Nursing Research: Reading, Using, and Creating Evidence
1. NURSING RESEARCH
READING, USING, AND CREATING EVIDENCE
FOURTH EDITION
JANET HOUSER, PHD, RN
Provost and Professor
Rueckert-Hartman College for Health Professions
Regis University
Denver, Colorado
JONES & BARTLETT
LEARNING
2
World Headquarters
Jones & Bartlett Learning
5 Wall Street
Burlington, MA 01803
978-443-5000
[email protected]
www.jblearning.com
Jones & Bartlett Learning books and products are available
through most
bookstores and online booksellers. To contact Jones & Bartlett
Learning
3. and such reference shall not be used for advertising or product
endorsement
purposes. All trademarks displayed are the trademarks of the
parties noted
herein. Nursing Research: Reading, Using, and Creating
Evidence, Fourth
Edition is an independent publication and has not been
authorized, sponsored,
or otherwise approved by the owners of the trademarks or
service marks
referenced in this product.
There may be images in this book that feature models; these
models do not
necessarily endorse, represent, or participate in the activities
represented in
3
the images. Any screenshots in this product are for educational
and instructive
purposes only. Any individuals and scenarios featured in the
case studies
throughout this product may be real or fictitious, but are used
for instructional
purposes only.
The author, editor, and publisher have made every effort to
provide accurate
information. However, they are not responsible for errors,
omissions, or for any
outcomes related to the use of the contents of this book and take
no
responsibility for the use of the products and procedures
4. described.
Treatments and side effects described in this book may not be
applicable to all
people; likewise, some people may require a dose or experience
a side effect
that is not described herein. Drugs and medical devices are
discussed that
may have limited availability controlled by the Food and Drug
Administration
(FDA) for use only in a research study or clinical trial.
Research, clinical
practice, and government regulations often change the accepted
standard in
this field. When consideration is being given to use of any drug
in the clinical
setting, the health care provider or reader is responsible for
determining FDA
status of the drug, reading the package insert, and reviewing
prescribing
information for the most up-to-date recommendations on dose,
precautions,
and contraindications, and determining the appropriate usage
for the product.
This is especially important in the case of drugs that are new or
seldom used.
15833-5
Production Credits
VP, Executive Publisher: David D. Cella
Executive Editor: Amanda Martin
Editorial Assistant: Emma Huggard
Senior Production Editor: Amanda Clerkin
Senior Marketing Manager: Jennifer Scherzay
Product Fulfillment Manager: Wendy Kilborn
Composition: S4Carlisle Publishing Services
6. Contents
Preface
Acknowledgments
Contributors
About the Author
Part I: An Introduction to Research
1 The Importance of Research as Evidence in Nursing
Research as Evidence for Nursing Practice
What Is Nursing Research?
Research: A Fundamental Nursing Skill
The Evolution of Research in Nursing
Contemporary Nursing Research Roles
Research Versus Problem Solving
Research as Evidence in Nursing Practice
Evidence-Based Practice
The Importance of Evidence-Based Practice in Nursing
How Can Evidence Be Used in Health Care?
Strategies for Implementing Evidence-Based Practice
Strategies for Overcoming Barriers
Reading Research for Evidence-Based Practice
7. Using Research in Evidence-Based Practice
Creating Evidence for Practice
Future Directions for Nursing Research
Summary of Key Concepts
For More Depth and Detail
References
2 The Research Process and Ways of Knowing
Introduction
The Research Process
Classification of Research by Philosophical Assumptions About
the Nature of the World
Choosing a Design
Classifications of Research by the Intent of the Researcher
Classifications of Research by the Nature of the Design
Classifications of Research by the Time Dimension
Reading Research for Evidence-Based Practice
Using Research in Evidence-Based Practice
Creating Evidence for Practice
Summary of Key Concepts
8. For More Depth and Detail
References
6
3 Ethical and Legal Considerations in Research
Introduction
Learning from the Past, Protecting the Future
International Guides for the Researcher
National Guidelines for the Nurse Researcher
The Ethical Researcher
Legal and Regulatory Guidelines for Conducting Research
Institutional Review Boards
Research Involving Animals
Research Misconduct
The HIPAA Privacy Rule
Reading Research for Evidence-Based Practice
Using Research in Nursing Practice
Creating Evidence for Practice
Summary of Key Concepts
9. For More Depth and Detail
References
Part II: Planning for Research
4 Finding Problems and Writing Questions
Introduction
Finding and Developing Research Problems
Developing the Research Question
Reading Research for Evidence-Based Practice
Using Research in Evidence-Based Practice
Creating Evidence for Practice
Summary of Key Concepts
For More Depth and Detail
References
5 The Successful Literature Review
An Introduction to the Literature Review
Purpose, Importance, and Scope of the Literature Review
Types of Literature Used in the Review
Searching for the Evidence in a Research Problem
Competencies for Information Literacy
10. Reading the Literature Review Section
Using Evidence-Based Literature in Nursing Practice
Creating a Strong Literature Review
Summary of Key Concepts
For More Depth and Detail
References
6 Selecting an Appropriate Research Design
Introduction
What Is a Design?
7
The Basis for Design Selection
The Design Decisions
Reading Research for Evidence-Based Practice
Using Research in Evidence-Based Practice
Creating Evidence for Practice
Summary of Key Concepts
For More Depth and Detail
11. References
Part III: Research Process
7 The Sampling Strategy
Introduction
Selection Strategy: How Were the Subjects Chosen?
The Sample Selection Strategy
Reading the Sampling Section of a Research Study
Using Research as Evidence for Practice
Creating an Adequate Sampling Strategy
Summary of Key Concepts
For More Depth and Detail
References
8 Measurement and Data Collection
Introduction
Measurement
The Measurement Strategy
Strategies to Minimize Measurement Error
Collecting Data Using Instruments
Data Management Procedures
12. Reading About Measurement and Data Collection
Using Measurements from a Research Study
Creating Measures and Collecting Data
Summary of Key Concepts
References
9 Enhancing the Validity of Research
Introduction
Minimizing Threats to Internal Validity
Factors That Jeopardize Internal Validity
Factors That Jeopardize External Validity
Balancing Internal and External Validity
Trustworthiness in Qualitative Research
Strategies to Promote the Validity of Qualitative Research
Reading a Research Study to Determine Validity
Using Valid Studies as Evidence for Nursing Practice
Creating a Valid Research Study
Summary of Key Concepts
8
13. For More Depth and Detail
References
Part IV: Research That Describes Populations
10 Descriptive Research Questions and Procedures
Introduction
Descriptive Research Studies
Characteristics of a Descriptive Design
Describing Groups Using Surveys
Describing Groups Relative to Time
Describing the Responses of Single Subjects
Designs That Describe Relationships
Reading Descriptive Research
Using Descriptive Research in Evidence-Based Nursing Practice
Creating Descriptive Research
Summary of Key Concepts
References
11 Summarizing and Reporting Descriptive Data
Introduction
An Overview of Descriptive Data Analysis
14. Understanding Levels of Measurement
Identifying Shape and Distribution
Describing the Center and Spread
Common Errors in Summarizing Data
Reading the Descriptive Data Section in a Research Study
Using Descriptive Data Analysis in Practice
Creating Descriptive Data Summaries for a Research Study
Reporting Descriptive Results
Summary of Key Concepts
References
Part V: Studies that Measure Effectiveness
12 Quantitative Questions and Procedures
Introduction
Quantitative Research Questions
Characteristics of a Quantitative Design
The Gold Standard: Experimental Design
More Common: Quasi-Experimental Designs
Designs That Focus on Intact Groups
15. Time-Series Designs
Reading Quantitative Research
Using Quantitative Research in Evidence-Based Nursing
Practice
Generalizing the Results of Quantitative Studies
Creating Quantitative Research
9
Summary of Key Concepts
References
13 Analysis and Reporting of Quantitative Data
Introduction
Some General Rules of Quantitative Analysis
Types of Quantitative Analysis
An Overview of Quantitative Analysis
Selecting the Appropriate Quantitative Test
Reading the Analysis Section of the Research Report
Using Quantitative Results as Evidence for Practice
Creating a Quantitative Analysis
16. Summary of Key Concepts
References
Part VI: Research That Describes the Meaning of an Experience
14 Qualitative Research Questions and Procedures
An Introduction to Qualitative Research
Characteristics of Qualitative Research Methods
Enhancing the Trustworthiness of Qualitative Studies
Classifications of Qualitative Traditions
Reading Qualitative Research Studies
Using Qualitative Research Studies as Evidence
Creating Qualitative Evidence
Summary of Key Concepts
References
15 Analyzing and Reporting Qualitative Results
Introduction to Qualitative Analysis
The Qualitative Analysis Process
Management and Organization of Data
Software for Qualitative Analysis
Reliability and Validity: The Qualitative Version
17. Reporting Qualitative Results
Reading the Qualitative Analysis Section of a Report
Using Qualitative Analysis in Nursing Practice
Creating Qualitative Analyses
Summary of Key Concepts
References
Part VII: Research Translation
16 Translating Research into Practice
Introduction
The Nurse’s Role in Knowledge Translation
Identifying Problems for Knowledge Translation
Communicating Research Findings
10
Finding and Aggregating Evidence
Models for Translating Research into Practice
Summary of Key Concepts
References
Glossary
18. Index
11
The Pedagogy
Nursing Research: Reading, Using, and Creating Evidence,
Fourth Edition
demonstrates how to use research as evidence for successful
nursing
practice. Fully updated and revised, this readerfriendly new
edition provides
students with a fundamental understanding of how to appraise
and utilize
research, translating it into actionable guidelines for practice.
Organized
around the different types of research that can be used in
evidence-based
practice, it addresses contemporary methods including the use
of technology
in data collection, advice for culturally competent research, and
suggestions
for accessing hard-to-reach subjects. Additionally, it explores
both quantitative
and qualitative traditions and encourages students to read, use,
and
participate in the research process. The pedagogical aids that
appear in most
chapters include the following:
12
19. 13
14
15
16
17
18
Preface
This nursing research text is based on the idea that research is
essential for
nurses as evidence for practice. Its contents are intended to be
relevant for
nursing students, and practicing nurses who must apply
evidence to practice.
All nurses should be able to read research, determine how to use
it
appropriately in their practice, and participate in the research
process in some
way during their careers as professionals. This text is intended
20. to support all
these efforts.
Evidence-based practice is one of the most exciting trends in
nursing practice
to emerge in decades. However, its integration into daily
practice requires a
solid understanding of the foundations of research design,
validity, and
application. This text is intended as a reader-friendly approach
to a complex
topic so that beginners can grasp the fundamentals of appraising
research,
experienced nurses can use research in practice, and practicing
nurses can
gain skills to create bedside research projects or participate
effectively on
research teams.
This text is presented in an uncluttered, straightforward manner.
Although it
uses many bulleted lists to make the material visually
interesting, the sidebars,
figures, and tables are limited to those that illustrate truly
important concepts.
This format allows the reader to grasp the information quickly
and to navigate
the text efficiently. Margin notes provide definitions of new
terms when they
first appear, and the Gray Matter features offer information
about key concepts
that are of particular importance.
This text differs in its approach from traditional texts in that it
does not focus
primarily on interpreting inferential research; rather, it seeks to
21. impart a
fundamental understanding of all types of research that may be
used as
evidence. It adds depth by considering the use of qualitative
research in
nursing practice—a natural fit with this holistic profession. This
text also
addresses contemporary concerns for today’s nurses, including
ethical and
legal issues. Although both ethics and legal issues are
mentioned in many
research texts, a full chapter is devoted to these topics in this
text so that the
intricacies of these issues can be thoroughly considered.
The integrated discussion of both the quantitative and the
qualitative traditions
is another unique facet of this text’s coverage of the research
process. Most
nurse researchers have learned to appreciate the need to
consider all
paradigms when approaching a research question; separating the
two
approaches when discussing the fundamental interests of
researchers results
in a polarized view. Intuitively, nurses know that the lines
between quantitative
19
and qualitative designs are not always so clear in practice and
that they should
consider multiple ways of knowing when evaluating research
questions. The
22. planning process covered here helps the novice researcher
consider the
requirements of both approaches in the context of sampling,
measurement,
validity, and other crucial issues they share. Detailed
descriptions of the
procedures for each type of design are given attention in
separate chapters.
The chapters are organized around the types of research
processes that make
up the evidence base for practice. The first section of the text
provides
information that is applicable to all research traditions, whether
descriptive,
quantitative, or qualitative. Part I provides an overview of
issues relevant to all
researchers: understanding the way research and practice are
related, the
ways that knowledge is generated, and legal and ethical
considerations. Part II
describes the processes that go into planning research. The
chapters in Part
III consider the various decisions that must be made in each
phase of the
research process.
The evidence generated by descriptive, survey, and qualitative
designs is
placed in the context of both the definition of evidence-based
practice and
application in practice guidelines. In Parts IV, V, and VI, each
major
classification of research is explored in depth through review of
available
designs, guidelines for methods and procedures, and discussion
23. of
appropriate analytic processes. Brief examples of each type of
research are
provided, along with notes explaining the features demonstrated
in each case
in point. Finally, Part VII details the models and processes used
to translate
research into clinical practice.
Many chapters begin with a feature called “Voices from the
Field” that relates
a real-life story of a nurse’s experience with the research
process, illustrating
the way that the material covered in that chapter might come to
life. The main
content for each chapter is broken into five parts:
A thorough review of the topic under consideration is presented
first. This
review lays out the fundamental knowledge related to the topic.
Next, the nurse isa guided to consider the aspects of a study that
should
be appraised when reading research. All nurses—regardless of
their
experience—should be able to read research critically and apply
it
appropriately to practice, and the second section of each chapter
addresses this skill. Added features include advice on where to
look for the
key elements of a research paper, the wording that might be
used to
describe them, and specific things to look for during the
evaluation
process. Evaluation checklists support this process.
The third section of the chapter focuses on using research in
practice. This
24. section supports the nurse in determining if and how research
findings
might be used in his or her practice.
20
The fourth section is intended for nurses who may be involved
with teams
that are charged with creating research or who may plan bedside
research
projects to improve practice. This section gives practical advice
and
direction about the design and conduct of a realistic, focused
nursing
research project.
The final section of each chapter contains summary points and a
critical
appraisal exercise so that the nurse can immediately apply the
chapter
concepts to a real research report.
All of these features are intended to help the reader gain a
comprehensive
view of the research process as it is used to provide evidence
for professional
nursing practice. The use of this text as a supportive resource
for learning and
for ongoing reference in clinical practice has been integrated
into the design of
each element of the text. The goal is to stimulate nurses to read,
use, and
participate in the process of improving nursing practice through
the systematic
use of evidence. Accomplishing this goal improves the
25. profession for all of us.
21
Acknowledgments
It is a bit misleading to conclude that a text is produced solely
by the person
whose name appears on the cover. Help and support are needed
from many
people on both professional and personal fronts to complete a
project of this
size. The help of editorial staff is always welcome; advice from
Amanda Martin
was invaluable in merging the interests of writing with those of
producing a
book that others will want to read. I appreciate Amanda
Clerkin’s calm and
steady approach after our sixth manuscript together, and I’ve
learned a lot
from reading Jill Hobbs’s edits, which I must begrudgingly
admit make my
writing much better.
My family—my husband, Floyd; my sisters, Anne and Ande; my
niece, Stef;
and mini-me, Amanda—provided me with enough
encouragement to keep
going, even as they reminded me there is life beyond the pages
of a book.
I must thank Regis University profusely for providing me with
inspirational
colleagues and a place that supports my work. Pat Ladewig, as
always,
26. provided pragmatic advice and guidance from her impressive
experience
publishing her own texts. My contributors and reviewers each
provided a
unique viewpoint and helped me discover the best way to ensure
that students
“get it.”
Writing always makes me realize how much I miss my mom,
Marty, who
encouraged me to publish from the time she surreptitiously sent
one of my
poems to Highlights magazine when I was 9 years old. She was
proud of that
poem, framed the issue, and had my grandmother embroider it
on a pillow.
Seeing this book in print would have impressed her only
slightly more, but I
know she’s smiling.
22
Contributors
Michael Cahill, MS, CPHQ
Parker Adventist Hospital
Parker, Colorado
Summarizing and Reporting Descriptive Data
Sheila Carlon, PhD, RHIA, FAHIMA
Regis University
Denver, Colorado
Ethical and Legal Considerations in Research
Phyllis Graham-Dickerson, PhD, RN, CNS
27. Regis University
Denver, Colorado
Qualitative Research Questions and Procedures Analyzing and
Reporting Qualitative Results
LeeAnn Hanna, PhD, RN, CPHQ, FNAHQ
HCA, TriStar Centennial Medical Center
Nashville, Tennessee
Finding Problems and Writing Questions
Kimberly O’Neill, MS, MLIS
Dayton Memorial Library, Regis University
Denver, Colorado
The Successful Literature Search
23
About the Author
Janet Houser, PhD, RN
Regis University
Dr. Janet Houser is currently Provost at Regis University in
Denver, Colorado.
Prior to her appointment, she was Dean of the Rueckert-
Hartman College for
Health Professions and the Vice Provost for Resource Planning.
Dr. Houser has a BSN, an MN in Maternal-Child Health, an MS
in healthcare
administration, and a PhD in applied statistics and research
methods. She has
taught nurses, administrators, pharmacists, and physical therapy
30. KEY TERMS
Blinded
Evidence-based practice
Evidence-based practice guideline
External validity
Journal club
Magnet status
National Institute of Nursing Research (NINR)
Nursing process
Nursing research
Outcomes measurement
Peer review
Principal investigator
Quality improvement
Randomized controlled trial
Replication
Systematic review
28
31. Research as Evidence for Nursing Practice
The practice of nursing is deeply rooted in nursing knowledge,
and nursing
knowledge is generated and disseminated through reading,
using, and
creating nursing research. Professional nurses rely on research
findings to
inform their practice decisions; they use critical thinking to
apply research
directly to specific patient care situations. The research process
allows nurses
to ask and answer questions systematically that will ensure that
their decisions
are based on sound science and rigorous inquiry. Nursing
research helps
nurses in a variety of settings answer questions about patient
care, education,
and administration. It ensures that practices are based on
evidence, rather
than eloquence or tradition.
VOICES FROM THE FIELD
I was working as the clinical nurse specialist in a busy surgical
intensive
care unit (ICU) when we received a critically ill patient. He was
fresh
from cardiac surgery and quite unstable; he needed multiple
drugs and
an intra-aortic balloon pump just to maintain his perfusion
status. The
patient was so sick that we were not able to place him on a
special bed
for pressure relief. For the first 24 hours, we were so busy
trying to
keep him alive that we did not even get a chance to turn him.
32. Approximately 36 hours into his ICU admission, he was stable
enough
to place on a low-air-loss mattress for pressure-ulcer
prevention. When
we were finally able to turn him, we noted he had a small stage
II
pressure ulcer on his coccyx. Despite the treatments that we
used, the
pressure ulcer evolved into a full-thickness wound. The patient
recovered from his cardiac surgical procedure but,
unfortunately,
required surgeries and skin grafts to close the pressure ulcer
wound.
The experience I had with this patient prompted me to review
the
evidence-based practice (EBP) guidelines we had in place to
prevent
pressure ulcers in critically ill patients. I wanted to make sure
we could
prevent this kind of incident from happening again, but I had a
lot of
questions. Could we preventively place high-risk patients on
low-air-
loss mattresses while they were still in the perioperative
service? Did
we even know which patients were at risk for pressure ulcers?
Which
assessment tools did nurses use to assess the patient’s risk?
When a
high-risk patient was identified, which interventions did the
nurses use
to prevent pressure ulcers? How were the ulcers treated once
they
appeared?
33. I was fortunate that my chief nursing officer (CNO) was a
strong
advocate for EBP, and she encouraged me to initiate an EBP
review of
pressure ulcer prevention and treatment. Specifically, I wanted
to find
out which nursing interventions were supported by research
evidence
when we were trying to prevent pressure ulcers in the surgical
ICU. As
29
part of my review, I contacted other inpatient units at the
hospital to
determine what they were doing.
I discovered that the surgical ICU was no different from the
other
inpatient units in this regard: There was no standard, evidence-
based
nursing practice for pressure ulcer prevention. Units were not
consistently using the same skin assessment tools, so it was
difficult to
objectively communicate risk from one unit to another. The
tools we
were using were not necessarily based on research. It was clear
that
we needed to identify the best available evidence and devise a
protocol.
We started by establishing an evidence-based skin care council
for the
34. hospital. This team consisted of bedside nurses from all
inpatient units
and the perioperative service. Initially the council reviewed the
hospital’s current nursing skin assessment forms, and we
conducted a
review of the literature on pressure ulcer prevention and
interventions.
We discovered the Agency for Healthcare Research and Quality
(AHRQ) guidelines on pressure ulcer prevention and
treatment—a key
source of evidence for healthcare practices.
Over the course of the next year, we revised our nursing policy
and
procedure, incorporating the AHRQ evidence into a treatment
guideline.
This guideline included a procedure for skin assessment and
nursing
documentation, and pressure ulcer assessment and treatment
decision
algorithms. We reviewed skin-care products and narrowed down
the list
of products to those that were supported by evidence. One
algorithm
helped staff make selections between products that maximized
prevention and treatment. Another algorithm guided nurses in
the use
of therapeutic surfaces (e.g., low-air-loss mattresses) to prevent
pressure ulcers. To monitor our progress, we began quarterly
pressure
ulcer prevalence studies. As part of the implementation, we
scheduled
a skin-care seminar featuring a national expert on skin care.
At the beginning of our EBP skin-care journey, our facility’s
pressure
35. ulcer prevalence was 9%. Since implementing our EBP skin-
care
initiatives, it has dropped by two …
Association of the Hospital Readmissions Reduction Program
With Mortality Among Medicare Beneficiaries Hospitalized
for Heart Failure, Acute Myocardial Infarction, and Pneumonia
Rishi K. Wadhera, MD, MPP, MPhil; Karen E. Joynt Maddox,
MD, MPH; Jason H. Wasfy, MD, MPhil; Sebastien Haneuse,
PhD;
Changyu Shen, PhD; Robert W. Yeh, MD, MSc
IMPORTANCE The Hospital Readmissions Reduction Program
(HRRP) has been associated with
a reduction in readmission rates for heart failure (HF), acute
myocardial infarction (AMI), and
pneumonia. It is unclear whether the HRRP has been associated
with change in patient mortality.
OBJECTIVE To determine whether the HRRP was associated
with a change in patient mortality.
DESIGN, SETTING, AND PARTICIPANTS Retrospective
cohort study of hospitalizations for HF,
AMI, and pneumonia among Medicare fee-for-service
beneficiaries aged at least 65 years
across 4 periods from April 1, 2005, to March 31, 2015. Period
1 and period 2 occurred before
the HRRP to establish baseline trends (April 2005-September
2007 and October
2007-March 2010). Period 3 and period 4 were after HRRP
announcement (April 2010 to
September 2012) and HRRP implementation (October 2012 to
March 2015).
36. EXPOSURES Announcement and implementation of the HRRP.
MAIN OUTCOMES AND MEASURES Inverse probability–
weighted mortality within 30 days of
discharge following hospitalization for HF, AMI, and
pneumonia, and stratified by whether
there was an associated readmission. An additional end point
was mortality within 45 days of
initial hospital admission for target conditions.
RESULTS The study cohort included 8.3 million
hospitalizations for HF, AMI, and pneumonia,
among which 7.9 million (mean age, 79.6 [8.7] years; 53.4%
women) were alive at discharge.
There were 3.2 million hospitalizations for HF, 1.8 million for
AMI, and 3.0 million for pneumonia.
There were 270 517 deaths within 30 days of discharge for HF,
128 088 for AMI, and 246 154 for
pneumonia. Among patients with HF, 30-day postdischarge
mortality increased before the
announcement of the HRRP (0.27% increase from period 1 to
period 2). Compared with this
baseline trend, HRRP announcement (0.49% increase from
period 2 to period 3; difference in
change, 0.22%, P = .01) and implementation (0.52% increase
from period 3 to period 4;
difference in change, 0.25%, P = .001) were significantly
associated with an increase in
postdischarge mortality. Among patients with AMI, HRRP
announcement was associated with a
decline in postdischarge mortality (0.18% pre-HRRP increase vs
0.08% post-HRRP
announcement decrease; difference in change, −0.26%; P = .01)
and did not significantly change
after HRRP implementation. Among patients with pneumonia,
37. postdischarge mortality was
stable before HRRP (0.04% increase from period 1 to period 2),
but significantly increased after
HRRP announcement (0.26% post-HRRP announcement
increase; difference in change, 0.22%,
P = .01) and implementation (0.44% post-HPPR implementation
increase; difference in change,
0.40%, P < .001). The overall increase in mortality among
patients with HF and pneumonia was
mainly related to outcomes among patients who were not
readmitted but died within 30 days of
discharge. For all 3 conditions, HRRP implementation was not
significantly associated with an
increase in mortality within 45 days of admission, relative to
pre-HRRP trends.
CONCLUSIONS AND RELEVANCE Among Medicare
beneficiaries, the HRRP was significantly
associated with an increase in 30-day postdischarge mortality
after hospitalization for HF and
pneumonia, but not for AMI. Given the study design and the
lack of significant association of
the HRRP with mortality within 45 days of admission, further
research is needed to
understand whether the increase in 30-day postdischarge
mortality is a result of the policy.
JAMA. 2018;320(24):2542-2552. doi:10.1001/jama.2018.19232
Editorial page 2539
Supplemental content
Author Affiliations: Author
affiliations are listed at the end of this
article.
39. tients with heart failure, acute myocardial infarction, and pneu-
monia, beginning in 2012.1 After the announcement of the
HRRP, readmission rates among Medicare beneficiaries de-
clined for target conditions nationwide.2,3 Recently, how-
ever, policy makers and physicians have raised concern that
the HRRP may have also had unintended consequences
that adversely affected patient care, potentially leading to in-
creased mortality.4,5 For instance, the financial penalties im-
posed by the HRRP may have inadvertently pushed some phy-
sicians to avoid indicated readmissions, potentially diverted
hospital resources and efforts away from other quality im-
provement initiatives, or worsened quality of care at resource-
poor hospitals that are often penalized by the program. How-
ever, it is also possible that the same mechanisms by which
some hospitals have reduced readmissions, such as im-
proved coordination and transitions of care, resulted in reduc-
tions in mortality.
Understanding whether the HRRP has been associated
with changes in mortality at the patient level is important as
policy makers evaluate this program, particularly given the
ongoing expansion of the HRRP to include other conditions6
and the almost $2 billion in financial penalties that have been
imposed on hospitals since 2012.7 This study aims to answer
3 questions. First, compared with past trends, was the
announcement or implementation of the HRRP associated
with a change in mortality within 30 days of discharge fol-
lowing hospitalization for heart failure, acute myocardial
infarction, or pneumonia? Second, was the HRRP associated
with a change in the distribution of patients who experienced
death and no readmission, readmission and no death, read-
mission and death, or no death and no readmission during
the 30 days after discharge? Third, was the HRRP associated
with a change in mortality within 45 days of hospital admis-
sion for target conditions?
40. Methods
Institutional review board approval, including waiver of the
requirement of participant informed consent because the data
were deidentified, was provided by the Beth Israel Deacon-
ess Medical Center.
Study Cohort
We used Medicare Provider Analysis and Review files to iden-
tify hospital admissions and discharges at short-term acute care
hospitals from April 1, 2005, through March 31, 2015, with
a principal discharge diagnosis of heart failure, acute myo-
cardial infarction, or pneumonia. Study cohorts were de-
fined using International Classification of Diseases, Ninth
Revision, Clinical Modification codes used in the publicly re-
ported CMS readmission and mortality measures.8-10 We in-
cluded Medicare beneficiaries aged 65 years or older in the
analysis. We excluded patients who were discharged against
medic al advice, were not enrolled in Medic are fee-for-
service for at least 30 days after discharge (absent death),
or were enrolled in Medicare fee-for-service for less than 1
year before hospitalization. Transfers to other hospitals were
linked to a single index hospitalization. To examine 30-day
postdischarge outcomes, we also excluded patients who
died during hospitalization. Comorbidities were defined using
CMS hierarchical condition categories based on Medicare
claims up to 1 year before hospitalization.11 Specifically, we
used
covariates in the CMS risk-adjustment models for heart fail-
ure, acute myocardial infarction, and pneumonia,12-14 as has
been done in previous studies.2,15 The race/ethnicity of all pa-
tients was identified based on claims files and was desig-
nated into the following fixed categories: white, black, or other.
Race/ethnicity was included as a covariate in the analysis be-
cause it is associated with mortality for target conditions.16
41. Study Periods
We identified 4 nonoverlapping study periods of equal dura-
tion for index hospitalization. We chose to evaluate differ-
ences in outcomes between time periods, rather than annual
trends, for 2 reasons. First, we were interested in changes in
outcomes among time periods defined by their relationship to
the announcement and implementation of the HRRP, rather
than within-period trends. Second, this strategy avoids as-
sumptions on how the HRRP imposes its effect on different pa-
tient groups (eg, assumptions on main effects and interaction
terms) and of a linear relationship between outcomes and time
and continuous confounders in a conventional logistic or mul-
tinominal regression model.
We identified 2 study periods before the HRRP was estab-
lished to examine baseline trends in outcomes. The first study
period included hospitalizations from April 2005 to Septem-
ber 2007 (period 1) and the second included hospitalizations
from October 2007 to March 2010 (period 2). Two periods af-
ter the HRRP was established were also included: 1 following
the initial announcement of HRRP with passage of the ACA
from April 2010 through September 2012 (period 3) and the
other between October 2012 and March 2015 (period 4), which
Key Points
Question Was the announcement and implementation of the
Hospital Readmissions Reduction Program (HRRP) associated
with
an increase in patient-level mortality?
Findings In this retrospective cohort study that included
approximately 8 million Medicare beneficiary fee-for-service
hospitalizations from 2005 to 2015, implementation of the
HRRP
was associated with a significant increase in trends in 30-day
43. death, (3) readmission and no death, and (4) no readmission
and no death. These subgroup outcomes were examined to try
to provide mechanistic insights on the relationship between
readmission and mortality. To fully assess trends in mortality
related to a complete clinical episode, 45-day patient mortal-
ity rates following admission (postadmission mortality) were
also evaluated, because efforts to reduce readmissions could
potentially encompass care during hospitalization and might
influence discharge timing and location of death. This mea-
sure included varying hospital lengths of stay and captured
both in-hospital and 30-day postdischarge deaths for the ma-
jority of the cohort.
Statistical Analysis
To account for a potential imbalance in case mix between study
periods, a propensity score approach (ie, the probability of
being in a specific period given the demographics and comor-
bidities of the patient and calendar month of hospitalization)
was used to standardize populations among periods. Patient
demographics, comorbidities, and seasonal indicators (calen-
dar month) from period 4 were used as a reference to re-
weight observed outcomes in all other study periods. Logis-
tic regression models were fit on data from periods 1 and 4 to
obtain a propensity score for period 1. The propensity score was
then used to weight the outcomes in period 1, generating event
rates through inverse probability weighting (IPW) that would
have been observed if period 1 had the same case mix as pe-
riod 4. Similarly, separate logistic regression models were fit
to data from periods 2 and 4 and periods 3 and 4 to provide
IPW-adjusted event rates in periods 2 and 3, respectively. This
approach allowed the calculated distribution of each out-
come in each of the 4 periods to be based on the same case mix
(ie, the case mix from period 4).20 Because the primary aim
was to understand the association of the HRRP with mortal-
ity at the individual level, we did not examine hospital-level
44. effects in the analysis.
To establish the change in rates of outcomes after the an-
nouncement of the HRRP, the change in event rates between
periods 2 and 3 was calculated. Similarly, the change in rates
of outcomes between periods 3 and 4 was also calculated to
examine the change in outcomes between the announce-
ment and the implementation of the HRRP (Figure 1).
To isolate the association between the HRRP and the out-
comes, we sought to remove secular trends for each out-
come. To do so, the change in outcomes between periods 1 and
2 was computed to establish a baseline trend in outcomes be-
fore the announcement and implementation of the HRRP. This
difference was then subtracted from the change in outcomes
after the announcement of the HRRP (between periods 2 and
3) to account for trends that were unrelated to the HRRP. Simi-
larly, the baseline difference was also subtracted from the
change in outcomes after the implementation of the HRRP, be-
tween periods 3 and 4.
Additional Analyses
Several sensitivity analyses were performed. First, patients
enrolled in hospice were excluded because greater use of
hospice care at the end of life might shift deaths that previ-
ously occurred within a hospital to the postdischarge setting
over time.21,22 Second, because 1 hospitalization was ran-
domly selected for patients that experienced multiple hospi-
talizations in a given study period, the main analysis was
repeated using the first hospitalization for each patient in
each study period as well as all hospitalizations for each
Figure 1. Study Periods and Analytic Approach in a Study of the
Association Between the Hospital
Readmissions Reduction Program (HRRP) and Mortality
45. Period 1
(April 2005-
September 2007)
Period 2
(October 2007-
March 2010)
Period 3
(April 2010-
September 2012)
Period 4
(October 2012-
April 2015)
Baseline change
in mortality before
HRRP announcement
Difference in change in mortality prior
to HRRP (A) compared with change
after HRRP announcement (B)
Difference in change in mortality before HRRP (A)
compared with change after HRRP implementation (C)
Change in mortality
after HRRP
announcement
Change in mortality
47. approach used.
More details on the methodologic approach are provided
in the Supplement. Significance testing was performed using
z tests, with standard error estimates that accounted for in-
verse probability weighting. Statistical tests were 2-sided at a
significance level of .05. The false discovery rate (FDR) based
multiple comparison procedure was used to assess the statis-
tical significance of the difference in the change in mortality-
related end points (eg, aggregate mortality, mortality with or
without readmission) at the FDR level of 0.05.23,24 Analyses
were performed using SAS version 9.4 (SAS Institute).
Results
There were 8 326 688 Medicare fee-for-service hospitaliza-
tions for heart failure, acute myocardial infarction, and pneu-
monia from April 1, 2005, to March 31, 2015, among which
7 948 937 patients were alive at hospital discharge. The mean
(SD) age of the study population was 79.6 (8.7) years,
4 246 45 4 partic ipants (53.4%) were women, 6 802 296
(85.6%) were white, and 738 198 (9.3%) were black. There
were 3.2 million hospitalizations for heart failure, 1.8 million
for acute myocardial infarction, and 3.0 million for pneumo-
nia and, overall, there were 270 517 deaths from heart failure,
128 088 deaths from ac ute myoc ardial infarction, and
246 154 deaths from pneumonia within 30 days of discharge.
Baseline patient demographics were similar among the 4
study periods; comorbidities are shown in Table 1 for patients
alive at discharge. Observed trends in 30-day postdischarge
and 45-day postadmission outcomes for target conditions are
shown in Figure 2 and eTables 1 and 2 in the Supplement.
HRRP and 30-Day Postdischarge Mortality
Among patients with heart failure, IPW-adjusted postdis-
charge mortality (Figure 3A and eTable 3 in the Supplement)
increased before the announcement or implementation
48. of the HRRP (0.27% increase from period 1 to period 2;
Table 2). Relative to this baseline trend, the announcement
of the HRRP was significantly associated with an increase in
postdischarge mortality (0.49% increase from period 2 to
period 3; 0.22% difference between the change from period
1 to period 2 and period 2 to period 3; P = .01). An analysis
stratified by whether there was an associated readmission
showed that this change was entirely driven by a significant
increase in mortality without readmission (0.27% increase
from period 1 to period 2 vs 0.53% increase from period 2 to
period 3; 0.26% difference between the change from period
1 to period 2 and period 2 to period 3; P < .001). In addition,
HRRP implementation was significantly associated with an
increase in postdischarge mortality overall relative to base-
line trends (0.52% increase from period 3 to period 4; 0.25%
difference between the change from period 1 to period 2 and
period 3 to period 4; P = .001), which was also explained by
an increase in death without readmission.
In contrast, among patients with acute myocardial infarc-
tion (Figure 3B), HRRP announcement was significantly asso-
ciated with a decline in postdischarge mortality (Table 2;
0.18% increase from period 1 to period 2 vs 0.08% decrease
from period 2 to period 3; −0.26% difference between the
change from period 1 to period 2 and period 2 to period 3;
P = .01). Compared with baseline trends, HRRP implementa-
tion was not associated with a significant change in mortality
(0.15% increase from period 3 to period 4; −0.03% difference
between the change from period 1 to period 2 and period 3 to
period 4; P = .69).
Postdischarge mortality among patients with pneumonia
(Figure 3C) was relatively stable before the HRRP (0.04%
increase from period 1 to period 2), but increased signifi-
cantly after announcement of the HRRP (Table 2; 0.26%
49. increase from period 2 to period 3; 0.22% difference between
the change from period 1 to period 2 and period 2 to period 3;
P = .01). This overall change was driven by an increase in
patients who were not readmitted but died within 30 days of
discharge (0.09% increase from period 1 to period 2 vs 0.32%
increase from period 2 to period 3; 0.23% difference between
the change from period 1 to period 2 and period 2 to period 3;
P = .003). In addition, compared with baseline trends, HRRP
implementation was also significantly associated with an
increase in mortality overall (0.44% increase from period 3 to
period 4; 0.40% difference between the change from period 1
to period 2 and period 3 to period 4; P < .001) and among
stratified mortality outcomes of death and no readmission
(0.09% from period 1 to period 2 vs 0.38% from period 3 to
period 4; 0.30% difference between the change from period 1
to period 2 and period 3 to period 4; P < .001) and readmis-
sion and death (0.05% decrease from period 1 to period 2 vs
0.05% increase from period 3 to period 4; 0.11% difference
between the change from period 1 to period 2 and period 3 to
period 4; P = .003).
All P values less than .05 for the 18 comparisons involv-
ing 3 end points (total mortality, mortality without readmis-
sion, and mortality with readmission), 2 differences in change
(post-HRRP announcement trends and post-HRRP implemen-
tation trends compared with pre-HRRP trends) and 3 condi-
tions (heart failure, acute myocardial infarction, and pneu-
monia) were also significant at the FDR level of 0.05 (Table 2).
Other 30-Day Postdischarge Outcomes
Inverse probability-weighted readmissions without death
within 30 days declined significantly following the announce-
ment and implementation of the HRRP compared with the
years preceding the HRRP for all 3 target conditions (Table 2).
Trends across study periods in rates of patients who were not
readmitted and were alive within 30 days of discharge are also
51. http://www.jama.com/?utm_campaign=articlePDF%26utm_medi
um=articlePDFlink%26utm_source=articlePDF%26utm_content
=jama.2018.19232
rates steadily increased before the announcement of the HRRP
(Table 2; 0.15% increase from period 1 to period 2). Compared
with this baseline trend, the HRRP announcement was sig-
nificantly associated with an increase in mortality (0.42% in-
crease from period 2 to period 3; 0.27% difference between the
change from period 1 to period 2 and period 2 to period 3;
P = .01). However, mortality did not significantly change af-
ter HRRP implementation (0.32% increase from period 3 to pe-
riod 4; 0.17% difference between the change from period 1 to
period 2 and period 3 to period 4; P = .06).
Postadmission mortality declined among patients hospi-
talized for acute myocardial infarction before the announce-
ment of the HRRP (0.24% decline from period 1 to period 2), a
trend that did not significantly change after the HRRP an-
nouncement (0.35% decline from period 2 to period 3; −0.12%
difference between the change from period 1 to period 2 and
period 2 to period 3; P = .39). Following the HRRP implemen-
tation, postadmission mortality continued to decline (0.44%
from period 3 to period 4), but did not significantly differ from
baseline trends (−0.21% difference between the change from
period 1 to period 2 and period 3 to period 4: P = .06).
Among patients hospitalized for pneumonia, postadmis-
sion mortality was relatively stable before the HRRP (0.05%
increase from period 1 to period 2), and did not significantly
change after the HRRP announcement (0.15% decline from pe-
riod 2 to period 3; −0.20% difference between the change from
period 1 to period 2 and period 2 to period 3; P = .07) and
imple-
52. mentation (0.14% increase from period 3 to period 4; 0.09%
difference between the change from period 1 to period 2 and
period 3 to period 4; P = .30).
Table 1. Baseline Characteristics of Patients Discharged After
Hospitalization for Heart Failure,
Acute Myocardial Infarction, or Pneumoniaa
Participants, %
Period 1
(April 2005-
September 2007)
Period 2
(October 2007-
March 2010)
Period 3
(April 2010-
September 2012)
Period 4
(October 2012-
March 2015)
Hospitalizations 2 283 774 2 011 915 1 857 337 1 795 911
Demographics
Age, mean (SD), y 79.5 (8.5) 79.7 (8.7) 79.7 (8.9) 79.6 (9.0)
Women 54.4 53.7 53.1 52.2
Men 45.6 46.3 46.9 47.8
Race/ethnicity
54. Depression 8.4 8.1 8.5 8.2
Functional disability 2.9 3.2 3.4 3.3
Liver disease 1.0 1.0 1.1 1.3
Malnutrition 4.6 6.5 7.7 8.2
Psychiatric disorder 2.8 3.2 3.3 3.2
Kidney failure 14.0 18.2 21.2 21.9
Respiratory failure 6.6 8.7 10.2 11.5
Substance abuse 6.9 6.6 7.0 7.3
Trauma 7.5 7.4 7.2 6.6
Length of stay, mean (SD), d 5.6 (4.9) 5.5 (4.8) 5.2 (4.5) 5.1
(4.4)
Abbreviation: COPD, chronic
obstructive pulmonary disease.
a Data are reported as percentages
unless otherwise noted. HRRP
announcement was in April 2010
and implementation was
in October 2012.
b Race/ethnicity denoted as Asian,
Hispanic, North American Native,
other, or unknown.
Research Original Investigation Association of the Hospital
57. conditions, has reduced
hospital readmissions and led to unintended consequences. Our
analyses of Florida
hospital administrative data between 2008 and 2014 find that
the HRRP resulted in
a reduction in the likelihood of readmissions by 1% to 2% for
traditional Medicare
(TM) beneficiaries with heart failure, pneumonia, or chronic
obstructive pulmonary
disease. Readmission rates for Medicare Advantage (MA)
beneficiaries and privately
insured patients with heart attack and heart failure decreased
even more than TM
patients with the same target condition (e.g., for heart attack,
the likelihood for TM
beneficiaries to be remitted is 2.2% higher than MA
beneficiaries and 2.3% higher
than privately insured patients). We do not find any evidence of
cost-shifting, delayed
readmission, or selection on discharge disposition or patient
income. However,
the HRRP reduced the likelihood of Hispanic patients with
target conditions being
admitted by 2% to 4%.
Keywords
Medicare, readmissions, hospital, discharge
This article, submitted to Medical Care Research and Review on
30 June 2017, was revised and accepted
for publication on November 6, 2017.
1Florida International University, Miami, FL, USA
2Harvard Medical School, Boston, MA, USA
Corresponding Author:
58. Min Chen, College of Business, Florida International
University, 11200 SW 8th Street,
Miami, FL 33199, USA.
Email: [email protected]
744611MCRXXX10.1177/1077558717744611Medical Care
Research and ReviewChen and Grabowski
research-article2017
https://us.sagepub.com/en-us/journals-permissions
https://journals.sagepub.com/home/mcr
mailto:[email protected]
http://crossmark.crossref.org/dialog/?doi=10.1177%2F10775587
17744611&domain=pdf&date_stamp=2017-12-03
644 Medical Care Research and Review 76(5)
Introduction
Hospital readmissions are common and costly. In 2011, the U.S.
Medicare program
paid for 1.8 million 30-day readmissions with a total cost of $24
billion (Hines, Barrett,
Marguerit, Jiang, Joanna, & Steiner, 2014). Some readmissions
could be prevented
with better quality of care (Axon & Williams, 2011), and the
Medicare Payment
Advisory Commission (MedPAC) estimates that a 10%
reduction in avoidable read-
missions would save the Medicare program at least $1 billion
(MedPAC, 2013). To
achieve both better outcomes for patients and greater savings
for Medicare, the
Affordable Care Act (ACA) created the Hospital Readmissions
Reduction Program
59. (HRRP), which applies financial penalties to acute care
hospitals with higher-than-
expected readmission rates among Medicare fee-for-service
(FFS) beneficiaries in the
30-days following discharge for certain target conditions.
Since October 2012, the HRRP has targeted three conditions:
acute myocardial
infarction (AMI), congestive heart failure, and pneumonia.
Beginning in October
2014, total hip or knee replacement and chronic obstructive
pulmonary disease
(COPD) were also included in the program. The Centers for
Medicare and Medicaid
Services (CMS) calculates the average risk-adjusted, 30-day
hospital-readmission
rates for patients with each targeted condition and penalizes
hospitals that perform
worse than the national average. For Fiscal Year (FY) 2013, the
maximum penalty for
a hospital with excess readmissions was 1% of its total
Medicare base payment. The
penalty went up to 2% of the Medicare base payment for FY
2014, and 3% for FY
2015 forward (CMS, 2016).
New Contribution
Prior studies have examined the initial three target conditions
(i.e., AMI, heart failure,
and pneumonia) and suggested that the HRRP has lowered 30-
day readmissions
among Medicare FFS beneficiaries (Carey & Lin, 2015;
Gerhardt et al., 2013;
Zuckerman, Sheingold, Orav, Ruhter, & Epstein, 2016). Using
Medicare FFS claims
60. data, two recent articles compared the changes in readmission
rates by hospital penalty
status and confirmed that hospitals with the lowest pre-HRRP
performance had the
greatest improvement (Desai et al., 2016; Wasfy et al., 2017).
How readmissions
change among Medicare Advantage beneficiaries and privately
insured patients, how-
ever, is still somewhat unclear and vitally important. Because
the HRRP penalties only
apply to traditional Medicare patients, one way that a hospital
could recoup lost
Medicare reimbursements as a result of excess readmissions
would be to readmit more
privately insured or Medicare Advantage patients. In this study,
we exploit a state-
based all-payer dataset (through 2014) to examine the overall
impact of the HRRP on
readmissions among traditional Medicare, Medicare Advantage,
and privately insured
patients, respectively. We examine not only the aforementioned
three originally tar-
geted conditions but also the two new penalty conditions (i.e.,
COPD and total hip or
knee replacement).
Furthermore, we explore several other potential consequences of
the HRRP across
targeted and nontargeted conditions. First, we examine the
impact of the HRRP on
Chen and Grabowski 645
readmissions post–30 days to detect if the HRRP has simply
61. delayed readmissions.
Next, we examine whether the HRRP led to any “cherry
picking” of low-risk patients
at admission. Finally, we examine whether the HRRP led to
increased skilled nursing
facility (SNF) or home health agency (HHA) discharges.
Conceptual Framework
The HRRP is a very direct policy instrument. Hospitals are
financially penalized for
excess 30-day readmissions for the target conditions.
Medicare’s goal in implementing
the HRRP was to encourage hospitals to reduce 30-day
readmissions through better
hospital care. In response to the HRRP, we hypothesize that
hospitals will lower read-
missions for these target conditions assuming the cost of
reducing readmissions is
below the amount of the readmission penalty. We also assume
that hospitals want to
avoid any negative reputation effects associated with being
penalized (Winborn,
Alencherril, & Pagán, 2014), which might lead them to lower
readmissions even if the
cost of doing so exceeds the readmission penalty.
Because the HRRP is a relatively blunt policy, we expect it to
incent hospitals to
change their behaviors in both intended and unintended ways. In
terms of unintended
consequences, strong potential exists for what economists term
the multitasking prob-
lem in which providers direct their efforts toward those metrics
for which they might
be penalized while shirking on those metrics for which they are
62. not penalized. Under
the HRRP, hospitals would have the incentive to push any
readmissions out past day
30 when they are no longer penalized for the readmission.
Critics have suggested that
hospitals might dodge the HRRP penalties by increasingly
placing returning patients
within 30 days of discharge on observation status (Himmelstein
& Woolhandler, 2015;
Noel-Miller & Lind, 2015). Observation stays are billed as
outpatient services rather
than readmissions to acute care and would not be counted in the
HRRP penalty calcu-
lation. Between 2006 and 2013, observation stays increased by
96% for Medicare
patients (MedPAC, 2015). One recent study, however, did not
find a statistically
significant increase in observation stays for targeted versus
nontargeted conditions
(Zuckerman et al., 2016).
Another unintended consequence would be to discharge patients
with a low risk of
readmission to costlier postacute care settings because the
hospitals are only at risk for
readmissions under the HRRP and not postdischarge spending.
Thus, at the margin,
hospitals have the incentive to increase discharges to home
health and skilled nursing
facilities for the HRRP target conditions if such discharges
would help hospitals reduce
readmission rates. From Medicare’s perspective, spending on
these postacute services
would likely more than offset any potential savings from
decreased 30-day
readmissions.
63. Finally, the HRRP’s readmission measures adjust for
demographic characteristics
associated with higher rates of hospital readmissions (such as
age) and severity.
However, they do not allow risk adjustment based on patients’
race, ethnicity, or socio-
economic status. Because patients with low socioeconomic
status are found to have
higher readmission rates than the overall population (Hu,
Gonsahn, & Nerenz, 2014),
646 Medical Care Research and Review 76(5)
hospitals may respond to the omission of these risk factors by
selecting patients on
race and socioeconomic status associated with lower rates of
hospital readmissions.
Method
Data and Outcome Variables
We construct our hospital admissions and readmissions
measures using the State
Inpatient Discharge data, collected and maintained by the
Florida Agency for Health
Care Administration. The data contain detailed information on
all inpatient stays in
Florida from Quarter 1 of 2008 to Quarter 4 of 2014 and a
unique patient identifier that
allows us to track a patient’s historical visits across hospitals
over time. In addition, we
used Medicare Hospital Compare data released in July 2009 (for
64. the period July 2005–
June 2008) to examine baseline risk-adjusted readmission rates
at the inpatient pro-
spective payment system (IPPS) hospitals in the United States.
We adapt methods used in the prior studies to construct index
hospitalization and
30-day all-cause readmission at the patient level. Specifically,
we code index hospital-
izations as stays in which no inpatient discharge had occurred
within the previous 30
days. Hence, a hospitalization is either an index stay or a
readmission. We then iden-
tify target conditions by the principal diagnosis or procedure of
the index hospitaliza-
tion, using Healthcare Cost and Utilization Project’s (HCUP’s)
Clinical Classifications
Software (CCS). CCS is a tool that collapses diagnosis and
procedure codes from the
International Classification of Diseases, 9th Revision, Clinical
Modification
(ICD-9-CM).1 We used the single level CCS diagnosis code 100
for AMI, 108 for
heart failure, 122 for pneumonia, and 127 for COPD. The CCS
procedure code used
for total hip and knee replacement is 152-153. The ICD-9 codes
used to identify total
hip and knee replacement are 81.51 (primary hip replacement)
and 81.54 (primary
knee replacement). In addition, we follow the prior literature
(Carey & Lin, 2015;
Mellor, Daly, & Smith, 2016) and select gastrointestinal
conditions with Medicare
Severity Diagnosis Related Group (MS-DRG) codes 329-331,
377-379, and 391-392
to be our control group of Medicare index hospitalizations.2
65. Similarly, we define two additional indicator variables when
readmission occurred
within 45 days or 60 days, respectively, and compared them to
the 30-day readmission
to identify if readmission occurred within 31 to 45 days or 31 to
60 days. Finally, we
use the disposition of the patient at discharge to code dummy
variables indicating
whether the patient was discharged to an SNF or HHA.
Control Variables
To control for heterogeneity associated with changes in
readmission and other out-
comes over time, our models include a rich set of patient-level
covariates. The covari-
ates include demographics such as sex, age group, race, primary
payer, income
category, and rural/urban location. We also constructed time-
varying clinical measures
for severity adjustment, including (1) indicators of high severity
with major
Chen and Grabowski 647
complications/comorbidities based on the MS-DRG codes and
(2) the number of
comorbid conditions compiled from a set of 29 binary variables
identifying coexisting
medical conditions that are not directly related to the main
reason for index admission
(refer to HCUP’s Elixhauser Comorbidity Software for
details).3
66. We identify and exclude certain index hospitalizations
following the rules specified
in the technical reports of constructing 30-day all cause
readmission rates prepared for
CMS: (1) hospitalizations during which patients died, (2)
discharged against medical
advice, and (3) discharged or transferred to another acute care
facility. For AMI admis-
sions, we also excluded cases with same-day discharges. The
analysis sample contains
951,215 index admissions from 156 hospitals.
Statistical Analysis
We use a difference-in-differences (DD) method to compare
changes in outcomes of
patients in the treatment group before and after the HRRP
relative to changes in out-
comes of the control group. The treatment group consists of
Medicare FFS beneficia-
ries aged at least 65 years old and with one of the five HRRP
target conditions as the
primary diagnosis for their index admission. For each condition
we use three different
comparison groups for a total of 15 models. The first
comparison group consists of
hospital admissions among Medicare FFS patients aged 65 years
and older and with
gastrointestinal conditions as their primary diagnosis. The
second comparison group
includes hospitalizations of each of the five target conditions
among Medicare
Advantage patients aged 65 years and older. The third
comparison group comprises
privately insured patients with those five target conditions.
67. We estimate the following model:
Y Post Treatment Post
Treatment X Hospi
iht t i t
i it
= +
+ + +
∗ +α µ µ
µ β
1 2
3 � ttalh iht+ε
(1)
where Yiht is an indicator for a study outcome for patient i at
hospital h in time period
t. More specifically, we first examine if the patient was
readmitted within 30 days of
discharge and if there is any delayed readmission after 30 days
but within 45 or 60
days of discharge. We also examine if the patient was
discharged to a costlier postacute
care setting (i.e., a SNF or a HHA). Finally, we examine
whether the HRRP reduced
the likelihood of admitting minority patients or lower income
patients. Minority
patients are indicated by whether the patient is Black or of
Hispanic ethnicity. We iden-
68. tify a patient to be in a lower income region if the patient
resides in a ZIP code wherein
the estimated annual median household income is in the bottom
two quartiles. Each of
these outcome measures represents a separate regression.
Postt is a dummy variable set to 1 if the observation is from the
posttreatment
period in either the treatment or a comparison group. We use
2008-2009 as the pre-
HRRP period and 2012-2014 as the post-HRRP period for AMI,
heart failure, and
pneumonia. For the two newly added conditions (i.e., COPD and
total hip or knee
replacement), we use 2014 as the post-HRRP period. Treatmenti
indicates whether the
648 Medical Care Research and Review 76(5)
index admission was a hospitalization targeted by the HRRP,
and equals zero if the
index admission was part of a comparison group. The
interaction effect of Postt *
Targeti represents our key variable of interest, the DD estimate
of the impact of the
HRRP. Xit is a vector that captures the time-varying patient
characteristics (listed in
Table 1). The hospital fixed effects (Hospitalh) are used to
control for the unobserved,
time-invariant differences across hospitals.
Thus, we use pre-HRRP levels for the target admissions and
concurrent changes
from the precontract to postcontract period in the nontarget
69. admissions to establish
counterfactuals that would be expected in the absence of HRRP
program, and we esti-
mate changes that differed from this expectation (i.e., the
differential change or the
change attributable to the HRRP). For all the regression
analyses, the standard errors
are clustered at the level of the hospital to allow for an arbitrary
covariance matrix
within the clusters.
Because penalties are based on whether a hospital’s readmission
rate exceeds the
national average, hospitals with a baseline readmission rate
above the threshold are at
greater risk of the penalty and thus have stronger incentives to
improve. In July 2009,
the CMS Hospital Compare website began to publicly report
IPPS hospitals’ perfor-
mance in 30-day readmission rates for AMI, heart failure, and
pneumonia, respectively.
For each IPPS hospital with more than 25 cases, its performance
is classified into three
categories: “better than U.S. national rate,” “no different than
U.S. National Rate,” or
“worse than U.S. national rate.” We use the national rate for the
period July 2005 to
June 2008 obtained from CMS’s Hospital Compare data as the
baseline threshold rate
and compare the hospital specific average 30-day readmission
rates to the national
average to define if a hospital is “at risk” for any penalty.4
Given that penalties are
based on a hospital’s past 3-year average readmission
performance, partial responses
might be observed immediately after ACA passage but before
70. penalties go into effect.
Using historic readmission rates prior to ACA passage allows us
to test the full effects
of the HRRP. To examine how the impact of HRRP varies
across hospitals with differ-
ent risks of facing the penalty, we divide the sample into two
groups based on whether
patients were admitted into a hospital with its baseline
readmission rate above the
threshold rate, and then we re-estimate the DD model on both of
the subsamples.
We further compare this DD estimate of patients treated at
hospitals at risk for
HRRP penalties versus those patients treated at hospitals not at
risk for penalties. More
formally, we estimate the triple difference model (DDD)
specified below:
Y Post Target Risk Post Target Post
Risk
iht t i h t i t
h
= + + +
+
∗ ∗ ∗
∗
α µ µθ 1 2
µµ γ
71. γ γ β
3 1
2 3
Target Risk Post
Target Risk X Year
i h t
i h it t iht
∗ +
+ + + + +ε
(2)
Compared with Equation (1), the added variable Riskh is an
indicator variable that
specifies whether a hospital is at risk for HRRP penalties, which
equals to 1 if hospital
h’s baseline readmission rate is above the national average and
0 otherwise. The inter-
action effect of Postt * Targeti * Riskh represents our key
variable of interest, the triple
difference estimate of the impact of the HRRP.
649
T
a
b
136. ns
a
re
in
b
ra
ck
et
s.
650 Medical Care Research and Review 76(5)
As noted above, the DDD approach implicitly assumes that
hospitals at-risk and
not at-risk for the HRRP share the same readmission shocks in a
given hospital and
year that are unrelated to the HRRP policy. The DD approach,
which instead used
as controls the within-hospital readmission shocks among
patients not included in
the HRRP program, may actually be preferable. Because little
basis exists for distin-
guishing these approaches ex ante, these models are probably
best viewed as com-
plementary approaches for exploring the validity of this study’s
key results.
We conduct additional analyses to explore potential sources of
bias. We compare
trends in each outcome between the targeted and nontargeted
137. admissions during the
pre-HRRP period. Similar pre-HRRP trends would support our
assumption that
changes from the pre-HRRP to post-HRRP periods would have
been similar for the
target and nontarget conditions in the absence of the HRRP
program. Considering that
CMS began publicly reporting hospital performance in July
2009 and hospitals might
start to respond by changing their behavior since then, we
restrict the pre-HRRP period
to be the first two quarters of 2009 and reestimated all the
specifications using the
alternative sample and the results stay robust.
Results
We observe several notable trends when examining the 30-day
all cause readmis-
sions by condition from 2008 to 2014 (see Figure 1). First, the
30-day readmission
rates of FFS patients followed similar trends from 2008 to 2009
across the five target
conditions and gastrointestinal condition. Second, the 30-day
readmission rates of
FFS patients with each of the five target conditions decreased or
stayed relatively
stable from 2012 to 2014, while the FFS patients with
gastrointestinal conditions
Figure 1. Thirty-day all-cause readmission trend by condition.
Note. HF = heart failure; PN = pneumonia; HIP = total hip or
knee replacement; GI = gastrointestinal
conditions.
138. Chen and Grabowski 651
experienced an increase in their 30-day readmission rate during
the same time
period. Finally, within the same condition, the 30-day
readmission rates of FFS and
Medicare Advantage patients followed similar trends from 2008
to 2009. We there-
fore use 2008-2009 as the pre-HRRP comparison period. When
comparing across
payers for a given target condition, we also observe that
Medicare readmissions
rates were consistently higher than the rates for privately
insured patients. Table 1
reports the descriptive statistics of the whole sample and by
each of the five target
conditions as well as the gastrointestinal condition. Compared
with the national
average, our sample has slightly higher 30-day all-cause
readmission rates in AMI
and heart failure and comparable readmission rates in
pneumonia, COPD, and total
hip or knee replacement.
We next examine the DD estimates on HRRP targeted
admissions using three dif-
ferent comparison groups (see Table 2). Compared with
Medicare FFS patients with
gastrointestinal conditions as the primary diagnosis, there was a
1% to 2% decrease in
30-day readmissions for comparable heart failure, pneumonia,
and COPD patients.
However, when compared with Medicare Advantage patients
with the same target
condition, we observe a statistically significant increase in 30-
139. day Medicare FFS read-
mission for AMI, heart failure, and pneumonia. Similarly, when
compared with
privately insured patients, 30-day readmissions for Medicare
FFS patients admitted
with AMI and heart failure increased. The results reveal that
although the HRRP tar-
geted Medicare FFS patients only, hospital readmission rates
declined substantially in
the MA and privately insured population after the HRRP,
especially among cardiac
related admissions. This may suggest that there are spillover
effects from the HRRP
extending to MA and privately insured patients. We then restrict
our attention to MA
and privately insured patients admitted with one of the five
HRRP target conditions
and compare changes in their readmissions to those of MA and
privately insured
Table 2. Difference-in-Differences (DID) Estimates of the
Effect of the Hospital
Readmissions Reduction Program on Medicare FFS 30-Day
Readmissions.
Medicare FFS patients
with GI conditions as
control
Medicare FFS with
Medicare Advantage,
same condition as control
Medicare FFS with
140. private insurance, same
condition as control
DID impact DID impact DID impact
(1) (2) (3)
(1) Heart attack −0.004 (0.005) 0.022*** (0.008) 0.023***
(0.007)
(2) Heart failure −0.007** (0.004) 0.012** (0.005) 0.020***
(0.008)
(3) Pneumonia −0.006* (0.003) 0.010* (0.006) 0.007 (0.005)
(4) Chronic obstructive
pulmonary disease
−0.018** (0.005) 0.005 (0.007) −0.006 (0.008)
(5) Total hip or knee
angioplasty
0.013 (0.008) 0.01 (0.011) 0.002 (0.007)
Note. FFS = fee-for-service; GI = gastrointestinal. All models
include control variables listed in Table 1 as well as hospital
fixed effects. The standard errors are clustered at hospital level.
Robust standard errors in parentheses.
*p < .1. **p < .05. ***p < .01.
652 Medical Care Research and Review 76(5)
patients with gastrointestinal conditions, respectively. The DD
estimates reported in
Appendix Table A1 confirmed that after passage of the HRRP,
141. hospitals reduced car-
diac-related readmissions not only for Medicare FFS patients,
but also for MA and
privately insured patients.
Next, we reran the DD estimation conditional on hospitals’
baseline readmission
performance (see Table 3). Compared with admissions with
gastrointestinal condi-
tions, index hospitalizations with target conditions at a hospital
“at risk” for penalties
had statistically significant lower …
EVIDENCE-
BASED CARE
SHEET
Author
Hillary Mennella, DNP, ANCC-BC
Cinahl Information Systems, Glendale, CA
Reviewers
Darlene Strayer, RN, MBA
Cinahl Information Systems, Glendale, CA
Jocelyn Cajanap-Gantman, RN, MSN,
FNP, CNS
Sepsis Coordinator, Glendale Adventist
Medical Center
Nursing Executive Practice Council
143. readmissions are estimated at
$26 billion, with potentially preventable readmissions
accounting for approximately $17
billion of that cost(1)
› All healthcare providers are responsible for identifying patient
discharge needs and
developing a thorough discharge plan to reduce the risk for
hospital readmissions(3)
› In accordance with the legislative passing of the Affordable
Care Act (ACA), the
United States Centers for Medicare and Medicaid Services
(CMS) established the
Hospital Readmissions Reduction Program (HRRP) to decrease
the frequency of hospital
readmissions of Medicare beneficiaries. Effective October 1,
2012, the provisions of the
HRRP permit the CMS to reduce payments to hospitals under
the inpatient prospective
payment system (IPPS) for readmission rates that are reviewed
by CMS and determined
to be excessive. The HRRP adjusts hospital reimbursement
based on the data for
excessive readmissions following patient admissions for acute
myocardial infarction
(AMI), congestive heart failure (CHF), pneumonia (PN),
coronary artery bypass graft
(CABG), chronic obstructive lung disease (COPD), stroke, and
complications related to
readmissions for total hip/knee replacements
(THR/TKR)(4,9,12)
• Under the HRRP, readmission is defined by the CMS as “an
admission to a subsection
hospital within 30 days of a discharge from the same or another
144. subsection hospital”(4)
–The hospital readmission rate is calculated from the date of
discharge, plus 30 days.
For example, for a patient who is discharged on October 1, the
last day for the
postdischarge follow-up period is October 31
–The CMS recognizes 30 days as an industry standard that is
strongly influenced by the
quality of care
–Hospital readmissions exclude those involving a patient’s
death in the hospital,
enrollment in the Medicare fee-for-service program, hospital
admission after at least
30 days post-hospital discharge, and planned hospital
readmission (i.e., a nonacute
readmission for a scheduled procedure)
–CMS does not consider preventability when calculating
readmission rates(11)
- A standard for identifying and defining what is considered to
be a preventable
readmission does not exist(11)
–Hospital readmission rates are assigned a “yes/no” readmission
status regardless of the
number of readmissions for a patient during the 30-day
postdischarge time period
› Each year during the period 2003–2004, according to billing
claims from the CMS, an
estimated 2.3 million Medicare beneficiaries were readmitted to
the hospital within 30
145. days of discharge. Investigators in a study of 11,855,702
Medicare beneficiaries reported
that(8)
• 19.6% were discharged from the hospital and rehospitalized
within 30 days
–Of this group, 50.25% did not have a bill for a follow-upvisit
to a physician’s office
between the time of discharge to the community and
rehospitalization
• 34% were discharged from the hospital and rehospitalized
within 30 days
–An estimated 10% were planned hospital readmissions
• 67.1% of Medicare beneficiaries with medical conditions who
were discharged from the hospital were rehospitalized or
died within the first year after discharge
• 51.5% of Medicare beneficiaries who were discharged from
the hospital after surgical procedures were rehospitalized or
died within the first year after discharge
–Of this group, 70.5% were rehospitalized with a medical
condition
› The CMS began reporting the 30-day mortality rates for the
quality outcome measures for AMI and CHF in 2007 and
for PN in 2008. These quality outcome measures are publicly
reported in an effort by CMS to increase transparency and
accountability of hospitals for patient care services and
treatment(2,12)
• The CMS recommends that hospitals review their 30-day
146. mortality outcome measures in conjunction with their 30-day
hospital readmission data in order to modify the quality and
type of care provided to reduce hospital readmissions
› The financial penalties for the HRRP were calculated by the
CMS using data from July 2008 through June 2011 for the
readmission rates for all hospitalizations for AMI, PN, and
CHF; these rates were adjusted for age, gender, patient frailty,
and coexisting medical conditions and compared with the actual
readmission rates over the same period of time using a
methodology that is endorsed by the National Quality Forum
(NQF)(4,9,12)
• A hospital’s calculated readmission rate for MI, PN, CHF,
COPD, CABG, stroke, and THR/TKR is the performance
measure of that hospital’s readmission rate compared with the
national average for a hospital’s set of patients with the same
applicable conditions
• For the fiscal year 2013, hospital readmission rates were
calculated from data on discharges from July 1, 2008, through
June 30, 2011
• In the fiscal year 2014, an estimated 80% of hospitals were
penalized, at a cost of $428 million(1)
• Kaiser Health News (KHN) reported 4 out of 5 hospitals were
penalized for readmissions based on patient discharge data
analyzed between July 2013 and June 2016(13)
–The average penalty between October 1, 2017 and September
30, 2018 is expected to be 0.73% for each payment
Medicare makes per patient(13)
› The CMS levied financial penalties of up to 1% of hospital
reimbursement rates for readmission of Medicare beneficiaries.
147. The financial penalties increased to 2% in 2014 and to a
maximum of 3% in 2015(9)
› The CMS 30-day hospital readmission measures are federally
mandated to be publicly available under the Hospital Inpatient
Quality Reporting Program(12)
› The CMS provides hospitals with Hospital-Specific Reports
(HSRs) under the Hospital Inpatient Quality Reporting
(IQR) program to promote hospital quality improvement efforts.
The HSRs provide detailed information on a hospital’s
readmission rates, discharge data, and specific risk factor
data(4,12)
› Investigators analyzing the publicly available data from July
2008 through June 2011 for 3,282 hospitals found that
large hospitals, teaching hospitals, and safety-net hospitals (i.e.,
a hospital system that provides care to a large number of
uninsured or low-income patients) had higher readmission rates
compared with small hospitals and non-teaching hospitals.
Of this sample, 2,189 hospitals, or 66.7% of hospitals, will
receive financial penalties as a result of the HRRP.
Investigators
call for additional research to determine why large hospitals,
teaching hospitals, and safety-net hospitals have higher
readmission rates than small and non-teaching hospitals(10)
› Researchers evaluating the impact of community factors on
hospital readmission rates noted that a large portion of
readmission rates is affected by the characteristics of the local
healthcare community (e.g., quality of nursing homes, access
to primary care), specifically the county where the hospital is
located. This suggests that penalizing hospitals with high
readmission rates might not be equitable and that interventions
aimed at community-based readmission reduction strategies
might result in improved outcomes(6)