Patient Experience Journal Patient Experience Journal
Volume 8
Issue 2
The Impact of Inequity & Health
Disparities on the Human Experience
Article 6
2021
Sociodemographic characteristics and patient and family Sociodemographic characteristics and patient and family
experience survey response biases experience survey response biases
Lauren N. Brinkman
Cincinnati Children's Hospital Medical Center
Myra S. Saeed
Cincinnati Children's Hospital Medical Center
Andrew F. Beck
Cincinnati Children's Hospital Medical Center
Michael C. Ponti-Zins
Cincinnati Children's Hospital Medical Center
Ndidi I. Unaka
Cincinnati Children's Hospital Medical Center
See next page for additional authors
Follow this and additional works at: https://pxjournal.org/journal
Part of the Health Services Administration Commons, Health Services Research Commons, and the
Quality Improvement Commons
Recommended Citation Recommended Citation
Brinkman LN, Saeed MS, Beck AF, Ponti-Zins MC, Unaka NI, Burkhardt MC, Meinzen-Derr J, Hanke SP.
Sociodemographic characteristics and patient and family experience survey response biases.
Patient
Experience Journal
. 2021; 8(2):18-25. doi: 10.35680/2372-0247.1536.
This Research is brought to you for free and open access by Patient Experience Journal. It has been accepted for
inclusion in Patient Experience Journal by an authorized editor of Patient Experience Journal.
Sociodemographic characteristics and patient and family experience survey Sociodemographic characteristics and patient and family experience survey
response biases response biases
Cover Page Footnote Cover Page Footnote
We would like to thank the thousands of families and patients who took the time to give us feedback on
their care and experiences, without which this study would not be possible. This article is associated with
the Policy & Measurement lens of The Beryl Institute Experience Framework
(https://www.theberylinstitute.org/ExperienceFramework). You can access other resources related to this
lens including additional PXJ articles here: http://bit.ly/PX_PolicyMeasure
Authors Authors
Lauren N. Brinkman, Myra S. Saeed, Andrew F. Beck, Michael C. Ponti-Zins, Ndidi I. Unaka, Mary C.
Burkhardt, Jareen Meinzen-Derr, and Samuel P. Hanke
This research is available in Patient Experience Journal: https://pxjournal.org/journal/vol8/iss2/6
Patient Experience Journal
Volume 8, Issue 2 – 2021, pp. 18-25
Patient Experience Journal, Volume 8, Issue 2 – Special Issue: The Impact of Inequity & Health Disparities on the Human Experience
© The Author(s), 2021. Published in association with The Beryl Institute.
Downloaded from www.pxjournal.org 18
Research
Sociodemographic characteristics and patient and family experience survey
response biases
Lauren N. Brinkman, Cincinnati Children's Hospital Medical Center, lauren.brinkman@cchmc.org
Myra S. Saeed, Cincinnati Children's Hospital Medical Center, [email protected]
Andrew F. Beck, Cincinnati Children's Hospital Medical Center, andrew.beck1@cchmc.org
Michael C. Ponti-Zins, Cincinnati Children's Hospital Medical Center, Michael.Ponti-Zins@cchmc.org
Ndidi I. Unaka, Cincinnati Children's Hospital Medical Center, [email protected]
Mary C. Burkhardt, Cincinnati Children's Hospital Medical Center, Mary.Burkhardt@cchmc.org
Jareen Meinzen-Derr, Cincinnati Children's Hospital Medical Center & Univ. of Cincinnati, j[email protected]
Samuel P. Hanke, Cincinnati Children's Hospital Medical Center, Samuel.Hanke@cchmc.org
Abstract
Enhancing Patient and Family Experience (PFE) is vital to the delivery of quality healthcare services. Sociodemographic
differences affect health outcomes and experiences, but research is limited on biases in PFE survey methodology. We
sought to assess survey participation rates across sociodemographic characteristics. This retrospective study analyzed a
health system’s ambulatory PFE survey data, collected January 1 – July 31, 2019. Outcomes of interest were rates of survey
response, completion, and comments. Predictors included respondent-reported race, ethnicity, language, and measure of
social deprivation attached to a respondent’s home address. Addresses were geocoded to census tracts. The tract’s degree
of socioeconomic deprivation was defined using the Deprivation Index (DPI). Associations between outcomes and
predictors were assessed using the Chi square test. 77,627 unique patient encounters were analyzed. Patients were
predominantly White (76%); 5% were Hispanic; and 1% were Spanish-speaking. The overall response, completion, and
comment rates were 20.1%, 17.6%, and 4.1%, respectively. There were significant differences across assessed
sociodemographic characteristics in response, completion, and comment rates. White patients were most likely to respond,
complete, and leave a comment. Spanish-speaking respondents and those living in the most deprived areas were more
likely to respond and complete the survey, but less likely to comment than English-speaking respondents and those living
in less deprived areas, respectively. PFE survey participation differs across a range of sociodemographic characteristics,
potentially introducing noteworthy biases. Health systems should minimize differences in how they collect feedback and
account for potential biases when responding to experience data.
Keywords
Measurement, patient experience, pediatrics, survey response bias, social determinants of health
Introduction
Patient and Family Experience (PFE) surveys are used by
healthcare organizations to assess perceptions of
healthcare delivery and quality within their systems.
However, survey participation biases may affect survey
results and should be considered during interpretation. For
instance, patients’ and families’ response rates vary based
on their perceptions of a provider or experience.
Additionally, while a single patient may score a provider or
experience similarly across multiple questions, ratings may
differ considerably from patient to patient, or family to
family.
1
Furthermore, social and economic inequities
influence health outcomes.
2, 3
Emerging evidence suggests
these differences may impact patient experience.
4
Patient
specific characteristics, such as health status,
4
English
proficiency,
5
age, education level,
4,6
race, and employment
status
6
affect a patient’s ability to participate in their
healthcare
5,6
and likely influence how different patients (or
family member respondents) report on their experience.
4
The varied characteristics and experiences of patients and
families are important for healthcare systems to
understand fully in order to implement systemic changes
and drive improvement. Still, most work assessing
potential PFE survey participation biases has been
conducted among adults. In a study of adults, Tyser et al.
found differences in survey participation based on ages
from 18 and up, sex, insurance type, and orthopedic care
visit types.
7
A similar study, conducted in a pediatric
subspecialty division, found response biases stemming
from demographic differences in race, and insurance status
as well.
8
There is a paucity of research into the impact of
sociodemographic characteristics on patient experience
Sociodemographic characteristics and survey response bias, Brinkman et al.
19 Patient Experience Journal, Volume 8, Issue 1 – 2021
within pediatrics.
9
Since survey participation differences
can affect the generalizability and usefulness of PFE
surveys results,
1
such differences are critical to understand
participation biases. Such data is lacking especially in the
context of pediatric healthcare, leaving a gap in the
understanding of potential participation biases in PFE data
sets. Thus, we sought to enumerate response, completion,
and comment rates to our PFE survey across key
sociodemographic characteristics.
Patients and Methods
PFE Survey Methodology
This retrospective study took place at Cincinnati
Children’s Hospital Medical Center (CCHMC) using data
collected between January 1, 2019 and July 31, 2019.
CCHMC is a large, urban, free-standing, academic, acute
care, children’s hospital. For patients contributing more
than one visit during the study period, one visit was
selected at random. This study analyzed 77,627 encounters
from ambulatory medical and surgical specialty divisions
where surveys where administered. The study was
reviewed and deemed exempt by the CCHMC Institutional
Review Board.
PFE surveys are administered following clinical encounters
using the NRC Health Real-Time survey platform.
10
Families are automatically contacted within 24-48 hours of
an ambulatory visit requesting completion of the PFE
survey. This outreach is initially via email, using an email
address provided at the time of clinical encounter. If no
response is received within 24 hours of the email, a second
outreach via an interactive voice response (IVR) phone call
is attempted. If there is still no response following an
additional 24 hours, a second email is sent. If there is no
email on file, or the email is a CCHMC institutional email,
all three outreach attempts are made via an IVR phone
call. Families have up to 14 days to respond and complete
the survey before being identified as a survey non-
responder.
Survey Exclusion/Inclusion Methodology
Several exclusion criteria exist as part of the NRC Real-
Time platform based on demographic, diagnostic, and
prior utilization factors. To minimize survey fatigue, no
survey is attempted if a patient received a survey from any
other clinical encounter within three days of the
ambulatory encounter or from the same ambulatory
division within the previous 30 days. The survey is only
administered in English or Spanish based on a patient’s
self-reported primary language. Additional exclusions exist
for encounters connected to an international address,
those with incomplete or erroneous contact information,
and patients in custody of the state or county (<1% of
encounters). Finally, to prevent accidental disclosure and
to protect patient privacy, patients evaluated for a set of
“sensitive diagnoses” chosen a priori (e.g., pregnancy,
suspected or verified abuse, and visits related to sexually
transmitted infections) were excluded.
PFE Survey Elements
Each survey consisted of 14 questions that assess common
PFE priorities, including access to care, communication
with the care team, continuity and coordination of care,
information and education, respect, environment of care,
and overall rating of experience. Responses to the
questions were anchored to either a three-point or four-
point scale, or a 10-point Likert scale for the overall rating
of the provider question. Questions and response options
are in Figure 1. After these questions, respondents have
the option of leaving a free response comment.
Outcomes and Sociodemographic Variables
The outcomes of interest were PFE survey response,
completion, and comment rates. A survey response was
defined as answering at least one of the Likert-based
questions. Completion was defined as answering all the
Likert-based multiple response questions on the survey.
Commenting was defined as inclusion of any free-text
comment.
The patient level sociodemographic variables included
were race, ethnicity, primary language, and socioeconomic
deprivation. These sociodemographic characteristics were
extracted from the electronic medical record. Race was
categorized as White, Black or African American,
Multiracial, other, or unknown. Other consisted of Asian,
Hispanic, and other. Ethnicity was categorized as Hispanic
or non-Hispanic, while language was segmented into
English, Spanish, or Other/Unknown. Race, ethnicity, and
preferred language are self-reported by patients and
families during the clinical encounter and entered into the
electronic health record. Patients can self-identify as both
Hispanic race and Hispanic ethnicity.
Socioeconomic deprivation was estimated using the
Deprivation Index (DPI) measured at the census tract
level. This measure is used to approximate respondent
socioeconomic status and characterize the context in
which they live. The DPI is calculated from variables
present within the U.S. Census American Community
Survey (ACS) such as poverty, educational attainment, and
access to health insurance.
11
For this analysis, we grouped
respondents into quartiles by their DPI scores with higher
DPI indicating higher deprivation. The DPI was obtained
by extracting the patient street address from the electronic
health record at the time of the encounter which was then
geocoded and connected to the corresponding census tract
geography. Once the census tract was identified, it was
then linked to the DPI value.
Sociodemographic characteristics and survey response bias, Brinkman et al.
Patient Experience Journal, Volume 8, Issue 1 – 2021 20
Statistical Analysis
Descriptive statistics enumerated distributions of key
variables. We then assessed bivariate relationships between
participation rates and the range of sociodemographic
predictor variables. Finally, we assessed the relationship
between survey method participation and the range of
sociodemographic predictor variables. Associations
between PFE survey participation outcomes and
predictors were then assessed using the Chi-square
goodness of fit test.
Figure 1. PFE Survey Questions and Response Options
Question Text & Order
Question
Response
Options
Did this provider explain things about your child's health in a way that was easy to understand?
Yes definitely,
Yes somewhat,
No
Did this provider listen carefully to you?
Did you talk with this provider about any questions or concerns you had about your child's
health?
Yes, No
Did this provider give you easy to understand information about these health questions or
concerns?
Yes definitely,
Yes somewhat,
No
Did this provider seem to know the important information about your child's medical history?
Did this provider show respect for what you had to say?
Did this provider spend enough time with your child?
Rotate 1 of 3:
For this visit, were you able to get an appointment as soon as you needed?
Did clerks and receptionists treat you with courtesy and respect?
During your visit, did your child see this provider within 20 minutes of the appointment time? (wait time
includes time spent in the waiting room and exam room)
Would you recommend this provider's office to your family and friends?
Using any number from 0 to 10, where 0 is the worst provider possible and 10 is the best provider
possible, what number would you use to rate this provider?
Experience
0-10
Using any number from 0 to 10, where 0 is the worst facility possible and 10 is the best facility
possible, what number would you use to rate this facility?
Rating 0-10
Magnet Questions (Rotate 2 of 8)*:
Did nurses treat you with courtesy and respect?
Yes, definitely;
Yes, mostly; Yes,
somewhat; No
Did nurses listen carefully to you?
Did nurses explain things in a way you could understand?
Did you have confidence and trust in the nurses treating you
Did the staff do everything they could to help you with your pain?
Did you have enough input or say in your care?
Were you comfortable talking with nurses about your worries or concerns?
Was there good communication between the different doctors and nurses?
Is there anyone you would like to recognize or anything you would like to say about your
experience?
Open Question
* Magnet questions are for nursing Magnet Recognition. There are eight options, and two are randomly assigned per survey.
Sociodemographic characteristics and survey response bias, Brinkman et al.
21 Patient Experience Journal, Volume 8, Issue 1 – 2021
Results
Study Population
Survey outreach was attempted for 140,994 ambulatory
encounters during the study period, representing 82,294
unique patients. A total of 4,667 patient encounters were
excluded due to an inability to map reported street address
to a census tract. This resulted in 77,627 unique patient
encounters for the analysis (Figure 2). Patients involved in
the included encounters were predominantly White (76%)
and non-Hispanic (95%). (Table 1). Those excluded due to
unmappable street addresses were slightly more likely to
identify as White (78% vs. 76%) and slightly less likely to
identify as Black (9% vs. 12%). There was no statistical
difference between those who were included and those
who were excluded with respect to language and ethnicity.
Figure 2. PFE Survey Inclusion/Exclusion Classifications
Sociodemographic characteristics and survey response bias, Brinkman et al.
Patient Experience Journal, Volume 8, Issue 1 – 2021 22
Response, Completion, and Comment Rates
Overall, the survey response, completion, and comment
rates were 20.1%, 17.6%, and 4.1% respectively (Table 2).
Response rates differed across each sociodemographic
variable. White families responded 19.6% of the time
compared to 19.3% of Black and 18.8% of Multiracial
families. Hispanic (30.3%) and Spanish speaking families
(40.1%) had higher response rates than the overall cohort.
Those that were identified as living in more deprived areas
were more likely to respond compared to families living in
less deprived areas per the DPI. All differences were
statistically significant (p <.0001).
Completion rates differed across racial, ethnic, and
language groups, but not by DPI. Among racial groups,
differences in completion rates were even more disparate
than for response rates. White families (17.6%) were more
likely to complete the survey compared to Black (15.6%)
or Multiracial families (16.1%). Similar to the response
rates, Hispanic (24.9%) and Spanish- speaking families
(32.5%) had higher completion rates than the overall
cohort.
In addition to being more likely to respond to and
complete a survey, White families (4.3%) were also the
most likely racial group to leave a comment. While
Hispanic and Spanish-speaking families were more likely
to respond to and complete a survey, their comment rates
(3.1%, 1.8% respectively) were significantly lower
compared to non-Hispanic (4.2%) and English-speaking
families (3.9%). Comment rates decreased as social
deprivation within the DPI increased.
Families that had lower email response rates were less
likely to have a valid email address on file with the
hospital. (Table 3). Responses via email were over four
times more likely to include a comment compared to those
via IVR. White families were the most likely to respond to
the survey via email (35.5%) while Black families were less
likely (19.3%). Similarly, Non-Hispanic and English-
speaking families had much higher rates of utilizing email
as their survey method. Rates of participation via email
decreased as the social deprivation within the DPI
increased.
In addition to being more likely to respond via IVR,
Hispanic (5.1%) and Spanish-speaking families (2.9%)
were less likely to leave a comment on IVR surveys.
Similarly, more deprivation in the DPI coincided with
lower comment rates via IVR surveys. The higher rates of
responses via IVR in Hispanic families, Spanish-speaking
families, or families experiencing higher levels of
deprivation could be attributed to the of lack of valid email
(Table 3).
Table 1. Ambulatory PFE Survey Sociodemographic Groupings
Sociodemographic Characteristic
Encounters (%)
Race
White
59,374 (76%)
Black or African
American
9,525 (12%)
Multiracial
3,503 (5%)
Other*
3,788 (5%)
Unknown
1,437 (2%)
Ethnicity
Hispanic
3,901 (5%)
Non-Hispanic
73,082 (94%)
Unknown
644 (1%)
Language
English
47,578 (61%)
Spanish
1,054 (1%)
Not reported / Other
28,995 (37%)
DPI**
Encounters (DPI range)
Quartile 1
19,500 (0-0.28)
Quartile 2
19,480 (0.29-0.34)
Quartile 3
19,296 (0.35-0.43)
Quartile 4
19,351 (0.44-1)
Total
77,627
*“Other” Race represents “Hispanic” 46%, “Asian” 43%, and “Other” 11%
**DPI (Deprivation Index) is a composite measure of socioeconomic deprivation
based on census tract level data collected in the American Community Survey.
Quartile 1 represents the least deprived census tracts.
Sociodemographic characteristics and survey response bias, Brinkman et al.
23 Patient Experience Journal, Volume 8, Issue 1 – 2021
Discussion
Our study illustrates variation across sociodemographic
groups in how they participate in PFE surveys. This
significant variation was noted in the overall response
rates, completion rates, and likelihood to leave a comment.
Black and Multiracial respondents had lower than average
response and completion rates compared to White and
other race respondents. Black, Multiracial, Hispanic, and
Spanish-speaking respondents, along with those classified
as most deprived census tracts, were less likely to complete
the survey upon responding. Our study adds to the body
of literature given the paucity of data on how different
populations respond in different ways when asked to
complete PFE surveys after encounters within healthcare
systems.
For both hospital procedures and providers, post-visit
PFE survey responses can inform areas of improvement
12
and enhance insights on what patients and families
experience when they seek care. Given the weight placed
on PFE surveys, it is essential that we understand the
potential biases – based on responses and lack thereof –
the survey results may introduce. This was the rationale for
our study, to connect a range of patient and family
sociodemographic characteristics to participation
behaviors to more fully understand possible biases. These
biases may provide insights into disparities within the
healthcare system, identifying barriers to care, distrust in
the system, and ultimately factors at the root of differences
in health outcomes. More proximally, our findings will
inform survey methodology and limit the degree to which
interventions in response to PFE surveys do not focus
solely on those most likely to share their experience
through the survey.
The variation we found may imply that there are
differences in the ways people take the survey. Differences
in response, completion, and comment rates were partially
attributable to differences in survey methods across the
sociodemographic groups. Our survey is administered
after discharge via email or IVR. The different rates of lid
email addresses across demographic groups may influence
comment rates. There was an increase in response rate as
social deprivation increased within the DPI. Previous
studies have shown variation in response rates of PFE
surveys when completed prior to discharge via a tablet.
Surveying prior to discharge by in-clinic tablets was
effective in increasing response rates, particularly for those
that were non-White, publicly insured, and with lower
levels of education, therefore increasing representation
within the survey respondents.
13
Survey fatigue may also play a role in the variation in
completion and comment participation rates. In the PFE
survey, 14 multiple-choice questions preceded the open
response question. The survey length may influence survey
completion and desire to leave comments. In our analysis,
the decreasing completion rate with increasing social
deprivation (which includes estimate of literacy level)
supports this theory.
Table 2. Ambulatory PFE Survey Response Rates by Sociodemographic Grouping
Sociodemographic
Characteristic
Response Rate
P value
Completed Rate
P value
Comment Rate
P Value
Race
White
19.6%
<.0001*
17.6%
<.0001*
4.3%
<.0001*
Black or African
American
19.3%
15.6%
3.2%
Multiracial
18.8%
16.1%
4.0%
Other
30.1%
24.2%
3.8%
Unknown
24.2%
20.2%
5.0%
Ethnicity
Hispanic
30.3%
<.0001*
24.9%
<.0001*
3.1%
<.001*
Non-Hispanic
19.6%
17.2%
4.2%
Unknown
22.8%
20.5%
5.9%
Language
English
18.3%
<.0001*
16.2%
<.0001*
3.9%
<.0001*
Spanish
40.1%
32.5%
1.8%
Not Reported/
Other
22.4%
19.4%
4.5%
DPI**
Quartile 1
19.7%
<.0001*
17.7%
0.0665*
4.8%
<.0001*
Quartile 2
19.1%
17.0%
4.4%
Quartile 3
20.2%
17.6%
3.9%
Quartile 4
21.5%
18.2%
3.5%
Total
20.1%
17.6%
4.1%
* All p values calculated using a chi squared test.
Sociodemographic characteristics and survey response bias, Brinkman et al.
Patient Experience Journal, Volume 8, Issue 1 – 2021 24
The consequences of biased surveys may be significant
within institutions as patient experience ratings can drive
change, pointing to areas of potential improvement.
Comments may be weighted particularly highly and
deemed useful by management.
14
Here, the comment rates
demonstrated the most potential for bias within our survey
given the discrepancy between groups leaving comments.
This is especially significant as free response comments are
a major source of service recovery, which is when the
institution reaches out to families personally after a
negative experience in the healthcare system. With fewer
comments, organizations might be unable to address
service failures as they are less likely to know when and
how something went wrong in the experience.
Interventions biased towards those groups who leave
detailed comments may leave underrepresented groups
even further marginalized in the healthcare system.
Limitations
This study was limited by established exclusion criteria that
may influence the generalizability of the results. In an
attempt to reduce survey fatigue, exclusions included limits
of surveys to some patients and families with frequent
utilization of ambulatory services. This may lead to
underrepresentation of populations such as medically
complex children who see multiple specialties. However,
the random sampling at a patient level should minimize
the impact of these exclusions. Second, we only compared
the sociodemographic characteristics of those families who
responded to the survey. We did not compare the
characteristics of those who did not respond to survey.
Additionally, this is a single institution study, the results of
which may not be generalizable to other centers with a
different demographic patient mix. While our findings
have statistical significance, the clinical significance is still
unknown. Further evaluation of our findings to determine
clinical significance is necessary.
Table 3. Ambulatory PFE Survey Participation Methods by Sociodemographic Characteristics
Sociodemographic
Characteristic
Percent
with a
Valid
Email
P
Value
Percent
Respond-
ing via
Email
P
Value
Percent of
Email
Respon-
dents
Leaving a
Comment
P
Value
Percent
Respond
-ing via
IVR
P
Value
Percent
of IVR
Respon-
dents
Leaving a
Commen
t
P
Value
Race
White
67.7%
<.0001
*
35.5%
<.0001
*
43.3%
0.330
*
64.5%
<.000
1*
10.2%
<.01*
Black or
African
American
57.6%
19.3%
47.9%
80.7%
8.9%
Multiracial
63.8%
30.6%
47.8%
69.4%
9.9%
Other
52.1%
17.1%
43.1%
82.9%
6.2%
Unknown
49.1%
24.1%
54.8%
75.9%
9.8%
Ethnicity
Hispanic
45.8%
<.0001
*
12.0%
<.0001
*
46.5%
0.734
*
88.0%
<.000
1*
5.1%
<.0001
*
Non-
Hispanic
66.3%
33.4%
43.8%
66.5%
10.1%
Unknown
59.2%
32.7%
50.0%
67.3%
14.1%
Language
English
68.2%
<.0001
*
31.7%
<.0001
*
45.1%
0.292
*
68.3%
<.000
1*
10.6%
<.0001
*
Spanish
18.9%
2.8%
58.3%
97.2%
2.9%
Not
Reported/
Other
61.8%
33.9%
42.5%
66.1%
8.9%
DPI**
Quartile 1
72.3%
<.0001
*
40.0%
<.0001
*
43.4%
0.812
*
60.0%
<.000
1*
11.6%
<.0001
*
Quartile 2
67.2%
35.4%
45.5%
64.5%
11.1%
Quartile 3
64.0%
30.0%
43.3%
70.0%
8.6%
Quartile 4
57.1%
22.8%
43.7%
77.2%
8.0%
Total
65.2%
31.8%
44.0%
68.2%
9.6%
* All p values calculated using a chi squared test.
Sociodemographic characteristics and survey response bias, Brinkman et al.
25 Patient Experience Journal, Volume 8, Issue 1 – 2021
Conclusion
Sociodemographic differences drive disparities in a range
of health outcomes and, in parallel, influence patient
experience. Our understanding of the patient family
experience has been heavily influenced by PFE surveys.
This study identifies significant variation in PFE survey
participation, completion and comment rates based on
sociodemographic characteristics. Pediatric healthcare
systems should exercise caution in interpreting PFE survey
results and work to minimize this variation and potential
bias in future surveys.
References
1. Mazor KM, Clauser BF, Field T, et al. A
demonstration of the impact of response bias on the
results of patient satisfaction surveys. HSR: Health
Services Research. 2002; 37(5): 1403-1417.
2. Institute of Medicine 2003. Unequal Treatment:
Confronting Racial and Ethnic Disparities in Health
Care. Washington, DC: The National Academies
Press.
3. Braveman P, Gottlieb L. The social determinants of
health: it's time to consider the causes of the causes.
Public Health Reports. 2014. 129 (2): 19-31.
4. Sandager M, Freil M, Knudsen JL. Please tick the
appropriate box: Perspectives on patient reported
experience. Patient Experience Journal 2016; 3(1): p. 63-
79.
5. Arthur KC, Mangione-Smith R, Meischke H, et al.
Impact of English proficiency on care experiences in
a pediatric emergency department. . Acad Pediatr 2015;
15 (2): 218-224.
6. Sun R, Wu L, Barnett S, et al., Socio-demographic
predictors associated with capacity to engage in health
care. Patient Experience Journal 2019; 6(2): 35-41.
7. Tyser AR, Abtahi AM, McFadden M, et al. Evidence
of non-response bias in the Press-Ganey patient
satisfaction survey. BMC Health Serv Res 2016; 16(a):
350.
8. Nieman CL, Benke JR, Ishman SL, et al. Whose
experience is measured? A pilot study of patient
satisfaction demographics in pediatric otolaryngology.
Laryngoscope 2014. 124(1): 290-4.
9. Batbaatar E, Dorjdagva J, Luvsannyam A, et al.
Determinants of patient satisfaction: a systematic
review. Perspect Public Health 2017. 137(2): 89-101.
10. NRC Health Real-time feedback website. Accessed
July 9, 2020. https://nrchealth.com/platform/real-
time-feedback/.
11. Brokamp C, Beck AF, Goyal NK, et al. Material
community deprivation and hospital utilization during
the first year of life: an urban population-based cohort
study. Ann Epidemiol 2019; 30: 37-43.
12. Aharony L, Strasser, S. Patient satisfaction: What we
know about and what we still need to explore. Medical
Care Review 1993. 50(1): 49-79.
13. Toomey SL, Elliott MN, Zaslavsky AM, et al.
Improving response rates and representation of hard-
to-reach groups in family experience surveys. Acad
Pediatr 2019. 19(4): 446-453.
14. Riiskjaer E, Ammentorp J, Kofoed PE. The value of
open-ended questions in surveys on patient
experience - number of comments and perceived
usefulness from a hospital perspective. International
Journal for Quality in Health Care 2012. 24(5):509–516.
