcomes of Care
Statistically significant findings of studies examining associations of CAHPS ratings with clinical
outcomes are summarized in Tables 3a and 3b, indicating the direction of the relationships. We
identified ten studies investigating clinical outcomes and six found significant associations.
59,61–
17
63,66,67
Studies examining clinical outcomes were completed between 2012-2020 and were a combination
of cross-sectional (n=7),
60–62,64,66–68
prospective cohort (n=1),
59
retrospective cohort (n=1),
63
and matched
case-control (n=1).
65
In studies of individual patients, the number of participants ranged from 600
64
to
10,144.
63
In studies of facilities, the number of participating facilities ranged from 200
60
to 4,535.
68
All ten studies examining clinical outcomes investigated the association of patient-provider
communication with outcomes of care and five found significant associations with at least one clinical
outcome. Specifically, higher ratings of patient-provider communication were associated with higher
self-reported physical health scores, higher self-reported mental health scores, fewer emergency room
visits, fewer inpatient hospital stays, and lower odds of prolonged hospitalization.
59,61,66,67
Higher ratings
of patient-provider communication were also associated with increased likelihoods of being prescribed
statins and aspirin for atherosclerotic cardiovascular disease, which are guideline management
therapies for this condition.
66
However, while Dottino et al. found that higher ratings of patient-provider communication were
associated with lower odds of inpatient mortality,
61
Shirk et al. found no association between patient-
provider communication and inpatient mortality.
67
In addition, while Dy et al. found that higher ratings
of patient-provider communication were associated with decreased odds of 30-day hospital
readmission,
62
Yang et al. found that ratings of patient-provider communication were not associated
with 30-day readmission for acute myocardial infarction, chronic obstructive pulmonary disease,
hip/knee surgery, pneumonia, or stroke.
68
Seven studies investigating associations with clinical outcomes found that some outcomes were
not significantly associated with patient-provider communication. Specifically, odds of kidney transplant
failure at 30 days and 1 year, inpatient postoperative complication, inpatient medical complication,
inpatient surgical complication, inpatient care team complication, inpatient nursing complication, breast
cancer stage at diagnosis, HbA1c level, LDL-c level, systolic blood pressure, development of hospital
18
acquired pressure ulcer, and nonhome discharge were all found to have no significant association with
ratings of patient-provider communication.
60,61,63–65,67
Higher ratings of customer service and getting care quickly were associated with earlier stage of
breast cancer at diagnosis. However, ratings of getting needed care and getting needed prescription
drugs were not associated with stage of breast cancer at diagnosis.
63
Quality Outcomes of Care
Statistically significant findings of studies examining associations of CAHPS ratings with quality
outcomes are summarized in Tables 4a and 4b. We identified nine studies investigating quality
outcomes and eight found significant associations.
69,71–77
Studies examining quality outcomes were
completed between 2001-2019 and were a combination of cross-sectional (n=7)
69,72–77
and prospective
cohort (n=2).
70,71
In studies of individual patients, the number of participants ranged from 236
71
to
26,432.
77
In studies of facilities or health plans, the number of participants ranged from 233 health
plans
69
to 3,176 dialysis facilities.
75
All CAHPS composite domains were studied, with patient-provider
communication being the most studied (n=7)
70–76
and getting needed prescription drugs being the least
studied (n=1).
77
Schneider et al. found that higher ratings of customer service were associated with
improvements in four HEDIS quality of care metrics: effective continuation of antidepressant
medication, follow-up within 30 days after hospitalization for mental illness, LDL-C testing after an acute
cardiovascular event, and mammography.
69
Additionally, Elder et al. found that higher ratings of
customer service were associated with lower global ratings of one’s personal physician.
72
However,
Winpenny et al. found no significant association between ratings of customer service and physician
global rating.
77
Higher ratings of care coordination, getting care quickly, getting needed care, and getting
needed prescription drugs were positively associated with physician global rating.
73,77
In addition, higher
19
ratings of getting needed care were associated with higher percentages of patients receiving eye
examinations for diabetes (HEDIS quality metric).
69
However, Hays et al. did not find that higher ratings
of getting care quickly or getting needed care were significantly associated with physician global rating.
73
Seven studies examined associations of patient-provider communication ratings with quality
outcomes and six found significant associations. Higher patient-provider communication ratings were
significantly associated with higher physician global rating, higher health care global rating, higher care
facility star rating, increased medication refill adherence, and higher Hospital CAHPS (HCAHPS) pain
management score.
71–76
Conversely, there were no associations between patient-provider
communication ratings and antidepressant prescription medication adherence metrics (including
primary non-adherence, early non-persistence, and new prescription medication gap), medication refill
adherence for oral hypoglycemic medication and blood pressure medications, or quality incentive
program score (a measure developed by the Centers for Medicare & Medicaid Services as a pay-for-
performance metric for end stage renal disease patients and facilities).
70,75,76
Study Quality
A summary table of study quality, evaluated as the potential risk of selection bias,
misclassification, random error, residual confounding, and external validity is included in Appendix C.
There was a moderate lack of external validity, which we defined as the extent to which studies could be
generalized to the US adult patient population, as most studies focused on highly specific populations
(e.g., Medicare beneficiaries, breast cancer patients) and all participants must have been enrolled in a
health plan with a usual source of care to participate in the CAHPS survey.
Many of the included studies had relatively low response rates, raising some concern for the
potential of participation bias. While nonresponse may not be the best way to measure nonresponse
bias, nonresponse to CAHPS surveys is consistently higher in several patient groups tending to have
poorer outcomes of care, such as racial/ethnic minorities, patients in poorer health states, patients who
20
would tend to rate their care experiences lower, and patients more likely to experience poorer health
outcomes.
39,78–82
Nonetheless, case mix adjustment and nonresponse weighting, performed by several
included studies, may eliminate some bias introduced by selective participation in CAHPS surveys.
79,80
Although we included all studies where the unit of analysis was health plans or facilities, five out
of nine studies with this design did not control for individual patient characteristics, limiting our ability
to rule out residual confounding bias in these studies. Moreover, given the ecological nature of data at
the health plan or facility level, we are not able to draw conclusions at the individual level.
DISCUSSION
We found 19 studies that examined associations of CAHPS patient experiences with clinical
health and health care quality outcomes. While findings were mostly consistent regarding the
associations of improved outcomes with improved CAHPS patient experiences, results demonstrated
some inconsistencies.
Patient-provider communication was the most studied composite rating, with 17 included
studies investigating its role. Moreover, among the ten studies of clinical outcomes, nine exclusively
analyzed associations with patient-provider communication. Eleven studies investigating patient-
provider communication found significant associations with outcomes (five studies investigating clinical
outcomes and six studies investigating quality outcomes) and among most studies investigating the
same outcome, consistent findings demonstrated a significant association with patient-provider
communication. This trend was seen for five clinical outcomes (self-reported physical health, self-
reported mental health, emergency room visit frequency, inpatient hospital stay frequency, and
prolonged hospitalizations) and one quality outcome (CAHPS physician global rating).
Ratings of patient-provider communication are known to be driven by physician empathy and
respect.
83–85
Thus, associations with patient-provider communication may be partly explained by
components of care resulting from increased physician understanding and recognition of individual
21
patient conditions, such as individualized care plans and patient autonomy in clinical encounters.
Despite other CAHPS patient experience measures being less studied, aspects of communication such as
empathy and continuity, are also known to be key drivers of the overall patient experience,
83,86,87
suggesting that there may be additional connections between less-studied CAHPS patient experience
measures and outcomes of care which have yet to be investigated.
Nonetheless, there were some inconsistencies between studies regarding the associations
between patient-provider communication and the clinical outcomes of 30-day hospital readmission and
inpatient mortality. Both studies investigating 30-day readmission were conducted with hospitals as the
unit of analysis; however, the study finding a significant association with patient-provider
communication rating controlled for both patient and hospital characteristics,
62
while the study finding
no association only controlled for hospital characteristics.
68
The latter may not have been fully risk
adjusted and results may have been affected by some residual confounding. In addition, the studies
investigating inpatient mortality were conducted in two different populations of oncologic surgery
patients; highlighting how relationships between patient experiences with care and outcomes may vary
in different patient and diagnoses groups.
Additional discrepancies between studies investigating associations of getting care quickly,
getting needed care, and customer service ratings further suggest that the relationships between
patient experiences and outcomes may not be equal for all patients. Elder et al. found no associations
between ratings of getting care quickly and getting needed care and physician global rating,
72
while Hays
et al. and Winpenny et al. found these associations to be significant.
73,77
In addition, Elder et al. found
that customer service ratings were significantly associated with physician global rating,
72
while
Winpenny et al. found no significant association between these measures.
77
Elder et al. was conducted
with a sample of African American men largely between the ages of 35-64, while Hays et al. and
Winpenny et al. were conducted with samples of Medicare beneficiaries consisting of over 70% white
22
patients. Inconsistencies in findings may suggest differences in the relationships between patient
experiences and outcomes among patients with different sociodemographic characteristics that have
yet to be explored and understood.
Only Farias et al. specifically examined the associations of racial/ethnic differences in patient
experiences with care and found that better patient experiences with customer service and getting care
quickly significantly increased the odds of earlier stage at breast cancer diagnosis, though only among
non-Hispanic black patients.
63
These contrasting findings provide additional evidence that patients of
different racial/ethnic groups experience care in different ways and that these experiences differentially
relate to care outcomes.
While publication year of included studies ranged from 2001-2020, 63% of included studies
were published in 2016 or later; demonstrating a recent increase in research examining how CAHPS
patient experience ratings relate to outcomes of care. As this trend continues, research must also
investigate how these relationships vary for a wider variety of CAHPS measures and for different patient
populations. To enable these efforts, health plans, providers, and clinical practices should actively work
to develop linked and shared data sources that combine CAHPS patient experience survey data with
clinical and quality outcomes of care. Databases like the SEER-CAHPS and the Medical Expenditure
Panel Survey (MEPS) have illustrated the extensive implications of this type of data source.
29,88
Yet,
more work is needed to establish similar data sources which collect information on diverse patient
populations and link this information to CAHPS survey responses.
In addition, for the clinical outcomes of kidney transplant failure, inpatient complications
(including the development of hospital acquired pressure ulcers as well as non-specific complications),
specific health metrics (such as HbA1C, LDL-c, and systolic blood pressure), and nonhome discharge, and
for the quality outcome of quality incentive program score, no included studies found associations with
CAHPS patient experience measures. As research in this field accelerates, it will also be important for
23
investigators to seek to understand why patient experiences with care might relate to certain outcomes
and not others.
Previous reviews which have included a variety of measures of patient experience, including
Anhang Price et al. and Rathert et al. also found support for associations between patient experiences
and outcomes of care.
21,53
However, the findings of this systematic review expand on and update
previous work by informing how patient experiences specifically captured by the validated and widely
used CAHPS tool relate to a variety of clinical and quality outcomes of care. In fact, all included studies
that investigated clinical outcomes were published after previous reviews; further highlighting the
growing interest in understanding how patient experiences with care relate to clinical outcomes. In
addition, all included studies connecting CAHPS patient experience measures with CAHPS physician
global rating, as a quality measure, were published after previous reviews; reflecting a more recent
tendency for quality improvement programs to utilize CAHPS global ratings as key measures of care
quality.
52
Strengths and Limitations
Our review has strengths that warrant consideration. Using multiple databases, our review was
extremely thorough in searching for articles featuring a vast array of clinical and quality of care
outcomes potentially associated with CAHPS patient experiences. Our review also has a few limitations
that must be considered in the interpretation of our findings. First, there is the potential that retrieval
of studies was incomplete and that all relevant studies are not included in this review; however, this is
unlikely given our search of three separate databases. Furthermore, our search did not return a large
number of studies with similar outcomes with which quantitative meta-analysis could be conducted; this
limits the extent to which we can draw conclusions based on the summary of these findings but points
to the need for further research in this area.
24
Conclusion
A great deal of effort is currently underway to improve CAHPS patient experiences with care, as
they are tied closely to national rankings of care, public reporting of star ratings for Medicare, and
Medicare reimbursement.
47–51
We found that research investigating associations between CAHPS
patient experiences and outcomes often substantiated the use of CAHPS ratings as independent
measures of patient-reported quality. However, several inconsistent findings point to the need for more
rigorous research solidifying these associations and investigating how associations differ for various
patient groups and outcomes of care.
25
Figure 2. Conceptual framework describing theorized, denoted by solid lines, and hypothesized,
denoted by dashed line, associations between specific and actionable measures of CAHPS patient
experiences with care and clinical and quality outcomes of care, moderated by associations with
aspects of health and health care.
26
Figure 3. PRISMA flow diagram.
Full-text articles assessed
for eligibility
(n = 99)
Records identified through
database searching
(n = 3,040)
Screening Included
Eligibility
Identification
Records after duplicates removed
(n = 1,174)
Records screened
(n = 1,174)
Records excluded
(n = 1,075)
Full-text articles excluded,
with reasons
(n = 80)
- CAHPS Patient Experience
not analyzed via CAHPS
Composite Domains (n = 34)
- Abstract only (n = 27)
- Clinical/Quality measure was
not outcome variable (n = 15)
- Not empirical research (n = 2)
- Pediatric population (n = 1)
- Hospice population (n = 1)
Studies included in
qualitative synthesis
(n = 19)
27
Table 2. Description of included studies. (n=19)
Study Design Aim Sampling Data Source
Level of
Analysis
CAHPS Patient
Experience
Domains Studied Outcomes Studied Findings
Studies investigating impact on clinical outcomes of care (n=10)
Chen et al.,
2018
Prospective
Cohort
To define the
association between
PPC and a broad
spectrum of health
outcomes for hepato-
pancreato-biliary
(HPB) patients in a
representative US
population.
N = 1,951
adult
patients
with HPB
disease and
a usual
source of
care
Medical
Expenditure
Panel Survey
(MEPS)
Individual
patient
- PPC - Physical and
mental health
status
- Number of
emergency room
visits
- Total number of
hospitalizations
- Annual healthcare
expenditures
- Annual out-of-
pocket
expenditures
1. Compared to patients who
rated PPC as poor, patients
rating PPC as average had
0.59 times the odds of a poor
SF-12 physical health score
(95% CI= 0.35, 0.98), 0.51
times the odds of a poor SF-
12 mental health score (95%
CI= 0.28, 0.91), 0.57 the odds
of ≥2 ER visits in one year
(95% CI= 0.37, 0.89), and did
not have significantly
different odds of ≥2 inpatient
hospital stays in one year.
Chakrabarti
et al., 2016
Cross
Sectional
To determine the
relationship between
clinical outcomes in
kidney
transplantation and
objective
assessments of
patient satisfaction
with inpatient care in
the parent hospital.
N = 200
hospitals
housing
kidney
transplant
centers
Scientific
Registry of
Transplant
Recipients,
American
Hospital
Association
Annual Survey,
CMS Hospital
Compare
Hospital - PPC (HCAHPS
Communication
with Doctors)
- Kidney transplant
center
performance level
(defined by rates of
transplant failure)
at 30 days and 1
year post-
transplant
1. Patient experiences
measured by the doctor
communication composite
were not predictive of center-
specific odds of kidney
transplant failure at 30 days
or at 1 year.
28
Table 2. continued
Study Design Aim Sampling
Data
Source
Level of
Analysis
CAHPS Patient
Experience
Domains Studied
Outcomes
Studied Findings
Dottino et
al., 2019
Cross
Sectional
To evaluate the
association between
hospital-level data
from the HCAHPS
survey and
postoperative
outcomes in
gynecologic cancer
patients using the
National Inpatient
Sample (NIS)
database.
N = 17, 509
inpatient
admissions for
gynecologic
oncology-
specific
surgeries from
651 hospitals
NIS, CMS
Hospital
Compare
Individual
patient
outcome,
hospital-
level
HCAHPS
ratings
-PPC (HCAHPS
Communication
with Doctors)
- In-hospital
mortality
during surgical
admissions
- Prolonged
length of
hospital stay
- Medical
complications
- Surgical
complications
- Care team
complications
- Any
complications
1. Compared to hospitals in the
bottom tercile of HCAHPS score for
physician communication, hospitals
in the top tercile of HCAHPS score
for physician communication had
0.45 times the odds of inpatient
mortality (95% CI = 0.26, 0.79) and
0.70 times the odds of prolonged
hospitalization (95% CI = 0.54,
0.92); but did not have significantly
different odds of any postoperative
complication, medical complication,
surgical complication, or care team
complication.
Dy et al.,
2016
Cross
Sectional
To evaluate the
relationship between
hospital-level quality
measures of
processes of care and
patient outcomes for
inpatient heart
failure.
N = 895
hospitals
reporting
HCAHPS data
to Press Ganey
Press
Ganey,
CMS
Hospital
Compare
Hospital - PPC (HCAHPS
Communication
with Doctors)
- 30-day heart
failure
readmissions
- 30-day heart
failure
mortality
1. Higher ratings of physician
communication were associated
with fewer risk-adjusted heart
failure 30-day readmissions (β = -
0.87, p<0.05) but were not
significantly associated with risk-
adjusted heart failure 30-day
mortality (p>0.05).
29
Table 2. continued
Study Design Aim Sampling Data Source
Level of
Analysis
CAHPS
Patient
Experience
Domains
Studied
Outcomes
Studied Findings
Farias et
al., 2020
Retrospective
cohort
To identify
racial/ethnic
differences in patient
experiences with
medical care and to
determine the
association between
their medical care
experiences and stage
at diagnosis.
N = 10,144
women ≥ 65
years with
breast cancer
and Medicare
SEER-CAHPS Individual
patient
- Customer
Service
- Getting Care
Quickly
- Getting
Needed Care
- Getting
Needed
Prescription
Drugs
- PPC
- Breast
cancer stage
at diagnosis
1. Amongst non-Hispanic white,
non-Hispanic black, Hispanic, and
Asian patients, only non-Hispanic
black patients were found to have
significantly greater odds of lower
stage at diagnosis of breast cancer
with enhanced customer service
ratings (OR=1.04, 95% CI= 1.01,
1.06) and with enhanced ratings of
getting care quickly (OR=1.03, 95%
CI= 1.01, 1.05).
2. Enhanced ratings of getting
needed care, getting needed
prescription drugs, and PPC were
not found to significantly affect
odds of lower stage at diagnosis of
breast cancer for any racial/ethnic
group.
Fernandez
et al., 2012
Cross
sectional
To assess whether 3
aspects of culturally
competent care are
associated with
glycemic, lipid, and
blood pressure control
among ethnically
diverse patients with
diabetes.
N = 600 adult
patients with
type 2
diabetes and
a regular
primary care
physician
Immigration,
Culture, and
Health Care
(ICHC) Study
Individual
patient
- PPC (CAHPS-
CC Positive
Behaviors)
- Glycemic
control
- LDL
cholesterol
- Systolic
blood
pressure
control
1. Odds of poor HbA1c, LDL-c, and
systolic blood pressure were not
significantly influenced by patient
ratings of PPC.
30
Table 2. continued
Study Design Aim Sampling Data Source
Level of
Analysis
CAHPS Patient
Experience
Domains Studied
Outcomes
Studied Findings
Odom-
Maryon
et al.,
2019
Matched
case-
control
To determine
whether hospital-
level HCAHPS
patient satisfaction
scores are
associated with the
occurrence of
hospital acquired
pressure ulcers
(HAPUs) among
Medicare patients
with stroke.
N = 297 cases who
developed HAPUs
not present on
admission for
stroke; N = 4,158
matched controls
admitted for stroke
who did not develop
HAPUs
Medicare
Beneficiary
Summary files,
Medicare
Provider
Analysis and
Review files,
American
Hospital
Association
Annual Survey,
CMS Hospital
Compare
Individual
patient
outcome,
hospital-
level
HCAHPS
ratings
- PPC (HCAHPS
Communication
with Doctors)
- HAPU
development
1. Increase in quartile of
hospital HCAHPS
communication with
physicians score was not
significantly associated with
odds of developing a HAPU.
Shirk et
al., 2016
Cross
sectional
To determine
whether patient
experience as
defined by the
HCAHPS survey
correlates with
surgical quality for
those patients
undergoing surgery
for urologic
malignancies.
N = 46,988 inpatient
hospital admissions
for prostate cancer
and prostatectomy
(63.4%), kidney
cancer and
nephrectomy
(28.8%), or bladder
cancer and
cystectomy (7.8%)
NIS, CMS
Hospital
Compare
Individual
patient
outcome,
hospital-
level
HCAHPS
ratings
- PPC (HCAHPS
Communication
with Doctors)
- Inpatient
mortality
- Prolonged
hospitalization
- Nonhome
discharge
- Any inpatient
complications
- Medical
complications
- Surgical
complications
- Nursing
complications
1. Compared to patients
treated in hospitals in the
bottom tercile of hospital
HCAHPS physician
communication score, patients
treated in hospitals in the top
tercile had 0.84 times the odds
of prolonged hospitalization
(95% CI = 0.71, 0.99), but did
not have significantly different
odds of inpatient mortality,
nonhome discharge, any
complication, medical
complication, surgical
complication, or nursing
complication.
31
Table 2. continued
Study Design Aim Sampling Data Source
Level of
Analysis
CAHPS Patient
Experience
Domains Studied Outcomes Studied Findings
Yang
et al.,
2018
Cross
sectional
To determine the
relationship between
patient perceptions of
hospitalization care
and hospital
readmissions.
N = 4,535
Medicare certified
all acute care and
critical access
hospitals reporting
to CMS Hospital
Compare
CMS Hospital
Compare
Hospital - PPC (HCAHPS
Communication
with Doctors)
- 30-day
readmission rate
for six clinical
conditions: acute
myocardial
infarction, chronic
obstructive
pulmonary disease,
heart failure,
hip/knee surgery,
pneumonia, stroke
1. For hospitals in the lowest
quartile, middle two
quartiles, and highest
quartile of readmissions,
doctor communication was
not significantly associated
with 30-day readmission for
acute myocardial infarction,
chronic obstructive
pulmonary disease, hip/knee
surgery, pneumonia, or
stroke.
Studies investigating impact on quality outcomes of care (n=9)
Bauer
et al.,
2014
Prospective
cohort
To determine whether
shared decision-
making, patient-
provider trust, or
communication are
associated with early
stage and ongoing
antidepressant
adherence.
N = 1,523 KPNC
members, aged 30-
75, with diabetes
and newly
prescribed
antidepressant
medication
Diabetes
Study of
Northern
California
(DISTANCE)
Individual
patient
- PPC - Primary non-
adherence to anti-
depressants
- Early non-
persistence to anti-
depressants
- New prescription
medication gap
1. Enhanced PPC experience
was not significantly
associated with primary non-
adherence of
antidepressants, early non-
persistence with
antidepressants, or new
prescription medication gap
time.
Chang
et al.,
2006
Prospective
cohort
To investigate the
relationship between
patient-reported
global ratings of
health care and the
quality of providers’
communication and
technical quality of
care.
N = 236 managed
care members
Medical
record review
and
telephone
quality-of-
care
interview
Individual
patient
- PPC - Patient global
rating of health
care
1. Enhanced PPC experience
was a significant predictor of
highest global rating of
overall health care (β=1.73,
p<0.01).
32
Table 2. continued
Study Design Aim Sampling Data Source
Level of
Analysis
CAHPS Patient
Experience
Domains Studied
Outcomes
Studied Findings
Elder
et al.,
2015
Cross
sectional
To examine African
American male health
care experience using
nationally
representative data
from the CAHPS.
N = 12,074
African
American adult
male
respondents to
CAHPS surveys
CAHPS Adult
Commercial
Health Plan
Survey 3.0
Individual
patient
- Customer
Service
- Getting Care
Quickly
- Getting Needed
Care
- PPC
- Global
rating of
personal
physician
1. Compared to those who rated
customer service with their health plan
as a big problem, those who rated
customer service as a small problem
did not have significantly different
experience with their personal
physician, and those who rated
customer service as not a problem had
significantly worse experience with
their personal physician (β=-0.380,
p=0.014).
2. Compared to those who reported
never having good PPC, those who
reported sometimes, usually, or
always having good PPC had
significantly better experience with
their personal physician (sometimes:
β=0.937, p=0.001; usually: β=1.17,
p<0.001; always: β=1.32, p<0.001).
3. No significant associations were
found between experience with
personal physician and the various
levels of both getting care quickly and
getting needed care.
Gupta
et al.,
2009
Cross
sectional
To analyze HCAHPS data
regarding hospital
performance on pain
care.
N = 2,429
hospitals
reporting
HCAHPS data to
CMS
Hospital
Quality
Alliance
Hospital - PPC (HCAHPS
Communication
with Doctors)
- Patient
level of
pain
control
1. Narrative results describe that,
“patients who feel they have good
communication with their doctors
…appear to be the greatest indicator
of pain relief.”
Hays
et al.,
2018
Cross
sectional
To evaluate whether the
association of care
coordination with the
global rating of one’s
personal doctor varies
by number of chronic
conditions and self-
rated health.
N = 242,871
Medicare
beneficiaries
with a personal
doctor
CAHPS
Medicare
Survey
Individual
patient
- Care
Coordination
- Getting Care
Quickly
- Getting Needed
Care
- PPC
- Global
rating of
personal
physician
1. Care coordination, getting care
quickly, getting needed care, and PPC
were all significant predictors of global
rating of personal physician (care
coordination: β=1.79, p<0.0001;
getting care quickly: β=0.41, p<0.0001;
getting needed care: β=0.23,
p=0.0202; PPC: β=8.11, p<0.0001).
33
Table 2. continued
Study Design Aim Sampling
Data
Source
Level of
Analysis
CAHPS Patient
Experience
Domains
Studied Outcomes Studied Findings
Kshirsagar et
al., 2019
Cross
sectional
To examine the
relationship of ICH-
CAHPS survey results
with Quality Incentive
Program (QIP) scores
and star ratings for
dialysis clinics in the
United States.
N = 3,176 dialysis
facilities reporting
star ratings, QIP
score, and ICH-
CAHPS data
CMS
Dialysis
Facility
Compare
Dialysis
Facility
- PPC
(ICH-CAHPS
Patients-
nephrologists’
communication
and caring)
- QIP score
-Star ratings
1. Higher ratings of
patients-nephrologists’
communication and caring
were not significantly
associated with QIP score.
2. Each 10% point increase
in patients-nephrologists’
communication and caring
was associated with 1.24
times the odds of higher
star rating (95% CI = 1.14,
1.34).
Ratanawongsa
et al., 2013
Cross
sectional
To investigate whether
patient assessments of
health care provider
communication were
associated with
objective measures of
poor adherence for
cardiometabolic
medications using
pharmacy utilization
data among a diverse
sample of fully insured
persons with diabetes.
N = 9,377 KPNC
members who were
30-75 years of age
and dispensed 1 or
more
cardiometabolic
medication in the
12 months
preceding survey
participation
DISTANCE Individual
patient
- PPC - Medication refill
adherence for:
cardiometabolic
medications,
diabetes
medications, lipid-
lowering
medications,
hypertension
medications
1. For each 10 point
increase in PPC score, the
prevalence of poor refill
adherence for all
cardiometabolic
medications combined
decreased by 0.9% (95% CI
= 0.2%, 1.7%, p = 0.01) and
for lipid-lowering
medications, specifically,
decreased by 0.8% (95% CI
= 0%, 1.6%, p = 0.04).
However, PPC score was
not significantly associated
with the relative risk of
poor refill adherence
specifically for oral
hypoglycemic medications
or with poor refill
adherence specifically for
blood pressure
medications.
34
Table 2. continued
Study Design Aim Sampling Data Source
Level of
Analysis
CAHPS
Patient
Experience
Domains
Studied Outcomes Studied Findings
Schneider
et al.,
2001
Cross
sectional
To assess the
relationships between
enrollee reports on
the quality of health
plans as measured by
CAHPS 2.0 survey and
the clinical quality of
care measured by the
Medicare Health
Effectiveness Data
and Information Set
(HEDIS).
N = 233
Medicare
health
plans
Health Care
Financing
Administration
(HCFA)
Health
plan
- Customer
Service
- Getting
Needed
Care
- HEDIS Measures,
including: Mammography,
Eye examinations for
diabetes, β-blocker after
myocardial infarction, LDL-
C testing after acute
cardiovascular event,
Follow-up within 30 days
after hospitalization for
mental illness, Effective
continuation of
antidepressant medication
1. Health plan customer service
rating was a significant predictor
of the Mammography, LDL-C
Testing After Acute Cardiovascular
Event, Follow-up Within 30 Days
after Hospitalization for Mental
Illness, and Effective Continuation
of Antidepressant Medication
HEDIS measures (mammography:
β=26.01, p=0.0001; LDL-C:
β=52.01, p=0.001; mental illness
follow-up: β=100.66, p=0.0001;
antidepressant continuation:
β=38.04, p=0.02).
2. Getting needed care rating was
a significant predictor of the Eye
Examinations for Diabetes HEDIS
measure (β=59.7, p=0.002).
35
Table 2. continued
Study Design Aim Sampling
Data
Source
Level of
Analysis
CAHPS Patient
Experience
Domains
Studied
Outcomes
Studied Findings
Winpenny
et al., 2017
Cross
sectional
To examine the
relationship
between physician
advice to quit
smoking and
patient care
experiences.
N = 26,432
Medicare
beneficiaries who
were smokers
with an office visit
in the last 6
months
CAHPS
Medicare
Survey
Individual
patient
- Care
Coordination
- Customer
Service
- Getting Care
Quickly
- Getting
Needed Care
- Getting
Needed
Prescription
Drugs
- Global
rating of
personal
physician
1. Patient ratings of care coordination,
getting care quickly, getting needed care,
and getting needed prescription drugs were
all significant predictors of patient global
ratings of their physician (care coordination:
β=0.41, p<0.001; quick care: β=0.01,
p<0.001; needed care: β=0.05, p<0.001;
needed prescription drugs: β=0.02, p<0.05).
Patient ratings of customer service were not
significant predictors of patient global
ratings of their physician.
Abbreviations: CAHPS, Consumer Assessment of Healthcare Providers and Systems; PPC, Patient-Provider Communication; HCAHPS, Hospital Consumer Assessment of
Healthcare Providers and Systems; CMS, Centers for Medicare & Medicaid Services; CAHPS-CC, Consumer Assessment of Healthcare Providers and Systems – Cultural
Competency; KPNC, Kaiser Permanente Northern California; ICH-CAHPS; In-Center Hemodialysis Consumer Assessment of Healthcare Providers and Systems
36
Table 3a. Summary of significant findings involving clinical outcomes (1-5) of care (n=6).
CAHPS Patient
Experience
Domain
Physical Health
Score
Mental Health
Score
ER Visits in One
Year
Inpatient Hospital
Stays in One Year
Inpatient
Mortality
Customer
Service
Getting Care
Quickly
Patient-Provider
Communication
Chen et al., 2018 (+)
Okunrintemi et al.,
2017 (+)
Chen et al., 2018 (+)
Okunrintemi et al.,
2017 (+)
Chen et al., 2018 (-)
Okunrintemi et al.,
2017 (-)
Chen et al., 2018 (-)
Okunrintemi et al.,
2017 (-)
Dottino et
al., 2019 (-)
(-): indicates negative association found
(+): indicates positive association found
Table 3b. Summary of significant findings involving clinical outcomes (6-10) of care (n=6).
CAHPS Patient
Experience
Domain
Prolonged
Hospitalization
30-Day
Hospital
Readmission
Stage at
Cancer
Diagnosis
Statin Use for
Atherosclerotic
Cardiovascular
Disease
Aspirin Use for
Atherosclerotic
Cardiovascular
Disease
Customer
Service
Farias et al.,
2020 (-)
Getting Care
Quickly
Farias et al.,
2020 (-)
Patient-Provider
Communication
Dottino et al., 2019 (-)
Shirk et al., 2016 (-)
Dy et al.,
2016 (-)
Okunrintemi et al.,
2017 (+)
Okunrintemi et al.,
2017 (+)
(-): indicates negative association found
(+): indicates positive association found
37
Table 4a. Summary of significant findings involving quality outcomes (1-5) of care (n=8).
CAHPS Patient
Experience
Domain
Global Rating
of Health Care
Global Rating of
Physician
Care Facility Star
Rating
Effective
Communication of
Antidepressant
Medication
a
Eye Examinations
for Diabetes
a
Care
Coordination
Hays et al., 2018 (+)
Winpenny et al.,
2017 (+)
Customer
Service
Elder et al., 2015 (-)
Schneider et al.,
2001 (+)
Getting Care
Quickly
Hays et al., 2018 (+)
Winpenny et al.,
2017 (+)
Getting Needed
Care
Hays et al., 2018 (+)
Winpenny et al.,
2017 (+)
Schneider et al.,
2001 (+)
Getting Needed
Prescription
Drugs
Winpenny et al.,
2017 (+)
Patient-
Provider
Communication
Chang et al.,
2006 (+)
Elder et al., 2015 (+)
Hays et al., 2018 (+)
Kshirsagar et al.,
2019 (+)
a: Indicates Health Effectiveness Data and Information Set (HEDIS) quality of care measure.
b: Indicates study findings are of unknown significance.
(-): indicates negative association found
(+): indicates positive association found
Table 4b. Summary of significant findings involving quality outcomes (6-10) of care (n=8).
CAHPS Patient
Experience
Domain
Follow-Up within 30
Days after
Hospitalization for
Mental Illness
a
LDL-C Testing after
Acute Cardiovascular
Event
a
Mammography
a
Medication Refill
Adherence
HCAHPS Pain
Management
Score
Care
Coordination
Customer
Service
Schneider et al.,
2001 (+)
Schneider et al.,
2001 (+)
Schneider et al.,
2001 (+)
Getting Care
Quickly
Getting Needed
Care
Getting Needed
Prescription
Drugs
Patient-
Provider
Communication
Ratanawongsa
et al., 2013 (+)
Gupta et al.,
2009 (+)
b
a: Indicates Health Effectiveness Data and Information Set (HEDIS) quality of care measure.
b: Indicates study findings are of unknown significance.
(-): indicates negative association found
(+): indicates positive association found
38
CHAPTER 2: Investigating the Associations Between Patient Race and Ethnicity, Patient Experiences
with Care, and Health Care Use Following Colorectal Cancer Diagnosis Among Medicare Beneficiaries
with Comorbid Chronic Conditions
INTRODUCTION
In the US, 56% of colorectal cancer diagnoses are among patients aged 65 and older and nearly
64% of patients in this age group have more than one chronic condition.
1,3,4
Yet, after diagnosis of
colorectal cancer, patients with pre-existing chronic conditions are less likely to receive comprehensive
and appropriate care for their chronic conditions compared to their counterparts who are free of a
colorectal cancer diagnosis.
6,10,89,90
Subsequently, evidence suggests that following a cancer diagnosis,
patients are more likely to use acute health care services, such as hospitalizations.
91,92
Disparities in care for chronic conditions following cancer diagnosis are compounded for racial
and ethnic minority patients, with non-Hispanic Black and Hispanic patients being less likely than non-
Hispanic White patients to receive comprehensive diabetes care and preventive care in the aftermath of
a cancer diagnosis, yet being more likely to seek acute care services.
13,14,93,94
Indeed, Du et al. not only
found that racial and ethnic minority patients with colorectal cancer were less likely than their non-
Hispanic White counterparts to receive appropriate care for their chronic conditions, but that racial and
ethnic minority patients with colorectal cancer were also more likely to have chronic conditions in
addition to colorectal cancer.
95
Nonetheless, patients, primary care physicians, and oncologists consistently agree that the
biggest barriers to caring for chronic conditions in the aftermath of a cancer diagnosis involve limited
communication between patients, primary care providers, and oncologists regarding the specific needs
of those with cancer and comorbid chronic conditions; limited physician knowledge across a breadth of
specialties (e.g., limited oncologist knowledge of care for chronic conditions and limited primary
physician knowledge of cancer treatment and related side effects); and a lack of defined roles for each
39
member of a patient’s care team.
96,97
Standardized and validated Medicare Consumer Assessment of
Healthcare Providers and Systems (CAHPS) surveys measure patients perceptions of potential barriers to
care like those which contribute to declines in appropriate care use for chronic conditions following a
cancer diagnosis. Specifically, Medicare CAHPS surveys allow beneficiaries to report their experiences
with elements of their health care such as the quality of communication with their primary physician,
transfer of medical records and information between members of the health care team, and ease of
accessing care from multiple members of the health care team. Among a multitude of positive health
care outcomes, high quality patient experiences reported on CAHPS surveys are associated with
appropriate medication use for chronic conditions, appropriate routine and preventive care use in the
setting of chronic conditions, less frequent ER visits, less frequent inpatient hospital stays, and lower
likelihood of hospital readmission.
22,98
However, patient experiences reported on CAHPS surveys are not equal for all patients.
Specifically, while the CAHPS survey tool has been validated and demonstrated as statistically reliable
across racial and ethnic groups in real world settings, there are consistent differences in patient-
reported CAHPS experiences with care by race and ethnicity.
23–25,99
In addition, patients in poorer overall
health (e.g., those with multimorbidity) tend to report disparate experiences with their care compared
to their counterparts in better overall health states.
28,100,101
Furthermore, as shown by Orr et al. in a
study of Medicare beneficiaries, differences in patient-reported experiences with health care by patient
race and ethnicity may be modified by patient comorbidity load.
26
Therefore, it is important to understand how patient-reported experiences with health care may
relate to care use among older adults with colorectal cancer and comorbid chronic conditions, as well as
how experiences with care may modify associations between patient race and ethnicity and health
services use. Among a cohort of Medicare beneficiaries with colorectal cancer and comorbid chronic
conditions, we aim to analyze associations between patient race and ethnicity, CAHPS patient
40
experiences with care, and 1) use of outpatient evaluation and management visits for chronic conditions
following colorectal cancer diagnosis, 2) change in use of outpatient evaluation and management visits
for chronic conditions from before to after colorectal cancer diagnosis, 3) emergency department use
following colorectal cancer diagnosis, and 4) hospitalizations following colorectal cancer diagnosis.
METHODS
Our study cohort was identified from Surveillance, Epidemiology, and End Results (SEER)-CAHPS
data. The SEER-CAHPS database links nationally representative cancer registry data from the National
Cancer Institute’s SEER program with Medicare beneficiary claims files and CAHPS patient experience
survey data available from the Centers for Medicare & Medicaid Services (CMS). A full description of this
data linkage has been described elsewhere.
29
Medicare beneficiaries were included in our study cohort if they were diagnosed with non-
distant colorectal cancer from 2001-2017 and were at least 67 years old at diagnosis. Patients diagnosed
with metastatic colorectal cancer were excluded from this study due to a more serious disease state in
which health services use may not be directly comparable to those diagnosed with less severe, non-
metastatic disease. To allow for use of data in Medicare fee-for-service claims files, participants must
have had continuous Medicare parts A and B coverage (with no part C coverage) in the 24 months
before and after colorectal diagnosis. Participants with any Medicare part C coverage were excluded, as
services received through part C are not included in claims data within the SEER-CAHPS data linkage.
Participants must have also completed a CAHPS patient experience survey within 36 months before or
after colorectal cancer, to capture experiences of care relatively close to the period of colorectal
diagnosis and first course of treatment. Lastly, participants must have had a pre-existing diagnosis of
diabetes, hyperlipidemia, and/or hypertension prior to colorectal cancer diagnosis and must have
survived to at least 24 months following colorectal cancer diagnosis. A complete summary of participant
selection for this study cohort is included in Figure 4.
41
Comorbid chronic condition status
Pre-existing diagnoses of diabetes, hyperlipidemia, and/or hypertension were chosen as eligible
chronic conditions for study inclusion as they are the three most prevalent chronic conditions from the
CMS list of chronic conditions which are both sensitive to ambulatory care and which present a relatively
low risk of short-term mortality.
42
We identified diagnoses of these conditions prior to colorectal cancer
diagnosis using Medicare claims files and CMS Chronic Conditions Warehouse algorithms.
43
A complete
description of how these conditions were identified is included in Appendix D.
Patient race and ethnicity
Self-reported CAHPS survey data, SEER registry data, and Medicare beneficiary enrollment data
were used to determine patient race and ethnicity. Using CAHPS data, patient ethnicity (Hispanic or non-
Hispanic) was first determined. Subsequently, race (non-Hispanic White, non-Hispanic Black, non-
Hispanic Asian) was determined for all non-Hispanic respondents. If race and ethnicity information was
not provided via CAHPS survey, then this information was abstracted from SEER registry data. If race and
ethnicity information was not available in CAHPS or SEER data, then this information was abstracted
from Medicare beneficiary enrollment files.
CAHPS patient experiences with care
CAHPS composite measures of patient experiences with care coordination, getting care quickly,
getting needed care, and physician communication were included in this research. These measures were
chosen for inclusion in this research as they ask patients about the care received from providers and
health care facilities, as opposed to their Medicare health plan. Each composite measure is reported on
a scale from 0-100 and is calculated by combining responses to several individual survey questions
which comprise each measure. Appendix E includes a complete description of the individual questions
which comprise each composite measure of patient experience included in this research. For the
purposes of this research, CAHPS composite measures were dichotomized as “less than excellent”
42
(scores less than 90) or “excellent” (scores greater than equal to 90). As individual CAHPS composite
measure scores tend to cluster at the upper ends of the ratings scales, this method of dichotomizing
scores when they are used as an independent variable follows guidelines established in previous work
investigating CAHPS survey data.
102–104
Covariates
Covariates included in multivariable analyses included those factors which may confound the
relationships between race and ethnicity and health services use and/or patient experiences with care
and health services use. These covariate factors included age at colorectal cancer diagnosis, sex (male,
female), self-reported education level (less than high school, high school, greater than high school),
area-level poverty concentration (0%-<10%, 10%-<20%, 20%+), SEER region (West, Midwest, Northeast,
South), months from colorectal cancer diagnosis to CAHPS survey completion, year of colorectal cancer
diagnosis, stage of colorectal cancer at diagnosis (in situ or localized, regional), health care services use
in the 24 months before diagnosis (dependent on individual analyses, complete description below), self-
reported general health status (fair or poor versus good, very good, or excellent), self-reported mental
health status (fair or poor versus good, very good, or excellent), modified Charlson Comorbidity Index
score in the 12 months prior to colorectal cancer diagnosis (0, 1, 2+), and diagnoses of diabetes,
hyperlipidemia, and/or hypertension.
Health services use
Health services use investigated in this study included use of outpatient evaluation and
management visits for chronic conditions in the 24 months before and after colorectal cancer diagnosis,
all-cause emergency department use in the 24 months following colorectal cancer diagnosis, and all-
cause hospitalizations in the 24 months following colorectal cancer diagnosis. A complete description of
how each measure of health services use was identified in Medicare claims files is included in Appendix
F. Following established guidelines of care for diabetes (follow-up visits “at least two to four
43
times/year”), hyperlipidemia (follow-up visits “every three to 12 months as needed”), and hypertension
(follow-up visits at least every “three to six months”), outpatient evaluation and management visits for
chronic conditions were categorized as fewer than 3 visits in the 24 months following colorectal cancer
diagnosis (representing an underuse of care compared to guidelines) or at least 3 visits in the 24 months
following colorectal cancer diagnosis (representing sufficient use of care compared to guidelines).
105–107
Outpatient evaluation and management visits for chronic conditions were defined as dates of service in
fee-for-service claims for which patients received outpatient care with a primary or secondary diagnosis
code for diabetes, hyperlipidemia, and/or hypertension, as well as a procedure code for evaluation and
management. We identified all visits with primary or secondary diagnosis codes for chronic conditions,
as the primary diagnosis code was often listed as the patient’s colorectal cancer diagnosis. To describe
changes in outpatient evaluation and management visits for chronic conditions from 24 months before
to 24 months after colorectal diagnosis, four change categories were then defined as follows: 1) less
than 3 visits both before and after colorectal cancer diagnosis (low care use), 2) less than 3 visits before
colorectal diagnosis and at least 3 visits after colorectal cancer diagnosis (increased care use), 3) at least
3 visits before colorectal cancer diagnosis and less than 3 visits after colorectal cancer diagnosis
(decreased care use), and 4) at least 3 visits both before and after colorectal cancer diagnosis (sufficient
care use).
All-cause emergency department visits in the 24 months following colorectal cancer diagnosis
were identified in fee-for-service claims as any claim with a billing code for emergency department use.
For our analyses, all-cause emergency department use was then dichotomized as either no all-cause
emergency department visits or at least one all-cause emergency department visit in the 24 months
following colorectal cancer diagnosis.
All-cause hospitalizations in the 24 months following colorectal cancer diagnosis were identified
in fee-for-service claims as any inpatient hospital claims. For our analyses, this measure was
44
dichotomized into either no all-cause hospitalizations or at least one all-cause hospitalization in the 24
months following colorectal cancer diagnosis.
Statistical Analyses
Descriptive statistics used chi-squared (difference in proportions), ANOVA (difference in means),
and Kruskal-Wallis (difference in medians) tests to analyze patient characteristics by both race and
ethnicity and CAHPS patient experiences with care. Where individual cell counts were less than five, the
Fisher’s exact test or Montecarlo estimate of the Fisher’s exact test was used to test difference in
proportions. For each outcome describing care use after diagnosis only (i.e., outpatient evaluation and
management visits in the 24 months following colorectal cancer diagnosis, emergency department visits
in the 24 months following colorectal cancer diagnosis, and hospitalizations in the 24 months following
colorectal cancer diagnosis) simple survey weighted multivariable logistic regression models were used
to assess associations between CAHPS patient experiences with care and care use. Subsequently, for
each of these three outcomes simple survey weighted multivariable logistic regression models stratified
by experiences with care analyzed associations between patient race and ethnicity and care use
following colorectal cancer diagnosis by experiences with care. As sample sizes were insufficient to
complete similar multivariable analyses of the relationships between patient race and ethnicity,
experiences with care, and change in outpatient evaluation and management visits for chronic
conditions from before to after colorectal cancer diagnosis, descriptive statistics used chi-squared,
ANOVA, and Kruskal-Wallis tests to analyze patient characteristics and experiences with care by changes
in care use.
Simple survey weights included in all multivariable analyses of CAHPS survey data were provided
in SEER-CAHPS data and “were generated to produce estimates that are representative of the
beneficiary populations in the units of the original design…, for the fee-for-service sample, these were
states.”
108
Stratified models used a Bonferroni corrected alpha significance level, where the standard
45
alpha level of 0.05 was divided by the total number of multiple comparisons to calculate a Bonferroni
corrected alpha level for each analysis. Otherwise, all p-values for two-sided tests of statistical
significance were considered statistically significant at p=0.05. All analyses were completed using SAS
version 9.4 (SAS Institute Inc., Cary, NC, USA).
RESULTS
Descriptive statistics pertaining to our study cohort will first be presented. Subsequently, results
will be summarized by individual patient experience measures. Specifically, we will first summarize
associations between experiences with care coordination and use of outpatient evaluation and
management visits for chronic conditions, emergency department use, and hospitalizations following
colorectal cancer diagnosis. Next, and in turn, we will summarize findings regarding associations with
the patient experiences of getting care quickly, getting needed care, and physician communication and
our outcomes of outpatient evaluation and management visits for chronic conditions, emergency
department use, and hospitalizations following colorectal cancer diagnosis.
There were 2,613 SEER-CAHPS participants with colorectal cancer included in this research,
including 84.5% non-Hispanic White participants, 5.7% non-Hispanic Black participants, 5.0% Hispanic
participants, and 4.6% non-Hispanic Asian participants. Bivariate descriptive analyses by patient race
and ethnicity showed that, compared to non-Hispanic White and non-Hispanic Asian participants, non-
Hispanic Black and Hispanic participants tended to have higher rates of less than high school education
(51.0% and 52.3%, respectively, versus 21.6% and 22.7%, respectively) and lower rates of greater than
high school education (22.8% and 24.6%, respectively, versus 42.7% and 45.4%, respectively). Compared
to all other racial and ethnic groups, a lower proportion of Non-Hispanic White patients (74.0% versus
greater than 81%) had at least three outpatient evaluation and management visits for chronic conditions
in the 24 months before colorectal cancer diagnosis. In addition, among all racial and ethnic groups,
non-Hispanic Asian patients had the lowest proportion with any all-cause hospitalizations prior to
46
colorectal cancer diagnosis (30.3% versus greater than 40% for all other groups) and non-Hispanic Black
patients had the highest proportion with any all-cause hospitalizations prior to colorectal cancer
diagnosis (55.0% versus greater than 49% for all other groups). Non-Hispanic Asian patients also had the
lowest proportion with a modified Charlson comorbidity index score of 2+ (11.8% versus greater than
21% for all other racial and ethnic groups), while non-Hispanic Black patients had the highest proportion
with a modified Charlson comorbidity index score of 2+ (31.5% versus less than 24% for all other racial
and ethnic groups). Of all racial and ethnic groups, non-Hispanic Black participants reported the highest
rate of fair or poor general health (49.7%) and Hispanic participants reported the highest rate of fair or
poor mental health (25.6%). Non-Hispanic Black participants had the highest rates of diabetes (55.0%)
and hypertension (93.3%), while non-Hispanic Asian participants had the highest rate of hyperlipidemia
(79.8%) out of all racial and ethnic groups (Table 5).
Bivariate descriptive analyses by experiences with care showed that, compared to all other
racial and ethnic groups, slightly higher proportions of non-Hispanic Black participants reported
excellent experiences with getting care quickly (28.6% versus less than 24.5% for all others) and
physician communication (76.3% versus less than 69.5% for all others), while, out of all racial and ethnic
groups, non-Hispanic Asian participants had the lowest rates of reporting excellent experiences with
getting care quickly (15.9% versus greater than or equal to 23.0% for all others), getting needed care
(54.4% versus greater than 64.0% for all others), and physician communication (61.7% versus greater
than 68.5% for all others). Use of outpatient evaluation and management visits for chronic conditions,
all-cause emergency department visits, and all-cause hospitalizations in the 24 months prior to
colorectal cancer diagnosis did not tend to be correlated with experiences with care. Compared to those
reporting good, very good, or excellent general health, greater proportions of participants reporting fair
or poor general health reported less than excellent experiences with care coordination (51.3% versus
41.8%), getting needed care (38.7% versus 28.4%), and physician communication (35.2% versus 28.0%).
47
In addition, compared to those reporting good, very good, or excellent mental health, greater
proportions of patients reporting fair or poor mental health reported less than excellent experiences
with care coordination (50.9% versus 43.9%), getting needed care (39.9% versus 31.1%), and physician
communication (35.9% versus 30.0%) (Tables 6a and 6b).
Associations with the patient experience of care coordination
In simple survey weighted multivariable logistic regression models, compared to those reporting
less than excellent experience, those reporting excellent experience with care coordination were 33%
more likely to have at least three outpatient evaluation and management visits for chronic conditions
(OR=1.33, 95% CI=1.29-1.37) and 32% less likely to have any all-cause hospitalizations (OR=0.68, 95%
CI=0.66-0.71) in the 24 months after colorectal cancer diagnosis. However, compared to those reporting
less than excellent experience, reporting excellent experience with care coordination was not associated
with lower odds of any all-cause emergency department visits following colorectal cancer diagnosis
(OR=1.09, 95% CI=1.07-1.12) (Table 7). Individual group sample sizes were not large enough to
investigate effect modification of experiences with care coordination in relationships between patient
race and ethnicity and health services use following colorectal cancer diagnosis.
Associations with the patient experience of getting care quickly
Compared to those reporting less than excellent experience, those reporting excellent
experience with getting care quickly were 16% more likely to have at least three outpatient evaluation
and management visits for chronic conditions following colorectal cancer diagnosis (OR=1.16, 95%
CI=1.14-1.18) and were 29% less likely to have any all-cause emergency department visits (OR=0.71, 95%
CI=0.70-0.72) during this time. However, those reporting excellent experience with getting care quickly
had no difference in odds of any all-cause hospitalizations following colorectal cancer diagnosis
compared to those reporting less than excellent experience (OR=0.98, 95% CI=0.96-1.00) (Table 8).
48
Among those reporting less than excellent experience with getting care quickly, compared to
non-Hispanic White patients, non-Hispanic Black were 80% more likely to have any all-cause
hospitalizations following colorectal cancer diagnosis (OR=1.80, 99.38% CI=1.66-1.94). However, among
those reporting excellent experience with getting care quickly, Non-Hispanic Black patients were 46%
less likely than non-Hispanic White patients to have any all-cause hospitalizations following colorectal
cancer diagnosis (OR=0.54, 99.38% CI=0.49-0.61). Regardless of experiences with getting care quickly,
compared to non-Hispanic White patients, non-Hispanic Asian patients were more likely to have at least
three outpatient evaluation and management visits for chronic conditions (less than excellent
experience: OR=2.82, 99.38% CI=2.63-3.03; excellent experience: OR=2.92, 99.38% CI=2.40-3.57) and
were less likely to have any all-cause emergency department visits (less than excellent experience:
OR=0.93, 99.38% CI=0.89-0.98; excellent experience: OR=0.54, 99.38% CI=0.49-0.60) or hospitalizations
following colorectal cancer diagnosis (less than excellent experience: OR=0.47, 99.38% CI=0.45-0.50;
excellent experience: OR=0.67, 99.38% CI=0.55-0.82). In addition, regardless of experiences with getting
care quickly, compared to non-Hispanic White patients, non-Hispanic Black and Hispanic patients were
consistently less likely to have any all-cause emergency department visits in the 24 months after
colorectal cancer diagnosis. However, Hispanic patients were more likely than non-Hispanic White
patients to have any all-cause hospitalizations following colorectal cancer diagnosis regardless of
experience with getting care quickly (less than excellent experience: OR=1.30, 99.38% CI=1.21-1.39;
excellent experience: OR=1.66, 99.38% CI=1.47-1.88) (Table 9).
Associations with the patient experience of getting needed care
Compared to those reporting less than excellent experience, those reporting excellent
experience with getting needed care were 29% more likely to have at least three outpatient evaluation
and management visits for chronic conditions (OR=1.29, 95% CI=1.27-1.31), 30% less likely to have any
all-cause emergency department visits (OR=0.70, 95% CI=0.69-0.71), and 15% less likely to have any all-
49
cause hospitalizations following colorectal cancer diagnosis (OR=0.85, 95% CI=0.83-0.86) (Table 10).
Among those reporting less than excellent experience with getting needed care, compared to non-
Hispanic White patients, non-Hispanic Black patients were over 2.5 times as likely to have any all-cause
emergency department visits (OR=2.71, 99.38% CI=2.47-2.98) and over 13 times as likely to have any all-
cause hospitalizations following colorectal cancer diagnosis (OR=13.34, 99.38% CI=9.84-18.09).
However, among those reporting excellent experiences with getting needed care, non-Hispanic Black
patients were 35% less likely than non-Hispanic White patients to have any all-cause emergency
department visits (OR=0.65, 99.38% CI=0.61-0.68) and 19% less likely than non-Hispanic White patients
to have any all-cause hospitalizations (OR=0.81, 99.38% CI=0.76-0.87). Regardless of experience with
getting needed care, compared to non-Hispanic White patients, Hispanic and non-Hispanic Asian
patients were consistently less likely to have any all-cause emergency department visits or
hospitalizations following colorectal cancer diagnosis. In addition, non-Hispanic Asian patients were
more likely than non-Hispanic White patients to have at least three outpatient evaluation and
management visits for chronic conditions regardless of experience with getting needed care (less than
excellent experience: OR=6.38, 99.38% CI=5.71-7.13; excellent experience: OR=2.48, 99.38% CI=2.23-
2.75) (Table 11).
Associations with the patient experience of physician communication
Compared to those reporting less than excellent experience with physician communication,
those reporting excellent experience with physician communication were 22% more likely to have at
least three outpatient evaluation and management visits for chronic conditions (OR=1.22, 95% CI=1.20-
1.24) and 18% less likely to have any all-cause emergency department visits (OR=0.82, 95% CI=0.81-0.83)
following colorectal cancer diagnosis. However, reporting excellent experience with physician
communication was not associated with decreased odds of any all-cause hospitalizations following
colorectal cancer diagnosis (OR=1.19, 95% CI=1.17-1.21) (Table 12).
50
Among those reporting less than excellent experience with physician communication, compared
to non-Hispanic White patients, non-Hispanic Black patients were over twice as likely to have any all-
cause emergency department visits (OR=2.29, 99.38% CI=2.08-2.53) and over 6.5 times as likely to have
any all-cause hospitalizations following colorectal cancer diagnosis (OR=6.74, 99.38% CI=5.37-8.46).
However, among those reporting excellent experience with physician communication, non-Hispanic
Black patients were 34% less likely than non-Hispanic White patients to have any all-cause emergency
department visits (OR=0.66, 99.38% CI=0.63-0.69) and 7% less likely than non-Hispanic White patients to
have any all-cause hospitalizations following colorectal cancer diagnosis (OR=0.93, 99.38% CI=0.87-0.99).
Moreover, among those reporting less than excellent experience with physician communication,
Hispanic patients were 36% less likely than non-Hispanic White patients to have at least three
outpatient evaluation and management visits for chronic conditions (OR=0.64, 99.38% CI=0.58-0.71) and
over 3.5 times as likely to have any all-cause hospitalizations following colorectal cancer diagnosis
(OR=3.53, 99.38% CI=3.02-4.13). However, among those reporting excellent experience with physician
communication, Hispanic patients were 34% more likely than non-Hispanic White patients to have at
least three outpatient evaluation and management visits for chronic conditions (OR=1.34, 99.38%
CI=1.26-1.42) and 7% less likely than non-Hispanic White patients to have any all-cause hospitalizations
following colorectal cancer diagnosis (OR=0.93, 99.38% CI=0.87-0.99). Lastly, among those reporting less
than excellent experience with physician communication, non-Hispanic Asian patients were 34% more
likely than non-Hispanic White patients to have any all-cause hospitalizations following colorectal cancer
diagnosis (OR=1.34, 99.38% CI=1.22-1.48). However, among those reporting excellent experience with
physician communication, non-Hispanic Asian patients were 80% less likely than non-Hispanic White
patients to have any all-cause hospitalizations following colorectal cancer diagnosis (OR=0.20, 99.38%
CI=0.19-0.22). Regardless of experience with physician communication, non-Hispanic Asian patients
were more likely to have at least three outpatient evaluation and management visits for chronic
51
conditions (less than excellent experience: OR=5.39, 99.38% CI=4.69-6.20; excellent experience:
OR=2.31, 99.38% CI=2.12-2.51) and were less likely to have any all-cause emergency department visits
compared to non-Hispanic White patients (less than excellent experience: OR=0.61, 99.38% CI=0.57-
0.66; excellent experience: OR=0.67, 99.38% CI=0.64-0.71) (Table 13).
Bivariate associations with change in use of outpatient care following colorectal cancer diagnosis
Bivariate analysis of patient experiences and patient characteristics associated with changes in
use of outpatient evaluation management visits for chronic conditions from before to after colorectal
cancer diagnosis revealed that change in use did not tend to be correlated with experiences with care.
However, change in use of outpatient care for chronic conditions from before to after colorectal cancer
diagnosis was statistically significantly correlated with patient race and ethnicity; diagnoses of diabetes,
hypertension, and hyperlipidemia; and modified Charlson Comorbidity Index score (p<0.0001 for all).
Specifically, compared to all other racial and ethnic groups, non-Hispanic White patients had the lowest
proportion with sufficient use of outpatient care both before and after colorectal cancer diagnosis
(63.2%), while non-Hispanic Asian patients had the highest proportion with consistent sufficient use of
care (80.7%). In addition, compared to those with lower modified Charlson Comorbidity Index scores,
those with a modified Charlson Comorbidity Index score of 2+ had the greatest proportion of a
consistent use of sufficient outpatient care for chronic conditions (72.4%), while those with a modified
score of 0 had the greatest proportion of never having sufficient use of outpatient care (17.5%) (Table
14).
DISCUSSION
Among SEER enrolled Medicare fee-for-service beneficiaries who completed a CAHPS patient
experience survey within three years before or after colorectal cancer diagnosis and had diabetes,
hyperlipidemia, and/or hypertension prior to their colorectal cancer diagnosis, excellent experiences
with care coordination, getting care quickly, getting needed care, and physician communication were
52
associated with increased odds of sufficient outpatient care use for chronic conditions following
colorectal cancer diagnosis; excellent experiences with getting care quickly, getting needed care, and
physician communication were associated with decreased odds of emergency department use following
colorectal cancer diagnosis; and excellent experiences with care coordination and getting needed care
were associated with decreased odds of hospitalization following colorectal cancer diagnosis. Compared
to non-Hispanic White patients and to those reporting less than excellent experiences, excellent
experiences with getting care quickly, getting needed care, and physician communication mitigated
some racial and ethnic disparities in care among non-Hispanic Black patients. In addition, compared to
non-Hispanic White patients and to those reporting less than excellent experiences, excellent
experiences with physician communication mitigated some racial and ethnic disparities in care among
Hispanic and non-Hispanic Asian patients.
Patient experiences with access to care and health services use
Our findings suggest that improving patient experiences with getting needed care may be
impactful in not only promoting appropriate use of outpatient care, but also in reducing emergency
department and inpatient care use among older adults with colorectal cancer and comorbid chronic
conditions. In a similar manner, previous research has consistently demonstrated links between
sufficient access to necessary care and more appropriate health services use. For example, previous
work has found that older adults who report being unable to easily schedule appointments for primary
care visits are more likely to use the emergency department and that oncology patients with unplanned
hospitalizations are likely to report that it is easier to access hospital care than outpatient care.
109,110
In
addition, among SEER-CAHPS patients with leukemia and lymphoma, research has shown that having an
emergency department visit after cancer diagnosis is associated with lower scores of patient experience
with getting needed care.
111
Yet, our findings also showed that improvements to the patient experience
of getting care quickly may promote sufficient outpatient care use and decreased emergency
53
department use. Indeed, among patients with multimorbidity, poor timeliness of outpatient and
preventive care is a known facilitator of increased emergency department use.
112
Accordingly,
interventions which seek to improve access to necessary, comprehensive, and timely care among older
adults with colorectal cancer and comorbid chronic conditions may be important to both increase
patient experiences of getting needed care and getting care quickly as well as promote appropriate use
of health services following colorectal cancer diagnosis.
Patient experience with physician communication and health services use
Our findings suggest that improvements to the patient experience of physician communication
may also promote sufficient use of outpatient care for chronic conditions and decreased emergency
department use following colorectal cancer diagnosis. While associations between patient experiences
of provider communication and decreased acute care use are well-established, research continues to
examine how patient experience with physician communication can be optimized across patient
groups.
59,66
Specifically, work analyzing results of clinical interventions has found that interventions
which focus on efforts such as team based primary care, provider shadow coaching, longitudinal
continuity of care in primary care, and cultural competency in inpatient settings may each be impactful
in improving scores of patient experience with physician communication.
113–116
However, successful
interventions to improve patient experiences with physician communication must also account for the
many factors which patients consider important to excellent communication, including factors specific
to individual patient backgrounds and needs as well as factors specific to individual patient health
states.
117–120
Thus, interventions demonstrated to improve patient experience with physician
communication may be further strengthened by prioritizing patient identified needs for physician
communication in the design and implementation of these interventions. As our findings suggest, these
efforts may be beneficial in both promoting adequate use of outpatient care as well as limiting use of
54
emergency department care among older adults with colorectal cancer and comorbid chronic
conditions.
Patient experience with care coordination and health services use
Lastly, we found that improving patient experience with care coordination may promote
sufficient use of outpatient care and decreased hospital use following colorectal cancer diagnosis. The
benefits of well-designed and executed care coordination for those with multiple chronic conditions are
well known. In fact, a scoping review found that patient-centered care, comprehensive care
management, and interdisciplinary team approaches are key components of high quality health care for
those with multiple chronic conditions, while a systematic review found that among older adults with
multiple chronic conditions, use of care coordination strategies are associated with improvements in
several objective clinical measures of health.
121,122
Furthermore, experiences of care coordination among
patients with cancer and comorbid chronic conditions may be especially important in promoting
appropriate health services use, as this population must rely on both their primary and specialist
physicians to be knowledgeable about care for chronic conditions, care for colorectal cancer, and how
this care can be successfully managed concurrently. While the roles of primary and specialty physicians
in concurrent care for colorectal cancer and comorbid chronic conditions remains understudied, several
studies have investigated the complex relationship between primary and specialty physicians in the
provision of cancer treatment. For example, systematic reviews of this work have found that on the
provider side of care, efforts are needed to improve communication between primary and specialty
providers and to ascribe specific care roles to provider; while on the patient side of care, communication
from providers and access to medical information as well as definitive roles for individual providers are
important to effective care coordination in the setting of cancer treatment.
123,124
Therefore, for older
adults with colorectal cancer and comorbid chronic conditions, an emphasis on improving patient
experiences with care coordination by strengthening the individual and collective roles of primary and
55
specialist physicians may be especially important in promoting appropriate use of health services
following colorectal cancer diagnosis.
Improving patient experiences with care among racial and ethnic minority patients
Our findings suggest that improvements in experiences of getting care quickly, getting needed
care, and physician communication may be effective in mitigating racial and ethnic disparities in
sufficient outpatient care use, emergency department use, and hospitalizations among non-Hispanic
Black compared to non-Hispanic White patients. While these findings are unique to our study, previous
work has found that racial differences in provider communication can account for some of the non-
Hispanic Black-non-Hispanic White disparity in preparedness for serious illness and that racial
differences in fragmentation of health care account for some of the non-Hispanic Black-non-Hispanic
White disparity emergency department visit frequency.
125,126
Accordingly, to focus on improving patient
experiences among non-Hispanic Black older adults with colorectal cancer and comorbid chronic
conditions it may be important to understand specific communication and access needs of this patient
group which can be targeted for improvement. For example, qualitative work has shown that, unlike
non-Hispanic White patients, non-Hispanic Black patients describe being generally disappointed with
physician communication, requiring patient advocates in clinical settings to assist with communication,
and having to request medical information, as opposed to automatically being told all relevant medical
information by physicians.
118
To seek racial and ethnic equity in care experiences and outcomes of
health services use, work such as this should be used to design and implement interventions to improve
experiences with care by allowing patient voice and patient-described needs for care to direct the focus
and content of quality improvement work.
Furthermore, our findings suggest that improving patient experiences of physician
communication among Hispanic and non-Hispanic Asian patients may mitigate racial and ethnic
disparities in sufficient use of outpatient care among Hispanic compared to non-Hispanic White patients,
56
as well as disparities in hospitalizations among both Hispanic and non-Hispanic Asian patients compared
to non-Hispanic White patients. For both Hispanic and non-Hispanic Asian oncology patients, patient
language preference, provider language concordance, and provider cultural competence act as
determinants of patient experiences of communication with physicians.
119,127
In addition, previous work
has found that clinical settings with a higher level of cultural competency also tend to have patients who
report higher scores of experiences with physician communication.
128
Thus, it may be important for
healthcare providers and systems seek to make improvements to Hispanic and non-Hispanic Asian
experiences of physician communication by addressing linguistic and cultural barriers that patients are
tasked with overcoming when communicating with their physicians. These interventions will not only be
important to promote racial and ethnic equity in experiences of physician communication, but may also
be important to promote appropriate health services use among Hispanic and non-Hispanic Asian older
adults with colorectal cancer and comorbid chronic conditions.
Patient experiences with care coordination among racial and ethnic minority patients
Due to sample size limitations, we were unable to examine racial and ethnic associations with
health services use by patient experiences with care coordination. The Medicare CAHPS composite
measure of patient experience with care coordination is the most recently introduced composite
measure and questions comprising this measure were added to Medicare CAHPS surveys beginning in
2012. Thus, we had access to limited responses regarding experiences with care coordination and
sample sizes were insufficient to carry out our analyses when this data was divided into individual
patient racial and ethnic groups. Nonetheless, the importance of racial and ethnic specific patient
experiences with care coordination in the context of health services use for chronic conditions following
colorectal cancer diagnosis remains an important research question. Much of managing care for both
chronic conditions and cancer treatment deals directly with care coordinated between primary care
providers, chronic condition specialists, and oncologists. Thus, it remains important for future work to
57
attempt to identify racial and ethnic differences in experiences of care coordination among this patient
population, as well as how these differences may relate to differential outcomes of health services use.
Changes in use of outpatient care for chronic conditions following colorectal cancer diagnosis
In addition, sample size limitations prevented us from studying factors associated with changes
in care use after colorectal cancer diagnosis, compared to before diagnosis, in multivariable models.
Indeed, while results of bivariate descriptive analyses did not reveal any relationships between changes
in use of outpatient evaluation and management visits for chronic conditions and patient experiences
with care, these analyses did demonstrate that patient race and ethnicity, modified Charlson
Comorbidity Index score, and individual chronic condition diagnoses were associated with patterns of
outpatient care use before and after colorectal cancer diagnosis. Nonetheless, our unadjusted findings
contrasted with findings of research demonstrating that those with comorbid chronic conditions tend to
experience declines in care for chronic conditions following cancer diagnosis.
7,9
On the contrary, our
unadjusted findings suggested that the majority of our cohort with comorbid chronic conditions had
sufficient use of outpatient care for chronic conditions both before and after colorectal cancer diagnosis,
with the proportion of patients receiving consistently sufficient care increasing with comorbidity load.
However, our research was not investigating follow-up procedures which are specific to individual
chronic condition diagnosis (e.g., eye exams for those with diabetes) as previous research has done, so
our contrasting findings may be, in part, due to our broader definition of care for chronic conditions.
Thus, while our descriptive findings regarding changes in outpatient care use from before to after
colorectal cancer diagnosis highlight potentially interesting and informative findings in designing health
care interventions which promote sufficient use of outpatient care for chronic conditions among older
adults with colorectal cancer, it will be important that future work aim to appropriately control for
potential confounding factors to fully understand these trends.
58
Strengths and Limitations
This research has several strengths that support the validity of our findings. First, this research
makes use of nationally representative and population-based SEER-CAHPS data. This data allows for the
use of Medicare claims data to provide accurate accounting of health services use among this patient
population, as well as CAHPS survey data to provide highly validated records of patient-reported quality
of care. In addition, we were able to control for several individual sociodemographic, clinical, and tumor
factors due to the wealth of induvial patient information available in SEER-CAHPS data. Given the use of
CAHPS survey data as a main independent variable in this research, we were also able to account for
variations due to differential survey response by including a measure of simple survey weighting in all
multivariable analyses.
However, there are also limitations of this research that are important to consider in the
interpretation of our findings. First, as with all observational studies, there will always be a risk of
residual confounding due to unrecognized confounding factors. While we were able to control for
several factors that may act as potential confounders in our multivariable analyses, we remained unable
to control for certain potentially confounding factors that are not captured in SEER-CAHPS data, such as
patient-provider language concordance and patient language preference, distance from patients' homes
to primary and acute care services, health literacy, and ability to afford needed health care, among
other potentially confounding factors.
In addition, modeling of our outcomes was limited by the constraints of SEER-CAHPS data;
however, thorough efforts were taken to accurately and appropriately measure and report all outcomes
such that the results can be reliably translated into recommendations for clinical practice. A matrix fully
describing steps taken in the modeling choices of outcome variables is included in Appendix G, however
a summary of this process for the outcome of outpatient evaluation and management visits for chronic
conditions is included here. Specifically, we first sought to model outpatient evaluation and
59
management visits for chronic conditions by examining the change in visit use from before to after
colorectal cancer diagnosis. Despite attempts to try different methods of categorizing this change as an
outcome, insufficient data led to multivariable models that would not converge when attempting to
analyze these changes. Therefore, we made the decision to examine visit use after colorectal cancer
diagnosis only. Subsequently, following clinical recommendations for patients with our three eligible
chronic conditions, we aimed to group patients into those who had no visits after diagnosis, those with
too few visits relative to recommendations, those with an appropriate number of visits relative to
recommendations, and those with too many visits relative to recommendations. Again, with four
outcome groups, our multivariable models would not converge, so we proceeded to collapse those with
no visits and those with too few visits into a single outcome group. However, when modelling this
outcome there were some effect estimates which seemed extremely large or extremely small, so we
continued to explore additional methods of categorizing our outcome of outpatient evaluation and
management visits for chronic conditions after colorectal cancer diagnosis. In addition, we determined
that patients grouped as having “too many visits” relative to recommendations may have, in fact, had
more severe disease that required more frequent visits, which would still be appropriate care for those
individual patients. Thus, we removed the idea of a separate group for “over users” from our outcome
categorization scheme. The final model was decided upon due to distinct categories that were both
clinically relevant and parsimonious, such that interpretation of findings could be effectively translated
into recommendations for clinical practice.
In addition to the choice of how to model our outcomes, care was also taken to appropriately
define these outcomes for inclusion in this research. To define outpatient evaluation and management
visits for chronic conditions, we allowed visit claims to contain either a primary or a secondary diagnosis
code for an eligible chronic condition, as the primary diagnosis code was often used for the colorectal
cancer diagnosis. This choice allowed us to confidently include only visits for chronic conditions, while
60
maintaining consideration for the multiple diagnoses of this unique patient population. Nonetheless,
primary and secondary diagnosis codes for emergency department visits and hospitalizations were
rarely for eligible chronic conditions. In addition, with respect to emergency department visits and
hospitalizations, any use of these services signals use of high-resource acute care services for grave
health condition. Therefore, we chose to define the use of emergency department and hospital as all-
cause visits to either location. Accordingly, the findings of this research may differ from those which we
may see if we were to specifically define emergency department and hospital use as only for ambulatory
care sensitive conditions or as potentially preventable/avoidable.
Lastly, we chose to include participants who completed a CAHPS patient experience survey from
three years before colorectal cancer diagnosis to three years after colorectal cancer diagnosis. While this
allowed us to form a relatively large sample of included participants, this also created a relatively broad
window of time within which experiences surrounding the time of colorectal cancer diagnosis and first
course of treatment were considered. Thus, several sensitivity analyses were performed with varying
windows of CAHPS completion as inclusion criteria for our study cohort. Specifically, we compared
results of multivariable analyses stratified by patient experiences with care among the following
adapted study cohorts: 1) including those who completed a CAHPS survey in only the three years before
colorectal cancer diagnosis (seeking to evaluate effects of reporting experiences with care before
colorectal cancer diagnosis occurs), 2) including those who completed a CAHPS survey in the two years
before or after diagnosis (seeking to evaluate effects of experiences with care reported in a closer
window surrounding colorectal cancer diagnosis), and 3) including those who completed a CAHPS survey
in the six to 24 months before or after diagnosis (seeking to evaluate the effects of experiences with
care outside of the initial colorectal cancer diagnosis and treatment period). Where individual group
sample sizes were sufficient to run reliable multivariable models, most variation in findings, when
compared to our original multivariable analyses, occurred due to changes in statistical significance of
61
findings which accompanied limited sample sizes in our three experimental cohorts. Due to limited
statistical power accompanying limited sample sizes as well as mostly consistent findings, we decided to
maintain the study cohort within the criteria of including all participants who responded to CAHPS
patient experience surveys within three years before or after colorectal cancer diagnosis.
Conclusion
In conclusion, older adults with colorectal cancer and comorbid chronic conditions experience
different associations between their experiences with care and health services use by race and ethnicity.
Nonetheless, among all patients in our study cohort, emphasis on high quality patient experiences of
accessing necessary and timely care, coordination of care between primary and specialty providers
(including oncologists as well as specialists caring for chronic conditions), and physician communication
appear to be important in ensuring that patients receive sufficient outpatient care for chronic conditions
and minimize use of high-resource acute care services following colorectal cancer diagnosis. In addition,
addressing racial and ethnic inequities in experiences of accessing necessary and timely care as well as
of physician communication among non-Hispanic Black patients may be an impactful way to reduce
racial and ethnic disparities in health services use following colorectal cancer diagnosis in this patient
population. Moreover, addressing inequities in patient experiences of physician communication among
Hispanic and non-Hispanic Asian patients may also help to reduce racial and ethnic disparities in
outpatient care for chronic conditions and hospitalizations following colorectal cancer diagnosis. Overall,
for older adult patients with colorectal cancer and comorbid chronic conditions, attention to
strengthening specific aspects of patient-perceived care quality among specific racial and ethnic groups
may be an effective tool for promoting appropriate use of health services following colorectal cancer
diagnosis.
62
Figure 4. Summary of selection of study cohort from SEER-CAHPS participants with colorectal cancer.
63
Table 5. Characteristics of SEER-CAHPS participants with at least 1 comorbid chronic condition and non-distant colorectal
cancer (CRC) diagnosed from 2001-2017 at age 67+ who met inclusion criteria. N (%).
Non-
Hispanic
White
n=2,208
Non-Hispanic
Black
n=149
Hispanic
n=130
Non-Hispanic
Asian
n=119 p-value
Age at CRC diagnosis, Mean (SD) 78.1 (6.5) 76.7 (6.4) 76.2 (5.7) 76.2 (6.4) <0.0001
Sex
Male 1,087 (49.2) 56 (37.6) 58 (44.6) 54 (45.4) 0.0334
Female 1,121 (50.8) 93 (62.4) 72 (55.4) 65 (54.6)
Education Level
Less than high school 476 (21.6) 76 (51.0) 68 (52.3) 27 (22.7) <0.0001
High school 789 (35.7) 39 (26.2) 30 (23.1) 38 (31.9)
Greater than high school 943 (42.7) 34 (22.8) 32 (24.6) 54 (45.4)
Area-level poverty concentration
0 - <10% 1,247 (60.5) 22 (16.1) 39 (31.2) 70 (58.8) <0.0001
10 - <20% 583 (28.3) 48 (35.0) 35 (28.0) 32 (26.9)
20%+ 232 (11.3) 67 (48.9) 51 (40.8) 17 (14.3)
SEER Region
West 726 (32.9) 20 (13.4) 75 (57.7) 108 (90.8) <0.0001
Midwest 296 (13.4) 19 (12.8) <5% <5%
Northeast 719 (32.6) 50 (33.6) <35% <10%
South 467 (21.2) 60 (40.3) <10% <5%
Months from CRC diagnosis to CAHPS survey
completion, Median (IQR) 1.8 (35.5) -1.6 (36.4) 6.9 (32.3) -1.2 (30.8) 0.1305
Year of CRC diagnosis, Median (IQR) 2008 (7) 2008 (6) 2008.5 (6) 2010 (7) 0.0954
Stage of CRC at diagnosis
In situ or Localized 1,365 (61.8) 101 (67.8) 76 (58.5) 71 (59.7) 0.3768
Regional 843 (38.2) 48 (32.2) 54 (41.5) 48 (40.3)
Outpatient visits for chronic conditions in 24
months prior to CRC diagnosis
Less than 3 574 (26.0) 18 (12.1) 24 (18.5) 16 (13.5) <0.0001
At least 3 1,634 (74.0) 131 (87.9) 106 (81.5) 103 (86.6)
Any all-cause emergency department visits
in 24 months prior to CRC diagnosis
No 1,115 (50.5) 67 (45.0) 66 (50.8) 70 (58.8) 0.1631
Yes 1,093 (49.5) 82 (55.0) 64 (49.2) 49 (41.2)
Any all-cause hospitalizations in 24 months
prior to CRC diagnosis
No 1,146 (51.9) 67 (45.0) 77 (59.2) 83 (69.8) 0.0002
Yes 1,062 (48.1) 82 (55.0) 53 (40.8) 36 (30.3)
Self-reported general health status
Fair or Poor 696 (32.7) 70 (49.7) 52 (41.9) 32 (28.1) <0.0001
Good, Very Good, or Excellent 1,435 (67.3) 71 (50.4) 72 (58.1) 82 (71.9)
Self-reported mental health status
Fair or Poor 205 (9.7) 25 (18.5) 32 (25.6) 18 (15.9) <0.0001
Good, Very Good, or Excellent 1,913 (90.3) 110 (81.5) 93 (74.4) 95 (84.1)
Modified Charlson Comorbidity Index Score
a
0 1,170 (53.0) 66 (44.3) 75 (57.7) 77 (64.7) 0.0025
1 558 (25.3) 36 (24.2) 24 (18.5) 28 (23.5)
2+ 480 (21.7) 47 (31.5) 31 (23.9) 14 (11.8)
Diagnosis of diabetes 867 (39.3) 82 (55.0) 68 (52.3) 57 (47.9) <0.0001
Diagnosis of hyperlipidemia 1,659 (75.1) 103 (69.1) 93 (71.5) 95 (79.8) 0.1710
Diagnosis of hypertension 1,905 (86.3) 139 (93.3) 112 (86.2) 98 (82.4) 0.0518
a. Modified Charlson Comorbidity Index excludes cancer and diabetes.
Individual cell counts less than 11 have been suppressed for patient privacy.
64
Table 6a. Distribution of self-reported CAHPS patients experiences with care coordination and getting care quickly by patient
characteristics for study cohort. N (%).
Care Coordination Getting Care Quickly
Less than
excellent
n=215
Excellent
n=263 p-value
Less than
excellent
n=1,623
Excellent
n=485 p-value
Age at colorectal cancer (CRC) diagnosis, Mean
(SD) 79.0 (6.8) 78.0 (6.6) 0.1120 77.8 (6.4) 77.8 (6.6) 0.8826
Sex
Male 106 (44.0) 135 (56.0) 0.6590 813 (80.2) 201 (19.8) 0.0008
Female 109 (46.0) 128 (54.0) 810 (74.0) 284 (26.0)
Race and ethnicity
Non-Hispanic White 180 (44.7) 223 (55.3) 0.5508 1,382 (77.0) 412 (23.0) 0.2097
Non-Hispanic Black <60% <45% 80 (71.4) 32 (28.6)
Hispanic <40% <65% 82 (75.9) 26 (24.1)
Non-Hispanic Asian 12 (44.4) 15 (55.6) 74 (84.1) 14 (15.9)
Education Level
Less than high school 36 (47.4) 40 (52.6) 0.5687 364 (72.1) 141 (27.9) 0.0006
High school 67 (41.6) 94 (58.4) 556 (75.8) 178 (24.3)
Greater than high school 112 (46.5) 129 (53.5) 703 (80.9) 166 (19.1)
Area-level poverty concentration
0 - <10% 119 (47.8) 130 (52.2) 0.7620 870 (78.2) 242 (21.8) 0.0864
10 - <20% 58 (43.9) 74 (56.1) 420 (74.2) 146 (25.8)
20%+ 29 (45.3) 35 (54.7) 240 (80.0) 60 (20.0)
SEER Region
West 105 (49.5) 107 (50.5) 0.0020 605 (79.5) 156 (20.5) 0.1439
Midwest 19 (38.8) 30 (61.2) 192 (78.1) 54 (22.0)
Northeast 61 (52.6) 55 (47.4) 497 (75.4) 162 (24.6)
South 30 (29.7) 71 (70.3) 329 (74.4) 113 (25.6)
Months from CRC diagnosis to CAHPS survey
completion, Median (IQR) 11.4 (32.3) 9.0 (29.7) 0.4334 1.9 (35.3) 3.4 (35.0) 0.0410
Year of CRC diagnosis, Median (IQR) 2013 (3) 2014 (3) 0.0168 2008 (7) 2007 (5) 0.0647
Stage of CRC at diagnosis
In situ or Localized 126 (43.5) 164 (56.6) 0.4034 1,021 (77.5) 297 (22.5) 0.5047
Regional 89 (47.3) 99 (52.7) 602 (76.2) 188 (23.8)
Outpatient visits for chronic conditions in 24
months prior to CRC diagnosis
Less than 3 51 (45.1) 62 (54.9) 0.9700 386 (79.3) 101 (20.7) 0.1750
At least 3 164 (44.9) 201 (55.1) 1,237 (76.3) 384 (23.7)
Any all-cause emergency department visits in 24
months prior to CRC diagnosis
No 107 (48.2) 115 (51.8) 0.1877 814 (77.1) 242 (22.9) 0.9208
Yes 108 (42.2) 148 (57.8) 809 (76.9) 243 (23.1)
Any all-cause hospitalizations in 24 months prior
to CRC diagnosis
No 122 (46.0) 143 (54.0) 0.6038 850 (77.8) 242 (22.2) 0.3384
Yes 93 (43.7) 120 (56.3) 773 (76.1) 243 (23.9)
Self-reported general health status
Fair or Poor 80 (51.3) 76 (48.7) 0.0521 549 (76.8) 166 (23.2) 0.6021
Good, Very Good, or Excellent 130 (41.8) 181 (58.2) 1,030 (77.8) 294 (22.2)
Self-reported mental health status
Fair or Poor 27 (50.9) 26 (49.1) 0.3284 185 (78.1) 52 (21.9) 0.7764
Good, Very Good, or Excellent 182 (43.9) 233 (56.1) 1,381 (77.2) 407 (22.8)
65
Table 6a. continued
Care Coordination Getting Care Quickly
Less than
excellent
n=215
Excellent
n=263 p-value
Less than
excellent
n=1,623
Excellent
n=485 p-value
Modified Charlson Comorbidity Index Score
a
0 102 (43.8) 131 (56.2) 0.8723 858 (76.6) 262 (23.4) 0.9041
1 52 (46.4) 60 (53.6) 423 (77.5) 123 (22.5)
2+ 61 (45.9) 72 (54.1) 342 (77.4) 100 (22.6)
Diagnosis of diabetes 94 (45.0) 115 (55.0) 0.9991 680 (77.0) 203 (23.0) 0.9869
Diagnosis of hyperlipidemia 167 (44.1) 212 (55.9) 0.4310 1,206 (76.6) 368 (23.4) 0.4856
Diagnosis of hypertension 201 (47.9) 219 (52.1) 0.0007 1,396 (76.4) 432 (23.6) 0.0816
a. Modified Charlson Comorbidity Index excludes cancer and diabetes.
Individual cell counts less than 11 have been suppressed for patient privacy.
66
Table 6b. Distribution of self-reported CAHPS patients experiences with getting needed care and physician communication by
patient characteristics for study cohort. N (%).
Getting Needed Care Physician Communication
Less than
excellent
n=630
Excellent
n=1,326 p-value
Less than
excellent
n=598
Excellent
n=1,355 p-value
Age at colorectal cancer (CRC) diagnosis, Mean
(SD) 77.8 (6.4) 77.7 (6.3)
0.6695
78.2 (6.5) 77.6 (6.4)
0.0442
Sex
Male 300 (30.9) 672 (69.1) 0.2060 286 (30.3) 657 (69.7) 0.7876
Female 330 (33.5) 654 (66.5) 657 (69.7) 698 (69.1)
Race and ethnicity
Non-Hispanic White 528 (31.5) 1,149 (68.5) 0.0462 508 (30.8) 1,142 (69.2) 0.1834
Non-Hispanic Black 29 (28.7) 72 (71.3) 27 (23.7) 87 (76.3)
Hispanic 34 (35.8) 61 (64.2) 32 (31.4) 70 (68.6)
Non-Hispanic Asian 36 (45.6) 43 (54.4) 31 (38.3) 50 (61.7)
Education Level
Less than high school 127 (29.5) 304 (70.5) 0.0009 145 (31.1) 321 (68.9) 0.3980
High school 194 (28.4) 490 (71.6) 222 (32.2) 468 (67.8)
Greater than high school 309 (36.7) 532 (63.3) 231 (29.0) 566 (71.0)
Area-level poverty concentration
0 - <10% 358 (3.6) 709 (66.5) 0.6794 322 (31.2) 710 (68.8) 0.6060
10 - <20% 159 (31.7) 343 (68.3) 167 (31.9) 357 (68.1)
20%+ 85 (31.5) 185 (68.5) 79 (28.5) 198 (71.5)
SEER Region
West 250 (34.9) 466 (65.1) 0.0235 208 (30.7) 469 (69.3) 0.1119
Midwest 67 (29.1) 163 (70.9) 85 (36.0) 151 (64.0)
Northeast 205 (33.8) 402 (66.2) 194 (30.9) 433 (69.1)
South 108 (26.8) 295 (73.2) 111 (26.9) 302 (73.1)
Months from CRC diagnosis to CAHPS survey
completion, Median (IQR) 4.2 (35.6) 0.6 (34.0) 0.0211 0.5 (35.8) 2.9 (34.6) 0.0566
Year of CRC diagnosis, Median (IQR) 2009 (6) 2007 (8) <0.0001 2007 (7) 2008 (6) 0.0190
Stage of CRC at diagnosis
In situ or Localized 385 (31.6) 835 (68.4) 0.4275 365 (29.8) 860 (70.2) 0.3056
Regional 245 (33.3) 491 (66.7) 233 (32.0) 495 (68.0)
Outpatient visits for chronic conditions in 24
months prior to CRC diagnosis
Less than 3 144 (31.8) 309 (68.2) 0.8270 129 (30.1) 299 (69.9) 0.8076
At least 3 486 (32.3) 1,017 (67.7) 469 (30.8) 1,056 (69.3)
Any all-cause emergency department visits in
24 months prior to CRC diagnosis
No 316 (31.7) 682 (68.3) 0.5984 298 (30.8) 669 (69.2) 0.8513
Yes 314 (32.8) 644 (67.2) 300 (30.4) 686 (69.6)
Any all-cause hospitalizations in 24 months
prior to CRC diagnosis
No 331 (32.7) 680 (67.3) 0.6030 321 (32.0) 681 (68.0) 0.1633
Yes 299 (31.6) 646 (68.4) 277 (29.1) 674 (70.9)
Self-reported general health status
Fair or Poor 264 (38.7) 418 (61.3) <0.0001 241 (35.2) 444 (64.8) 0.0011
Good, Very Good, or Excellent 348 (28.4) 877 (71.6) 340 (28.0) 875 (72.0)
Self-reported mental health status
Fair or Poor 89 (39.9) 134 (60.1) 0.0083 85 (35.9) 152 (64.1) 0.0652
Good, Very Good, or Excellent 520 (31.1) 1,151 (68.9) 494 (30.0) 1,155 (70.0)
67
Table 6b. continued
Getting Needed Care Physician Communication
Less than
excellent
n=630
Excellent
n=1,326 p-value
Less than
excellent
n=598
Excellent
n=1,355 p-value
Modified Charlson Comorbidity Index Score
a
0 319 (30.8) 717 (69.2) 0.2097 312 (30.7) 705 (69.3) 0.9983
1 164 (32.4) 343 (67.7) 158 (30.6) 359 (69.4)
2+ 147 (35.6) 266 (64.4) 128 (30.6) 291 (69.5)
Diagnosis of diabetes 264 (32.6) 546 (67.4) 0.7599 236 (28.3) 597 (71.7) 0.0585
Diagnosis of hyperlipidemia 473 (32.2) 996 (67.8) 0.9871 436 (29.8) 1,029 (70.2) 0.1538
Diagnosis of hypertension 551 (32.5) 1,143 (67.5) 0.4441 527 (30.9) 1,179 (69.1) 0.4940
a. Modified Charlson Comorbidity Index excludes cancer and diabetes.
Individual cell counts less than 11 have been suppressed for patient privacy.
68
Table 7. Results of simple survey weighted multivariable
a
logistic regression models analyzing the associations between
excellent experience with care coordination and 1) at least three outpatient evaluation and management visits for chronic
conditions, 2) any all-cause emergency department visits, and 3) any all-cause hospitalizations in the 24 months after
diagnosis of colorectal cancer (CRC) by CAHPS patient experiences with care. (n=427)
Excellent Experience with Care Coordination
(ref: less than excellent experience with care
coordination)
Adjusted OR (95% CI)
b
Outcome Model 1: Receipt of at least three outpatient evaluation and
management visits for chronic conditions 1.33 (1.29, 1.37)
Outcome Model 2: Any all-cause emergency department visits 1.09 (1.07, 1.12)
Outcome Model 3: Any all-cause hospitalizations 0.68 (0.66, 0.71)
a. Analyses adjusted for individual patient age at diagnosis, sex, race and ethnicity, education level, area-level poverty
concentration, SEER region, time from CAHPS survey completion to colorectal cancer diagnosis, year of colorectal cancer
diagnosis, stage of colorectal cancer at diagnosis, care use in the 24 months prior to colorectal cancer diagnosis (Model 1:
receipt of outpatient evaluation and management outpatient visits for chronic conditions, Model 2: receipt of all-cause
emergency department visits, Model 3: receipt of all-cause hospitalizations), self-reported general health status, self-
reported mental health status, modified Charlson Comorbidity Index score, and diagnoses of diabetes, hyperlipidemia, and
hypertension.
b. Bold results indicate p<0.05.
Table 8. Results of simple survey weighted multivariable
a
logistic regression models analyzing the associations between
excellent experience with getting care quickly and 1) at least three outpatient evaluation and management visits for chronic
conditions, 2) any all-cause emergency department visits, and 3) any all-cause hospitalizations in the 24 months after
diagnosis of colorectal cancer (CRC) by CAHPS patient experiences with care. (n=1,886)
Excellent Experience with Getting Care
Quickly
(ref: less than excellent experience with
getting care quickly)
Adjusted OR (95% CI)
b
Outcome Model 1: Receipt of at least three outpatient evaluation and
management visits for chronic conditions 1.16 (1.14, 1.18)
Outcome Model 2: Any all-cause emergency department visits 0.71 (0.70, 0.72)
Outcome Model 3: Any all-cause hospitalizations 0.98 (0.96, 1.00)
a. Analyses adjusted for individual patient age at diagnosis, sex, race and ethnicity, education level, area-level poverty
concentration, SEER region, time from CAHPS survey completion to colorectal cancer diagnosis, year of colorectal cancer
diagnosis, stage of colorectal cancer at diagnosis, care use in the 24 months prior to colorectal cancer diagnosis (Model 1:
receipt of outpatient evaluation and management outpatient visits for chronic conditions, Model 2: receipt of all-cause
emergency department visits, Model 3: receipt of all-cause hospitalizations), self-reported general health status, self-
reported mental health status, modified Charlson Comorbidity Index score, and diagnoses of diabetes, hyperlipidemia, and
hypertension.
b. Bold results indicate p<0.05.
69
Table 9. Results of simple survey weighted multivariable
a
logistic regression models stratified by patient experience with
getting care quickly analyzing associations between patient race and ethnicity and the odds of 1) at least three outpatient
evaluation and management visits for chronic conditions, 2) any all-cause emergency department visits, and 3) any all-cause
hospitalizations in the 24 months after diagnosis of colorectal cancer (CRC) .
n
Receipt of at least three
outpatient evaluation and
management visits for
chronic conditions
Adjusted OR (99.38% CI)
b
Receipt of any all-cause
emergency department
visits
Adjusted OR (99.38% CI)
b
Receipt of any all-cause
hospitalizations
Adjusted OR (99.38% CI)
b
Excellent experience with
getting care quickly 419
Non-Hispanic White 1.0 (ref)
1.0 (ref) 1.0 (ref)
Non-Hispanic Black 1.86 (1.51, 2.29) 0.84 (0.77, 0.92) 0.54 (0.49, 0.61)
Hispanic 0.44 (0.39, 0.50) 0.37 (0.34, 0.40) 1.66 (1.47, 1.88)
Non-Hispanic Asian 2.92 (2.40, 3.57) 0.54 (0.49, 0.60) 0.67 (0.55, 0.82)
Less than excellent
experience with getting care
quickly 1,467
Non-Hispanic White 1.0 (ref) 1.0 (ref) 1.0 (ref)
Non-Hispanic Black 0.95 (0.90, 1.01) 0.92 (0.88, 0.97) 1.80 (1.66, 1.94)
Hispanic 1.44 (1.36, 1.53) 0.90 (0.86, 0.94) 1.30 (1.21, 1.39)
Non-Hispanic Asian 2.82 (2.63, 3.03) 0.93 (0.89, 0.98) 0.47 (0.45, 0.50)
a. Analyses adjusted for individual patient age at diagnosis, sex, education level, area-level poverty concentration, SEER
region, time from CAHPS survey completion to colorectal cancer diagnosis, year of colorectal cancer diagnosis, stage of
colorectal cancer at diagnosis, care use in the 24 months prior to colorectal cancer diagnosis (Model 1: receipt of outpatient
evaluation and management outpatient visits for chronic conditions, Model 2: receipt of all-cause emergency department
visits, Model 3: receipt of all-cause hospitalizations), self-reported general health status, self-reported mental health status,
modified Charlson Comorbidity Index score, and diagnoses of diabetes, hyperlipidemia, and hypertension.
b. Bold results indicate p<0.00625 (Bonferroni corrected alpha level accounts for eight multiple comparisons per model).
Table 10. Results of simple survey weighted multivariable
a
logistic regression models analyzing the associations between
excellent experience with getting needed care and 1) at least three outpatient evaluation and management visits for chronic
conditions, 2) any all-cause emergency department visits, and 3) any all-cause hospitalizations in the 24 months after
diagnosis of colorectal cancer (CRC) by CAHPS patient experiences with care. (n=1,766)
Excellent Experience with Getting Needed
Care
(ref: less than excellent experience with
getting needed care)
Adjusted OR (95% CI)
b
Outcome Model 1: Receipt of at least three outpatient evaluation and
management visits for chronic conditions 1.29 (1.27, 1.31)
Outcome Model 2: Any all-cause emergency department visits 0.70 (0.69, 0.71)
Outcome Model 3: Any all-cause hospitalizations 0.85 (0.83, 0.86)
a. Analyses adjusted for individual patient age at diagnosis, sex, race and ethnicity, education level, area-level poverty
concentration, SEER region, time from CAHPS survey completion to colorectal cancer diagnosis, year of colorectal cancer
diagnosis, stage of colorectal cancer at diagnosis, care use in the 24 months prior to colorectal cancer diagnosis (Model 1:
receipt of outpatient evaluation and management outpatient visits for chronic conditions, Model 2: receipt of all-cause
emergency department visits, Model 3: receipt of all-cause hospitalizations), self-reported general health status, self-
reported mental health status, modified Charlson Comorbidity Index score, and diagnoses of diabetes, hyperlipidemia, and
hypertension.
b. Bold results indicate p<0.05.
70
Table 11. Results of simple survey weighted multivariable
a
logistic regression models stratified by patient experience with
getting needed care analyzing associations between patient race and ethnicity and the odds of 1) at least three outpatient
evaluation and management visits for chronic conditions, 2) any all-cause emergency department visits, and 3) any all-cause
hospitalizations in the 24 months after diagnosis of colorectal cancer (CRC).
n
Receipt of at least three
outpatient evaluation
and management visits
for chronic conditions
Adjusted OR (99.38% CI)
b
Receipt of any all-cause
emergency department
visits
Adjusted OR (99.38% CI)
b
Receipt of any all-cause
hospitalizations
Adjusted OR (99.38% CI)
b
Excellent experience with
getting needed care 1,194
Non-Hispanic White 1.0 (ref)
1.0 (ref) 1.0 (ref)
Non-Hispanic Black 0.93 (0.87, 1.00) 0.65 (0.61, 0.68) 0.81 (0.76, 0.87)
Hispanic 0.84 (0.79, 0.90) 0.67 (0.63, 0.70) 0.89 (0.83, 0.96)
Non-Hispanic Asian 2.48 (2.23, 2.75) 1.05 (0.99, 1.11) 0.55 (0.51, 0.59)
Less than excellent experience
with getting needed care 572
Non-Hispanic White 1.0 (ref) 1.0 (ref) 1.0 (ref)
Non-Hispanic Black 1.01 (0.92, 1.12) 2.71 (2.47, 2.98) 13.34 (9.84, 18.09)
Hispanic 2.39 (2.18, 2.62) 0.69 (0.65, 0.74) 0.80 (0.71, 0.89)
Non-Hispanic Asian 6.38 (5.71, 7.13) 0.66 (0.61, 0.71) 0.36 (0.33, 0.40)
a. Analyses adjusted for individual patient age at diagnosis, sex, education level, area-level poverty concentration, SEER
region, time from CAHPS survey completion to colorectal cancer diagnosis, year of colorectal cancer diagnosis, stage of
colorectal cancer at diagnosis, care use in the 24 months prior to colorectal cancer diagnosis (Model 1: receipt of outpatient
evaluation and management outpatient visits for chronic conditions, Model 2: receipt of all-cause emergency department
visits, Model 3: receipt of all-cause hospitalizations), self-reported general health status, self-reported mental health status,
modified Charlson Comorbidity Index score, and diagnoses of diabetes, hyperlipidemia, and hypertension.
b. Bold results indicate p<0.00625 (Bonferroni corrected alpha level accounts for eight multiple comparisons per model).
Table 12. Results of simple survey weighted multivariable
a
logistic regression models analyzing the associations between
excellent experience with physician communication and 1) at least three outpatient evaluation and management visits for
chronic conditions, 2) any all-cause emergency department visits, and 3) any all-cause hospitalizations in the 24 months after
diagnosis of colorectal cancer (CRC) by CAHPS patient experiences with care. (n=1,757)
Excellent Experience with Physician
Communication
(ref: less than excellent experience with physician
communication)
Adjusted OR (95% CI)
b
Outcome Model 1: Receipt of at least three outpatient evaluation and
management visits for chronic conditions 1.22 (1.20, 1.24)
Outcome Model 2: Any all-cause emergency department visits 0.82 (0.81, 0.83)
Outcome Model 3: Any all-cause hospitalizations 1.19 (1.17, 1.21)
a. Analyses adjusted for individual patient age at diagnosis, sex, race and ethnicity, education level, area-level poverty
concentration, SEER region, time from CAHPS survey completion to colorectal cancer diagnosis, year of colorectal cancer
diagnosis, stage of colorectal cancer at diagnosis, care use in the 24 months prior to colorectal cancer diagnosis (Model 1:
receipt of outpatient evaluation and management outpatient visits for chronic conditions, Model 2: receipt of all-cause
emergency department visits, Model 3: receipt of all-cause hospitalizations), self-reported general health status, self-
reported mental health status, modified Charlson Comorbidity Index score, and diagnoses of diabetes, hyperlipidemia, and
hypertension.
b. Bold results indicate p<0.05.
71
Table 13. Results of simple survey weighted multivariable
a
logistic regression models stratified by patient experience with
physician communication analyzing associations between patient race and ethnicity and the odds of 1) at least three
outpatient evaluation and management visits for chronic conditions, 2) any all-cause emergency department visits, and 3)
any all-cause hospitalizations in the 24 months after diagnosis of colorectal cancer (CRC).
n
Receipt of at least three
outpatient evaluation
and management visits
for chronic conditions
Adjusted OR (99.38% CI)
b
Receipt of any all-cause
emergency department
visits
Adjusted OR (99.38% CI)
b
Receipt of any all-cause
hospitalizations
Adjusted OR (99.38% CI)
b
Excellent experience with
physician communication 1,213
Non-Hispanic White 1.0 (ref)
1.0 (ref) 1.0 (ref)
Non-Hispanic Black 1.48 (1.38, 1.58) 0.66 (0.63, 0.69) 0.93 (0.87, 0.99)
Hispanic 1.34 (1.26, 1.42) 1.01 (0.96, 1.06) 0.93 (0.87, 0.99)
Non-Hispanic Asian 2.31 (2.12, 2.51) 0.67 (0.64, 0.71) 0.20 (0.19, 0.22)
Less than excellent experience
with physician communication 544
Non-Hispanic White 1.0 (ref) 1.0 (ref) 1.0 (ref)
Non-Hispanic Black 0.91 (0.82, 1.02) 2.29 (2.08, 2.53) 6.74 (5.37, 8.46)
Hispanic 0.64 (0.58, 0.71) 0.83 (0.77, 0.90) 3.53 (3.02, 4.13)
Non-Hispanic Asian 5.39 (4.69, 6.20) 0.61 (0.57, 0.66) 1.34 (1.22, 1.48)
a. Analyses adjusted for individual patient age at diagnosis, sex, education level, area-level poverty concentration, SEER
region, time from CAHPS survey completion to colorectal cancer diagnosis, year of colorectal cancer diagnosis, stage of
colorectal cancer at diagnosis, care use in the 24 months prior to colorectal cancer diagnosis (Model 1: receipt of outpatient
evaluation and management outpatient visits for chronic conditions, Model 2: receipt of all-cause emergency department
visits, Model 3: receipt of all-cause hospitalizations), self-reported general health status, self-reported mental health status,
modified Charlson Comorbidity Index score, and diagnoses of diabetes, hyperlipidemia, and hypertension.
b. Bold results indicate p<0.00625 (Bonferroni corrected alpha level accounts for eight multiple comparisons per model).
72
Table 14. Patient characteristics and experiences with care by change in use of outpatient care for chronic conditions in the 24
months following colorectal cancer (CRC) diagnosis, compared to in the 24 months prior to CRC diagnosis. N (%).
Low care
use
n=383
Increased care
use
n=251
Decreased care
use
n=274
Sufficient care
use
n=1,705 p-value
Experience with Care Coordination
Less than excellent 27 (12.6) 24 (11.2) 18 (8.4) 146 (67.9) 0.9400
Excellent 36 (13.7) 26 (9.9) 20 (7.6) 181 (68.8)
Experience with Getting Care Quickly
Less than excellent 223 (13.7) 163 (10.0) 184 (11.3) 1,053 (64.9) 0.0771
Excellent 59 (12.2) 42 (8.7) 40 (8.3) 344 (70.9)
Experience with Getting Needed Care
Less than excellent 87 (13.8) 57 (9.1) 72 (11.4) 414 (65.7) 0.7699
Excellent 173 (13.1) 136 (10.3) 140 (10.6) 877 (66.1)
Experience with Physician Communication
Less than excellent 75 (12.5) 54 (9.0) 73 (12.2) 396 (66.2) 0.4108
Excellent 172 (12.7) 127 (9.4) 131 (9.7) 925 (68.3)
Age at CRC diagnosis, Mean (SD) 78.0 (6.7) 77.0 (5.7) 78.4 (7.0) 77.8 (6.4) 0.0822
Sex
Male 186 (14.8) 133 (10.6) 128 (10.2) 813 (64.5) 0.4320
Female 197 (14.6) 118 (8.7) 146 (10.8) 892 (65.9)
Race and ethnicity
Non-Hispanic White 346 (15.7) 228 (10.3) 239 (10.8) 1,395 (63.2) <0.0001
Non-Hispanic Black <10% <10% 15 (10.1) 116 (77.9)
Hispanic 19 (14.6) <5% <15% 93 (71.5)
Non-Hispanic Asian <10% <10% <10% 96 (80.7)
Education Level
Less than high school 92 (14.2) 60 (9.2) 70 (10.8) 428 (65.9) 0.4050
High school 120 (13.4) 80 (8.9) 103 (11.5) 595 (66.3)
Greater than high school 171 (16.1) 111 (10.4) 101 (9.5) 682 (64.0)
Area-level poverty concentration
0 - <10% 203 (14.7) 140 (10.1) 140 (10.1) 899 (65.1) 0.6424
10 - <20% 108 (15.5) 58 (8.3) 68 (9.7) 465 (66.5)
20%+ 49 (13.3) 31 (8.4) 44 (11.9) 245 (66.4)
SEER Region
West 186 (19.9) 89 (9.5) 95 (10.2) 565 (60.4) <0.0001
Midwest 54 (16.9) 29 (11.6) 36 (11.3) 201 (62.8)
Northeast 78 (9.6) 69 (8.5) 80 (9.8) 590 (72.2)
South 65 (12.0) 64 (11.8) 63 (11.7) 349 (64.5)
Months from CRC diagnosis to CAHPS
survey completion, Median (IQR) 3.1 (36.3) -1.5 (34.2) 1.4 (34.3) 2.3 (35.3) 0.6676
Year of CRC diagnosis, Median (IQR) 2008 (8) 2008 (7) 2008 (7) 2009 (7) 0.3538
Stage of CRC at diagnosis
In situ or Localized 219 (13.5) 166 (10.3) 163 (10.1) 1,069 (66.1) 0.0835
Regional 164 (16.5) 85 (8.5) 111 (11.1) 636 (63.9)
Self-reported general health status
Fair or Poor 107 (12.6) 82 (9.7) 91 (10.7) 570 (67.1) 0.2592
Good, Very Good, or Excellent 259 (15.6) 160 (9.6) 173 (10.4) 1,074 (64.5)
Self-reported mental health status
Fair or Poor 41 (14.6) 27 (9.6) 35 (12.5) 177 (63.2) 0.7044
Good, Very Good, or Excellent 329 (14.8) 211 (9.5) 227 (10.2) 1,450 (65.4)
Modified Charlson Comorbidity Index Score
a
0 243 (17.5) 145 (10.4) 142 (10.2) 859 (61.8) <0.0001
1 88 (13.6) 66 (10.2) 65 (10.0) 430 (66.3)
2+ 52 (9.0) 40 (7.0) 67 (11.7) 416 (72.4)
73
Table 14. continued
Low care
use
n=383
Increased care
use
n=251
Decreased care
use
n=274
Sufficient care
use
n=1,705 p-value
Diagnosis of diabetes
No 311 (20.3) 185 (12.1) 188 (12.3) 850 (55.4) <0.0001
Yes 72 (6.7) 66 (6.1) 86 (8.0) 855 (79.2)
Diagnosis of hyperlipidemia
No 146 (22.3) 90 (13.7) 90 (13.7) 330 (50.3) <0.0001
Yes 237 (12.1) 161 (8.2) 184 (9.4) 1,375 (70.3)
Diagnosis of hypertension
No 139 (39.5) 56 (15.9) 41 (11.7) 116 (33.0) <0.0001
Yes 244 (10.8) 195 (8.6) 233 (10.3) 1,589 (70.3)
a. Modified Charlson Comorbidity Index excludes cancer and diabetes.
Individual cell counts less than 11 have been suppressed for patient privacy.
74
CHAPTER 3: Investigating Associations Between Patient Race and Ethnicity, Patient Experiences with
Care, and Receipt of Stage-Specific Guideline Concordant Treatment for Colorectal Cancer Among
Medicare Beneficiaries with Comorbid Chronic Conditions
INTRODUCTION
Patients with colorectal cancer and comorbid chronic conditions are less likely to receive
curative surgical resection, adjuvant chemotherapy, and surveillance colonoscopy for their cancer.
11,129–
133
Yet, despite potential complications that additional chronic conditions may present when striving to
provide guideline concordant treatment (GCT) to this population, several studies have demonstrated
that administration of GCT to this patient group may present little to no additional risks to patient safety
under certain circumstances, and that clinical outcomes may improve for this group when patients
receive GCT.
134–136
In fact, the National Comprehensive Cancer Network (NCCN) recommends that older
adults with multimorbidity still be considered as candidates for standard GCT on a case-by-case basis,
with careful consideration of comorbidity management and thorough discussion between providers and
patients.
137
Further complicating disparities in receipt of GCT for colorectal cancer among patients with
multimorbidity are underlying racial and ethnic disparities in receipt of GCT for colorectal cancer.
Compared to non-Hispanic White patients, non-Hispanic Black and Hispanic patients are more likely to
experience delays in receipt of GCT and non-Hispanic Black patients are less likely to receive surgical
resection, chemotherapy, radiotherapy, and surveillance colonoscopy after a diagnosis of colorectal
cancer.
16,18,138
To identify specific components of health care provision which can be targeted and
strengthened to address disparities in care and outcomes, such as those which characterize care for
racially and ethnically diverse patients with colorectal cancer and comorbid chronic conditions, patient
experiences with care can be measured as an integral component of patient-reported quality of care.
75
Patient experiences with care are derived from subjective patient reports of their interactions with their
healthcare providers and the healthcare system and are associated with objective measures of care
quality and outcomes of care.
21,22
Patient experiences with aspects of care like care coordination and
physician communication are especially important among patients with colorectal cancer and comorbid
chronic conditions, as NCCN guidelines specifically advise clear and thorough communication between
patients, primary care providers, and specialty providers in the process of providing care for colorectal
cancer that is as effective and safe as possible for each individual patient.
137
Yet, patient experiences with care are not equal for all patients, with consistent disparities in
experiences reported for patients of different races and ethnicities and for patients with
multimorbidity.
23,24,26
Therefore, it remains important to understand how racial and ethnic differences in
patient experiences with care are associated with differential outcomes for patients and how these
measures may be best leveraged as a tool to combat longstanding racial and ethnic disparities in health
care and clinical outcomes. To identify potential aspects of patient experiences with health care which
may be targeted to address disparities among older adults with both colorectal cancer and comorbid
chronic conditions, this research aims to understand how patient experiences with care relate to receipt
of stage specific GCT for colorectal cancer and how these relationships might change among patients of
different racial and ethnic groups.
METHODS
The Surveillance, Epidemiology, and End Results (SEER)-Consumer Assessment of Healthcare
Providers and Systems (CAHPS) database is a nationally representative and population-based data set
that includes linked data from the National Cancer Institute’s SEER cancer database, Medicare
beneficiary enrollment and claims files, and Medicare CAHPS patient experience surveys. A complete
description of this data has been described elsewhere.
29
76
SEER-CAHPS patients were included in this research if they were diagnosed with stage I-III
colorectal cancer from 2001 to 2017, were at least 67 years old at diagnosis, had continuous Medicare
parts A and B and no Medicare part C from 24 months before to 24 months after colorectal cancer
diagnosis, completed a CAHPS patient experience survey within three years before to three years after
diagnosis, and met diagnostic inclusion criteria for one of three eligible chronic conditions in the pre-
colorectal cancer diagnosis period. A complete flow chart of selection of SEER-CAHPS participants with
colorectal cancer is included in Figure 5.
Comorbid Chronic Conditions
Comorbid chronic condition defining diagnoses included diabetes, hyperlipidemia, and
hypertension, which constitute the three most prevalent Centers for Medicare & Medicaid Services
(CMS) defined chronic conditions which are sensitive to ambulatory care and which present a relatively
low risk of short term mortality.
42
Patients were identified to have diabetes, hyperlipidemia, and/or
hypertension through use of CMS chronic condition identification algorithms in Medicare claims files.
43
A
complete description of identification of chronic condition diagnoses is included in Appendix D.
Patient Race and Ethnicity
Information on individual patient race and ethnicity was derived from self-reported race and
ethnicity on Medicare CAHPS surveys. Patients were first determined to be either Hispanic or non-
Hispanic and, if non-Hispanic, were determined to be either non-Hispanic White, non-Hispanic Black, or
non-Hispanic Asian. If information on race and ethnicity was not self-reported by CAHPS survey, then
information on race and ethnicity was derived from SEER registry data, using the same process for
identifying individual race and ethnicity. If information was not reported on CAHPS or SEER, then race
and ethnicity information was derived from Medicare beneficiary enrollment files.
77
CAHPS Patient Experiences with Care
Medicare CAHPS surveys are widely validated tools to assess patient-reported quality of care
and understand patient views of the health care that they receive as they interact with their health care
providers and the health system.
35,36,46
Answers to individual survey questions are combined within
specific domains of care to create composite scores of patient experience with specific aspects of care
received; these composite scores range from 0-100. CAHPS composite scores investigated in this
research include those which pertain to aspects of health care that are directly received for both
colorectal cancer and chronic conditions, including experiences with care coordination, getting care
quickly, getting needed care, and physician communication. A complete description of the individual
survey questions which comprise each of these composite measures of experience is included in
Appendix E. As composite scores tend to cluster at the upper end of the rating scale, to use composite
scores as a main independent variable in our research we categorized scores as ‘Excellent’ if they were
greater than or equal to 90 and as ‘Less than excellent’ if they were less than 90, as previous work
(including previous work investigating associations between patient race and ethnicity and CAHPS
composite measures of experiences with care) has done.
102,104,139
Receipt of GCT
GCT for colorectal cancer was defined following NCCN stage specific guidelines for the
treatment of colorectal cancer. Specifically, for those diagnosed with stage I or II colorectal cancer, GCT
was defined as receipt of surgical resection within six months of colorectal cancer diagnosis, followed by
surveillance colonoscopy within six to 18 months of surgery. For patients diagnosed with stage III
colorectal cancer, GCT was defined as receipt of surgical resection followed by adjuvant chemotherapy
within 6 months of colorectal cancer diagnosis.
140
Determination of receipt of stage-specific GCT was
performed by identifying receipt of surgical resection within 6 months of colorectal cancer diagnosis in
either Medicare claims files or SEER registry information followed by identification of adjuvant and
78
surveillance procedures in Medicare claims files using approaches outlined in previous work.
141–144
A
complete summary of this approach is described in Appendix H.
Covariates
Covariates included in multivariable analyses included factors which could confound the
relationships between patient experiences with care and receipt of GCT and/or patient race and
ethnicity and receipt of GCT. These factors included patient age at diagnosis, sex (male, female), self-
reported education level (less than high school, high school, greater than high school), area-level
poverty concentration (0-<10%, 10-<20%, 20%+), SEER region (West, Midwest, Northeast, South),
months from colorectal cancer diagnosis to CAHPS survey completion, year of colorectal cancer
diagnosis, colorectal cancer stage at diagnosis (in situ or localized, regional), modified Charlson
Comorbidity Index score in the 12 months prior to colorectal cancer diagnosis (0, 1, 2+), and individual
diagnoses of diabetes, hyperlipidemia, and hypertension.
Statistical Analyses
Descriptive statistics were calculated by patient race and ethnicity, patient experience scores,
and receipt of GCT. Bivariate analyses used chi-squared (difference in proportions), ANOVA (difference
in means), and Kruskal-Wallis (difference in medians) tests. Where individual cell counts were less than
five, the Montecarlo estimate of the Fisher’s exact test was used to test difference in proportions.
Simple survey weighted multivariable logistic regression (adjusting for all covariate factors previously
listed) was performed to measure the associations between patient experiences with care and receipt of
GCT for colorectal cancer. Findings were considered statistically significant when p<0.05. Simple survey
weights were provided in SEER-CAHPS data and were used within this research to “produce estimates
that are representative of the beneficiary populations in the units of the original design,…for the fee-for-
service sample, these were states.”
108
Subsequently, simple survey weighted multivariable logistic
regression was used to assess the relationships between patient race and ethnicity and receipt of GCT
79
by patient experiences with care. To measure effect modification by patient experience, a product
interaction term between measures of patient experience and patient race and ethnicity was included in
multivariable models. Findings of these models were considered statistically significant when p<0.00625,
using a Bonferroni correction to the typical alpha level of 0.05 by accounting for a total of eight multiple
comparisons. All multivariable models were adjusted for all covariates previously listed. Analyses were
performed using SAS version 9.4 (SAS Institute Inc, Cary, NC, USA).
RESULTS
A total of 3,284 SEER-CAHPS patients met inclusion criteria for this research, including 2,817
non-Hispanic White patients (85.8%), 187 non-Hispanic Black patients (5.7%), 149 Hispanic patients
(4.5%), and 131 non-Hispanic Asian patients (4.0%). Bivariate descriptive analyses by patient race and
ethnicity showed that, compared to all other racial and ethnic groups, non-Hispanic White participants
tended to be slightly older at diagnosis of colorectal cancer (mean age=78.2 years versus less than 76.5
years for all other groups). Compared to non-Hispanic White and non-Hispanic Asian participants,
greater proportions of Non-Hispanic Black and Hispanic patients had less than high school education
(48.7% and 53.7%, respectively, versus 21.5% and 23.7%, respectively) and lived in areas with 20% or
more of residents living in poverty (46.0% and 40.6%, respectively, versus 10.6% and 16.8%,
respectively). Compared to all other racial and ethnic groups, a lower proportion of non-Hispanic Black
patients were diagnosed with regional stage colorectal cancer (32.6% versus greater than 39% for all
others). Among all racial and ethnic groups, non-Hispanic Black patients had the highest proportion with
a modified Charlson Comorbidity Index score of 2+ (31.6%) and non-Hispanic Asian patients had the
lowest proportion with a modified score of 2+ (13.0%). In addition, non-Hispanic Black patients had the
highest proportions of those with diabetes (54.6%) and hypertension (94.7%), while non-Hispanic Asian
patients had the highest proportion of those with hyperlipidemia (80.9%) (Table 15).
80
Bivariate descriptive analyses by experiences with care showed that, compared to all other
racial and ethnic groups, a smaller proportion of non-Hispanic Black patients reported excellent
experience with care coordination (<40% versus greater than 53.0% for all others); however, non-
Hispanic Asian patients had the smallest proportions reporting excellent experiences with getting care
quickly (14.3% versus greater than or equal to 22.0% for all others), getting needed care (57.5% versus
greater than greater than 63.0% for all others), and physician communication (61.8% versus greater than
68.0% for all others). Compared to those with high school education or less, a smaller proportion of
those with greater than high school education reported excellent experiences with getting care quickly
(18.9% versus greater than 23.5% for all others) and getting needed care (63.6% versus greater than
70.5% for all others) (Tables 16a and 16b).
Bivariate descriptive analyses by receipt of GCT showed that compared to those who did not
receive stage-specific GCT, those who received GCT tended to be younger (mean age of 75.9 versus 79.2
years), at least high school educated (33.3% of less those with less than high school education versus
greater than 37% for more educated groups), and diagnosed with in situ or localized stage colorectal
cancer (41.1% versus 33.5% of those diagnosed with regional stage). In addition, those with higher
modified Charlson Comorbidity Index scores had lower proportions who received GCT (score of 0:
41.5%, score of 1: 37.4%, score of 2+: 30.8%) (Table 17).
Experiences with care and receipt of GCT
In multivariable models, among our entire cohort there were no measures of experience with
care in which excellent experience was associated with greater likelihood of receiving stage-specific GCT
compared to less than excellent experience (Table 18).
Patient race and ethnicity and receipt of GCT by experiences with care
Among those reporting less than excellent experience with getting care quickly, non-Hispanic
Black patients were 7% less likely than non-Hispanic White patients to receive stage-specific GCT for
81
colorectal cancer (OR=0.93, 99.38% CI=0.89-0.96). However, among those reporting excellent
experience with getting care quickly, non-Hispanic Black patients were 45% more likely than non-
Hispanic White patients to receive GCT (OR=1.45, 99.38% CI=1.36-1.55). In addition, among those
reporting less than excellent experience with getting needed care, Hispanic patients were 7% less likely
than non-Hispanic White patients to receive GCT (OR=0.93, 99.38% CI=0.88-0.99). However, among
those reporting excellent experience with getting needed care, Hispanic patients were 23% more likely
than non-Hispanic White patients to receive GCT (OR=1.23, 99.38% CI=1.17-1.29). In addition, regardless
of experience with getting needed care, non-Hispanic Asian patients were consistently less likely than
non-Hispanic White patients to receive GCT (less than excellent experience: OR=0.64, 99.38% CI=0.60-
0.68; excellent experience: OR=0.93, 99.38% CI=0.88-0.98) (Table 19).
DISCUSSION
Among SEER enrolled Medicare fee-for-service beneficiaries who completed a CAHPS patient
experience survey within three years before or after colorectal cancer diagnosis and had comorbid
chronic conditions prior to colorectal cancer diagnosis, excellent (compared to less than excellent)
patient experiences with care coordination, getting care quickly, getting needed care, and physician
communication were not associated with increased odds of receiving GCT for colorectal cancer.
Analyses of these relationships by patient race and ethnicity were more complex and suggested that
experiences of getting care quickly and getting needed care may be targeted to aid in mitigating racial
and ethnic disparities in GCT for colorectal cancer among non-Hispanic Black and Hispanic patients,
respectively.
Experience with getting care quickly and receipt of GCT among non-Hispanic Black patients
Previous analyses of SEER-CAHPS data have revealed that non-Hispanic Black SEER-CAHPS
participants with colorectal cancer rate experiences with getting care quickly lower than non-Hispanic
White patients.
145
In addition, previous research has shown that non-Hispanic Black patients with
82
excellent experiences with getting care quickly are more likely to be diagnosed with earlier stage
colorectal cancer than non-Hispanic Black patients with less than excellent experiences with this
measure.
146
However, our work is the first to demonstrate that non-Hispanic Black disparities in
experiences with getting care quickly may be related to disparities in receiving stage-specific GCT for
colorectal cancer. Nonetheless, studies of patients with cancer have shown that patient race and
ethnicity may be one of the most influential factors mediating cancer treatment delays, which may both
be reflected in lower scores of patient experiences of the ability to get care quickly as well as contribute
to disparities in receipt of GCT.
138,147–150
Furthermore, this work has found that the strongest mediators
in relationships between minority race and ethnicity and cancer treatment delays may include
socioeconomic status, comorbidity load, and type of care facility delivering treatment.
138,148,150,151
While factors empirically mediating associations between patient race and ethnicity and
timeliness of cancer treatment are largely non-modifiable, health care providers and systems have had
success in mitigating racial and ethnic disparities in timely cancer care through interventions which
support patients in accessing timely care. However, while interventions to specifically improve patient
experiences of care timeliness are not well studied, improvements in objective timeliness of care may
logically lead to improvements in patient experiences of timeliness. One such intervention which was
successful in improving racial and ethnic disparities in timeliness of surgical treatment for lung cancer
utilized an anti-racist framework to implement timely reminders of individual patient and population-
based care needs on the provider side of care, as well as individual patient navigators on the patient side
of care. Implementation of this intervention resulted in both non-Hispanic Black and non-Hispanic White
patients being more likely to receive timely surgery for lung cancer and closed a treatment disparity gap
between non-Hispanic Black and non-Hispanic White patients.
152
Moreover, use of oncology patient
navigation programs, specifically, has been repeatedly shown to improve timeliness of cancer treatment
for racial and ethnic minority patients with cancer at a variety of sites.
153,154
Similar interventions may
83
be effective in promoting racial and ethnic equity in experiences of getting care quickly as well as receipt
of stage-specific GCT for colorectal cancer among non-Hispanic Black older adults with comorbid chronic
conditions.
Experience with getting needed care and receipt of GCT among Hispanic patients
We also found that improvements in patient experiences of getting needed care among Hispanic
patients may mitigate a racial and ethnic disparity in receiving GCT compared to non-Hispanic White
patients. Disparities in access to care for colorectal cancer among Hispanic patients compared to non-
Hispanic White patients are well-documented.
138,155,156
In addition, compared to non-Hispanic White
patients, Hispanic patients disproportionately receive care at less-resourced hospitals with worse overall
clinical and quality of care outcomes.
157–159
Yet, similar to experiences with getting care quickly, there is
little research demonstrating successful interventions for improving patient experiences with getting
needed care. However, by improving objective access to high quality necessary health care, patient
experiences with the ability to get needed care may feasibly improve as well. One scoping review
surveyed the literature regarding disparities in access to cancer care among Hispanic patients and
developed a set of recommendations for clinical practice which included building trust with the
healthcare system, building cultural competency within the healthcare system, increasing enrollment in
oncologic clinical trials, and increasing representation within the healthcare workforce.
160
Interventions
which focus on improving these elements of cancer care delivery among Hispanic patients may be
impactful in both improving Hispanic patient experiences of getting needed care as well as receipt of
stage-specific GCT for colorectal cancer among older adults Hispanic patients with colorectal cancer and
comorbid chronic conditions.
84
Patient experiences with care and receipt of GCT among older adults with colorectal cancer and
comorbid chronic conditions
Despite findings regarding racial and ethnic disparities in receipt of GCT among those with less
than excellent experiences with care, compared to those with excellent experiences with care, we did
not find that excellent experiences with care are associated with greater odds of receiving GCT among
older adults with colorectal cancer and comorbid chronic conditions overall. Snyder et al. found similar
results of no differences in adjusted mean patient experience scores between SEER-CAHPS participants
who did and did not receive GCT for colon cancer. However, our cohort of participants with
multimorbidity makes our findings unique when considering characteristics of this patient population.
Specifically, as previous work has demonstrated, over time those with a higher number of chronic
conditions are more likely to adopt a usual source of health care and are less likely to lose their usual
source of care.
161,162
This may suggest that participants in our study cohort were more likely to have a
usual source of care, including access to a trusted physician and/or care team to assist in pursuing
comprehensive treatment for their colorectal cancer. While the relationship between having a usual
source of care and receiving GCT for cancer is understudied, having a usual source of care is associated
with a range of beneficial health care outcomes, including, as one systematic review found, consistent
increases in use of preventive care services.
163
In addition, having a usual source of care is associated
with higher patient reported quality of care.
164,165
Thus, our findings of excellent experiences with care
not being associated with greater likelihoods of receiving GCT may be partially attributable to a higher
likelihood of having a usual source of care among this specific patient group, which may be acting as a
residual confounding factor in this research.
Similarly, patients with multimorbidity are also more likely to be enrolled in case management
or chronic care management programs, which assign highly trained care navigators to support individual
patient chronic condition management by assisting patients with fulfilling individual care needs. Thus,
85
increased enrollment in case management may also be acting as a residual confounding factor in our
research as participation in care management is associated with both patient experiences with care as
well as with mitigation of barriers in access to comprehensive and necessary care.
166,167
Strengths and Limitations
Our study has several strengths which increase the validity and reliability of this work. This
research was completed using nationally representative and population-based SEER-CAHPS data, which
allowed for a large and reasonably representative sample of US Medicare fee-for-service beneficiaries
with colorectal cancer and comorbid chronic conditions. In addition, use of this data allowed for
extraction of information regarding diagnoses of chronic conditions and receipt of GCT from Medicare
claims files. Medicare claims are considered to be a highly reliable source of information regarding
identification of specific diagnoses and treatments.
168,169
Given the broad and comprehensive individual
patient sociodemographic and clinical factors captured in SEER-CAHPS data, we were also able to control
for a thorough list of individual patient factors which may act as confounding factors between our
independent and dependent variables.
Our research also has some limitations which should be considered in the interpretation of our
findings. Despite inclusion of several covariate factors which could act as confounders in the findings of
our research, there is always the risk of residual confounding in observational studies, and we remained
unable to control for additional potential confounders, such as the type of hospital where colorectal
cancer treatment was received (e.g., community hospital versus comprehensive cancer center),
proximity to health care, and individual patient income level. In addition, to be eligible for this research,
participants needed to have continuous Medicare fee-for-service coverage throughout the study period,
which may limit the external validity of our findings. Nonetheless, given our relatively large sample,
results should remain generalizable to older adults in the US with chronic conditions who are
subsequently diagnosed with colorectal cancer.
86
We identified Hispanic participants as any participant categorized as Hispanic in self-reported
CAHPS data, hospital-reported SEER data, or Medicare-reported enrollment files. This definition of
Hispanic may introduce some misclassification bias into our research if participants in SEER or Medicare
are being labeled Hispanic based on last name, instead of actual ethnicity. Nonetheless, approximately
82% of Hispanic patients included in this research were identified through self-report of Hispanic
ethnicity on CAHPS surveys and approximately 7% of all included Hispanic participants did not report
any race or ethnicity on a CAHPS survey, so we are fairly confident that Hispanic ethnicity is being
accurately represented within our research.
Use of CAHPS survey data may introduce participation bias into the findings of our research
through differential non-response to questions about experiences with care. Missing patient experience
data due to non-response ranged from 31.0% missing responses regarding getting needed care to 82.7%
missing responses regarding care coordination. However, the high rate of nonresponse regarding
experiences with care coordination can be partly attributed to the fact that this composite measure was
not introduced until 2012. Response rates to CAHPS patient experience survey questions did not vary by
colorectal cancer stage or by modified Charlson Comorbidity Index score. However, response rates did
vary by patient race and ethnicity, with a lower proportion of non-Hispanic Black patients responding to
questions regarding care coordination, getting care quickly, and getting needed care; and a higher
proportion of Hispanic patients responding to questions regarding physician communication. Indeed,
non-Hispanic White Medicare beneficiaries tend to have lower rates of non-response to individual items
on Medicare CAHPS surveys and higher overall response rates to Medicare CAHPS surveys than racial
and ethnic minority Medicare beneficiaries.
170
Nonetheless, use of survey weighting in statistical
analyses allowed CAHPS survey responses to be more representative of the Medicare fee-for-service
enrolled populations in each individual US state where participants were sampled.
87
Given a small proportion of SEER-CAHPS patients who met inclusion criteria for this research and
who did not have one of our three comorbid chronic condition defining diagnoses (diabetes,
hyperlipidemia, hypertension), we were unable to compare GCT outcomes by comorbid chronic
condition status. Nonetheless, sensitivity analyses of the associations between chronic condition status
and odds of receiving GCT revealed that, individually, neither of: (1) having at least one comorbid
chronic condition, (2) having diabetes, (3) having hyperlipidemia, (4) having hypertension, nor (5) the
count of comorbid chronic conditions (0-3) were associated with receipt of GCT for colorectal cancer.
Even so, to account for any potential confounding that these factors may introduce into our research, all
multivariable analyses included covariate factors of diabetes, hyperlipidemia, hypertension, and
modified Charlson Comorbidity Index score.
Our findings must also be interpreted with the consideration that this research was cross-
sectional in nature, and we can not show any directional effect of patient experiences with care on
receipt of GCT. In addition, we included participants who responded to CAHPS patient experience
surveys from three years before diagnosis to three years after colorectal cancer diagnosis. The intent of
this inclusion criteria was to capture experiences with care close to the time of colorectal cancer
diagnosis and first course of treatment, while not being so restrictive as to limit the sample size which
we were able to achieve (and the corresponding statistical power of our analyses). To account for any
effects of differential time of CAHPS survey completion, we included a measure of months from
colorectal cancer diagnosis to survey completion as a covariate in all multivariable analyses. In addition,
we completed sensitivity analyses where our cohort was limited to only those who completed a CAHPS
survey (1) within three years before colorectal cancer diagnosis (as experiences reported before
colorectal cancer diagnosis may be more temporally related to GCT outcomes), (2) within two years
before or after colorectal cancer diagnosis (narrowing the participant inclusion period to create more of
a focus on experiences reported near the initial colorectal cancer diagnosis and treatment period), or (3)
88
within six months to three years before or after colorectal cancer diagnosis (excluding experiences
reported during the initial colorectal cancer diagnosis and treatment period). While sample sizes were
not sufficient to complete analyses when our cohort was limited to just those who reported experiences
before diagnosis or when our cohort excluded those who reported experiences in the initial colorectal
cancer diagnosis and treatment period, our final results were mostly consistent with those found when
limiting our cohort to just those who reported experiences two years before or after diagnosis.
Specifically, most discrepancy in findings occurred because findings lost their statistical significance in
our final analysis. Thus, it did not appear necessary to narrow our CAHPS inclusion window to
experiences reported within two years before or after diagnosis.
Conclusion
In conclusion, compared to less than excellent experiences, excellent experiences with care
coordination, getting care quickly, getting needed care, and physician communication were nots
associated with greater likelihoods of receiving GCT for colorectal cancer among Medicare fee-for-
service beneficiaries with comorbid chronic conditions. However, compared to non-Hispanic White
participants, there are disparities in receiving GCT among non-Hispanic Black and Hispanic patients
reporting less than excellent experiences with care. Efforts to improve experiences with getting care
quickly among non-Hispanic Black patients, and getting needed care among Hispanic patients may help
to alleviate racial and ethnic disparities in receipt of GCT among Medicare fee-for-service beneficiaries
with colorectal cancer and comorbid chronic conditions.
89
Figure 5. Selection of study cohort of SEER-CAHPS participants with colorectal cancer.
90
Table 15. Characteristics of SEER-CAHPS participants with pre-existing chronic conditions and non-distant colorectal cancer
(CRC) diagnosed between 2001-2017 at age 67+ who completed a CAHPS patient experience survey within three years of
CRC diagnosis and had continuous Medicare parts A and B (with no part C) coverage in the 2 years before and after CRC
diagnosis. N(%)
Non-Hispanic
White
n=2,817
Non-Hispanic
Black
n=187
Hispanic
n=149
Non-Hispanic
Asian
n=131 p-value
Age at CRC diagnosis, Mean (SD) 78.2 (6.6) 76.4 (6.1) 76.2 (5.6) 76.4 (6.4) <0.0001
Sex
Male 1,350 (47.9) 68 (36.4) 66 (44.3) 61 (46.6) 0.0194
Female 1,467 (52.1) 119 (63.6) 83 (55.7) 70 (53.4)
Education Level
Less than high school 606 (21.5) 91 (48.7) 80 (53.7) 31 (23.7) <0.0001
High school 984 (34.9) 46 (24.6) 31 (20.8) 40 (30.5)
Greater than high school 1,227 (43.6) 50 (26.7) 38 (25.5) 60 (45.8)
Area-level poverty concentration
0 - <10% 1,600 (61.2) 33 (19.0) 48 (33.6) 74 (56.5) <0.0001
10 - <20% 739 (28.3) 61 (35.1) 37 (25.9) 35 (26.7)
20%+ 277 (10.6) 80 (46.0) 58 (40.6) 22 (16.8)
SEER Region
West 793 (28.2) 21 (11.2) 78 (52.4) 114 (87.0) <0.0001
Midwest 310 (11.0) 20 (10.7) <5% <5%
Northeast 1,229 (43.6) 81 (43.3) 54 (36.2) 14 (10.7)
South 485 (17.2) 65 (34.8) <10% <5%
Months from CRC diagnosis to CAHPS
survey completion, Median (IQR) 0.4 (35.4) -1.8 (37.2) 6.8 (34.7) -1.2 (31.7) 0.1158
Year of CRC diagnosis, Median (IQR) 2009 (7) 2009 (6) 2009 (6) 2010 (6) 0.4168
Stage of CRC at diagnosis
In situ or Localized 1,704 (60.5) 126 (67.4) 88 (59.1) 78 (59.5) 0.2856
Regional 1,113 (39.5) 61 (32.6) 61 (40.9) 53 (40.5)
Modified Charlson Comorbidity Index
Score
a
0 1,491 (52.9) 84 (44.9) 88 (59.1) 83 (63.4) 0.0016
1 713 (25.3) 44 (23.5) 28 (18.8) 31 (23.7)
2+ 613 (21.8) 59 (31.6) 33 (22.2) 17 (13.0)
Diagnosis of diabetes 1,116 (39.6) 102 (54.6) 76 (51.0) 63 (48.1) <0.0001
Diagnosis of hyperlipidemia 2,149 (76.3) 131 (70.1) 110 (73.8) 106 (80.9) 0.1178
Diagnosis of hypertension 2,434 (86.4) 177 (94.7) 127 (85.2) 109 (83.2) 0.0070
a. Modified Charlson Comorbidity Index excludes cancer and diabetes.
Individual cell counts less than 11 have been suppressed for patient privacy.
91
Table 16a. Distribution of self-reported patient experiences with care coordination and getting care quickly by patient
characteristics. N (%).
Care Coordination Getting Care Quickly
Less than
excellent
n=298
Excellent
n=352 p-value
Less than
excellent
n=1,984
Excellent
n=583 p-value
Age at colorectal cancer (CRC) diagnosis, Mean
(SD) 79.1 (6.9) 77.9 (6.7) 0.0362 77.8 (6.4) 77.9 (6.6) 0.6670
Sex
Male 146 (45.3) 176 (54.7) 0.7981 977 (80.4) 239 (19.7) 0.0005
Female 152 (46.3) 176 (53.7) 1,007 (74.5) 344 (25.5)
Race and ethnicity
Non-Hispanic White 254 (45.3) 307 (54.7) 0.4459 1,702 (77.2) 503 (22.8) 0.1048
Non-Hispanic Black <65% <40% 95 (72.0) 37 (28.0)
Hispanic <45% <60% 96 (78.1) 27 (22.0)
Non-Hispanic Asian 14 (46.7) 16 (53.3) 84 (85.7) 14 (14.3)
Education Level
Less than high school 49 (45.4) 59 (54.6) 0.5537 447 (72.2) 172 (27.8) <0.0001
High school 89 (43.0) 118 (57.0) 668 (76.2) 209 (23.8)
Greater than high school 160 (47.8) 175 (52.2) 869 (81.1) 202 (18.9)
Area-level poverty concentration
0 - <10% 166 (49.3) 171 (50.7) 0.5269 1,068 (78.0) 301 (22.0) 0.1869
10 - <20% 79 (44.1) 100 (55.9) 514 (75.2) 170 (24.9)
20%+ 41 (48.8) 43 (51.2) 279 (79.7) 71 (20.3)
SEER Region
West 112 (48.1) 121 (51.9) 0.0009 645 (79.0) 172 (21.1) 0.2726
Midwest 20 (39.2) 31 (60.8) 201 (77.3) 59 (22.7)
Northeast 136 (51.7) 127 (48.3) 797 (77.4) 233 (22.6)
South 30 (29.1) 73 (70.9) 341 (74.1) 119 (25.9)
Months from CRC diagnosis to CAHPS survey
completion, Median (IQR) 10.2 (31.0) 7.6 (30.4) 0.2378 0.6 (35.2) 1.0 (35.7) 0.1003
Year of CRC diagnosis, Median (IQR) 2014 (3) 2014 (4) 0.0315 2009 (7) 2008 (6) 0.0195
Stage of CRC at diagnosis
In situ or Localized 174 (45.2) 211 (54.8) 0.6879 1,235 (78.0) 348 (22.0) 0.2643
Regional 124 (46.8) 141 (53.2) 749 (76.1) 235 (23.9)
Modified Charlson Comorbidity Index Score
a
0 140 (44.2) 177 (55.8) 0.6902 1,041 (76.8) 315 (23.2) 0.8011
1 76 (46.9) 86 (53.1) 519 (77.8) 148 (22.2)
2+ 82 (48.0) 89 (52.1) 424 (77.9) 120 (22.1)
Diagnosis of diabetes 138 (48.1) 149 (51.9) 0.3087 826 (77.1) 245 (22.9) 0.8663
Diagnosis of hyperlipidemia 239 (45.6) 285 (54.4) 0.8059 1,499 (77.1) 446 (22.9) 0.6392
Diagnosis of hypertension 272 (47.7) 298 (52.3) 0.0105 1,706 (76.8) 516 (23.2) 0.1168
a. Modified Charlson Comorbidity Index excludes cancer and diabetes.
Individual cell counts less than 11 have been suppressed for patient privacy.
92
Table 16b. Distribution of self-reported patient experiences with getting needed care and physician communication by patient
characteristics. N (%).
Getting Needed Care Physician Communication
Less than
excellent
n=755
Excellent
n=1,601 p-value
Less than
excellent
n=730
Excellent
n=1,651 p-value
Age at colorectal cancer (CRC) diagnosis, Mean
(SD) 77.9 (6.4) 77.8 (6.4) 0.7752 78.4 (6.6) 77.6 (6.4) 0.0076
Sex
Male 365 (31.6) 792 (68.5) 0.6103 347 (30.6) 788 (69.4) 0.9302
Female 390 (32.5) 809 (67.5) 383 (30.7) 863 (69.3)
Race and ethnicity
Non-Hispanic White 642 (31.5) 1,397 (68.5) 0.0827 628 (30.9) 1,405 (69.1) 0.1102
Non-Hispanic Black 33 (28.0) 85 (72.0) 31 (23.1) 103 (76.9)
Hispanic 39 (36.8) 67 (63.2) 37 (31.6) 80 (68.4)
Non-Hispanic Asian 37 (42.5) 50 (57.5) 34 (38.2) 55 (61.8)
Education Level
Less than high school 153 (29.1) 373 (70.9) 0.0003 177 (30.9) 395 (69.1) 0.5210
High school 227 (28.4) 573 (71.6) 262 (31.9) 559 (68.1)
Greater than high school 375 (36.4) 655 (63.6) 291 (29.5) 697 (70.6)
Area-level poverty concentration
0 - <10% 427 (32.9) 873 (67.2) 0.8286 401 (31.4) 876 (68.6) 0.7233
10 - <20% 189 (31.7) 408 (68.3) 196 (31.2) 433 (68.8)
20%+ 103 (33.4) 205 (66.6) 94 (29.1) 229 (70.9)
SEER Region
West 261 (34.3) 501 (65.8) 0.0144 228 (31.5) 495 (68.5) 0.0525
Midwest 70 (28.9) 172 (71.1) 91 (36.7) 157 (63.3)
Northeast 314 (33.7) 618 (66.3) 296 (30.2) 684 (69.8)
South 110 (26.2) 310 (73.8) 115 (26.7) 315 (73.3)
Months from CRC diagnosis to CAHPS survey
completion, Median (IQR) 4.4 (35.2) -1.8 (34.4) 0.0003 -0.7 (35.6) 1.6 (35.3) 0.0515
Year of CRC diagnosis, Median (IQR) 2009 (6) 2007 (7) <0.0001 2008 (8) 2009 (7) 0.0364
Stage of CRC at diagnosis
In situ or Localized 464 (31.8) 995 (68.2) 0.7469 447 (30.3) 1,028 (69.7) 0.6324
Regional 291 (32.4) 606 (67.6) 283 (31.2) 623 (68.8)
Modified Charlson Comorbidity Index Score
a
0 380 (30.6) 861 (69.4) 0.1542 373 (30.3) 860 (69.8) 0.8433
1 196 (32.2) 413 (67.8) 193 (30.6) 437 (69.4)
2+ 179 (35.4) 327 (64.6) 164 (31.7) 354 (68.3)
Diagnosis of diabetes 321 (33.0) 651 (67.0) 0.3935 295 (29.2) 717 (70.9) 0.1697
Diagnosis of hyperlipidemia 576 (32.2) 1,214 (67.8) 0.8058 543 (30.0) 1,270 (70.1) 0.1801
Diagnosis of hypertension 652 (32.0) 1,384 (68.0) 0.9534 637 (30.7) 1,438 (69.3) 0.9135
a. Modified Charlson Comorbidity Index excludes cancer and diabetes.
Individual cell counts less than 11 have been suppressed for patient privacy.
93
Table 17. Distribution of patient characteristics by receipt of stage-specific guideline concordant treatment for colorectal
cancer (CRC). N (%).
Did not receive stage-specific
guideline concordant treatment for
colorectal cancer
n=2,038
Received stage-specific guideline
concordant treatment for
colorectal cancer
n=1,257 p-value
Age at CRC diagnosis
Mean (SD) 79.2 (6.8) 75.9 (5.5) <0.0001
67-<75 595 (53.6) 516 (46.4) <0.0001
75-<85 1,056 (65.7) 551 (34.3)
85+ 491 (85.1) 86 (14.9)
Sex
Male 907 (58.4) 646 (41.6) 0.0001
Female 1,131 (64.9) 611 (35.1)
Race and ethnicity
Non-Hispanic White 1,754 (62.3) 1,063 (37.7) 0.5810
Non-Hispanic Black 113 (60.4) 74 (39.6)
Hispanic 92 (61.7) 57 (38.3)
Non-Hispanic Asian 74 (56.5) 57 (43.5)
Education Level
Less than high school 544 (66.8) 271 (33.3) 0.0002
High school 693 (62.8) 410 (37.2)
Greater than high school 801 (58.2) 576 (41.8)
Area-level poverty concentration
0 - <10% 1,090 (61.9) 671 (38.1) 0.9762
10 - <20% 538 (61.5) 337 (38.5)
20%+ 272 (62.0) 167 (38.0)
SEER Region
West 623 (61.6) 389 (38.4) 0.0018
Midwest 193 (57.6) 142 (42.4)
Northeast 901 (65.2) 482 (34.9)
South 321 (56.8) 244 (43.2)
Months from CRC diagnosis to
CAHPS survey completion, Median
(IQR) 1.2 (35.9) -0.2 (34.0) 0.9624
Year of CRC diagnosis, Median
(IQR) 2009 (7) 2009 (7) 0.0027
Stage of CRC at diagnosis
In situ or Localized 1,179 (58.9) 824 (41.1) <0.0001
Regional 859 (66.5) 433 (33.5)
Modified Charlson Comorbidity
Index Score
a
0 1,022 (58.5) 726 (41.5) <0.0001
1 513 (62.6) 307 (37.4)
2+ 503 (69.2) 224 (30.8)
Diagnosis of diabetes 850 (62.4) 513 (37.6) 0.6119
Diagnosis of hyperlipidemia 1,533 (61.2) 974 (38.9) 0.1387
Diagnosis of hypertension 1,777 (62.2) 1,081 (37.8) 0.3259
a. Modified Charlson Comorbidity Index excludes cancer and diabetes.
94
Table 18. Results of survey weighted multivariable
a
logistic regression models analyzing the odds of receiving stage-specific
guideline concordant treatment for colorectal cancer (CRC) by CAHPS patient experiences with care.
n Adjusted OR (95% CI)
b
Experience with care coordination 596
Less than excellent 1.0 (ref)
Excellent 1.01 (0.99, 1.03)
Experience with getting care quickly 2,393
Less than excellent 1.0 (ref)
Excellent 0.92 (0.91, 0.92)
Experience with getting needed care 2,198
Less than excellent 1.0 (ref)
Excellent 0.84 (0.83, 0.85)
Experience with physician communication 2,250
Less than excellent 1.0 (ref)
Excellent 0.98 (0.97, 1.00)
a. Analyses adjusted for individual patient age at diagnosis, sex, race and ethnicity, education level, area-level poverty
concentration, SEER region, months from CRC diagnosis to CAHPS survey completion, year of colorectal cancer diagnosis,
stage of colorectal cancer at diagnosis, modified Charlson Comorbidity Index score, and diagnoses of diabetes,
hyperlipidemia, and hypertension.
b. Bold results indicate p<0.05.
Table 19. Results of survey weighted multivariable
a
logistic regression models analyzing the odds of receiving stage-specific
guideline concordant treatment for colorectal cancer (CRC) by CAHPS patient experiences with care and patient race and
ethnicity. Adjusted OR (99.38% CI)
b
n
Non-Hispanic
White
Non-Hispanic
Black Hispanic
Non-Hispanic
Asian
Experience with care coordination
c
596
Excellent 1.0 (ref) 0.35 (0.31, 0.41) 0.86 (0.79, 0.94) 0.25 (0.22, 0.28)
Less than excellent 1.0 (ref) 2.39 (2.19, 2.61) 0.91 (0.82, 1.02) 3.46 (3.13, 3.82)
Experience with getting care quickly
c
2,393
Excellent 1.0 (ref) 1.45 (1.36, 1.55) 1.04 (0.96, 1.11) 0.42 (0.37, 0.48)
Less than excellent 1.0 (ref) 0.93 (0.89, 0.96) 1.06 (1.02, 1.10) 1.12 (1.07, 1.17)
Experience with getting needed care
c
2,198
Excellent 1.0 (ref) 1.09 (1.05, 1.14) 1.23 (1.17, 1.29) 0.93 (0.88, 0.98)
Less than excellent 1.0 (ref) 1.05 (0.99, 1.12) 0.93 (0.88, 0.99) 0.64 (0.60, 0.68)
Experience with physician
communication
c
2,250
Excellent 1.0 (ref) 0.94 (0.90, 0.97) 1.05 (1.01, 1.10) 0.63 (0.59, 0.67)
Less than excellent 1.0 (ref) 1.23 (1.15, 1.33) 1.36 (1.28, 1.45) 1.92 (1.80, 2.04)
a. Analyses adjusted for individual patient age at diagnosis, sex, education level, area-level poverty concentration, SEER
region, months from CRC diagnosis to CAHPS survey completion, year of colorectal cancer diagnosis, stage of colorectal
cancer at diagnosis, modified Charlson Comorbidity Index score, and diagnoses of diabetes, hyperlipidemia, and
hypertension.
b. Bold results indicate p<0.00625 (alpha level set using Bonferroni correction to accommodate eight multiple comparisons
analyzed in interaction model).
c. p-value for product interaction term between patient experience with care and patient race and ethnicity <0.0001.
95
CONCLUSION
In our study of the roles of patient experiences with care and patient race and ethnicity in the
care of older adult colorectal cancer patients with comorbid chronic conditions, we found several
associations between excellent experiences with care and more appropriate health care use. In addition,
we identified several instances where improvements in experiences with care among patients of specific
racial and ethnic groups may be associated with improvements in disparate care use among racial and
ethnic minority patients, compared to non-Hispanic White patients. We examined health care use after
diagnosis of colorectal cancer by studying four types of care received: outpatient visits for chronic
conditions, all-cause emergency department visits, all-cause hospitalizations, and stage-specific
guideline concordant treatment for colorectal cancer. Among older adults with colorectal cancer and
comorbid chronic conditions, we hypothesized that excellent experiences with care would be associated
with increased likelihood of sufficient use of outpatient care for chronic conditions, decreased
likelihoods of any all-cause emergency department visits or hospitalizations, and increased likelihood of
receiving guideline concordant treatment for colorectal cancer. In addition, we hypothesized that
excellent experiences with care among racial and ethnic minority patients, compared to non-Hispanic
White patients, would be associated with decreased likelihood of disparate care use.
For all participants in our study cohort, we found that excellent experiences with care
coordination, getting care quickly, getting needed care, and physician communication were associated
with increased likelihoods of sufficient outpatient care use for chronic conditions following colorectal
cancer diagnosis and were not associated with increased likelihoods of receiving guideline concordant
treatment for colorectal cancer. Additionally, we found that: (1) excellent experience with care
coordination was associated with a decreased likelihood of any all-cause hospitalizations; (2) excellent
experience with getting care quickly was associated with a decreased likelihood of any all-cause
emergency department visits; (3) excellent experience with getting needed care was associated with
96
decreased likelihoods of any all-cause emergency department visits and hospitalizations; and (4)
excellent experience with physician communication was associated with a decreased likelihood of any
all-cause emergency department visits.
Compared to non-Hispanic White patients and to those reporting less than excellent
experiences with care, we found that: (1) among non-Hispanic Black patients, excellent experience with
getting care quickly improved disparate likelihoods of sufficient outpatient care and receipt of guideline
concordant treatment and reduced a disparate likelihood of hospitalization, excellent experience with
getting needed care reduced disparate likelihoods of emergency department visits and hospitalizations,
and excellent experience with physician communication improved a disparate likelihood of sufficient
outpatient care and reduced disparate likelihoods of emergency department visits and hospitalizations;
(2) among Hispanic patients, excellent experience with getting needed care improved a disparate
likelihood of guideline concordant treatment and excellent experience with physician communication
improved a disparate likelihood of sufficient outpatient care as well as reduced a disparate likelihood of
hospitalization; and (3) among non-Hispanic Asian patients, excellent experience with physician
communication reduced a disparate likelihood of hospitalization. Thus, for older adults with colorectal
cancer and comorbid diabetes, hyperlipidemia, and/or hypertension, efforts to improve patient
experiences with care among all patients (as well as targeted efforts to improve specific experiences
among racial and ethnic patient subpopulations) may be impactful in seeking to promote appropriate
health services use as well as racial and ethnic equity in health services use following colorectal cancer
diagnosis.
Implications for improving patient experience with care coordination
One service which has been shown to improve both the patient experience of care coordination
as well as objective care coordination of older adults with multimorbidity is chronic care management
offered by Medicare. Chronic care management was introduced for Medicare beneficiaries in 2015 and
97
implements structured follow up with multimorbid patients, as well as standards of care coordination
that providers must meet during both regular follow-up visits and across healthcare provision for
beneficiaries enrolled in chronic care management services.
171
Participants in chronic care management
report improved coordination of care received from various members of their healthcare team,
improved satisfaction with continuity of care from consistent healthcare providers, and easier access to
their healthcare team.
172
However, as a relatively new program offered by Medicare, limited work has
trialed chronic care management as a tool to build care coordination between primary care and
oncology practices. Nonetheless, one pilot project in Vermont sought to implement such a model and
found that, despite several barriers and areas for improvement, implementation of a chronic care
management model specifically for patients with cancer has promise in providing necessary structure to
care coordination between primary care providers and oncologists. This trial project included social
workers, nurse chronic care coordinators, primary care physicians, oncologists, and palliative care
physicians in the care of patients with cancer. Based on the results of their pilot, the authors of the
study suggested that future work aim to delegate specific roles and responsibilities to various healthcare
providers and members of the healthcare team and to routinize and simplify electronic sharing of
patient information between the entire care team.
173
Despite limited evidence from implementation studies of chronic care management among
patients with cancer, qualitative work studying patient experiences with care coordination among
patients with cancer and multimorbidity has revealed specific elements which may be important to
emphasize in chronic care management services for those with cancer. Specifically, a systematic review
among this patient population showed that patients appreciate care coordination which addresses
“conflicts between comorbidity management and cancer treatment”, which patients described as being
overcome through “improved communication between providers and improved physician knowledge of
individual patients’ needs and preferences”. Moreover, the authors of this systematic review posited
98
four major themes of patient-described care coordination needs, including: 1) communication (both
between patients and providers, as well as between providers), 2) defined roles for each member of the
care team, 3) clinical and care plan information that is accessible to patients, and 4) care that is
individualized to specific patient needs.
124
While Medicare chronic care management services may have strong promise in improving
patient perceptions of care coordination, our findings suggest that implementation of chronic care
management for those with colorectal cancer and comorbid chronic conditions must also directly
consider patient perspectives on factors which make care coordination successful. As qualitive work
suggests, important elements of care coordination for this patient population may vary based on
individual patient needs, values, and background and should, therefore, be tailored on both population
and individual levels.
Implications for improving patient experiences with getting care quickly and getting needed care
Common to the patient experiences of getting care quickly and getting needed care are the
experiences of being able to access care that is high quality, comprehensive to individual needs, and
timely. One well-researched and way to improve experiences with access to care and promote racial and
ethnic equity in access to care, is the use of patient navigator programs. Freeman first described a model
of patient navigation as a health care delivery tool to reduce cancer disparities over 25 years ago.
174
While his original model of patient navigation demonstrated strong promise in reducing racial and
ethnic disparities in cancer care, numerous subsequent health care interventions and studies have
demonstrated the wide applicability, feasibility, and utility of patient navigator programs in promoting
increased access to health care (including among vulnerable patient groups and those with
multimorbidity).
167
In a more recent reflection on the evolution of patient navigation as a health care delivery tool,
Freeman described nine principles of patient navigation: (1) “a patient-centered health care service
99
delivery model”, (2) “integrates a fragmented health care system for the individual patient”, (3) “the
core function is the elimination of barriers to timely care across all phases of the health care
continuum”, (4) “should be defined with a clear scope of practice that distinguishes the role and
responsibilities of the navigator from that of all other providers”, (5) “cost-effective and commensurate
with the training and skill necessary”, (6) “determination of who should navigate should be based on the
level of skill required at a given phase of navigation”, (7) “define the point at which navigation begins
and the point at which navigation ends”, (8) “need to navigate patients across disconnected systems of
care”, and (9) “ require coordination”.
175
In addition, since Freeman’s first patient navigation pilot
project, the American Cancer Society (ACS) has been a champion of the patient navigation model for
patients with cancer. In their most recent patient navigation position statement, the ACS reiterated their
support for patient navigation models as an aspect of care delivery that “can and should be a routine
component of the cancer care continuum”. In this recent publication of the ACS, they cite next steps in
the development of patient navigation programs, and in seeking equity through the use of patient
navigation programs, as working to develop sustainable models of care founded in secure
reimbursement programs for services rendered by patient navigators and standardized training
programs for patient navigators.
176
However, to focus on patient experiences of access to care, patient navigator programs must
also focus specifically on patient perceived drivers of successfully accessing needed care in a timely
manner. Further, as the findings of our study suggest, patient perceptions should be drawn from specific
racial and ethnic patient groups to promote equity in care. For example, among a cohort of non-Hispanic
Black patients in the southern US, perceived barriers to care included lack of available appointment
times, financial barriers to care, and limited literacy with regards to accessing health care.
177
In addition,
among a cohort of non-Hispanic Black women with breast cancer, barriers to timely cancer treatment
included things like mistrust and fear of the healthcare system, financial barriers, transportation
100
barriers.
178
Among Hispanic patients with cancer, major healthcare needs revealed through qualitative
studies include support receiving and understanding cancer information, communicating care needs,
overcoming language barriers and financial anxieties, accessing needed care and diagnostic information
in a timely manner, and dealing with mental health challenges.
179–181
Moreover, among both non-
Hispanic Black and Hispanic patients, a key facilitator of accessing necessary and timely care which was
identified by patients was access to skilled and culturally competent care navigators to support patients
in managing their health conditions and accessing needed care.
178,179
Thus, to seek equity in the patient experiences of getting care quickly and getting needed care
among non-Hispanic Black and Hispanic patients with colorectal cancer and comorbid chronic
conditions, interventions such as patient navigator programs should seek to not only be culturally and
linguistically appropriate, but to be informed by the specific needs which are most prevalent among this
patient population. Moreover, utilization of patient navigator programs which encompass all nine of
Freeman’s principles of patient navigation, and which account for modern challenges of patient
navigation, such as secure funding and reimbursement sources as well as sustainably maintaining a
skilled workforce of patient navigators, may be an important aspect of building racial and ethnic equity
in timely and appropriate access to health care for older adults with colorectal cancer and comorbid
chronic conditions.
Implications for improving patient experience with physician communication
We can also look to findings of qualitative work to better understand patient perspectives on
what makes physician communication successful. Specifically, qualitative research has shown that
patients with cancer value physician communication which considers time, place, and patient mindset
and background when discussing high-impact diagnoses and treatments; is direct and honest; delivered
appropriately for individual health literacy and concordant with individual language preference; directly
involves family and caretakers; invites patient questions and concerns; validates patient symptoms; and
101
is caring, empathic, and respectful.
117–120
However, our study also found that excellent experiences with
physician communication are associated with mitigation of racial and ethnic disparities in health care
use among non-Hispanic Black, Hispanic, and non-Hispanic Asian patients. Qualitative research seeking
to understand drivers of patient perceptions of physician communication among non-Hispanic Black
patients has found that there are several determinants of patient perceived physician communication
which are unique to non-Hispanic Black patients with cancer, compared to non-Hispanic White patients
with cancer. Specifically, physician communication perceived by non-Hispanic Black patients with cancer
more commonly involves having to request and seek out medical information (as opposed to being told
relevant medical information without having to ask), relying on patient advocates to assist with
processing medical information communicated by providers, and experiencing disappointment with
communication from healthcare providers.
118
To improve patient experiences with physician communication for the broad group of older
adults with colorectal cancer and comorbid chronic conditions, health care providers and systems
should place a focus on physician communication training based in qualitative evidence regarding
elements of physician communication that patients with cancer most value. Yet, to promote racial and
ethnic equity in experiences with physician communication among this patient population, it will be
necessary to further develop physician communication training based on the needs of groups which
might benefit from improved experiences of communication the most, such as non-Hispanic Black,
Hispanic, and non-Hispanic Asian patients. Qualitative work focused on patients of these racial and
ethnic backgrounds can be used to inform racial and ethnic specific physician communication training.
However, more work is needed to completely understand determinants of experiences with physician
communication among racial and ethnic subsets of older adults with colorectal cancer and comorbid
chronic conditions.
102
Summary of recommendations for clinical practice
Based on individual findings discussed in the previous chapters, as well as collective findings
discussed in this section, we have developed a set of recommendations for clinical practice. These
recommendations are intended specifically for the care of older adults with colorectal cancer and
comorbid chronic conditions and may be impactful in promoting appropriate health services use
following colorectal cancer diagnosis in this vulnerable patient population. Our recommendations are as
follows:
• Guaranteed access to highly-trained and culturally and linguistically appropriate care navigation
which can support patients in building both health and healthcare literacy, as well as assist
patients in accessing necessary care in a timely manner.
• Structural changes to health care provision which lower barriers to care access, including:
o increased quantity of available and timely medical appointments,
o decreased costs of medical care,
o direct transportation to and from medical care at no charge to patients,
o implementation of measures which ensure patients have longitudinal continuity of care
with the same set of health care providers, and
o increased representation of racial and ethnic minority health care providers who can
bring increased cultural competency to clinical settings and help patients to feel
increased trust in their providers and the health care system.
• Establishment of systematic care coordination between primary care providers, oncologists,
specialist providers, and additional members of individual patients’ healthcare teams, including:
o clearly defined and standardized roles for each member of the care team, and
103
o care team training regarding effective medical management of patients with cancer and
additional chronic conditions as a collective health care team and from the perspectives
of individual members of the care team.
• Communication training for health care providers in which instructional material is informed by
patient perspectives of effective communication in clinical settings, and which is mandatory at
each career/training stage and with standardized frequency throughout the career.
Next steps and future directions
Despite our findings, and the suggestions for strengthening clinical practice that they implicate,
our study had limitations which also prompt recommendations for future research to better understand
the associations between patient experiences with care, patient race and ethnicity, and post-colorectal
cancer diagnosis health services use among older adults with comorbid chronic conditions. Specifically,
we were limited to studying older adults with continuous fee-for-service Medicare coverage, and no
Medicare Advantage coverage. Future research should seek to understand relationships among older
adults who may have different combinations of Medicare coverage and/or inconsistent Medicare
coverage. In addition, this research included only the most common chronic conditions which are
amenable to medical care as eligible comorbid chronic conditions. Future research should seek
innovative ways to broaden this definition of multimorbidity. This could include use of a scale of
comorbidities (such as the Charlson Comorbidity Index included as a covariate in this research), use of
an illness burden index (such as the forthcoming SEER-CAHPS illness burden index), or broader use of
distinct diagnoses of chronic comorbid conditions.
In addition, Medicare CAHPS responses are derived from an annual stratified sample (based on
Medicare plans and geography) of Medicare beneficiaries, and survey completion rates are known to be
disparate in racial and ethnic minority groups. Therefore, future administration of the Medicare CAHPS
survey may seek to attempt purposive sampling of racial and ethnic minority beneficiaries to increase
104
our collective understanding of experiences with healthcare among racial and ethnic minority Medicare
beneficiaries. Use of purposive sampling by patient race and ethnicity may also help to increase
available data among racial and ethnic groups which had too little current data to be included in our
research, such as American Indian and Alaska Native populations and mixed-race populations.
Expanding the cohort of SEER-CAHPS participants eligible for this work, through efforts like
increasing racial and ethnic minority participation in CAHPS surveys and broadening the definition of
eligible chronic conditions may also enable the completion of analyses which we were not able to
complete in this work due to sample size limitations. These analyses included multivariable analyses of
associations between patient experiences with care, patient race and ethnicity, and change in use of
outpatient care for chronic conditions following colorectal cancer diagnosis as well as analyses
measuring associations of patient race and ethnicity with health services use following colorectal cancer
diagnosis by experiences with care coordination. While we were unable to complete this work under the
constraints of our current study cohort, these research questions remain important to explore to
develop a more complete understanding of associations between patient experiences with care, patient
race and ethnicity, and health services use following colorectal diagnosis among older adults with
comorbid chronic conditions.
In addition, our findings that excellent experiences with care were not associated with increased
odds of receiving stage-specific guideline concordant treatment for colorectal cancer may be
complicated by residual confounding that may exist in these analyses due to our inability to control for
an increased use of systemic healthcare supports such as case management and having a usual source of
healthcare among older adults with multimorbidity. Future work should seek to further explore and
verify these associations with attempts to control for these potentially confounding factors.
Our research also identified racial and ethnic disparities in associations between experiences
with care and health services use following colorectal cancer diagnosis; however, we were not able to
105
identify the root causes of these disparities. Future work, including an increased emphasis on qualitative
work like that of Ochoa et al., should seek to more comprehensively understand the causes of disparities
in associations between experiences with care and clinical and quality outcomes of care so that
attempts to seek and promote racial and ethnic health equity can be most effective.
85
In addition, future work should seek to broaden our understanding between racial and ethnic
differences in patient experiences with care and outcomes among older adults with colorectal cancer
and comorbid chronic conditions. To broaden these understandings, it will be important to study how
experiences and racial and ethnic differences in experiences are associated with outcomes that were not
studied in this research, like mortality and overall cost of care.
In conclusion, our study identified several aspects of patient experiences with care which are
associated with health services use, as well as with racial and ethnic disparities in this use, among older
adults with colorectal cancer and comorbid chronic conditions. To promote appropriate use of, and
racial and ethnic equity in use of, post-colorectal cancer diagnosis health services among this
population, we recommend targeted (to both this vulnerable patient population, as well as to
individually identified racial and ethnic subgroups facing disparities in care) implementation of health
care delivery programs which improve patient experiences of care coordination, the abilities to get care
quickly and get needed care, and physician communication.
106
REFERENCES
1. Siegel RL, Sandeep Wagle N, Cercek A, Smith RA, Ahmedin J. Colorectal cancer statistics, 2023. CA
Cancer J Clin. Published online March 2023:1-22. doi:10.3322/CAAC.21772
2. Siegel RL, Miller KD, Goding Sauer A, et al. Colorectal cancer statistics, 2020. CA Cancer J Clin.
2020;70(3):145-164. doi:10.3322/caac.21601
3. Wang R, Wang MJ, Ping J. Clinicopathological Features and Survival Outcomes of Colorectal
Cancer in Young Versus Elderly. Med (United States). 2015;94(35):e1402.
doi:10.1097/MD.0000000000001402
4. Boersma P, Black LI, Ward BW. Prevalence of Multiple Chronic Conditions Among US Adults,
2018. Prev Chronic Dis. 2020;17. doi:10.5888/pcd17.200130
5. Renzi C, Kaushal A, Emery J, et al. Comorbid chronic diseases and cancer diagnosis: disease-
specific effects and underlying mechanisms. Nat Rev Clin Oncol. 2019;16(12):746-761.
doi:10.1038/S41571-019-0249-6
6. Snyder CF, Frick KD, Herbert RJ, et al. Quality of care for comorbid conditions during the
transition to survivorship: Differences between cancer survivors and noncancer controls. J Clin
Oncol. 2013;31(9):1140-1148. doi:10.1200/JCO.2012.43.0272
7. Pinheiro LC, Soroka O, Kern LM, Leonard JP, Safford MM. Diabetes care management patterns
before and after a cancer diagnosis: A SEER-Medicare matched cohort study. Cancer.
2020;126(8):1727-1735. doi:10.1002/cncr.32728
8. Pinheiro LC, Kaur H, Nilo D, Safford MM, Derosa AP, Kern LM. Determining the impact of a cancer
diagnosis on diabetes management: A systematic literature review. Am J Clin Oncol Cancer Clin
Trials. 2019;42(11):870-883. doi:10.1097/COC.0000000000000612
9. Griffiths RI, Keating NL, Bankhead CR. Quality of diabetes care in cancer: A systematic review. Int
J Qual Heal Care. 2019;31(2):75-88. doi:10.1093/intqhc/mzy124
10. Earle CC, Neville BA. Under use of necessary care among cancer survivors. Cancer.
2004;101(8):1712-1719. doi:10.1002/CNCR.20560
11. Boakye D, Nagrini R, Ahrens W, Haug U, Günther K. The association of comorbidities with
administration of adjuvant chemotherapy in stage III colon cancer patients: a systematic review
and meta-analysis. Ther Adv Med Oncol. 2021;13. doi:10.1177/1758835920986520
12. Pinheiro LC, Soroka O, Kern LM, Higgason N, Leonard JP, Safford MM. Racial disparities in
diabetes care among incident breast, prostate, and colorectal cancer survivors: a SEER Medicare
study. J Cancer Surviv. Published online March 4, 2021. doi:10.1007/s11764-021-01003-z
13. Pinheiro LC, Soroka O, Kern LM, Leonard JP, Safford MM. Racial Disparities in Diabetes-Related
Emergency Department Visits and Hospitalizations Among Cancer Survivors. JCO Oncol Pract.
2022;18(6):e1023-e1033. doi:10.1200/OP.21.00684
107
14. Palmer N, Geiger A, Felder T, Lu L, Case L, Weaver K. Racial/Ethnic disparities in health care
receipt among male cancer survivors. Am J Public Health. 2013;103(7):1306-1313.
doi:10.2105/AJPH.2012.301096
15. Gorey KM, Haji-Jama S, Bartfay E, Luginaah IN, Wright FC, Kanjeekal SM. Lack of access to
chemotherapy for colon cancer: multiplicative disadvantage of being extremely poor,
inadequately insured and African American. BMC Heal Serv Res 2014 141. 2014;14(1):1-7.
doi:10.1186/1472-6963-14-133
16. Simpson D, Martínez M, Gupta S, et al. Racial disparity in consultation, treatment, and the impact
on survival in metastatic colorectal cancer. J Natl Cancer Inst. 2013;105(23):1814-1820.
doi:10.1093/JNCI/DJT318
17. Esnaola NF, Stewart AK, Feig BW, Skibber JM, Rodriguez-Bigas MA. Age-, race-, and ethnicity-
related differences in the treatment of nonmetastatic rectal cancer: A patterns of care study
from the national cancer data base. Ann Surg Oncol. 2008;15(11):3036-3047.
doi:10.1245/s10434-008-0106-9
18. Sanchez JI, Shankaran V, Unger JM, Madeleine MM, Selukar SR, Thompson B. Inequitable access
to surveillance colonoscopy among Medicare beneficiaries with surgically resected colorectal
cancer. Cancer. 2021;127(3):412-421. doi:10.1002/CNCR.33262
19. Andersen RM. Revisiting the behavioral model and access to medical care: does it matter? J
Health Soc Behav. 1995;36(1):1-10. doi:10.2307/2137284
20. Agency for Healthcare Research and Quality. About CAHPS. Published June 2021. Accessed
August 17, 2021. https://www.ahrq.gov/cahps/about-cahps/index.html
21. Anhang Price R, Elliott MN, Zaslavsky AM, et al. Examining the role of patient experience surveys
in measuring health care quality. Med Care Res Rev. 2014;71(5):522-554.
doi:10.1177/1077558714541480
22. Navarro S, Ochoa CY, Chan E, Du S, Farias AJ. Will improvements in patient experience with care
impact clinical and quality of care outcomes?: A systematic review. Med Care. 2021;59(9):843-
856. doi:10.1097/MLR.0000000000001598
23. Martino SC, Mathews M, Agniel D, et al. National racial/ethnic and geographic disparities in
experiences with health care among adult Medicaid beneficiaries. Health Serv Res. 2019;54(Suppl
1):287-296. doi:10.1111/1475-6773.13106
24. Halpern MT, Urato MP, Lines LM, Cohen JB, Arora NK, Kent EE. Healthcare experience among
older cancer survivors: Analysis of the SEER-CAHPS dataset. J Geriatr Oncol. 2018;9(3):194-203.
doi:10.1016/j.jgo.2017.11.005
25. Martino SC, Weinick RM, Kanouse DE, et al. Reporting CAHPS and HEDIS data by race/ethnicity
for medicare beneficiaries. Health Serv Res. 2013;48(2 PART1):417-434. doi:10.1111/j.1475-
6773.2012.01452.x
26. Orr N, Elliott MN, Burkhart Q, Haviland A, Weinick RM. Racial/Ethnic differences in Medicare
108
experiences and immunization: the role of disease burden. Med Care. 2013;51(9):823-831.
doi:10.1097/MLR.0b013e31829c8d77
27. Martino SC, Elliott MN, Haviland AM, Saliba D, Burkhart Q, Kanouse DE. Comparing the Health
Care Experiences of Medicare Beneficiaries with and without Depressive Symptoms in Medicare
Managed Care versus Fee-for-Service. Health Serv Res. 2016;51(3):1002-1020. doi:10.1111/1475-
6773.12359
28. Keenan PS, Elliott MN, Cleary PD, Zaslavsky AM, Landon BE. Quality Assessments by Sick and
Healthy Beneficiaries in Traditional Medicare and Medicare Managed Care. Med Care.
2009;47(8):882-888. doi:10.1097/MLR.0b013e3181a39415
29. Chawla N, Urato M, Ambs A, et al. Unveiling SEER-CAHPS®: A New Data Resource for Quality of
Care Research. J Gen Intern Med. 2015;30(5):641-650. doi:10.1007/s11606-014-3162-9
30. Number of Persons by Race and Hispanic Ethnicity for SEER Participants - SEER Registries.
Accessed August 11, 2021. https://seer.cancer.gov/registries/data.html#a6
31. SEER Program & Data - SEER-CAHPS. Accessed August 11, 2021.
https://healthcaredelivery.cancer.gov/seer-cahps/aboutdata/program.html
32. Population Characteristics - SEER Registries. Accessed August 11, 2021.
https://seer.cancer.gov/registries/characteristics.html
33. Harris-Kojetin LD, Fowler FJ, Brown JA, Schnaier JA, Sweeny SF. The use of cognitive testing to
develop and evaluate CAHPS
TM
1.0 core survey items. Med Care. 1999;37(3 SUPPL. 1).
doi:10.1097/00005650-199903001-00002
34. Hays RD, Shaul JA, Williams VSL, et al. Psychometric properties of the CAHPS
TM
1.0 survey
measures. Med Care. 1999;37(3 SUPPL. 1). doi:10.1097/00005650-199903001-00003
35. Cleary PD, Lubalin J, Hays RD, Short PF, Edgman-Levitan S, Sheridan S. Debating survey
approaches. Health Aff (Millwood). 1998;17(1):265-268. doi:10.1377/hlthaff.17.1.265
36. Hargraves JL, Hays RD, Cleary PD. Psychometric Properties of the Consumer Assessment of
Health Plans Study (CAHPS®) 2.0 Adult Core Survey. Health Serv Res. 2003;38(6 I):1509-1527.
doi:10.1111/j.1475-6773.2003.00190.x
37. Hays RD, Martino S, Brown JA, et al. Evaluation of a care coordination measure for the consumer
assessment of healthcare providers and systems (CAHPS®) medicare survey. Med Care Res Rev.
2014;71(2):192-202. doi:10.1177/1077558713508205
38. Medicare Advantage and Prescription Drug Plan (MA & PDP) CAHPS ® Survey: Quality Assurance
Protocols & Technical Specifications.; 2020. Accessed August 11, 2021.
https://www.cms.gov/files/document/ma-pdp-cahps-qapts-v11-complete-manual.pdf
39. Klein DJ, Elliott MN, Haviland AM, et al. Understanding nonresponse to the 2007 medicare CAHPS
survey. Gerontologist. 2011;51(6):843-855. doi:10.1093/geront/gnr046
109
40. Research Triangle Institute. Guidance Document: Missing Data in SEER-CAHPS. Accessed
November 5, 2021. https://healthcaredelivery.cancer.gov/seer-cahps/researchers/missing-data-
guidance.pdf
41. National Cancer Institute Division of Cancer Control & Population Sciences. How to Handle
Variables That Occur in Multiple Files in SEER-CAHPS. Published August 29, 2022. Accessed
January 30, 2023. https://healthcaredelivery.cancer.gov/seer-cahps/researchers/multiple-file-
variables.html
42. Centers for Medicare & Medicaid Services. Chronic Conditions Overview | CMS. Published
January 15, 2021. Accessed November 23, 2021. https://www.cms.gov/Research-Statistics-Data-
and-Systems/Statistics-Trends-and-Reports/Chronic-Conditions
43. Chronic Conditions Warehouse, Centers for Medicare & Medicaid Services. 27 CCW Chronic
Conditions Algorithms.; 2022. Accessed October 28, 2022.
https://www2.ccwdata.org/web/guest/condition-categories-chronic
44. Isaac T, Zaslavsky AM, Cleary PD, Landon BE. The Relationship between Patients’ Perception of
Care and Measures of Hospital Quality and Safety. Health Serv Res. 2010;45(4):1024-1040.
doi:10.1111/j.1475-6773.2010.01122.x
45. Quigley DD, Mendel PJ, Predmore ZS, Chen AY, Hays RD. Use of CAHPS® patient experience
survey data as part of a patient-centered medical home quality improvement initiative. J Healthc
Leadersh. 2015;7:41-54. doi:10.2147/JHL.S69963
46. Hays RD, Shaul JA, Williams VS, et al. Psychometric properties of the CAHPS 1.0 survey measures.
Consumer Assessment of Health Plans Study. Med Care. 1999;37(3 Suppl):MS22-31. Accessed
October 21, 2019. http://www.ncbi.nlm.nih.gov/pubmed/10098556
47. Adams Z, Harder B. Patient Experience and Patient Safety in the Specialty Rankings. U.S. News &
World Report. Published January 11, 2019. Accessed April 19, 2020.
https://health.usnews.com/health-news/blogs/second-opinion/articles/2019-01-11/patient-
experience-and-patient-safety-in-the-specialty-rankings
48. Medicare Advantage and Prescription Drug Plan CAHPS (MA and PDP CAHPS) | CMS. CMS.gov.
Published February 11, 2020. Accessed April 19, 2020. https://www.cms.gov/Research-Statistics-
Data-and-Systems/Research/CAHPS/MCAHPS
49. Survey of patients’ experiences (HCAHPS) star ratings. Medicare.gov. Accessed April 19, 2020.
https://www.medicare.gov/hospitalcompare/Data/HCAHPS-Star-Ratings.html
50. HOS and the Medicare Star Ratings . Medicare Health Outcomes Survey. Published February 18,
2020. Accessed April 19, 2020. https://www.hosonline.org/en/hos-and-the-star-ratings/
51. Squitieri L, Chung KC. Measuring Provider Performance for Physicians Participating in the Merit-
Based Incentive Payment System. Plast Reconstr Surg. 2017;140(1):217e-226e.
doi:10.1097/PRS.0000000000003430
52. Weinick RM, Quigley DD, Mayer LA, Sellers CD. Use of CAHPS patient experience surveys to
110
assess the impact of health care innovations. Jt Comm J Qual Patient Saf. 2014;40(9):418-427.
doi:10.1016/S1553-7250(14)40054-0
53. Rathert C, Wyrwich MD, Boren SA. Patient-centered care and outcomes: A systematic review of
the literature. Med Care Res Rev. 2013;70(4):351-379. doi:10.1177/1077558712465774
54. Anhang Price R, Elliott MN, Cleary PD, Zaslavsky AM, Hays RD. Should Health Care Providers be
Accountable for Patients’ Care Experiences? J Gen Intern Med. 2014;30(2):253-259.
doi:10.1007/s11606-014-3111-7
55. Otani K. Enrollees’ global rating process of health care with the national CAHPS® benchmarking
database. Health Care Manage Rev. 2006;31(3):205-212. doi:10.1097/00004010-200607000-
00006
56. Pan American Health Organization (PAHO), World Health Organization (WHO). Causes of Death
Considered Amenable to Health Care.; 2013.
https://www.paho.org/hq/dmdocuments/2013/annex-basic-indicators-2013.pdf.pdf
57. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-
analyses: The PRISMA statement. PLoS Med. 2009;6(7). doi:10.1371/journal.pmed.1000097
58. Elliott MN, Beckett MK, Chong K, Hambarsoomians K, Hays RD. How do proxy responses and
proxy-assisted responses differ from what medicare beneficiaries might have reported about
their health care? Health Serv Res. 2008;43(3):833-848. doi:10.1111/j.1475-6773.2007.00820.x
59. Chen Q, Beal EW, Schneider EB, Okunrintemi V, Zhang X feng, Pawlik TM. Patient-Provider
Communication and Health Outcomes Among Individuals with Hepato-Pancreato-Biliary Disease
in the USA. J Gastrointest Surg. 2018;22(4):624-632. doi:10.1007/s11605-017-3610-z
60. Chakrabarti AK, Sheetz KH, Katariya NN, et al. Do Patient Assessments of Hospital Quality
Correlate With Kidney Transplantation Surgical Outcomes? Transplant Proc. 2016;48(6):1986-
1992. doi:10.1016/j.transproceed.2016.03.038
61. Dottino JA, He W, Sun CC, et al. Centers for Medicare and Medicaid Services’ Hospital Consumer
Assessment of Healthcare Providers and Systems (HCAHPS) scores and gynecologic oncology
surgical outcomes. Gynecol Oncol. 2019;154(2):405-410. doi:10.1016/j.ygyno.2019.06.003
62. Dy SM, Chan KS, Chang HY, Zhang A, Zhu J, Mylod D. Patient perspectives of care and process and
outcome quality measures for heart failure admissions in US hospitals: How are they related in
the era of public reporting? Int J Qual Heal Care. 2016;28(4):522-528.
doi:10.1093/intqhc/mzw063
63. Farias AJ, Ochoa CY, Toledo G, Bang S-I, Hamilton AS, Du XL. Racial/ethnic differences in patient
experiences with health care in association with earlier stage at breast cancer diagnosis: findings
from the SEER-CAHPS data. Cancer Causes Control. 2020;31(1):13-23. doi:10.1007/s10552-019-
01254-3
64. Fernandez A, Seligman H, Quan J, Stern RJ, Jacobs EA. Associations between aspects of culturally
competent care and clinical outcomes among patients with diabetes. Med Care. 2012;50(9
111
SUPPL. 2):S74-9. doi:10.1097/MLR.0b013e3182641110
65. Odom-Maryon T, Hu HM, Tzeng HM. Associations between Hospital-Level Patient Satisfaction
Scores and Hospital-Acquired Pressure Ulcer Occurrences among Medicare Stroke Patients. J
Nurs Care Qual. 2019;34(4):364-369. doi:10.1097/NCQ.0000000000000396
66. Okunrintemi V, Spatz ES, Di Capua P, et al. Patient-Provider Communication and Health
Outcomes among Individuals with Atherosclerotic Cardiovascular Disease in the United States:
Medical Expenditure Panel Survey 2010 to 2013. Circ Cardiovasc Qual Outcomes. 2017;10(4).
doi:10.1161/CIRCOUTCOMES.117.003635
67. Shirk JD, Tan HJ, Hu JC, Saigal CS, Litwin MS. Patient experience and quality of urologic cancer
surgery in US hospitals. Cancer. 2016;122(16):2571-2578. doi:10.1002/cncr.30081
68. Yang L, Liu C, Huang C, Mukamel DB. Patients’ perceptions of interactions with hospital staff are
associated with hospital readmissions: A national survey of 4535 hospitals. BMC Health Serv Res.
2018;18(1). doi:10.1186/s12913-018-2848-9
69. Schneider EC, Zaslavsky AM, Landon BE, Lied TR, Sheingold S, Cleary PD. National quality
monitoring of Medicare health plans: the relationship between enrollees’ reports and the quality
of clinical care. Med Care. 2001;39(12):1313-1325. doi:10.1097/00005650-200112000-00007
70. Bauer AM, Parker MM, Schillinger D, et al. Associations between antidepressant adherence and
shared decision-making, patient-provider trust, and communication among adults with diabetes:
Diabetes Study of Northern California (DISTANCE). J Gen Intern Med. 2014;29(8):1139-1147.
doi:10.1007/s11606-014-2845-6
71. Chang JT, Hays RD, Shekelle PG, et al. Patients’ Global Ratings of Their Health Care Are Not
Associated with the Technical Quality of Their Care. Ann Intern Med. 2006;144(9):665.
doi:10.7326/0003-4819-144-9-200605020-00010
72. Elder K, Meret-Hanke L, Dean C, et al. How Do African American Men Rate Their Health Care? An
Analysis of the Consumer Assessment of Health Plans 2003-2006. Am J Mens Health.
2015;9(3):178-185. doi:10.1177/1557988314532824
73. Hays RD, Mallett JS, Haas A, et al. Associations of CAHPS Composites With Global Ratings of the
Doctor Vary by Medicare Beneficiaries’ Health Status. Med Care. 2018;56(8):736-739.
doi:10.1097/MLR.0000000000000942
74. Gupta A, Daigle S, Mojica J, Hurley RW. Patient perception of pain care in hospitals in the United
States. J Pain Res. 2009;2(2):157-164. doi:10.2147/jpr.s7903
75. Kshirsagar A V., Tabriz AA, Bang H, Lee SYD. Patient Satisfaction Is Associated With Dialysis
Facility Quality and Star Ratings. Am J Med Qual. 2019;34(3):243-250.
doi:10.1177/1062860618796310
76. Ratanawongsa N, Karter AJ, Parker MM, et al. Communication and medication refill adherence
the diabetes study of Northern California. JAMA Intern Med. 2013;173(3):210-218.
doi:10.1001/jamainternmed.2013.1216
112
77. Winpenny E, Elliott MN, Haas A, et al. Advice to Quit Smoking and Ratings of Health Care among
Medicare Beneficiaries Aged 65+. Health Serv Res. 2017;52(1):207-219. doi:10.1111/1475-
6773.12491
78. Groves RM, Peytcheva E. The Impact of Nonresponse Rates on Nonresponse Bias: A Meta-
Analysis. Public Opin Q. 2008;72(2):167-189. doi:10.1093/poq/nfn011
79. Elliott MN, Edwards C, Angeles J, Hambarsoomians K, Hays RD. Patterns of unit and item
nonresponse in the CAHPS® Hospital survey. Health Serv Res. 2005;40(6 II):2096-2119.
doi:10.1111/j.1475-6773.2005.00476.x
80. Elliott MN, Haviland AM, Kanouse DE, Hambarsoomian K, Hays RD. Adjusting for subgroup
differences in extreme response tendency in ratings of health care: impact on disparity
estimates. Health Serv Res. 2009;44(2 Pt 1):542-561. doi:10.1111/j.1475-6773.2008.00922.x
81. Dad T, Tighiouart H, Fenton JJ, et al. Evaluation of non-response to the In-Center Hemodialysis
Consumer Assessment of Healthcare Providers and Systems (ICH CAHPS) survey. BMC Health Serv
Res. 2018;18(1). doi:10.1186/s12913-018-3618-4
82. Dad T, Tighiouart H, Lacson E, Meyer KB, Weiner DE, Richardson MM. Long-term Clinical
Outcomes Among Responders and Nonresponders to the In-Center Hemodialysis Consumer
Assessment of Healthcare Providers and Systems (ICH CAHPS) Survey. Kidney Med.
2020;2(2):181-188. doi:10.1016/j.xkme.2019.12.002
83. Chaitoff A, Sun B, Windover A, et al. Associations Between Physician Empathy, Physician
Characteristics, and Standardized Measures of Patient Experience. Acad Med. 2017;92(10):1464-
1471. doi:10.1097/ACM.0000000000001671
84. Quigley DD, Elliott MN, Farley DO, Burkhart QQ, Skootsky SA, Hays RD. Specialties differ in which
aspects of doctor communication predict overall physician ratings. J Gen Intern Med.
2014;29(3):447-454. doi:10.1007/s11606-013-2663-2
85. Ochoa CY, Toledo G, Iyawe-Parsons A, Navarro S, Farias AJ. Multilevel Influences on Black Cancer
Patient Experiences With Care: A Qualitative Analysis. JCO Oncol Pract. 2021;17(5):e645-e653.
doi:10.1200/op.21.00011
86. Rodriguez HP, Rogers WH, Marshall RE, Safran DG. The Effects of Primary Care Physician Visit
Continuity on Patients’ Experiences with Care. Published online 2007. doi:10.1007/s11606-007-
0182-8
87. Chanfreau-Coffinier C, Washington DL, Chuang E, et al. Exploring the association of care
fragmentation and patient ratings of care quality: A mediation analysis of women Veterans’
experience with VA care. Health Serv Res. 2019;54(4):816-826. doi:10.1111/1475-6773.13153
88. Cohen SB. Design strategies and innovations in the medical expenditure panel survey. Med Care.
2003;41(7 Suppl). doi:10.1097/01.mlr.0000076048.11549.71
89. Lund JL, Gupta P, Amin KB, et al. Changes in chronic medication adherence in older adults with
cancer versus matched cancer-free cohorts. J Geriatr Oncol. 2021;12(1):72-79.
113
doi:10.1016/J.JGO.2020.04.012
90. Griffiths RI, McFadden EC, Stevens RJ, et al. Quality of diabetes care in breast, colorectal, and
prostate cancer. J Cancer Surviv. 2018;12(6):803-812. doi:10.1007/s11764-018-0717-5
91. Raval AD, Madhavan S, Mattes MD, Salkini M, Sambamoorthi U. Impact of Prostate Cancer
Diagnosis on Noncancer Hospitalizations Among Elderly Medicare Beneficiaries With Incident
Prostate Cancer. J Natl Compr Canc Netw. 2016;14(2):186-194. doi:10.6004/JNCCN.2016.0022
92. Chopra I, Dwibedi N, Mattes MD, Tan X, Findley P, Sambamoorthi U. Nonadherence to Statins
and Antihypertensives and Hospitalizations Among Elderly Medicare Beneficiaries With Incident
Cancer. J Natl Compr Canc Netw. 2017;15(11):1351-1360. doi:10.6004/JNCCN.2017.0168
93. Pinheiro LC, Soroka O, Kern LM, Higgason N, Leonard JP, Safford MM. Racial disparities in
diabetes care among incident breast, prostate, and colorectal cancer survivors: a SEER Medicare
study. J Cancer Surviv. 2022;16(1):52-60. doi:10.1007/S11764-021-01003-Z
94. Snyder C, Earle C, Herbert R, Neville B, Blackford A, Frick K. Preventive care for colorectal cancer
survivors: a 5-year longitudinal study. J Clin Oncol. 2008;26(7):1073-1079.
doi:10.1200/JCO.2007.11.9859
95. Du XL, Song L. Racial disparities in treatments and mortality among a large population-based
cohort of older men and women with colorectal cancer. Cancer Treat Res Commun. 2022;32.
doi:10.1016/J.CTARC.2022.100619
96. Cho J, Nilo D, Sterling MR, Kern LM, Safford MM, Pinheiro LC. Eliciting primary care and oncology
provider perspectives on diabetes management during active cancer treatment. Support Care
Cancer. 2021;29(11):6881-6890. doi:10.1007/S00520-021-06264-Z
97. Pinheiro LC, Cho J, Kern LM, et al. Managing diabetes during treatment for breast cancer:
oncology and primary care providers’ views on barriers and facilitators. Support Care Cancer.
2022;30(8):6901-6908. doi:10.1007/S00520-022-07112-4
98. Elliott MN, Adams JL, Klein DJ, et al. Patient-Reported Care Coordination is Associated with Better
Performance on Clinical Care Measures. J Gen Intern Med. 2021;36(12):3665-3671.
doi:10.1007/S11606-021-07122-8
99. Weinick RM, Elliott MN, Volandes AE, Lopez L, Burkhart Q, Schlesinger M. Using standardized
encounters to understand reported racial/ethnic disparities in patient experiences with care.
Health Serv Res. 2011;46(2):491-509. doi:10.1111/j.1475-6773.2010.01214.x
100. Meyers DJ, Rivera-Hernandez M, Kim D, Keohane LM, Mor V, Trivedi AN. Comparing the care
experiences of Medicare Advantage beneficiaries with and without Alzheimer’s disease and
related dementias. J Am Geriatr Soc. 2022;70(8):2344-2353. doi:10.1111/JGS.17817
101. Lines LM, Cohen J, Kirschner J, et al. Associations between illness burden and care experiences
among Medicare beneficiaries before or after a cancer diagnosis. J Geriatr Oncol. 2022;13(5):731-
737. doi:10.1016/J.JGO.2022.02.017
114
102. Arevalo M, Pickering TA, Vernon SW, Fujimoto K, Peskin MF, Farias AJ. Do breast cancer survivors
with a recent history of clinical depression report worse experiences with care? A retrospective
cohort study using SEER-CAHPS data. Cancer Med. 2022;00:1-12. doi:10.1002/CAM4.5031
103. Mohan CS, Rotter JS, Tan HJ, et al. The association between patient experience and healthcare
outcomes using SEER-CAHPS patient experience and outcomes among cancer survivors. J Geriatr
Oncol. 2021;12(4):623-631. doi:10.1016/J.JGO.2020.11.006
104. Lines LM, Cohen J, Halpern MT, Smith AW, Kent EE. Care experiences among dually enrolled older
adults with cancer: SEER-CAHPS, 2005–2013. Cancer Causes Control. 2019;30(10):1137-1144.
doi:10.1007/S10552-019-01218-7/TABLES/3
105. Hafida S, Ganda OP, Gabbay RA. CHAPTER 1. Clinical Guideline for Adults With Diabetes.; 2018.
Accessed December 22, 2022. https://www.ajmc.com/view/chapter-1-clinical-guideline-for-
adults-with-diabetes
106. Grundy SM, Stone NJ, Bailey AL, et al. 2018
AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA Guideline on the
Management of Blood Cholesterol: A Report of the American College of Cardiology/American
Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol. 2019;73(24):e285-
e350. doi:10.1016/j.jacc.2018.11.003
107. Rubenfire M. 2017 Guideline for High Blood Pressure in Adults - American College of Cardiology.
American College of Cardiology. Published May 7, 2018. Accessed December 22, 2022.
https://www.acc.org/latest-in-cardiology/ten-points-to-remember/2017/11/09/11/41/2017-
guideline-for-high-blood-pressure-in-adults
108. National Cancer Institute Division of Cancer Control & Population Sciences. Survey Weights.
Healthcare Delivery Research Program. Published June 15, 2022. Accessed January 26, 2023.
https://healthcaredelivery.cancer.gov/seer-cahps/researchers/weights.html
109. Afilalo M, Xue X, Colacone A, Jourdenais E, Boivin JF, Grad R. Association between access to
primary care and unplanned emergency department return visits among patients 75 years and
older. Can Fam Physician. 2022;68(8):599-606. doi:10.46747/CFP.6808599
110. Meisenberg BR, Karpman M, Tan J, Parlacoski S, Rhule J, Moxley J. The Perceptions of Cancer
Patients Regarding the Causes and Preventability of Unplanned Hospital Admissions. Am J Clin
Oncol. 2020;43(10):734-740. doi:10.1097/COC.0000000000000739
111. Fauer A, Wallner LP, Davis MA, Choi SW, Friese CR. Health care experiences for older adults
diagnosed with leukemia and lymphoma: Factors associated with emergency department use,
timeliness and access of health care. J Geriatr Oncol. 2021;12(2):250-255.
doi:10.1016/J.JGO.2020.09.008
112. Bone AE, Evans CJ, Henson LA, Etkind SN, Higginson IJ. Influences on emergency department
attendance among frail older people with deteriorating health: a multicentre prospective cohort
study. Public Health. 2021;194:4-10. doi:10.1016/J.PUHE.2021.02.031
113. Nguyen KH, Chien AT, Meyers DJ, Li Z, Singer SJ, Rosenthal MB. Team-Based Primary Care
115
Practice Transformation Initiative and Changes in Patient Experience and Recommended Cancer
Screening Rates. Inquiry. 2020;57. doi:10.1177/0046958020952911
114. Quigley DD, Elliott MN, Slaughter ME, Talamantes E, Hays RD. Follow-Up Shadow Coaching
Improves Primary Care Provider-Patient Interactions and Maintains Improvements When
Conducted Regularly: A Spline Model Analysis. J Gen Intern Med. 2023;38(1):221-227.
doi:10.1007/S11606-022-07881-Y
115. Katz DA, McCoy K, Sarrazin MV. Does improved continuity of primary care affect clinician-patient
communication in VA? J Gen Intern Med. 2014;29 Suppl 2(Suppl 2). doi:10.1007/S11606-013-
2633-8
116. Weech-Maldonado R, Elliott M, Pradhan R, Schiller C, Hall A, Hays RD. Can hospital cultural
competency reduce disparities in patient experiences with care? Med Care. 2012;50 Suppl(0).
doi:10.1097/MLR.0B013E3182610AD1
117. Song L, Hamilton JB, Moore AD. Patient-healthcare provider communication: perspectives of
African American cancer patients. Health Psychol. 2012;31(5):539-547. doi:10.1037/A0025334
118. Anderson JN, Graff JC, Krukowski RA, et al. “Nobody Will Tell You. You’ve Got to Ask!”: An
Examination of Patient-Provider Communication Needs and Preferences among Black and White
Women with Early-Stage Breast Cancer. Health Commun. 2021;36(11):1331-1342.
doi:10.1080/10410236.2020.1751383
119. Krok-Schoen JL, Fernandez K, Unzeitig GW, Rubio G, Paskett ED, Post DM. Hispanic breast cancer
patients’ symptom experience and patient-physician communication during chemotherapy.
Support Care Cancer. 2019;27(2):697-704. doi:10.1007/S00520-018-4375-Y
120. Odai-Afotey A, Kliss A, Hafler J, Sanft T. Defining the patient experience in medical oncology.
Support Care Cancer. 2020;28(4):1649-1658. doi:10.1007/S00520-019-04972-1
121. Poitras ME, Maltais ME, Bestard-Denommé L, Stewart M, Fortin M. What are the effective
elements in patient-centered and multimorbidity care? A scoping review. BMC Health Serv Res.
2018;18(1). doi:10.1186/S12913-018-3213-8
122. Kastner M, Cardoso R, Lai Y, et al. Effectiveness of interventions for managing multiple high-
burden chronic diseases in older adults: a systematic review and meta-analysis. CMAJ.
2018;190(34):E1004-E1012. doi:10.1503/CMAJ.171391
123. Dossett LA, Hudson JN, Morris MAM, et al. The Primary Care Provider (PCP)-Cancer Specialist
Relationship: A Systematic Review And Mixed Methods Meta-synthesis. CA Cancer J Clin.
2017;67(2):156. doi:10.3322/CAAC.21385
124. Hohmann NS, McDaniel CC, Mason SW, et al. Patient perspectives on primary care and oncology
care coordination in the context of multiple chronic conditions: A systematic review. Res Social
Adm Pharm. 2020;16(8):1003-1016. doi:10.1016/J.SAPHARM.2019.11.014
125. Bazargan M, Cobb S, Assari S, Bazargan-Hejazi S. Preparedness for Serious Illnesses: Impact of
Ethnicity, Mistrust, Perceived Discrimination, and Health Communication. Am J Hosp Palliat Care.
116
2022;39(4):461-471. doi:10.1177/10499091211036885
126. Kern LM, Ringel JB, Rajan M, et al. Ambulatory Care Fragmentation, Emergency Department
Visits, and Race: a Nationwide Cohort Study in the U.S. J Gen Intern Med. Published online 2022.
doi:10.1007/S11606-022-07888-5
127. Wang JHY, Adams IF, Pasick RJ, et al. Perceptions, expectations, and attitudes about
communication with physicians among Chinese American and non-Hispanic white women with
early stage breast cancer. Support Care Cancer. 2013;21(12):3315-3325. doi:10.1007/S00520-
013-1902-8
128. Weech-Maldonado R, Elliott M, Pradhan R, Schiller C, Hall A, Hays RD. Can hospital cultural
competency reduce disparities in patient experiences with care? Med Care. 2012;50
Suppl(0):S48-55. doi:10.1097/MLR.0b013e3182610ad1
129. Lemmens VEPP, Janssen-Heijnen MLG, Verheij CDGW, Houterman S, Repelaer Van Driel OJ,
Coebergh JWW. Co-morbidity leads to altered treatment and worse survival of elderly patients
with colorectal cancer. Br J Surg. 2005;92(5):615-623. doi:10.1002/bjs.4913
130. Etzioni DA, El-Khoueiry AB, Beart RW. Rates and predictors of chemotherapy use for stage III
colon cancer: A systematic review. Cancer. 2008;113(12):3279-3289. doi:10.1002/cncr.23958
131. Gross CP, McAvay GJ, Guo Z, Tinetti ME. The impact of chronic illnesses on the use and
effectiveness of adjuvant chemotherapy for colon cancer. Cancer. 2007;109(12):2410-2419.
doi:10.1002/cncr.22726
132. Sisler JJ, Seo B, Katz A, et al. Concordance with ASCO guidelines for surveillance after colorectal
cancer treatment: a population-based analysis. J Oncol Pract. 2012;8(4).
doi:10.1200/JOP.2011.000396
133. Ford ME, Sterba KR, Armeson K, Malek AM, Knight KD, Zapka J. Factors Influencing Adherence to
Recommended Colorectal Cancer Surveillance: Experiences and Behaviors of Colorectal Cancer
Survivors. J Cancer Educ. 2019;34(5):938-949. doi:10.1007/S13187-018-1398-5
134. Peltrini R, Imperatore N, Carannante F, et al. Age and comorbidities do not affect short-term
outcomes after laparoscopic rectal cancer resection in elderly patients. A multi-institutional
cohort study in 287 patients. Updates Surg. 2021;73(2):527-537. doi:10.1007/S13304-021-00990-
Z
135. Cronin D, Harlan L, Potosky A, Clegg L, Stevens J, Mooney M. Patterns of care for adjuvant
therapy in a random population-based sample of patients diagnosed with colorectal cancer. Am J
Gastroenterol. 2006;101(10):2308-2318. doi:10.1111/J.1572-0241.2006.00775.X
136. Lemmens V, Janssen-Heijnen M, Houterman S, et al. Which comorbid conditions predict
complications after surgery for colorectal cancer? World J Surg. 2007;31(1):192-199.
doi:10.1007/S00268-005-0711-8
137. Dotan E, Walter LC, Browner IS, et al. NCCN Guidelines Older Adult Oncology.; 2021. Accessed
November 4, 2021. https://www.nccn.org/professionals/physician_gls/pdf/senior.pdf
117
138. Obrochta CA, Murphy JD, Tsou MH, Thompson CA. Disentangling Racial, Ethnic, and
Socioeconomic Disparities in Treatment for Colorectal Cancer. Cancer Epidemiol Biomarkers Prev.
2021;30(8):1546-1553. doi:10.1158/1055-9965.EPI-20-1728
139. Mohan C, Rotter J, Tan H, et al. The association between patient experience and healthcare
outcomes using SEER-CAHPS patient experience and outcomes among cancer survivors. J Geriatr
Oncol. 2021;12(4):623-631. doi:10.1016/J.JGO.2020.11.006
140. Benson AB, Venook AP, Al-Hawary MM, et al. NCCN Guidelines Colon Cancer .; 2021. Accessed
November 4, 2021. https://www.nccn.org/professionals/physician_gls/pdf/colon.pdf
141. Hines RB, Jiban MJH, Specogna A V., et al. Surveillance Colonoscopy in Older Stage i Colon Cancer
Patients and the Association with Colon Cancer-Specific Mortality. Am J Gastroenterol.
2020;115(6):924-933. doi:10.14309/AJG.0000000000000537
142. Ramkumar N, Colla CH, Wang Q, O’Malley AJ, Wong SL, Brooks GA. Association of Rurality, Race
and Ethnicity, and Socioeconomic Status With the Surgical Management of Colon Cancer and
Postoperative Outcomes Among Medicare Beneficiaries. JAMA Netw Open. 2022;5(8):E2229247.
doi:10.1001/JAMANETWORKOPEN.2022.29247
143. Fisher DA, Princic N, Miller-Wilson LA, Wilson K, Fendrick AM, Limburg P. Utilization of a
Colorectal Cancer Screening Test Among Individuals With Average Risk. JAMA Netw Open.
2021;4(9):45390. doi:10.1001/JAMANETWORKOPEN.2021.22269
144. Méndez-Bailón M, Jiménez-García R, Muñoz-Rivas N, et al. Trends and Clinical Impact of
Gastrointestinal Endoscopic Procedures on Acute Heart Failure in Spain (2002–2017). J Clin Med.
2021;10(3):1-13. doi:10.3390/JCM10030546
145. Wardrop RC, Cass AL, Quinn SA, Wercholuk AN, Parikh AA, Snyder RA. Racial and ethnic
differences in patient ratings of colorectal and non-small-cell lung cancer care: A SEER-CAHPS
study. Cancer Causes Control. 2022;33(9):1125-1133. doi:10.1007/S10552-022-01606-6
146. Farias AJ, Toledo G, Ochoa CY, Hamilton AS. Racial/Ethnic Disparities in Patient Experiences With
Health Care in Association With Earlier Stage at Colorectal Cancer Diagnosis. Med Care.
2021;59(4):295-303. doi:10.1097/MLR.0000000000001514
147. Frankenfeld CL, Menon N, Leslie TF. Racial disparities in colorectal cancer time-to-treatment and
survival time in relation to diagnosing hospital cancer-related diagnostic and treatment
capabilities. Cancer Epidemiol. 2020;65. doi:10.1016/J.CANEP.2020.101684
148. Emerson MA, Golightly YM, Aiello AE, et al. Breast cancer treatment delays by socioeconomic and
health care access latent classes in Black and White women. Cancer. 2020;126(22):4957-4966.
doi:10.1002/CNCR.33121
149. Halpern MT, Holden DJ. Disparities in timeliness of care for U.S. Medicare patients diagnosed
with cancer. Curr Oncol. 2012;19(6). doi:10.3747/CO.19.1073
150. Kumar V, Alhaj-Moustafa M, Bojanini L, et al. Timeliness of Initial Therapy in Multiple Myeloma:
Trends and Factors Affecting Patient Care. JCO Oncol Pract. 2020;16(4):E341-E349.
118
doi:10.1200/JOP.19.00309
151. Grunvald MW, Underhill JM, Skertich NJ, et al. Mediating Factors Between Race and Time to
Treatment in Colorectal Cancer. Dis Colon Rectum. 2023;66(2):331-336.
doi:10.1097/DCR.0000000000002214
152. Charlot M, Stein JN, Damone E, et al. Effect of an Antiracism Intervention on Racial Disparities in
Time to Lung Cancer Surgery. J Clin Oncol. 2022;40(16):1755-1762. doi:10.1200/JCO.21.01745
153. Freund KM, Battaglia TA, Calhoun E, et al. Impact of patient navigation on timely cancer care: the
Patient Navigation Research Program. J Natl Cancer Inst. 2014;106(6). doi:10.1093/JNCI/DJU115
154. Roland KB, Milliken EL, Rohan EA, et al. Use of Community Health Workers and Patient
Navigators to Improve Cancer Outcomes Among Patients Served by Federally Qualified Health
Centers: A Systematic Literature Review. Heal equity. 2017;1(1):61-76.
doi:10.1089/HEQ.2017.0001
155. Hendren S, Chin N, Fisher S, et al. Patients’ barriers to receipt of cancer care, and factors
associated with needing more assistance from a patient navigator. J Natl Med Assoc.
2011;103(8):701-710. doi:10.1016/S0027-9684(15)30409-0
156. Haas JS, Brawarsky P, Iyer A, Fitzmaurice GM, Neville BA, Earle C. Association of area
sociodemographic characteristics and capacity for treatment with disparities in colorectal cancer
care and mortality. Cancer. 2011;117(18):4267-4276. doi:10.1002/CNCR.26034
157. Zhang W, Ayanian JZ, Zaslavsky AM. Patient characteristics and hospital quality for colorectal
cancer surgery. Int J Qual Heal care J Int Soc Qual Heal Care. 2007;19(1):11-20.
doi:10.1093/INTQHC/MZL047
158. Huang LC, Tran TB, Ma Y, Ngo J V., Rhoads KF. Factors that influence minority use of high-volume
hospitals for colorectal cancer care. Dis Colon Rectum. 2015;58(5):526-532.
doi:10.1097/DCR.0000000000000353
159. Epstein AJ, Gray BH, Schlesinger M. Racial and ethnic differences in the use of high-volume
hospitals and surgeons. Arch Surg. 2010;145(2):179-186. doi:10.1001/ARCHSURG.2009.268
160. Kronenfeld JP, Graves KD, Penedo FJ, Yanez B. Overcoming Disparities in Cancer: A Need for
Meaningful Reform for Hispanic and Latino Cancer Survivors. Oncologist. 2021;26(6):443-452.
doi:10.1002/ONCO.13729
161. Nothelle SK, Boyd C, Sheehan O, Wolff JL. Factors Associated With Loss of Usual Source of Care
Among Older Adults. Ann Fam Med. 2018;16(6):538-545. doi:10.1370/AFM.2283
162. Stransky ML. Two-Year Stability and Change in Access to and Reasons for Lacking a Usual Source
of Care Among Working-Age US Adults. Public Health Rep. 2017;132(6):660-668.
doi:10.1177/0033354917735322
163. Kim MY, Kim JH, Choi IK, Hwang IH, Kim SY. Effects of having usual source of care on preventive
services and chronic disease control: a systematic review. Korean J Fam Med. 2012;33(6):336-
119
345. doi:10.4082/KJFM.2012.33.6.336
164. Finney Rutten LJ, Agunwamba AA, Beckjord E, Hesse BW, Moser RP, Arora NK. The Relation
Between Having a Usual Source of Care and Ratings of Care Quality: Does Patient-Centered
Communication Play a Role? J Health Commun. 2015;20(7):759-765.
doi:10.1080/10810730.2015.1018592
165. Smith MA, Bartell JM. Changes in usual source of care and perceptions of health care access,
quality, and use. Med Care. 2004;42(10):975-984. doi:10.1097/00005650-200410000-00006
166. Campbell C, Craig J, Eggert J, Bailey-Dorton C. Implementing and measuring the impact of patient
navigation at a comprehensive community cancer center. Oncol Nurs Forum. 2010;37(1):61-68.
doi:10.1188/10.ONF.61-68
167. Budde H, Williams GA, Winkelmann J, Pfirter L, Maier CB. The role of patient navigators in
ambulatory care: overview of systematic reviews. BMC Health Serv Res. 2021;21(1).
doi:10.1186/S12913-021-07140-6
168. Lavery JA, Lipitz-Snyderman A, Li DG, Bach PB, Panageas KS. Identifying Cancer-Directed Surgeries
in Medicare Claims: A Validation Study Using SEER-Medicare Data. JCO Clin cancer informatics.
2019;3(3):1-24. doi:10.1200/CCI.18.00093
169. Brault MW, Landon BE, Zaslavsky AM. Validating Reports of Chronic Conditions in the Medicare
CAHPS Survey. Med Care. 2019;57(10):830-835. doi:10.1097/MLR.0000000000001187
170. DJ K, MN E, AM H, et al. Understanding nonresponse to the 2007 Medicare CAHPS survey.
Gerontologist. 2011;51(6):843-855. doi:10.1093/GERONT/GNR046
171. Aronson L, Bautista CA, Covinsky K. Medicare and Care Coordination: Expanding the Clinician’s
Toolbox. JAMA. 2015;313(8):797-798. doi:10.1001/JAMA.2014.18174
172. Wilson C, O’Malley AS, Bozzolo C, McCall N, Ma S. Patient Experiences with Chronic Care
Management Services and Fees: a Qualitative Study. J Gen Intern Med. 2019;34(2):250-255.
doi:10.1007/S11606-018-4750-X
173. Flieger SP, Thomas CP, Prottas J. Improving Interorganizational Coordination Between Primary
Care and Oncology: Adapting a Chronic Care Management Model for Patients With Cancer. Med
Care Res Rev. 2021;78(3):229-239. doi:10.1177/1077558719870699
174. Freeman H, Muth B, Kerner J. Expanding access to cancer screening and clinical follow-up among
the medically underserved. Cancer Pract. 1995;3(1):19-30. Accessed January 23, 2023.
https://pubmed.ncbi.nlm.nih.gov/7704057/
175. Freeman HP. The origin, evolution, and principles of patient navigation. Cancer Epidemiol
Biomarkers Prev. 2012;21(10):1614-1617. doi:10.1158/1055-9965.EPI-12-0982
176. Knudsen KE, Wiatrek DE, Greenwald J, McComb K, Sharpe K. The American Cancer Society and
patient navigation: Past and future perspectives. Cancer. 2022;128 Suppl 13(S13):2673-2677.
doi:10.1002/CNCR.34206
120
177. Maness SB, Low JH, Vu T. A Qualitative Analysis of Access to Healthcare Among African American
Adults in South Carolina. J Prim Care Community Health. 2023;14.
doi:10.1177/21501319231153593
178. Rose J, Oliver Y, Sage P, Dong W, Koroukian SM, Koopman Gonzalez S. Factors affecting timely
breast cancer treatment among black women in a high-risk urban community: a qualitative study.
BMC Womens Health. 2022;22(1). doi:10.1186/S12905-022-01938-0
179. Rodriguez GM, Leach M, Osorio J, et al. Exploring cancer care needs for Latinx adults: a
qualitative evaluation. Support Care Cancer. 2022;31(1):76. doi:10.1007/S00520-022-07518-0
180. Brown CM, Kanu C, Richards KM, Stevens L, Sasane R, McAneny B. Exploring access to care from
the perspective of patients with breast cancer: A qualitative study. Cancer Med.
2022;11(12):2455-2466. doi:10.1002/CAM4.4624
181. Sleight AG, Lyons KD, Vigen C, Macdonald H, Clark F. Supportive care priorities of low-income
Latina breast cancer survivors. Support Care Cancer. 2018;26(11):3851-3859.
doi:10.1007/S00520-018-4253-7
182. Tsui J, Hirsch JA, Bayer FJ, et al. Patterns in Geographic Access to Health Care Facilities Across
Neighborhoods in the United States Based on Data From the National Establishment Time-Series
Between 2000 and 2014. JAMA Netw open. 2020;3(5).
doi:10.1001/JAMANETWORKOPEN.2020.5105
121
Appendix A. Systematic review search strategy.
General strategy: Base search terms were connected to key words for either clinical or quality outcomes
of care by the Boolean operator “AND”. Base search terms were derived from keywords for CAHPS
surveys and CAHPS survey composite domains. A total of twelve individual searches were completed,
including 11 individual searches for clinical outcomes and 1 search for quality outcomes. The eleven
clinical outcomes searches including ten searches for each category of medical diagnoses deemed
amenable to health care and one search for general clinical outcomes. The ten categories of medical
diagnoses included infections; neoplasms; nutritional, endocrine and metabolic diseases; neurologic
disorders; cardiovascular diseases; diseases of the respiratory system; diseases of the digestive system;
diseases of the genitourinary system; maternal and perinatal mortality; and external causes.
Base Search Terms:
((CAHPS AND “customer service”) OR (CAHPS AND “care coordination”) OR (CAHPS AND
“provider communication”) OR (CAHPS AND “getting care quickly”) OR (CAHPS AND “getting
needed care”) OR (CAHPS AND “getting prescription drugs”) OR (HCAHPS AND “customer
service”) OR (HCAHPS AND “care coordination”) OR (HCAHPS AND “provider communication”)
OR (HCAHPS AND “getting care quickly”) OR (HCAHPS AND “getting needed care”) OR (HCAHPS
AND “getting prescription drugs”) OR (MCAHPS AND “customer service”) OR (MCAHPS AND
“care coordination”) OR (MCAHPS AND “provider communication”) OR (MCAHPS AND “getting
care quickly”) OR (MCAHPS AND “getting needed care”) OR (MCAHPS AND “getting prescription
drugs”) OR CAHPS OR HCAHPS OR MCAHPS OR “consumer assessment of healthcare providers
and systems”)
Quality Outcomes Search Terms (connected to base search terms using the Boolean operator “AND”):
(“quality of healthcare”/exp OR “quality assurance, health care”/exp OR “delivery of health
care”/exp OR “quality of care” OR “clinical quality” OR “health care quality” OR “quality of
health care” OR “care quality” OR “healthcare quality” OR “health care quality” OR “quality
assurance” OR “quality improvement” OR “quality indicators” OR “standard of care” OR HEDIS
OR “health effectiveness data and information set” OR “effectiveness of care” OR prevention OR
screening OR “medication management” OR “comprehensive care” OR “access to care” OR
“access to health care” OR “access to healthcare” OR “healthcare access” OR “health care
access” OR utilization OR safety OR timeliness OR “patient-centered” OR “diabetes care”)
Clinical Outcomes Search Terms (11 individual searches run, each connected to base search terms using
the Boolean operator “AND”):
1. Infections
("bacterial infections and mycoses"/exp OR choler* OR typh* OR paratyph* OR
salmonell* OR infecti* OR shigell* OR coli OR ETEC OR EPEC OR EIEC OR EHEC OR
campylobacter OR yersin* OR enteritis OR enterocolitis OR difficile OR "c. diff" OR
staph* OR botuli* OR perfringens OR vibrio OR cereus OR ameb* OR protozo* or
blanatidiasis OR giard* OR lambl* OR cryptospor* OR rotavir* or norwalk OR adenovir*
122
OR gastroenteritis OR colitis OR tuberculo* OR tulare* OR anthrax OR brucell* OR
glander* OR melioid* OR "rat-bite" OR "rat bite" OR spirill* OR strepto* OR erysipel* OR
pasteurell* OR "cat-scratch" OR "cat scratch" OR leprosy OR "Hansen's" OR
mycobacter* OR DMAC OR listeri* OR tetan* OR sepsis OR hemophilus OR gangrene OR
"Legionnaire's" OR "pontiac fever" OR "toxic shock" OR TSS OR "brazilian purpuric fever"
OR diphtheri* OR polio* OR "scarlet fever" OR chlamydi* OR chancroid OR granuloma
OR "sexually transmitted" OR STD OR STI OR "relapsing fever*" OR plasmodi* OR
malari* OR schistosom* OR dermatitis OR fluke OR opisthorchiasis OR clonorchiasis OR
dicroceliasis OR fascioliasis OR paragonimiasis OR echinococc* OR taeni* OR
cysticercosis OR cestode OR hymenolepiasis OR dipylidiasis OR onchocerciasis OR
filariasis OR wucheria OR brugia OR loiasis OR mansonelliasis OR trichinell* OR
hookworm OR ancylostomiasis OR necatoriasis OR ascariasis OR strongyloidiasis OR
trichuriasis OR enterobiasis OR helminth* OR anisakiasis OR capillariasis OR
trichostrongyliasis OR angiostrongyliasis OR parasit* OR "visceral larva migrans" OR
gnathostomiasis OR "parastrongylus contonensis" OR syndamiasis OR hirudiniasis OR
cellulitis OR lymphangitis OR meningitis OR encephalitis OR myelitis OR
encephalomyelitis OR "topical spastic paraplegia" OR meningoencephalitis OR
encephalopathy OR abscess OR whooping OR bordatella OR measles OR rubella OR
hepatitis OR "human immunodeficiency virus" OR HIV)
2. Neoplasms
("neoplasms by histologic type"/exp OR "neoplasms by site"/exp OR malignan* OR
neoplasm* OR cancer* OR Hodgkin OR "Hodgkin's" OR leukemia* OR AML OR CML OR
sarcoma* OR "acute panmyelosis with myelofibrosis" OR myelodysplastic OR
carcinoma* OR melanoma* OR benign OR hemangioma* OR lymphangioma* OR
leiomyoma*)
3. Nutritional, Metabolic and Endocrine Disorders
("thyroid diseases"/exp OR "iodine-deficiency" OR "iodine deficiency" OR hypothyroid*
OR "thyroid atrophy" OR "atrophy of thyroid" OR "atrophic thyroid" OR "myxedema
coma" OR "nontoxic goiter" OR "nontoxic single thyroid nodule" OR "nontoxic thyroid
nodule" OR thyrotoxicosis OR thyroiditis OR "hypersecretion of calcitonin" OR
"calcitonin hypersecretion" OR "diabetes mellitus"/exp OR diabet* OR "adrenal gland
diseases"/exp OR cushing* OR "nelson's" OR "ectopic ACTH" OR "ectopic
adrenocorticotropic hormone" OR "adrenogenital syndrome" OR "adrenogenital
disorder*" OR "adrenocortical hyperactivity" OR "adrenocortical overactivity" OR
"adrenocortical insufficiency" OR addison* OR "adrenomedullary hyperfunction" OR
"adrenomedullary overactivity" OR gierke* OR "glucose-6-phosphatase deficienc*" OR
"glucose 6 phosphatase deficienc*" OR pompe OR "alpha-1,4-glucosidase deficienc*"
OR " alpha 1,4 glucosidase deficienc*" OR cori OR "debranching enzyme deficienc*" OR
mcardle OR "glycogen phosphorylase deficienc*" OR galactosemia* OR "disorder* of
galactose metabolism" OR "galactose metabolism disorder")
4. Neurological Disorders
("epilepsy"/exp OR "seizure disorder*" OR epilep* OR "recurrent seizure*")
5. Cardiovascular Diseases
123
("cardiovascular diseases"/exp OR rheumatic OR "valve disease*" OR hypertensi* OR
angina OR "myocrdial infarct*" OR AMI OR STEMI OR NSTEMI OR hemopericardium OR
"atrial septal defect" OR ASD OR "ventricular septal defect" OR VSD OR "rupture of
cardiac wall" OR "cardiac wall rupture" OR "rupture of chordae tendinae" OR "rupture
of papillary muscle" OR "papillary muscle rupture" OR "atrial thrombosis" OR
"thrombosis of atrium" OR "thrombosis of auricular appendage" OR "auricular
thrombosis" OR "ventriuclar thrombosis" OR "thrombosis of ventricle" OR "ischemic
heart" OR "coronary thrombosis" OR "dressler's" OR atherosclero* OR "aneurysm of
heart" OR "coronary aneurysm" OR "cardiac aneurysm" OR "coronary artery aneurysm"
OR "coronary artery dissection" OR “dissection of coronary arter*" OR "ischemic
cardiomyopathy" OR "myocardial ischemia" OR "nontraumatic subarachnoid
hemorrhage" OR "nontraumatic intracerebral hemorrhage" OR "nontraumatic
intracranial hemorrhage" OR “nontraumatic extradural hemorrhage" OR "nontruamatic
subdural hemorrhage" OR "nontraumatic acute subdural hemorrhage" OR
"nontraumatic subacute subdural hemorrhage" OR "nontraumatic chronic subdural
hemorrhage" OR "cerebral infarct*" OR "precerebral arter* occlusion" OR "occlusion of
precerebral arter*" OR "precerebral arter* stenosis" OR "stenosis of precerebral arter*"
OR "cerebral arter* occlusion" OR "occlusion of cerebral arter*" OR "stenosis of cerebral
arter*" OR "cerebral arter* stenosis" OR "cerebrovascular disease*" OR "dissection of
cerebral arter*" OR "cerebral arter* dissection" OR "cerebral aneurysm" OR
"progressive vascular leukencephalopathy" OR "moyamoya" OR "intracranial venous
thrombosis" OR "thrombosis of intracranial ve*" OR "cerebral arteritis" OR
"cerebrovascular disorder*" OR "cerebral amyloid angiopathy" OR "peripheral vascular
disease*")
6. Diseases of the Respiratory System
("respiratory tract diseases"/exp OR nasopharyngitis OR "common cold" OR sinusitis OR
pansinusitis OR pharyngitis OR tonsillitis OR adenoiditis OR laryngitis OR tracheitis OR
laryngotracheitis OR supraglottitis OR croup OR epiglottitis OR "upper respiratory tract
infection*" OR laryngopharyngitis OR URI OR influenza OR bronchitis OR bronchiolitis
OR "respiratory tract infection" OR rhinitis OR "nasal polyp" OR "polyp of nasal cavity"
OR "polypoid sinus degeneration" OR "polyp of sinus" OR "nose abscess" OR "nasal
abscess" OR "abscess of nose" OR "nose furuncle" OR "furuncle of nose" OR "nasal
furuncle" OR "nose carbuncle" OR "nasal carbuncle" OR "carbuncle of nose" OR "cyst of
nose" OR "nasal cyst" OR "cyst of nasal sinus" OR "mucocele of nose" OR "nasal
mucocele" OR "mucocele of nasal sinus" OR "deviated nasal septum" OR "deviated
septum" OR "hypertrophy of nasal turbinate" OR "turbinate hypertrophy" OR
"hypertrophy of adenoids" OR "adenoid hypertrophy" OR "hypertrophy of tonsils" OR
"tonsil* hypertrophy" OR "pertonsillar abscess" OR "paralysis of vocal cords" OR "vocal
cord paralysis" OR "paralysis of larynx" OR "laryngeal paralysis" OR "polyp of vocal cord"
OR "vocal cord polyp" OR "polyp of larynx" OR "laryngeal polyp" OR "nodule* of vocal
cord" OR "vocal cord nodule*" OR "edema of larynx" OR "laryngeal edema" OR "stenosis
of larynx" OR "laryngeal stenosis" OR "retropharyngeal abscess" OR "parapharyngeal
abscess" OR "abscess of pharynx" OR "pharyngeal abscess" OR "upper respiratory tract
hypersensitivity" OR emphysema OR "chronic obstructive pulmonary disease" OR COPD
124
OR astma* OR bronchiectasis OR "coalworker's" OR pneumoconiosis OR byssinosis OR
"flax-dresser's" OR cannabinosis OR pneumonitis OR "farmer's lung" OR bagassosis OR
"bird fancier's" OR suberosis OR "maltworker's" OR "mushroom-worker's" OR "maple-
bark-stripper's" OR "air conditioner lung" OR "humidifier lung" OR "pulmonary edema"
OR "respiratory inflammation" OR "acute respiratory distress syndrome" OR ARDS OR
"pulmonary eosinophilia" OR "interstitial pulmonary disease*" OR "alveolar proteinosis"
OR "alveolar microlithiasis" OR "pulmonary hemosiderosis" OR "pulmonary fibrosis" OR
lymphangioleiomyomatosis OR "langerhans cell histiocytosis" OR "surfactant
mutation*" OR "interstitial lung disorder*" OR "interstitial lung disease*" OR pyothorax
OR "pleural effusion" OR "pleural plaque" OR pneumothorax OR "disease* of the
pleura" OR "disease* of pleura" OR "pleural disease*" OR "chylous effusion*" OR
fibrothorax OR hemothorax OR "pulmonary insufficiency" OR "subglottic stenosis" OR
"respiratory failure" OR "disease* of bronchus" OR "bronchial disease" OR "pulmonary
collapse" OR "disorders of the lung" OR "lung disorders" OR "mediastin* disease" OR
"disease* of the mediastinum" OR "disorder* of the diaphragm" OR "diaphragm
disorder*" OR "respiratory disorder*" OR bronchospasm OR atelectasis OR
mediastinitis)
7. Diseases of the Digestive System
("biliary tract diseases"/exp OR "digestive system fistula"/exp OR "pancreatic
diseases"/exp OR "gastroinstestinal diseases"/exp OR ulcer OR gastritis OR duodenitis
OR appendicitis OR "disease* of appendix" OR "hyperplasia of appendix" OR
"appendicular concretions" OR "diverticulum of appendix" OR fistula OR hernia OR
"intestinal obstruction" OR ileus OR intussusception OR volvulus OR "impaction of
intestine" OR "intestinal impaction" OR "intestinal adhesions" OR "intestinal bands" OR
cholelithiasis OR "calcul* of gallbladder" OR "gallbladder calcul*" OR “calcul* of bile
duct" OR "bile duct calcul*" OR cholecystitis OR "perforation of gallbladder" OR
"gallbladder perforation" OR "obstruction of gallbladder" OR "gallbladder obstruction"
OR "hydrops of gallbladder" OR "cholesterolosis of gallbladder" OR "gallbladder
disease*" OR "disease* of gallbladder" OR cholangitis OR "obstruction of the bile duct"
OR "bile duct obstruction" OR "perforation of bile duct" OR "bile duct perforation" OR
"spasm of sphincter of oddi" OR "biliary cyst" OR "biliary tract disease" OR pancreatitis
OR "cyst of pancreas" OR "pseudocyst of pancreas" OR "pancreatic disease*" OR
"postsurgical malabsorption" OR "postcholecystectomy syndrome")
8. Diseases of the Genitourinary System
("urination disorders"/exp OR "nephritis"/exp OR "adnexal diseases"/exp OR "uterine
diseases"/exp OR "vaginal diseases"/exp OR "vulvar diseases"/exp OR "nephritic
syndrome" OR hematuria OR "nephrotic syndrome" OR proteinuria OR "nephropathy"
OR "glomerular disorder*" OR "glomerular disease*" OR "reflux uropathy" OR
"obstructive uropathy" OR hydronephrosis OR hydroureter OR pyonephrosis OR
"vesicoureteral reflux" OR "urinary calcul*" OR "calcul* of kidney" OR "calcul* or ureter"
OR “calcul* in bladder" OR "bladder calcul*" OR "calcul* in urethra" OR "urethral
calculi" OR "calcul* of lower urinary tract" OR "urinary tract calcul*" OR "renal colic" OR
"kidney failure" OR "kidney disease" OR "end stage renal disease" OR ESRD OR
"impaired renal tubular function" OR "renal osteodystrophy" OR "nephrogenic diabetes
125
insipidus" OR "impaired renal tubular function" OR "atrophy of kidney" OR "renal
atrophy" OR "page kidney" OR "renal sclerosis" OR "small kidney" OR urethritis OR
"urethral stricture" OR "benign prostatic hyperplasia" OR BPH OR "nodular prostate" OR
salpingitis OR oophoritis OR "inflammatory disease* of the uterus" OR "inflammatory
disease* of the cervix" OR "pelvic inflammatory disease*" OR PID OR parametritis OR
"pelvic cellulitis" OR "pelvic peritonitis" OR "pelvic peritoneal adhesions" OR "bartholin's
cyst" OR "abscess of bartholin's gland" OR "bartholin's gland abscess" OR "disease* of
bartholin's gland" OR "bartholin's gland disease*" OR vaginitis OR vulvitis OR "abscess of
the vulva" OR "vulvar abscess" OR "ulceration of the vagina" OR "vaginal ulceration" OR
"ulceration of the vulva" OR "vulvar ulceration" OR "cervical dysplasia" OR "dysplasia of
the cervix" OR "leukoplakia of the cervix" OR "stricture of the cervix" OR "cervical
stricture" OR "stenosis of the cervix" OR "cervical stenosis" OR "incompetence of the
cervix" OR "cervical incompetence" OR "incompetent cervix" OR "hypertrophic
elongation of the cervix" OR "vulvar dysplasia" OR "leukoplakia of the vulva" OR
"atrophy of the vulva" OR "vulvar atrophy" OR "hypertrophy of the vulva" OR "vulvar
hypertrophy" OR "vulvar cyst" OR "postprocedural urethral stricture")
9. Maternal and Perinatal Mortality
(pregnancy or "hydatidiform mole" OR "abnormal product* of conception" OR abortion
OR "blighted ovum" OR "nonhydatidiform mole" OR "pre-eclampsia" OR preeclampsia
OR gestat* OR HELLP OR maternal OR gravida* OR childbirth OR puerper* OR
intrauterine OR IUD OR antenatal OR twin* OR triplet* OR quadruplet* OR fetus OR
fetal* OR malpresentation OR "unstable lie" OR breech OR "transverse lie" OR "oblique
lie" OR "face presentation" OR "brow presentation" OR "chin presentation" OR
malformation* OR "tumor of corpus uteri" OR "uterine tumor" OR hereditary OR
chromosom* OR isoimmunization OR polyhydramnios OR oligohydramnios OR
premature OR placenta* OR antepartum OR "false labor" OR "preterm labor" OR
"induction of labor" OR "induced labor" OR "forces of labor" OR "inadequate
contractions" OR "uterine inertia" OR "precipitate labor" OR "uterine contractions" OR
"long labor" OR "prolonged labor" OR "obstructed labor" OR intrapartum OR "perineal
laceration" OR "laceration of perineum" OR "obstetric trauma" OR "rupture of uterus"
OR "uterine rupture" OR postpartum OR "laceration of cervix" OR "cervical laceration"
OR "vaginal laceration" OR "laceration of vagina" OR "anesthesia during labor" OR
meconium OR cesarean OR "obstetric embolism" OR "obstetric air embolism" OR
"amniotic fluid" OR “obstetric thromboembolism" OR "obstetric pyemic embolism" OR
"obstetric septic embolism" OR lactation OR syphilis OR gonorrhea OR anencephaly OR
craniorachischisis OR iniencephaly OR encephalocele OR microcephaly OR
hydrocephalus OR "foramina of magendie" OR "foramina of luschka" OR "reduction
deform*" OR arhinencephaly OR holoprosencephaly OR "septo-optic dysplasia" OR
"cerebral cyst*" OR "spina bifida" OR amyelia OR "hypoplasia of spinal cord" OR
"dysplasia of spinal cord" OR diastematomyelia OR hydromyelia OR "arnold-chiari" OR
congenital OR "absence of lacrimal" OR agenesis OR "stenosis of lacrimal duct" OR
"stricture of lacrimal duct" OR anophthalmos OR microphthalmos OR macrophthalmos
OR "cystic eyeball" OR coloboma OR spherophakia OR "absence of iris" OR "blue sclera"
OR "absence of eustachian tube" OR "accessory auricle" OR macrotia OR microtia OR
126
"misshapen ear" OR "misplaced ear" OR "prominent ear" OR "sinus of branchial cleft"
OR "fistula of branchial cleft" OR "cyst of branchial cleft" OR "preauricular sinus" OR
"preauricular cyst" OR "webbing of neck" OR macrostomia OR microstomia OR
macrocheilia OR microcheilia OR "common arterial trunk" OR "double outlet right
ventricle" OR "double outlet left ventricle" OR "discordant ventriculoarterial
connection" OR "double inlet ventricle" OR "discordant atrioventricular connection" OR
"isomerism of atrial appendages" OR "ventricular septal defect" OR VSD OR "atrial
septal defect" OR ASD OR "atrioventricular septal defect" OR "tetralogy of fallot" OR
"aortopulmonary septal defect" OR atresia OR "ebstein's" OR "hypoplastic right heart
syndrome" OR "hypoplastic left heart syndrome" OR dextrocardia OR levocardia OR "cor
triatriatum" OR "pulmonary infundibular stenosis" OR "patent ductus arteriosus" OR
PDA OR "coarctation of aorta" OR "supravalvular aortic stenosis" OR "stenosis of
pulmonary artery" OR "interruption of aortic arch" OR "persistent left superior vena
cava" OR "anomalous pulmonary venous connection" OR "anomalous portal venous
connection" OR "portal vein-hepatic artery fistula" OR "aberrant left subclavian artery"
OR "aberrant origin of left subclavian artery" OR "anomalous origin of right subclavian
artery" OR perinatal OR "fissured nose" OR "notched nose" OR "cleft nose" OR "web of
larynx" OR "laryngeal hypoplasia" OR laryngocele OR "accessory lobe of lung" OR
"sequestration of lung" OR "ectopic tissue in lung" OR "dysplasia of lung" OR "anomaly
of pleura" OR "cleft palate" OR "cleft lip" OR ankyloglossia OR macroglossia OR
"esophageal web" OR "meckel's" OR "hirshsprung's" OR aplasia OR "hypoplasia of
gallbladder" OR "choledochal cyst" OR "cystic disease of liver" OR "hypoplasia of
pancrea*" OR "annular pancreas" OR "developmental ovarian cyst" OR embryo* OR
"accessory ovary" OR "ovarian streak" OR "doubling of uterus" OR "bicornate uterus"
OR "unicornate uterus" OR "doubling of vagina" OR "imperforate hymen" OR "fusion of
labia" OR "undescended test*" OR "ectopic testi*" OR "abdominal testis" OR "ectopic
perineal testis" OR "high scrotal testis" OR hypospadias OR "hypoplasia of test*" OR
"hypoplasia of scrotum" OR polyorchism OR "retractile testis" OR "indeterminate sex"
OR pseudohermaphrodit* OR "reduction defect*" OR "renal hypoplasia" OR "potter's"
OR "cystic kidney disease" OR "renal dysplasia" OR "polycystic kidney" OR "medullary
cystic kidney" OR "multicystic dysplastic kidney" OR "duplication of ureter" OR
"malposition of ureter" OR cecouretocele OR "accessory kidney" OR "lobulated kidney"
OR "fused kidney" OR "horseshoe kidney" OR "ectopic kidney" OR "hyperplastic kidney"
OR "giant kidney" OR epispadias OR exstrophy OR "stenosis of urethra" OR "stenosis of
bladder neck" OR talipes OR "pez cavus" OR dolichocephaly OR plagiocephaly OR
"pectus excavatum" OR "pectus carinatum" OR "discoid meniscus" OR polydactyly OR
"accessory finger*" OR "accessory thumb*" OR "accessory toe*" OR "syndactyly” OR
"fused fingers" OR "webbed fingers" OR "fused toes" OR "webbed toes" OR
polysyndactyly OR "lobster-claw hand" OR "split foot" OR phocomelia OR
"arthrogryposis multiplex congenita" OR craniosynostosis OR dysostosis OR
hypertelorism OR macrocephaly OR "klippel-feil" OR spondylolisthesis OR scoliosis OR
"cervical rib*" OR osteochondrodysplasia OR achondrogenesis OR thanatophoric OR
"short rib syndrome" OR chondrodysplasia OR achondroplasia OR "diastrophic
dysplasia" OR "chondroectodermal dysplasia" OR “spondyloepiphyseal dysplasia" OR
127
"osteogenesis imperfecta" OR "polyostotic fibrous dysplasia" OR enchondromatosis OR
"metaphyseal dysplasia" OR exomphalos OR gastroschisis OR "prune belly" OR "ehlers-
danlos" OR "epidermolysis bullosa" OR "xeroderma pigmentosum" OR "incontinentia
pigmenti" OR "ectodermal dysplasia" OR "absence of breast" or "absent breast" OR
"accessory breast" OR "absent nipple" OR "absence of nipple" OR "accessory nipple" OR
anonychia OR "enlarged nails" OR "hypertrophic nails" OR phakomatos* OR
neurofibromatosis OR "tuberous sclerosis" OR schwannomatos* OR dysmorph* OR
"marfan's" OR "situs inversus" OR "conjoined twins" OR "down syndrome" or trisom*
OR triploid* OR polyploid* OR monosom* OR isochromosom* OR angelman OR "vel-
cardio-facial" OR "williams syndrome" OR microdeletion* OR autosom* OR "turner's"
OR karyotype OR mosaicism OR klinefelter OR chimera OR hermaphrodit* OR "fragile x"
OR maternal OR "prolapsed cord" OR "compression of umbilical cord" OR
chorioamnionitis OR breech OR "vacuum-assisted vaginal delivery" OR "precipitate
delivery" OR "abnormal uterine contraction*" OR gestation* OR "low birth weight" OR
newborn OR preterm OR "high birth weight" OR neonat* OR "caput sucedaneum" OR
"erb's" OR "klumpke's" OR "birth trauma" OR "wilson-mikity" OR omphalitis OR "vasa
previa" OR "ruptured cord" OR kernicterus OR "inspissated bile syndrome" OR "anemia
of prematurity" OR "diabetic mother" OR "inspissated milk" OR "hypoxic ischemic
encephalopathy" OR "failure to thrive" OR "grey baby" OR stillbirth OR "wide cranial
sutures" OR "delayed separation of umbilical cord") AND (CAHPS OR HCAHPS OR
MCAHPS OR HHCAHPS OR “S-CAHPS” OR “consumer assessment of healthcare providers
and systems” ) AND (“customer service” OR “physician communication” OR “patient-
provider communication” OR “patient-physician communication” OR “physician-patient
communication” OR "provider-patient communication” OR “getting care quickly” OR
“timely care” OR timeliness OR “get care quickly” OR “getting needed care” OR “get
needed care” OR “getting needed prescription drugs” OR “get needed prescription
drugs” OR “getting needed prescriptions” OR “get needed prescriptions” OR “care
coordination” OR “doctor communication” OR “provider communication” OR “physician
communication” OR “getting information about prescription drugs” OR “get information
about prescription drugs” OR “getting information about prescriptions” OR “get
information about prescriptions”)
10. External Causes
("intraoperative complications"/exp OR "medical errors"/exp OR "health services
misuse"/exp OR "failure of sterile precautions" OR "failure in dos*" OR "incorrect
dilution" OR "overdos* of radiation" OR "inadvertent exposure to radiation" OR
"innapropriate temperature" OR underdos* OR "nonadministration of necessary drug"
OR "contaminated medical substances" OR "contaminated biological substances" OR
misadventure* OR "mismatched blood" OR "wrong fluid" OR "failure in suture" OR
"wrongly placed" OR "failure to introduce" OR "failure to remove" OR "wrong
procedure" OR "wrong operation" OR "wrong body part" OR "wrong side of body" OR
complicat*)
11. General Clinical Outcomes
(“clinical outcomes” OR morbidity OR “mortality amenable to health care” OR “mortality amenable to
healthcare” OR mortality OR adherence OR non-adherence OR “progression-free survival” OR
128
“progression free survival” OR survival OR readmission OR “adverse event*” OR “adverse effect*” OR
hospitalization*)
129
Appendix B. PRISMA 2009 Checklist. Page numbers refer to published manuscript.
22
Section/topic # Checklist item
Reported
on page #
TITLE
Title 1 Identify the report as a systematic review, meta-analysis, or both. 1
ABSTRACT
Structured
summary
2 Provide a structured summary including, as applicable: background; objectives; data
sources; study eligibility criteria, participants, and interventions; study appraisal and
synthesis methods; results; limitations; conclusions and implications of key findings;
systematic review registration number.
1
INTRODUCTION
Rationale 3 Describe the rationale for the review in the context of what is already known. 1
Objectives 4 Provide an explicit statement of questions being addressed with reference to
participants, interventions, comparisons, outcomes, and study design (PICOS).
2
METHODS
Protocol and
registration
5 Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address),
and, if available, provide registration information including registration number.
3
Eligibility
criteria
6 Specify study characteristics (e.g., PICOS, length of follow-up) and report
characteristics (e.g., years considered, language, publication status) used as criteria for
eligibility, giving rationale.
3
Information
sources
7 Describe all information sources (e.g., databases with dates of coverage, contact with
study authors to identify additional studies) in the search and date last searched.
2
Search 8 Present full electronic search strategy for at least one database, including any limits
used, such that it could be repeated.
SDC 1
Study
selection
9 State the process for selecting studies (i.e., screening, eligibility, included in systematic
review, and, if applicable, included in the meta-analysis).
3
Data
collection
process
10 Describe method of data extraction from reports (e.g., piloted forms, independently, in
duplicate) and any processes for obtaining and confirming data from investigators.
4
Data items 11 List and define all variables for which data were sought (e.g., PICOS, funding sources)
and any assumptions and simplifications made.
4
Risk of bias in
individual
studies
12 Describe methods used for assessing risk of bias of individual studies (including
specification of whether this was done at the study or outcome level), and how this
information is to be used in any data synthesis.
4
Summary
measures
13 State the principal summary measures (e.g., risk ratio, difference in means). NA
Synthesis of
results
14 Describe the methods of handling data and combining results of studies, if done,
including measures of consistency (e.g., I
2
) for each meta-analysis.
NA
Risk of bias
across
studies
15 Specify any assessment of risk of bias that may affect the cumulative evidence (e.g.,
publication bias, selective reporting within studies).
NA
Additional
analyses
16 Describe methods of additional analyses (e.g., sensitivity or subgroup analyses, meta-
regression), if done, indicating which were pre-specified.
NA
130
Appendix C. Quality appraisal of included studies.
Risk of Selection Bias
Risk of Misclassification of
Outcome Variable Risk of Random Error Risk of Residual Confounding Risk of lack of External Validity
Study Judgement Notes Judgement Notes Judgement Notes Judgement Notes Judgement Notes
Bauer et
al., 2014
LOW 62% response
rate was
relatively low,
but survey was
offered in
multiple
formats and
languages and
non-response
weighting
performed
LOW Measures of
medication
adherence
are
validated;
objective
measures of
pharmacy
utilization
used
LOW Large
sample size
(N=1,523)
with tight
interval
estimates
LOW An evidence
based directed
acyclic graph
was used to
identify
potential
confounders
and avoid over
adjustment of
unnecessary
variables
MODERATE Population
with access to
health
care/insurance
and
Chakrabarti
et al., 2016
LOW Data collected
from Scientific
Registry of
Transplant
Recipients,
American
Hospital
Association,
and CMS
Hospital
Compare are
nationally
representative
LOW Rates of
kidney
transplant
failure were
objectively
reported but
the Scientific
Registry of
Transplant
Recipients
LOW Relatively
large
sample
(N=200) of
kidney
transplant
centers with
tight
interval
estimates
HIGH Data are risk
adjusted for
hospital
characteristics,
but not for
patient
characteristics
MODERATE Data derived
only from
kidney
transplant
centers
131
Appendix C. continued
Risk of Selection Bias
Risk of Misclassification of
Outcome Variable Risk of Random Error
Risk of Residual Confounding Risk of lack of External
Validity
Study Judgement Notes Judgement Notes Judgement Notes Judgement Notes Judgement Notes
Chang et
al., 2006
HIGH 56%
participation
rate; the
analytic sample
had greater
proportions of
higher income
patients,
patients with
diabetes,
smokers and a
lower
proportion of
patients
experiencing
falls
LOW Quality of
care index
compiled by
participant
interview
and chart
review
LOW Despite
smaller
sample size
(N=236),
interval
estimates
are
relatively
tight
LOW General health
and
demographic
factors were
adjusted for
MODERATE Population of
vulnerable
elders
enrolled in
managed care
organizations
Chen et
al., 2018
LOW MEPS
participant
selection is
nationally
representative;
performed
non-response
weighting
LOW Clinical
outcomes
queried
from MEPS
full-year
consolidated
files
LOW Large
sample size
(N=1,951)
with tight
interval
estimates
LOW Socio-
demographic
factors and
disease-
related risk
factors were
adjusted for
MODERATE Population of
patients with
hepato-
pancreato-
biliary disease
and a usual
source of care
Dottino et
al., 2019
LOW Data collected
from the
National
Inpatient
Sample and
CMS Hospital
Compare are
nationally
representative
LOW Inpatient
hospital
outcomes
were
abstracted
from
discharge
data
LOW Large
sample size
(N=17,509)
with tight
interval
estimates
LOW Data are
adjusted for
both hospital
and patient
characteristics
MODERATE Population of
inpatient
gynecologic
oncology
patients
132
Appendix C. continued
Risk of Selection Bias
Risk of Misclassification of
Outcome Variable Risk of Random Error Risk of Residual Confounding
Risk of lack of External
Validity
Study Judgement Notes Judgement Notes Judgement Notes Judgement Notes Judgement Notes
Dottino et
al., 2019
LOW Data collected
from the
National
Inpatient
Sample and
CMS Hospital
Compare are
nationally
representative
LOW Inpatient
hospital
outcomes
were
abstracted
from
discharge
data
LOW Large sample
size
(N=17,509)
with tight
interval
estimates
LOW Data are
adjusted for
both hospital
and patient
characteristics
MODERATE Population of
inpatient
gynecologic
oncology
patients
Dy et al.,
2016
MODERATE Data excludes
hospitals which
did not report
an author
defined
threshold of
CAHPS survey
data (may
exclude more
under-
resourced
centers with
subsequently
worse
outcomes than
the centers
included in this
analysis)
LOW Heart failure
outcome
measures
were
collected
from CMS
Hospital
Compare
data
LOW Large sample
(N=895) of
hospitals
LOW Data are
adjusted for
both hospital
and patient
characteristics
MODERATE Population of
inpatient
heart failure
patients
133
Appendix C. continued
Risk of Selection Bias
Risk of Misclassification of
Outcome Variable Risk of Random Error Risk of Residual Confounding
Risk of lack of External
Validity
Study Judgement Notes Judgement Notes Judgement Notes Judgement Notes Judgement Notes
Elder et al.,
2015
MODERATE African
American male
respondents to
CAHPS survey
more highly
educated than
general African
American male
population
LOW Global
ratings of
quality
queried
from CAHPS
3.0 data
LOW Large sample
size
(N=12,074)
with tight
interval
estimates
MODERATE Factors related
to access and
utilization
were adjusted
for, but the
potential
confounders of
health literacy
and
masculinity
were not
included
MODERATE Population of
African
American
men
Farias et
al., 2020
LOW SEER cancer
registry
program is
nationally
representative
sample of
cancer patients
LOW Cancer
outcomes
queried
from SEER-
CAHPS data
LOW Large sample
size
(N=10,144)
with tight
interval
estimates
MODERATE Socio-
demographic
factors and
insurance type
were adjusted
for, but the
potential
confounder of
type of care
setting was not
included
MODERATE Population of
women ≥ 65
years old with
breast cancer
Fernandez
et al., 2012
LOW 91%
participation
rate with
surveys
administered
in both English
and Spanish
LOW Clinical
outcomes
abstracted
from
medical
record
review
LOW Moderately
large sample
size (N=600)
with
relatively
tight interval
estimates
LOW Socio-
demographic
factors and
known or
hypothesized
risk factors
associated
with diabetes
control were
adjusted for
MODERATE Population of
patients with
type 2
diabetes who
have a
primary care
physician and
are not
privately
insured
134
Appendix C. continued
Risk of Selection Bias
Risk of Misclassification of
Outcome Variable Risk of Random Error Risk of Residual Confounding
Risk of lack of External
Validity
Study Judgement Notes Judgement Notes Judgement Notes Judgement Notes Judgement Notes
Gupta et
al., 2009
LOW 60% of
Hospital
Quality
Alliance
reporting
hospitals
reported
HCAHPS data;
Performed
case mix
adjustment to
reduce
nonresponse
bias
LOW Pain control
collected
from
validated
HCAHPS
self-report
LOW Large sample
(N=2,429) of
hospitals
HIGH Information on
inclusion of
potentially
confounding
variables
included in
analysis is not
clear
LOW Data
considers all
hospitalized
patients who
completed a
CAHPS survey
Hays et al.,
2018
MODERATE 45% response
rate;
Performed
non-response
weighting
LOW Global
ratings of
quality
queried
from CAHPS
Medicare
data
LOW Large sample
size
(N=242,871)
with small
estimates of
standard
error
MODERATE General
health, type of
insurance, and
living situation
were adjusted
for but other
potentially
confounding
socio-
demographic
variables were
not included
MODERATE Population of
Medicare
beneficiaries
with a
personal
physician
135
Appendix C. continued
Risk of Selection Bias
Risk of Misclassification of
Outcome Variable Risk of Random Error Risk of Residual Confounding
Risk of lack of External
Validity
Study Judgement Notes Judgement Notes Judgement Notes Judgement Notes Judgement Notes
Kshirsagar et
al., 2019
HIGH 44% of all
dialysis
facilities met
selection
criteria of
reporting ICH-
CAHPS, Quality
Incentive
Program, and
star rating data
LOW Quality
Incentive
Program
ratings were
derived from
a mandatory
federal pay-
for-
performance
system; star
ratings
published by
CMS
LOW Large
sample
(N=3,176) of
dialysis
facilities
with tight
interval
estimates
HIGH Analyses
adjusted for
facility
characteristics
but not
patient
characteristics
MODERATE Data derived
from dialysis
facilities
Odom-
Maryon et
al., 2019
MODERATE Details on
selection of
subset of
stroke patients
from all stroke
patients with
both American
Hospital
Association
Annual Survey
data and
HCAHPS data
are unclear
LOW Occurrence
of hospital
acquired
pressure
ulcer was
abstracted
from
Medicare
Beneficiary
Summary
files
LOW Moderately
large sample
size (N=297
cases, 4,158
controls)
with tight
interval
estimates
LOW Analyses
adjusted for
hospital
characteristics
and cases
were matched
to controls
based on
patient
characteristics
MODERATE Population of
Medicare
stroke patients
Okunrintemi
et al., 2017
LOW MEPS
participant
selection is
nationally
representative;
Performed
non-response
weighting
LOW Clinical
outcomes
queried from
MEPS
LOW Large
sample size
(N=6,810)
with tight
interval
estimates
LOW Covariates in
statistical
analyses
included
socio-
demographic
information,
modifiable risk
factors, and
comorbid
disease
burden
MODERATE Population of
atherosclerotic
cardiovascular
disease
patients
136
Appendix C. continued
Risk of Selection Bias
Risk of Misclassification of
Outcome Variable Risk of Random Error Risk of Residual Confounding
Risk of lack of External
Validity
Study Judgement Notes Judgement Notes Judgement Notes Judgement Notes Judgement Notes
Ratanawongsa
et al., 2013
LOW 62% response
rate was
relatively low,
but survey was
offered in
multiple
formats and
languages;
Performed
non-response
weighting
LOW Poor refill
adherence
was
measured
using
pharmacy
utilization
data
LOW Large
sample size
(N=9,377)
with tight
interval
estimates
LOW Statistical
analyses
adjusted for
socio-
demographic
and medical
characteristics
MODERATE Population of
privately
insured
patients with
continuous
pharmacy
benefits
Schneider et
al., 2001
HIGH 81% overall
response rate,
with response
rates of
individual
health plans
ranging from
39-94%
LOW Quality of
care data
queried
from
Medicare
HEDIS data
LOW Large
sample
(N=233) of
Medicare
health plans
HIGH Health plan
utilization
information
and region
were adjusted
for, but not
patient
characteristics
MODERATE Medicare
population
Shirk et al.,
2016
LOW Data collected
from the
National
Inpatient
Sample and
CMS Hospital
Compare are
nationally
representative
LOW Inpatient
hospital
outcomes
were
abstracted
from
discharge
data
LOW Large
sample size
(N=46,988)
with tight
interval
estimates
LOW Analyses are
adjusted for
patient
characteristics,
surgical
characteristics,
and hospital
staffing
characteristics
MODERATE Population of
patients
undergoing
surgery for
urologic
malignancies
137
Appendix C. continued
Risk of Selection Bias
Risk of Misclassification of
Outcome Variable Risk of Random Error Risk of Residual Confounding
Risk of lack of External
Validity
Study Judgement Notes Judgement Notes Judgement Notes Judgement Notes Judgement Notes
Winpenny
et al.,
2017
MODERATE 48%
response
rate;
Performed
case-mix
adjustment
MODERATE Participant
self-report of
smoking
habits and
history of
smoking
cessation
advice from
medical
provider
LOW Large sample
size
(N=26,432)
with small
estimates of
standard
error
LOW Standard
Medicare
CAHPS case-
mix adjustors
were included
as well as
information
about smoking
history,
demographics,
and medical
history
MODERATE Population of
Medicare
smokers
Yang et
al., 2018
LOW Data
includes all
all acute care
and critical
access
hospitals
reported on
in CMS
Hospital
Compare
LOW 30-day
readmission
data was
collected from
CMS Hospital
Compare
LOW Large sample
(N=4,535) of
hospitals
HIGH Analyses
adjusted for
hospital
characteristics
but not patient
characteristics
MODERATE Population of
acute
myocardial
infarction,
chronic
obstructive
pulmonary
disease,
hip/knee
surgery,
pneumonia,
stroke
patients
Abbreviations: CMS, Centers for Medicare & Medicaid Services; MMEPS, Medical Expenditure Panel Survey; CAHPS, Consumer Assessment of Healthcare Providers and
Systems; SEER, Surveillance, Epidemiology and End Results; ICH=CAHPS, In-Center Hemodialysis Consumer Assessment of Healthcare Providers and Systems; HEDIS,
Healthcare Effectiveness Data and Information Set
138
Appendix D. Descriptions of how diagnoses of diabetes, hyperlipidemia, and hypertension were identified in Medicare claims files.
Chronic
Condition
Medicare Claims Files
Searched and Claim
Quantity Required
Time Frame of
Medicare Claims
Included in Search ICD-9 and ICD-10 Diagnosis Codes (any diagnosis on the claim)
Diabetes
1. MEDPAR (1 claim)
or,
2. HHA (1 claim) or,
3. Outpatient (2
claims) or,
4. Carrier (2 claims)
0-24 months
before colorectal
cancer (CRC)
diagnosis
249.00, 249.01, 249.10, 249.11, 249.20, 249.21, 249.30, 249.31, 249.40, 249.41, 249.50, 249.51, 249.60,
249.61, 249.70, 249.71, 249.80, 249.81, 249.90, 249.91, 250.00, 250.01, 250.02, 250.03, 250.10, 250.11,
250.12, 250.13, 250.20, 250.21, 250.22, 250.23, 250.30, 250.31, 250.32, 250.33, 250.40, 250.41, 250.42,
250.43, 250.50, 250.51, 250.52, 250.53, 250.60, 250.61, 250.62, 250.63, 250.70, 250.71, 250.72, 250.73,
250.80, 250.81, 250.82, 250.83, 250.90, 250.91, 250.92, 250.93, 357.2, 362.01, 362.02, 362.03, 362.04,
362.05, 362.06, 366.41, E08.00, E08.01, E08.10, E08.11, E08.21, E08.22, E08.29, E08.311, E08.319, E08.321,
E08.3211, E08.3212, E08.3213, E08.3219, E08.329, E08.3291, E08.3292, E08.3293, E08.3299, E08.331,
E08.3311, E08.3312, E08.3313, E08.3319, E08.339, E08.3391, E08.3392, E08.3393, E08.3399, E08.341,
E08.3411, E08.3412, E08.3413, E08.3419, E08.349, E08.3491, E08.3492, E08.3493, E08.3499, E08.351,
E08.3511, E08.3512, E08.3513, E08.3519, E08.3521, E08.3522, E08.3523, E08.3529, E08.3531, E08.3532,
E08.3533, E08.3539, E08.3541, E08.3542, E08.3543, E08.3549, E08.3551, E08.3552, E08.3553, E08.3559,
E08.359, E08.3591, E08.3592, E08.3593, E08.3599, E08.36, E08.37X1, E08.37X2, E08.37X3, E08.37X9,
E08.39, E08.40, E08.41, E08.42, E08.43, E08.44, E08.49, E08.51, E08.52, E08.59, E08.610, E08.618, E08.620,
E08.621, E08.622, E08.628, E08.630, E08.638, E08.641, E08.649, E08.65, E08.69, E08.8, E08.9, E09.00,
E09.01, E09.10, E09.11, E09.21, E09.22, E09.29, E09.311, E09.319, E09.321, E09.3211, E09.3212, E09.3213,
E09.3219, E09.329, E09.3291, E09.3292, E09.3293, E09.3299, E09.331, E09.3311, E09.3312, E09.3313,
E09.3319, E09.339, E09.3391, E09.3392, E09.3393, E09.3399, E09.341, E09.3411, E09.3412, E09.3413,
E09.3419, E09.349, E09.3491, E09.3492, E09.3493, E09.3499, E09.351, E09.3511, E09.3512, E09.3513,
E09.3519, E09.3521, E09.3522, E09.3523, E09.3529, E09.3531, E09.3532, E09.3533, E09.3539, E09.3541,
E09.3542, E09.3543, E09.3549, E09.3551, E09.3552, E09.3553, E09.3559, E09.359, E09.3591, E09.3592,
E09.3593, E09.3599, E09.36, E09.37X1, E09.37X2, E09.37X3, E09.37X9, E09.39, E09.40, E09.41, E09.42,
E09.43, E09.44, E09.49, E09.51, E09.52, E09.59, E09.610, E09.618, E09.620, E09.621, E09.622, E09.628,
E09.630, E09.638, E09.641, E09.649, E09.65, E09.69, E09.8, E09.9, E10.10, E10.11, E10.21, E10.22, E10.29,
E10.311, E10.319, E10.321, E10.3211, E10.3212, E10.3213, E10.3219, E10.329, E10.3291, E10.3292,
E10.3293, E10.3299, E10.331, E10.3311, E10.3312, E10.3313, E10.3319, E10.339, E10.3391, E10.3392,
E10.3393, E10.3399, E10.341, E10.3411, E10.3412, E10.3413, E10.3419, E10.349, E10.3491, E10.3492,
E10.3493, E10.3499, E10.351, E10.3511, E10.3512, E10.3513, E10.3519, E10.3521, E10.3522, E10.3523,
E10.3529, E10.3531, E10.3532, E10.3533, E10.3539, E10.3541, E10.3542, E10.3543, E10.3549, E10.3551,
E10.3552, E10.3553, E10.3559, E10.359, E10.3591, E10.3592, E10.3593, E10.3599, E10.36, E10.37X1,
E10.37X2, E10.37X3, E10.37X9, E10.39, E10.40, E10.41, E10.42, E10.43, E10.44, E10.49, E10.51, E10.52,
E10.59, E10.610, E10.618, E10.620, E10.621, E10.622, E10.628, E10.630, E10.638, E10.641, E10.649, E10.65,
E10.69, E10.8, E10.9, E11.00, E11.01, E11.10, E11.11, E11.21, E11.22, E11.29, E11.311, E11.319, E11.321,
E11.3211, E11.3212, E11.3213, E11.3219, E11.329, E11.3291, E11.3292, E11.3293, E11.3299, E11.331,
E11.3311, E11.3312, E11.3313, E11.3319, E11.339, E11.3391, E11.3392, E11.3393, E11.3399, E11.341,
E11.3411, E11.3412, E11.3413, E11.3419, E11.349, E11.3491, E11.3492, E11.3493, E11.3499, E11.351,
E11.3511, E11.3512, E11.3513, E11.3519, E11.3521, E11.3522, E11.3523, E11.3529, E11.3531, E11.3532,
E11.3533, E11.3539, E11.3541, E11.3542, E11.3543, E11.3549, E11.3551, E11.3552, E11.3553, E11.3559,
E11.359, E11.3591, E11.3592, E11.3593, E11.3599, E11.36, E11.37X1, E11.37X2, E11.37X3,
139
Appendix D. continued
Chronic
Condition
Medicare Claims Files
Searched and Claim
Quantity Required
Time Frame of
Medicare Claims
Included in Search ICD-9 and ICD-10 Diagnosis Codes (any diagnosis on the claim)
Diabetes
(continued) (as above) (as above)
E11.37X9, E11.39, E11.40, E11.41, E11.42, E11.43, E11.44, E11.49, E11.51, E11.52, E11.59, E11.610,
E11.618, E11.620, E11.621, E11.622, E11.628, E11.630, E11.638, E11.641, E11.649, E11.65, E11.69, E11.8,
E11.9, E13.00, E13.01, E13.10, E13.11, E13.21, E13.22, E13.29, E13.311, E13.319, E13.321, E13.3211,
E13.3212, E13.3213, E13.3219, E13.329, E13.3291, E13.3292, E13.3293, E13.3299, E13.331, E13.3311,
E13.3312, E13.3313, E13.3319, E13.339, E13.3391, E13.3392, E13.3393, E13.3399, E13.341, E13.3411,
E13.3412, E13.3413, E13.3419, E13.349, E13.3491, E13.3492, E13.3493, E13.3499, E13.351, E13.3511,
E13.3512, E13.3513, E13.3519, E13.3521, E13.3522, E13.3523, E13.3529, E13.3531, E13.3532, E13.3533,
E13.3539, E13.3541, E13.3542, E13.3543, E13.3549, E13.3551, E13.3552, E13.3553, E13.3559, E13.359,
E13.3591, E13.3592, E13.3593, E13.3599, E13.36, E13.39, E13.40, E13.41, E13.42, E13.43, E13.44, E13.49,
E13.51, E13.52, E13.59, E13.610, E13.618, E13.620, E13.621, E13.622, E13.628, E13.630, E13.638, E13.641,
E13.649, E13.65, E13.69, E13.8, E13.9
Hyperlipidemia
1. MEDPAR (1 claim)
or,
2. HHA (1 claim) or,
3. Outpatient (2
claims) or,
4. Carrier (2 claims)
0-12 months
before CRC
diagnosis 272.0, 272.1, 272.2, 272.3, 272.4, E78.0, E78.00, E78.01, E78.1, E78.2, E78.3, E78.4, E78.41, E78.49, E78.5
Hypertension
1. MEDPAR (1 claim)
or,
2. HHA (1 claim) or,
3. Outpatient (2
claims) or,
4. Carrier (2 claims)
0-12 months
before CRC
diagnosis
362.11, 401.0, 401.1, 401.9, 402.00, 402.01, 402.10, 402.11, 402.90, 402.91, 403.00, 403.01, 403.10,
403.11, 403.90, 403.91, 404.00, 404.01, 404.02, 404.03, 404.10, 404.11, 404.12, 404.13, 404.90, 404.91,
404.92, 404.93, 405.01, 405.09, 405.11, 405.19, 405.91, 405.99, 437.2, H35.031, H35.032, H35.033,
H35.039, I10, I11.0, I11.9, I12.0, I12.9, I13.0, I13.10, I13.11, I13.2, I15.0, I15.1, I15.2, I15.8, I15.9, I67.4,
N26.2
140
Appendix E. List of individual questions comprising each included composite measure of CAHPS patient experience. Unless noted, response options for each
question included: “Never”, “Sometimes”, “Usually”, or “Always”.
CAHPS Patient
Experience Measure Individual Questions Asked
1997-2005
a
2007-2019
Care Coordination
N/A
1. When you visited your personal doctor for a scheduled
appointment, how often did he or she have your medical
records or other information about your care?
2. When your personal doctor ordered a blood test, x-ray or
other tests for you, how often did someone from your
personal doctor’s office follow up to give you those results?
3. When your personal doctor ordered a blood test, x-ray or
other tests for you, how often did you get those results as
soon as you needed them?
4. How often did you and your personal doctor talk about all the
prescription medicines you were taking?
5. Did you get the help you needed from your personal doctor’s
office to manage your care among these different providers
and services?
b
6. How often did your personal doctor seem informed and up-to-
date about the care you got from specialists?
Getting Care Quickly 1. When you needed care right away for an illness or injury [or
condition], how often did you get care as soon as you [thought
you] wanted?
2. [Not counting the times you needed care right away,] how often
did you get an appointment for [health care / regular or routine
care / regular or routine healthcare] as soon as you wanted?
3. How often [were you taken to the exam room within / did you
wait in the doctor’s office or clinic more than] 15 minutes [of /
past] your appointment time [to see the person you went to see]?
1. When you needed care right away, how often did you get care
as soon as you thought you needed?
2. Not counting the times you needed care right away, how often
did you get an appointment for your health care at a doctor’s
office or clinic as soon as you thought you needed?
3. Wait time includes time spent in the waiting room and exam
room. How often did you see the person you came to see
within 15 minutes of your appointment time?
Getting Needed Care 1. How much of a problem, if any, was it to see a specialist that you
needed to see?
c
2. How much of a problem, if any, was it to get the care [,tests or
treatment] you or a doctor believed necessary?
c
3. Since you joined you health plan, how much of a problem, if any
was it to get a personal doctor or nurse you are happy with?
c
4. How much of a problem, if any, were delays in health care while
you waited for approval from Medicare?
c
1. How often was it easy to get appointments with specialists?
2. How often was it easy to get the care, tests, or treatment you
thought you needed through Medicare?
141
Appendix E. continued
CAHPS Patient
Experience Measure Individual Questions Asked
1997-2005
a
2007-2019
Doctor
Communication
1. How often did doctors or other health providers explain things in
a way you could understand?
2. How often did doctors or other health providers listen carefully
to you?
3. How often did doctors or other health providers show respect for
what you had to say?
4. How often did doctors or other health providers spend enough
time with you?
1. How often did your personal doctor explain things in a way
that was easy to understand?
2. How often did your personal doctor listen carefully to you?
3. How often did your personal doctor show respect for what
you had to say?
4. How often did your personal doctor spend enough time with
you?
a. Brackets included to indicate wording of various question versions.
b. Response choices for this question are “Yes, Definitely”, “Yes, Somewhat”, or “No”.
c. Response choices for this question are “A Big Problem”, “A Small Problem”, and “Not a Problem”.
142
Appendix F. Descriptions of how measures of health care services use were identified in Medicare claims files for the 24 months before and the 24 months after colorectal
cancer diagnosis.
Measure
Medicare
Claims Files
Searched Identification Criteria ICD-9 or ICD-10 Diagnosis Codes HCPCS Codes
Revenue Center
Codes
BETOS
Codes
Outpatient
evaluation and
management visits
for chronic
conditions
1. Outpatient
2. Carrier
1. Claim must have
evaluation and
management HCPCS
code or BETOS code
1-4
2. Primary or
secondary diagnosis
code must be for
diabetes,
hyperlipidemia, or
hypertension
249.00, 249.01, 249.10, 249.11, 249.20, 249.21,
249.30, 249.31, 249.40, 249.41, 249.50, 249.51,
249.60, 249.61, 249.70, 249.71, 249.80, 249.81,
249.90, 249.91, 250.00, 250.01, 250.02, 250.03,
250.10, 250.11, 250.12, 250.13, 250.20, 250.21,
250.22, 250.23, 250.30, 250.31, 250.32, 250.33,
250.40, 250.41, 250.42, 250.43, 250.50, 250.51,
250.52, 250.53, 250.60, 250.61, 250.62, 250.63,
250.70, 250.71, 250.72, 250.73, 250.80, 250.81,
250.82, 250.83, 250.90, 250.91, 250.92, 250.93,
357.2, 362.01, 362.02, 362.03, 362.04, 362.05,
362.06, 366.41, E08.00, E08.01, E08.10, E08.11,
E08.21, E08.22, E08.29, E08.311, E08.319,
E08.321, E08.3211, E08.3212, E08.3213,
E08.3219, E08.329, E08.3291, E08.3292,
E08.3293, E08.3299, E08.331, E08.3311,
E08.3312, E08.3313, E08.3319, E08.339,
E08.3391, E08.3392, E08.3393, E08.3399,
E08.341, E08.3411, E08.3412, E08.3413,
E08.3419, E08.349, E08.3491, E08.3492,
E08.3493, E08.3499, E08.351, E08.3511,
E08.3512, E08.3513, E08.3519, E08.3521,
E08.3522, E08.3523, E08.3529, E08.3531,
E08.3532, E08.3533, E08.3539, E08.3541,
E08.3542, E08.3543, E08.3549, E08.3551,
E08.3552, E08.3553, E08.3559, E08.359,
E08.3591, E08.3592, E08.3593, E08.3599, E08.36,
E08.37X1, E08.37X2, E08.37X3, E08.37X9, E08.39,
E08.40, E08.41, E08.42, E08.43, E08.44, E08.49,
E08.51, E08.52, E08.59, E08.610, E08.618,
E08.620, E08.621, E08.622, E08.628, E08.630,
E08.638, E08.641, E08.649, E08.65, E08.69, E08.8,
E08.9, E09.00, E09.01, E09.10, E09.11, E09.21,
E09.22, E09.29, E09.311, E09.319, E09.321,
E09.3211, E09.3212, E09.3213, E09.3219,
E09.329, E09.3291, E09.3292, E09.3293,
E09.3299, E09.331, E09.3311, E09.3312,
99201, 99202,
99203, 99204,
99205, 99211,
99212, 99213,
99214, 99215,
99241, 99242,
99243, 99244,
99245, 99354,
99387, 99397 NA
M1A,
M1B,
M5A,
M5B,
M5C,
M5D, M6
143
Appendix F. continued
Measure
Medicare
Claims Files
Searched Identification Criteria ICD-9 or ICD-10 Diagnosis Codes HCPCS Codes
Revenue Center
Codes
BETOS
Codes
Outpatient
evaluation and
management visits
for chronic
conditions
(continued) (as above) (as above)
E09.3313, E09.3319, E09.339, E09.3391,
E09.3392, E09.3393, E09.3399, E09.341,
E09.3411, E09.3412, E09.3413, E09.3419,
E09.349, E09.3491, E09.3492, E09.3493,
E09.3499, E09.351, E09.3511, E09.3512,
E09.3513, E09.3519, E09.3521, E09.3522,
E09.3523, E09.3529, E09.3531, E09.3532,
E09.3533, E09.3539, E09.3541, E09.3542,
E09.3543, E09.3549, E09.3551, E09.3552,
E09.3553, E09.3559, E09.359, E09.3591,
E09.3592, E09.3593, E09.3599, E09.36, E09.37X1,
E09.37X2, E09.37X3, E09.37X9, E09.39, E09.40,
E09.41, E09.42, E09.43, E09.44, E09.49, E09.51,
E09.52, E09.59, E09.610, E09.618, E09.620,
E09.621, E09.622, E09.628, E09.630, E09.638,
E09.641, E09.649, E09.65, E09.69, E09.8, E09.9,
E10.10, E10.11, E10.21, E10.22, E10.29, E10.311,
E10.319, E10.321, E10.3211, E10.3212, E10.3213,
E10.3219, E10.329, E10.3291, E10.3292,
E10.3293, E10.3299, E10.331, E10.3311,
E10.3312, E10.3313, E10.3319, E10.339,
E10.3391, E10.3392, E10.3393, E10.3399,
E10.341, E10.3411, E10.3412, E10.3413,
E10.3419, E10.349, E10.3491, E10.3492,
E10.3493, E10.3499, E10.351, E10.3511,
E10.3512, E10.3513, E10.3519, E10.3521,
E10.3522, E10.3523, E10.3529, E10.3531,
E10.3532, E10.3533, E10.3539, E10.3541,
E10.3542, E10.3543, E10.3549, E10.3551,
E10.3552, E10.3553, E10.3559, E10.359,
E10.3591, E10.3592, E10.3593, E10.3599, E10.36,
E10.37X1, E10.37X2, E10.37X3, E10.37X9, E10.39,
E10.40, E10.41, E10.42, E10.43, E10.44, E10.49,
E10.51, E10.52, E10.59, E10.610, E10.618,
E10.620, E10.621, E10.622, E10.628, E10.630,
E10.638, E10.641, E10.649, E10.65, E10.69, E10.8,
E10.9, E11.00, E11.01, E11.10, E11.11, E11.21,
E11.22, E11.29, E11.311, E11.319, E11.321, (as above) (as above)
(as
above)
144
Appendix F. continued
Measure
Medicare
Claims Files
Searched Identification Criteria ICD-9 or ICD-10 Diagnosis Codes HCPCS Codes
Revenue Center
Codes
BETOS
Codes
Outpatient
evaluation and
management visits
for chronic
conditions
(continued) (as above) (as above)
E11.3211, E11.3212, E11.3213, E11.3219,
E11.329, E11.3291, E11.3292, E11.3293,
E11.3299, E11.331, E11.3311, E11.3312,
E11.3313, E11.3319, E11.339, E11.3391,
E11.3392, E11.3393, E11.3399, E11.341,
E11.3411, E11.3412, E11.3413, E11.3419,
E11.349, E11.3491, E11.3492, E11.3493,
E11.3499, E11.351, E11.3511, E11.3512,
E11.3513, E11.3519, E11.3521, E11.3522,
E11.3523, E11.3529, E11.3531, E11.3532,
E11.3533, E11.3539, E11.3541, E11.3542,
E11.3543, E11.3549, E11.3551, E11.3552,
E11.3553, E11.3559, E11.359, E11.3591,
E11.3592, E11.3593, E11.3599, E11.36, E11.37X1,
E11.37X2, E11.37X3, E11.37X9, E11.39, E11.40,
E11.41, E11.42, E11.43, E11.44, E11.49, E11.51,
E11.52, E11.59, E11.610, E11.618, E11.620,
E11.621, E11.622, E11.628, E11.630, E11.638,
E11.641, E11.649, E11.65, E11.69, E11.8, E11.9,
E13.00, E13.01, E13.10, E13.11, E13.21, E13.22,
E13.29, E13.311, E13.319, E13.321, E13.3211,
E13.3212, E13.3213, E13.3219, E13.329,
E13.3291, E13.3292, E13.3293, E13.3299,
E13.331, E13.3311, E13.3312, E13.3313,
E13.3319, E13.339, E13.3391, E13.3392,
E13.3393, E13.3399, E13.341, E13.3411,
E13.3412, E13.3413, E13.3419, E13.349,
E13.3491, E13.3492, E13.3493, E13.3499,
E13.351, E13.3511, E13.3512, E13.3513,
E13.3519, E13.3521, E13.3522, E13.3523,
E13.3529, E13.3531, E13.3532, E13.3533,
E13.3539, E13.3541, E13.3542, E13.3543,
E13.3549, E13.3551, E13.3552, E13.3553,
E13.3559, E13.359, E13.3591, E13.3592,
E13.3593, E13.3599, E13.36, E13.39, E13.40,
E13.41, E13.42, E13.43, E13.44, E13.49, E13.51,
E13.52, E13.59, E13.610, E13.618, E13.620,
E13.621, E13.622, E13.628, E13.630, E13.638, (as above) (as above)
(as
above)
145
Appendix F. continued
Identification Criteria
ICD-9 or ICD-
10 Diagnosis
Codes HCPCS Codes Revenue Center Codes BETOS Codes
Outpatient
evaluation and
management visits
for chronic
conditions
(continued) (as above) (as above)
E13.641, E13.649, E13.65, E13.69, E13.8, E13.9,
272.0, 272.1, 272.2, 272.3, 272.4, E78.0, E78.00,
E78.01, E78.1, E78.2, E78.3, E78.4, E78.41,
E78.49, E78.5, 362.11, 401.0, 401.1, 401.9,
402.00, 402.01, 402.10, 402.11, 402.90, 402.91,
403.00, 403.01, 403.10, 403.11, 403.90, 403.91,
404.00, 404.01, 404.02, 404.03, 404.10, 404.11,
404.12, 404.13, 404.90, 404.91, 404.92, 404.93,
405.01, 405.09, 405.11, 405.19, 405.91, 405.99,
437.2, H35.031, H35.032, H35.033, H35.039, I10,
I11.0, I11.9, I12.0, I12.9, I13.0, I13.10, I13.11,
I13.2, I15.0, I15.1, I15.2, I15.8, I15.9, I67.4, N26.2 (as above) (as above)
(as
above)
All-cause emergency
department visits
5
1. MEDPAR
2. Outpatient
1. For MEDPAR claims:
charge amount for
emergency room
services provided
must be greater than
zero
2. For Outpatient
claims: must have
emergency
department revenue
center code NA NA
For outpatient
claims only: 0450,
0451, 0452, 0453,
0454, 0455, 0456,
0457, 0458, 0459,
0981 NA
All-cause
hospitalizations 1. MEDPAR
1. Any MEDPAR
inpatient
hospitalization claim NA NA NA NA
1. https://www.ama-assn.org/system/files/2019-06/cpt-office-prolonged-svs-code-changes.pdf
2. https://medicaid.ms.gov/wp-content/uploads/2021/10/CPT-Category-II-Codes-Chart.pdf
3. https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-
Reports/MedicareFeeforSvcPartsAB/downloads/BETOSDescCodes.pdf
4. https://resdac.org/cms-data/variables/evaluation-and-management-beneficiary-payments
5. https://resdac.org/articles/how-identify-hospital-claims-emergency-room-visits-medicare-claims-data
146
Appendix G. Matrix of outcome modelling decision processes.
Conceptual Motivation Modelling Notes
Categorization of Outpatient Visits
Before and after colorectal cancer (CRC)
diagnosis:
1. 0 visits before/0 visits after
2. more visits before/fewer visits after
3. fewer visits before/more visits after
4. X number of visits before/X number of
visits after
a
• Captures change in care use after CRC diagnosis, compared
to before.
• Considers patients who never used care as a separate user
group. • Multivariable models would not converge.
Before and CRC diagnosis: (where
inappropriate care defined as <3 or >9 visits
and appropriate care defined as 3-9 visits)
1. inappropriate care
before/inappropriate care after
2. inappropriate care before/appropriate
care after
3. appropriate care before/inappropriate
care after
4. appropriate care before/appropriate
care after
• Captures use of care both and after diagnosis to make
conclusions about changes in care use. • Multivariable models would not converge.
After CRC diagnosis only:
1. 0 visits
2. 1-2 visits
3. 3-9 visits
4. 10+ visits
• Considers patients who never used care as a separate user
group.
• Divides “ever-users” into those with an underuse of care
relative to recommendations, those with appropriate use
relative to recommendations, and those with an overuse of
care relative to recommendations. • Multivariable models would not converge.
After CRC diagnosis only:
1. <3 visits
2. 3-9 visits
3. 10+ visits
• Compared to above model, collapses those who never used
care with those who underused care relative to
recommendations.
• Some effect estimates and confidence intervals were
quite large or quite small.
After CRC diagnosis only:
1. 0 visits
2. 1-2 visits or 10+ visits
3. 3-9 visits
• Considers patients who never used care as a separate user
group.
• Divides “ever-users” into those with an “inappropriate use”
of care relative to recommendations and those with
“appropriate use” relative to recommendations.
• Some effect estimates and confidence intervals were
quite small.
• When stratifying models (as opposed to using a product
interaction term), multivariable models would not
converge.
147
Appendix G. continued
Conceptual Motivation Modelling Notes
After CRC diagnosis only:
1. <3 or >9 visits
2. 3-9 visits
• Compared to above model, collapses those who never used
care with those who “inappropriately used” care relative to
recommendations.
• Interpretation of models was difficult, as those grouped
as “overusers” (grouped in “inappropriate use” group)
may just have more severe disease and may require a
higher frequency of visits.
After CRC diagnosis only:
1. 0 visits
2. 1-2 visits
3. 3+ visits
• Considers patients who never used care as a separate user
group.
• Divides “ever-users” into those with a sufficient number of
visits and those with less than a sufficient number of visits.
• Modelling of three outcome groups makes for
complicated interpretations that were not easily
translated into recommendations for clinical practice.
Categorization of Emergency Department Visits After CRC Diagnosis
1. 0 visits
2. 1-2 visits
3. 3+ visits
• Considers patients who never used care as a separate user
group.
• Divides “ever-users” into those with lower and higher
frequency of use.
• Interpretation of models was difficult, as cut points for
lower and higher users lacked clinical basis or
relevance.
Categorization of Hospitalizations After CRC Diagnosis
1. 0 visits
2. 1-2 visits
3. 3+ visits
• Considers patients who never used care as a separate user
group.
• Divides “ever-users” into those with lower and higher
frequency of use.
• Interpretation of models was difficult, as cut points for
lower and higher users lacked clinical basis or
relevance.
a. Methodology abstracted from Tsui et el., 2020.
182
148
Appendix H. Descriptions of how stage-specific guideline concordant treatment for colorectal cancer (CRC) was identified in Medicare claims files.
CRC Stage
at
Diagnosis
CRC Stage-
Specific
Treatment
Medicare
Claims Files
Searched
Time Frame
of Medicare
Claims
Included in
Search
ICD-9 or ICD-10 Diagnosis
Codes (any diagnosis on the
claim) CPT Codes HCPCS Codes
Revenue
Center
Codes
In situ,
Localized,
and
Regional
Surgical
Resection
1. MEDPAR
2. Outpatient
3. Carrier
0-6 months
after CRC
diagnosis
17.3, 17.31, 17.32, 17.33, 17.34,
17.35, 17.36, 45.7, 45.71, 45.72,
45.73, 45.74, 45.75, 45.76,
45.79, 45.8, 45.81, 45.82, 45.83,
48.4, 48.40, 48.41, 48.42, 48.43,
48.49, 48.5, 48.50, 48.51, 48.52,
48.59, 48.6, 48.61, 48.62, 48.63,
48.64, 48.65, 48.69, 0DTE4ZZ,
0DTF4ZZ, 0DTG4ZZ, 0DTH4ZZ,
0DTK4ZZ, 0DTL4ZZ, 0DTM4ZZ,
0DTN4ZZ, 0DTE8ZZ, ODTF8ZZ,
0DTG8ZZ, 0DTH8ZZ, 0DTK8ZZ,
0DTL8ZZ, 0DTM8ZZ, 0DTN8ZZ,
0DTEFZZ, 0DTFFZZ, 0DTGFZZ,
0DTLFZZ, 0DTMFZZ, 0DTNFZZ,
0DTE0ZZ, 0DTF0ZZ, 0DTG0ZZ,
0DTH0ZZ, ODTK0ZZ, 0DTL0ZZ,
0DTM0ZZ, 0DTN0ZZ, 0DTE7ZZ,
0DTF7ZZ, 0DTG7ZZ, 0DTH7ZZ,
0DTH7ZZ, 0DTK7ZZ, 0DTL7ZZ,
0DTM7ZZ, 0DTN7ZZ
44110, 44111, 44130,
44139, 44140, 44141,
44143, 44144, 44145,
44146, 44147, 44150,
44151, 44152, 44153,
44155, 44156, 44157,
44158, 44160, 44202,
44203, 44204, 44205,
44206, 44207, 44208,
44210, 44211, 44212,
44238, 44239, 44950,
44955, 44960, 44970,
44979, 45110, 45111,
45112, 45113, 45114,
45116, 45119, 45120,
45121, 45123, 45126,
45130, 45135, 45160,
45170, 45171, 45172,
45190, 45378, 45383,
45384, 45385, 45392,
45395, 45397 NA NA
In situ and
Localized
Surveillance
Colonoscopy
1. MEDPAR
2. Outpatient
3. Carrier
6-18 months
after surgical
resection for
CRC
45.21, 45.22, 45.23, 45.25,
45.27, 45.41, 45.42, 45.43,
48.24, 48.36, 0DJD8ZZ,
0DBE8ZX, 0DBH8ZX,
0DBN8ZX, 0DDE8ZX,
0DDH8ZX, 0DDN8ZX,
0D9E8ZX, 0D9H8ZX,
0D9N8ZX, 0DBP8ZX,
0D9P8ZX, 0DBP8ZZ
44388, 44389, 44390,
44394, 44397, 44401,
44402, 44403, 44404,
44405, 44406, 44407,
44408, 45355, 45378,
45379, 45380-45393,
45398
G0105, G0121,
G0120, G6019,
G6020, G6024,
G6025 NA
149
Appendix H. continued
CRC Stage
at
Diagnosis
CRC Stage-
Specific
Treatment
Medicare
Claims Files
Searched
Time Frame
of Medicare
Claims
Included in
Search
ICD-9 or ICD-10 Diagnosis
Codes (any diagnosis on the
claim) CPT Codes HCPCS Codes
Revenue
Center
Codes
Regional
Adjuvant
Chemotherapy
1. Outpatient
2. Carrier
After
surgical
resection for
CRC and
within 6
months of
CRC
diagnosis NA
36640, 51720, 96400,
96405, 96406, 96408,
96410, 96412, 96414,
96420, 96422, 96423,
96425, 96440, 96445,
96450, 96501, 96504,
96505, 96508, 96510,
96511, 96512, 96520,
96524, 96530, 96538,
96540, 96542, 96545,
96549, 96550, 96555
C8953, C8954,
C8955, G0355,
G0356, G0359,
G0361, J9190,
J0640, J9263,
J8520, J8521,
Q0083, Q0084,
Q0085, S9329,
S9330, S9331,
WW089,
WW090,
WW091,
WW093,
WW094, WW096
0331,
0332,
0335
Abstract (if available)
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Racial/ethnic differences in colorectal cancer patient experiences, health care utilization and their association with mortality: findings from the SEER-CAHPS data
PDF
Racial/ethnic differences in patient experiences with health care in association with the Gleason’s score level at prostate cancer diagnosis: findings from the SEER-CAHPS data
PDF
Disparities in colorectal cancer survival among Latinos in California
PDF
The impact of the COVID-19 pandemic on cancer care delivery
PDF
Disparities in gallbladder, intra-hepatic bile duct, and other biliary cancers among multi-ethnic populations: a California Cancer Registry study
PDF
Racial/ethnic variation in care preferences and care outcomes among United States hospice enrollees
PDF
Prostate cancer disparities among Californian Latinos by country of origin: clinical characteristics, incidence, treatment received and survival
PDF
Racial and ethnic disparities in delays of surgical treatment for breast cancer
PDF
Melanoma in children, adolescents, and young adults
PDF
Exploring geography and multi-level factors associated with disparities in HPV vaccination among Latino adolescents in Los Angeles County
PDF
Is stress exposure enough? Race/ethnic differences in the exposure and appraisal of chronic stressors among older adults
PDF
Childhood cancer survivorship: parental factors associated with survivor's follow-up care behavior and mental health
PDF
Carcinogenic exposures in racial/ethnic groups
PDF
Chronic eye disease epidemiology in the multiethnic ophthalmology cohorts of California study
PDF
Genes and environment in prostate cancer risk and prognosis
PDF
Exploring end-of-life care experiences and disparities among Hispanic populations in the United States
PDF
Ancestral/Ethnic variation in the epidemiology and genetic predisposition of early-onset hematologic cancers
PDF
A biodemographic approach to understanding sociodemographic disparities in kidney functioning on three dimensions: individual, population, and cross-national
PDF
Healthcare affordability among adults with liver disease from 2004-2018 in the United States: a national cohort study
PDF
The association between Echocardiography use in the Intensive Care Unit and 28-day mortality among patients with septic shock: a systematic review
Asset Metadata
Creator
Navarro, Stephanie
(author)
Core Title
Investigating racial and ethnic disparities in patient experiences with care and health services use following colorectal cancer diagnosis among older adults with comorbid chronic conditions
School
Keck School of Medicine
Degree
Doctor of Philosophy
Degree Program
Epidemiology
Degree Conferral Date
2023-05
Publication Date
04/06/2023
Defense Date
03/01/2023
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
CAHPS,colorectal cancer,guideline concordant treatment,health services use,multimorbidity,OAI-PMH Harvest,patient experiences with care,racial and ethnic disparities
Format
theses
(aat)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Farias, Albert J. (
committee chair
), Barzi, Afsaneh (
committee member
), Pickering, Trevor (
committee member
), Stern, Mariana C. (
committee member
), Tsui, Jennifer (
committee member
)
Creator Email
stephaan@usc.edu,stephnav110@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-oUC112952492
Unique identifier
UC112952492
Identifier
etd-NavarroSte-11577.pdf (filename)
Legacy Identifier
etd-NavarroSte-11577
Document Type
Dissertation
Format
theses (aat)
Rights
Navarro, Stephanie
Internet Media Type
application/pdf
Type
texts
Source
20230406-usctheses-batch-1017
(batch),
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Repository Email
cisadmin@lib.usc.edu
Tags
CAHPS
colorectal cancer
guideline concordant treatment
health services use
multimorbidity
patient experiences with care
racial and ethnic disparities