University of Southern California Dissertations and Theses

Investigation of health system performance: effects of integrated triple element method of high reliability, patient safety, and care coordination

(USC Thesis Other)

Investigation of health system performance: effects of integrated triple element method of high reliability, patient safety, and care coordination

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content i

Investigation of Health System Performance:
Effects of Integrated Triple Element Method of
High Reliability, Patient Safety, and Care Coordination

By
Sanaz Massoumi

Dissertation Submitted to the Faculty of the Graduate School of the
University of Southern California
In Partial Fulfillment of the Requirements for the Degree of
DOCTOR OF PHILOSOPHY
(Industrial and Systems Engineering Department)
December 2017

PhD Dissertation Committee:
Professor Najmedin Meshkati, Committee Chair
Professor Mansour Rahimi, Committee Member
Professor Philip D. Lumb, Committee Member
Professor Alexander Capron, Committee Member
ii

©Copyright by
Sanaz Massoumi
2017

iii

I dedicate this work to the true angels of my life:
my mom, my dad, and my family,
for their infinite love and multifarious devotion.

iv

Acknowledgements
I would like to express my deepest appreciation to my committee chair, Professor Najmedin
Meshkati, whose commitment to excellence in teaching and research is unparalleled. He has
always believed in me throughout my endeavors as a Ph.D. student. Without his guidance and
persistent help this dissertation would have not been possible. He has been an enormous inspiration
for me and everyone around, in our personal and technical lives. Dr. Meshkati has the attitude and
the substance of a genius and continually and convincingly conveyed a spirit of adventure in regard
to research and innovation.
My entire Ph.D. advisory committee have played such a crucial role in the success of this research
that it is nearly impossible for me to put the depth of my gratitude towards them in words. Professor
Rahimi has always been a tremendous support. He patiently and brilliantly mentored me to learn
how the academic environment works with respect to teaching and research. I highly appreciate
all he has taught me. Dr. Lumb has had the most pivotal role in the success of this research. He
has been the most enlightening guidance for me to explore the field of healthcare. I had the honor
of being constantly guided by Dr. Capron who always made room in his super busy schedule to
meet with me and help me with different aspects of the project.
There have been other truly supportive mentors and educators that I would like to pay utmost
respect and tribute to. Dr. Stephanie Hall has been the most empowering force behind the progress
of this research. She truly made it possible for me to proceed with all the practical parts of the
work. Dr. Jamal Abedi is a true treasure. He has always extended his invaluable support with
respect to all the experimental design, statistical, and data processing parts of this research. Dr.
Carol Peden's impact on my technical life has been incredible and is sure to last forever.
Her meticulous and innovative way of dealing with technical matters has undoubtedly made her a
v

role model for me. Dr. Greg Placencia has been always there for me. The fruitful discussions with
him, as an excellent researcher and a brilliant mind, have always illuminated my way ahead. I had
the privilege to work closely with Dr. David Belson. As one of the technical pioneers of the field,
he was the one who made it possible for me to experience and observe how the world of clinical
practice works.
Many special thanks to Dr. Katherine Sullivan that made it possible for me to have hands-on
experience with the practical side of healthcare. I wish her the greatest success in all her highly
significant efforts, as I witnessed through years, to elevate the quality of care. Dr. Brian Prestwich
is such an inspiration. It has been a real honor to have the chance to know him. He is truly
encouraging and supportive and I cannot thank him enough for it.
There are some angel figures that have unconditionally loved me and persistently supported me in
all stages of life. They literally and figuratively took my hands and showed me how to take steps.
I owe them for all I have achieved. What they have done for me is not something that my words
could do justice in thanking them for, but I just want them to know that I love them with every
ounce of my being. My mother and father fully dedicated their lives to my wellbeing and success.
Nothing can match what they have done for me. They made me who I am and I am proud of it.
They have been so caring, so loving, so empowering, and so patient all along and I owe them so
very much. The love of my life, Med, has been the true force that enabled me to do the Ph.D.
research. He is so encouraging, so inspiring, so supportive, so affectionate, and so dependable, and
his support means everything to me. I just admire and adore Dr. Med Nariman so much. My super
talented and brilliant brother, Soroush, has been always extremely supportive and kind. He
has always made me feel special and I am thankful to him for that. I love him so dearly and wish
him a life and career as special as he is.
vi

Maman Aghdas, I know you lived your whole life to see this day, but I am sorry that I was late to
fulfill your dream. I could never wrap my mind around your passing. Your love was true love,
pure love, and selfless love. I am sure you look down at me and smile like you always did when
you took me in your arms and put me to sleep. My infinite special thanks to Baba Mammad, Khaleh
Shirin, Khaleh Maryam, and their lovely families. They played such a crucial role in my life. I
always looked up to them and learned so much from them. They are terrific and I cannot put the
level of my love towards them in words.
Throughout the years, I have been supported by so many phenomenal, kind, intelligent, and
supportive friends, family members, teachers, professors, and coworkers that I owe them special
thanks. I most owe it to my best friend Negin Tootoonchian who has always been closer than a
sister to me and is an embodiment of unconditional love. I learned so much technically and
otherwise from my friend in need and friend indeed, Dr. Yasaman Dehghani Abbasi. I was always
inspired by her commitment to excellence in everything she has done. Dr. Maryam Tabibzadeh
has been always there for me unconditionally. She is truly an angel inside and out with a superb
intellect. I also want to acknowledge some super kind-hearted individuals in my life: Saeedeh
Ahmadi, Leila Navaee, Leila Saki, and their lovely families. They are most amazing and helpful.
I want to pay a special thanks to all faculty and staff at the Daniel J. Epstein Department of
Industrial and Systems Engineering at USC for their continual support throughout my journey as
a graduate student.

vii

Abstract
It is called a ‘Systems of Systems’, as healthcare is a purposeful amalgamation of several distinct
systems, each of which are designed to serve some particular purpose. When bonded together,
these systems give rise to a colossal system that serves new purposes, pursues new causes, and
faces new challenges. To perform the new functionalities, healthcare requires new system elements
that create interdisciplinary interactions and collaborations among subdivisions of the healthcare
delivery system.
In May 2014, the President’s Council of Advisors on Science and Technology published a report
titled “Better Health Care and Lower Costs”. The authors heavily emphasized improvement
through Systems Engineering. They claimed systems engineering approaches have reduced waste
and increased reliability in many industries. These methods helped generate remarkable positive
results in the few healthcare organizations that have implemented such concepts. Systems
engineering approaches have transformed healthcare at much larger scales. System tools and
methods can ensure that care is reliably safe and centered on patients and their families. They help
to eliminate inefficient processes. Applying systems engineering approaches can help healthcare
organizations to re-design their operations with safety measures embedded, to reduce
redundancies, to standardize work processes, and to account for human factors. Moreover, they
help with identifying the infrastructures needed to measure and monitor the health outcomes and
the relationships among different subsystem elements. In addition, systems engineering
approaches foster continuous improvement tools to help with the handling of healthcare
multifaceted processes.
Considering healthcare as a hypercomplex system and bearing in mind the aforementioned
recommendations, the endeavors of this study have been to apply systems engineering approaches
viii

to integrate a set of system elements to form a package of system modules and to evaluate the
impact of that package on the performance of health systems. The proposed package comprises
High Reliability Organizations principles, Patient Safety factors, and Care Coordination activities.
The bundle of High Reliability, Patient Safety, and Care Coordination symbolizes a three pillar
groundwork for a healthy care delivery system. They complement each other to synchronize health
systems with high effectiveness. They share similar subset elements that reinforce the impact of
the package on the health systems’ performance while they have their own unique features. They
collectively provide a comprehensive patient centered care system where patients receive reliable
and safe care within and among all providers. The three system elements work as three silos, and
this study has integrated them to evaluate the effectiveness of their combination on the
performance of the healthcare delivery systems.
To conduct this analysis, Keck Hospital of the University of Southern California was selected and
the data for the two years of 2013 and 2015 was collected on a monthly basis for a set of 20 harm
reduction measures that indicate the performance of the hospital. Deliberated changes were
implemented with new strategies in Keck Hospital starting in 2014 and gave health outcomes that
were visible and evident starting in 2015. That was the reason for the two years of 2013 and 2015
being used in the study, as a form of Before and After states of Keck Hospital’s performance. The
modifications in hospital strategies in regards to delivering the care were categorized in three
brackets of High Reliability, Patient Safety, and Care Coordination with respect to the formats and
structures of the new approaches and policies.
The Multivariate Analysis of Variance (MANOVA) technique was used to run an overall statistical
test on the set of identified variables to determine how the independent variable influences the
responses and outcomes of the dependent variables. The independent variable is the proposed
ix

package of system elements consist of High Reliability, Patient Safety, and Care Coordination,
while the dependent variables are the performance measures of Keck Hospital. In addition to
MANOA, Discriminant Analysis, Reliability Test, Factor Analysis, and Individual Analysis of
Variance (ANOVAs), as a set of statistical tests, were used to create a more comprehensive
investigation and to deliver an insightful conclusion.
The MANOVA result rejected the null hypothesis indicating that overall Keck Hospital’s 2015
performance surpassed its 2013 performance as a result of the proposed package of High
Reliability, Patient Safety, and Care Coordination. In the Discriminant Analysis, the result found
a group of predictor variables that predicted the outcome at a statistically significant level with a
very strong model. The results of performing individual ANOVAs indicated that the majority of
measures demonstrated a significant improvement from 2013 to 2015.
The collective outcome resulting from this series of statistical analyses is a manifestation that the
addition of the proposed package of High Reliability, Patient Safety, and Care Coordination in a
health system is truly positively impactful. This combination of the triple system elements offer
major benefits to healthcare delivery systems including, but not limited to, improving the
performance, enhancing patients’ health outcomes, reducing the number and severity of adverse
events, promoting an effective and powerful teamwork and practical team behaviors, empowering
staff and patients, developing a purposeful Safety Culture, encouraging continuous improvement,
embracing systematic changes, and reducing the fragmentation of care. They collectively provide
a solid foundation for an effective and efficient healthcare system where patients receive a safe
and reliable care at the right time from the right set of professionals.

x

Table of Contents
Acknowledgments……………………………………………………………..…………………iv
Abstract……………………………………………………………………..……………………vii
List of Figures………………………………………………………………………...…………..xi
List of Tables………………………………………………………………………...…………...xii
Abbreviations………………………………………………………………………...………….xiii
I. Introduction……………………………………………………………………………………...1
II. Significance, Novelty, and Impact of this Research…………………………………………….7
III. Literature Review…………………………………………………………………………….10
A. High Reliability Organization Principles……………………………………………...11
B. Patient Safety………………………………………………………………………….34
C. Care Coordination……………………………………………………………………..44
IV. Methodology…………………………………………………………………………………49
A. Research Design………………………………………………………………………49
B. Data Collection and Analysis………………………………………………………….55
C. Harmonizing the Scale………………………………………………………………...57
V. Results……...…………………………………………………………………………………59
A. MANOVA…………………………………………………………………………….68
B. Discriminant Analysis………………………………………………………………...72
C. Reliability Test………………………………………………………………………...76
D. Factor Analysis………………………………………………………………………..76
E. Individual ANOVA……………………………………………………………………79
VI. Discussion……………………………………………………………………………………83
A. Statistical Analyses……………………………………………………………………84
B. Triple Element Model……………………………..…………………………………..86
C. Limitations of this Research…………………………………………………………87
D. Future Studies…………………………………………………………………………88
VII. Conclusion…………………………………………………………………………………..89
VIII. Bibliography………………………………………………………………………………..91
IX. Appendices…………………………………………………………………………………...99

xi

List of Figures
Figure 1: Key Features of Affordable Care Act.............…………………………………………..3
Figure 2: Linked Aims of Improvement..………………………………………………………….4
Figure 3: HRO, Patient Safety, and Care Coordination Three Pillar Groundwork………………..10
Figure 4: HRO, Patient Safety, and Care Coordination Package…………………………………10
Figure 5: State of Mindfulness………………………………………………...……...…………..18
Figure 6: Building Blocks of High Reliability……………………………………………………24
Figure 7: Workplace Safety………………………………………………………………………38
Figure 8: Elements of Successful Care Coordination…………………………………………....48
Figure 9: Care Coordinated Model………………………………………………………………48
Figure 10: Performance Comparison Chart………………………………………………………59
Figure 11: Monthly Performance with Tend Lines…………………………………………….…60
Figures 12: Distribution of Data……………………………………………………………….…61
Figure 13: SPSS Normality Test Results..…………………...……………………………….…..63
Figure 14: Separate-Group Graphs……………………………………………………...…….….75
Figure 15: Scree-Plot……………………………………………………………………….….…78
Figure 16: Pictograph of Triple Element Method………………………………………………...86

xii

List of Tables
Table 1: High Reliability Health Care Maturity Model………………………………………….26
Table 2: Design of Experiment Techniques………………………………………………………51
Table 3: MANOVA Test Statistics……………………………………………………………….52
Table 4: List of Dependent Variable Measures…………………………………………………...56
Table 5: Levene’s Test of Homogeneity………………………………………………….……....64
Table 6: Summary of Normality and Homogeneity Test Results…………………………………65
Table 7: Final Selection of Dependent Variables…………………………………………………66
Table 8: Descriptive Statistics………………………………………...………………………….68
Table 9: Box’s Test………………………………………………………...……………………..69
Table 10: Levene’s Test of Equality of Error Variances………………………………………….70
Table 11: Multivariate Tests of MANOVA…………………………………...………………….71
Table 12: Pearson Correlations……………………………………………………………...……72
Table 13: Discriminant Scores…………………………………………………………….……..73
Table 14: Structure Matrix……………………………………………………………………….74
Table 15: Classification Results…………………………………………………………...……..75
Table 16: Reliability Test…………………………………………………………...……………76
Table 17: KMO and Bartlett’s Test……………………………………………………….………77
Table 18: Factor Analysis - Total Variance Explained…………………………………..……….77
Table 19: Rotated Component Matrix………………………………………………………...….79
Table 20: ANOVA and t-Test…………………………………………………………………….80
Table 21: Individual ANOVAs…………………………………………………………….……..81
xiii

Abbreviations
OECD: Organization for Economic Co-operation and Development……………………………...1
ACA: Affordable Care Act……………………………………………………………………..….2
HRO: High Reliability Organization…………………………………………...………………….3
USC: University of Southern California…………………………………………………………...4
DOE: Design of Experiment……………………………………………………………………….5
MANOVA: Multivariate Analysis of Variance……………………..……………………………..5
ANOVA: Analysis of Variance……………………………………………………………………5
SPSS: IBM Statistical Package for the Social Sciences……………….…………………………...6
HRT: High Reliability Team……………………………………………………………………..11
IOM: Institute of Medicine……………………………………………………………………….12
RPI: Robust Process Improvement……………………………………………………………….17
AHRQ: Agency for Healthcare Research and Quality…………………………………………....36
HEDIS: Healthcare Effectiveness Data and Information Set……………………………………..48
IRB: Institutional Review Board…………………..……………………………………………..55
HCAHPS: Hospital Consumer Assessment of Healthcare Providers and Systems……………….56
CMS: Center for Medicare and Medicaid Services...................................................................…..56

1

I. Introduction
There is no doubt that our current healthcare system in the U.S. is not strong enough to prevent
medical errors, while there are also severe shortcomings in providing the right treatments to
patients in need at the right time and with a high efficiency. At this juncture, only half the time do
patients receive proper care, and that translates to wasting more than 50% of our resources. These
flaws are system driven and hence effective interdisciplinary collaborations in health organizations
can significantly improve the outcome of the care for patients and ensure their safety (Garman,
Leach, & Spector, 2006).
The public health goal for the 21
st
century is to shift the focus from mortality and morbidity to a
social model of health. Thus, the primary outcome of both public health and health service delivery
is to ensure that the entire population experiences a healthy lifestyle and greater expectancy rather
than a premature death or a longer life with chronic illness and disability. At this point, the U.S.
spends 18% of its Gross Domestic Product on health related expenses, and still the U.S. population
is in the lower half of the spectrum of the Organization for Economic Co-operation and
Development (OECD) countries ranking for major public health outcomes, such as infant
mortality, infection control, and overall health status of families, women, and children. In 1957
the annual health spending per person in America was $157 and in 2012 it was $8,953, yet we
have not been able to achieve better health outcomes in accordance with the increased expenditure
(Brink, 2014).
The Commonwealth Fund published a report in April 2014 containing data on the performance of
the U.S. healthcare system in different states, and mentioned that between 2007 and 2012 some
important measures of access to care for adults and care coordination had deteriorated while the
cost had been on a rising slope (Radley et al., 2014). Measures such as primary and preventive
2

care for adults, obesity, and health related quality of life have worsened, and the disparity in
peoples’ healthcare experience across and within states have increased (Radley et al., 2014). They
stated the most important dimensions of care in four categories of “Access and Affordability”,
“Prevention and Treatment”, “Potentially Avoidable Hospital Use and Cost”, and “Healthy Lives”
(Radley et al., 2014).
Healthcare is a very large system that can be described as a “Systems of Systems”. This is a
scenario where a combination of multiple systems, which have been designed for a specific
purpose, are united, so an enormous system emerges to serve new causes and to overcome new
challenges (Committee on Human-System Design Support for Changing Technology, 2007). This
is why new system elements are required to create interdisciplinary communication and interaction
among subdivisions of the healthcare delivery system. Implementing a structured, evidence-based
systems engineering method can be extremely beneficial, as has been shown in other industries,
and can redesign the care processes to integrate policies and procedures with personnel, human,
and organizational factors to provide better health at lower cost (Kaplan et al., 2013).
In 2010 the Affordable Care Act (ACA) was established, and the three key features are consumer
protection, access to healthcare, and lowering healthcare costs, while improving the quality of care,
which have been the basis for many changes in care delivery models. There however is still missing
infrastructure to fully implement these main principles.
3

Figure 1: Key Features of Affordable Care Act. US Department of Health and Human Services.
Source: http://www.hhs.gov/healthcare/facts/timeline/index.html

To overcome these deficiencies and in alignment with the major tenets of ACA, this study suggests
a systematic approach to analyze the impact of a package of system elements on the performance
of hospitals and to assess the influence of such additions on the enhancement of the outcomes of
care for patients as they receive it in health systems. The proposed package of system elements
consist of High Reliability Organization (HRO) principles, Patient Safety factors, and Care
Coordination activities integrated and combined together in one set. This study aims to assess and
analyze the impact of the addition of this new bundle of system elements on the overall
performance of hospitals.
Thomas Nolan stated that “The will to provide ever better care and services, a constant flow of
ideas on ways to improve and innovate care and services for better outcomes at lower costs, and
the ability to execute tests of change and to implement plans and to operate a smooth-flowing and
effective delivery system is needed for a health system to work well” (Nelson et al., 2008). P.
Batalden and F. Davidoff also mentioned that “Necessary component for a health system to work
4

well is for everyone in the health system to help achieve better patient outcomes, better system
performance, and better professional development” (Batalden & Davidoff, 2007; Nelson et al.,
2008).

Figure 2: Linked aims of improvement (Batalden & Davidoff, 2007).

Healthcare systems include not only hospitals, as major points of delivering care, but also clinics,
mental institutes, rehabilitation centers, and private doctors’ offices, etc. The focus of this research
is hospitals and the data was collected from Keck Hospital of the University of Southern California
(USC) in a before/after scenario. In 2013 Keck Hospital performed their operations based on their
basic fundamentals of providing care. Later in 2014 the executive management decided to
incorporate, step by step, new measures to monitor and policies to embed in their daily operations,
with the aim of providing a higher quality of care to their patients and achieving superior
performance. To attain this goal, Keck Hospital followed a pattern of embedding HRO principles,
Patient Safety factors, and Care Coordination activities into their processes. The deliberate changes
instituted at Keck Hospital were hence categorized into three brackets of High Reliability, Patient
Safety, and Care Coordination by the investigator.
5

The data were collected from 2013 as the “before state” and 2015 when the new measures and
policies where effective and in place as the “after state”. Data for 20 identified measures were
gathered for a total of 24 months (12 months for each selected period of one year). After several
phases of analyzing the data and with the help and guidance of subject matter experts (Dr.
Stephanie L. Hall, Chief Medical Officer of Keck Hospital of USC and USC Norris Cancer
Hospital and Dr. Philip D. Lumb Chair and Professor of the Department of Anesthesiology of
Keck Hospital of USC) eight of those measures that best represented and characterized the
performance of Keck Hospital were identified to conduct the statistical analysis.
Although Design of Experiment (DOE) was not used in its entirety, since an experiment was not
piloted at Keck Hospital, this is a label to explain the method that was utilized to conduct this
analysis. This is a conceptual framework that was applied to the structure of the data to develop a
systematic method in leading this assessment. The reason to select DOE was that this is a
methodical technique to identify causal relationships between variables that impact a process and
the outgoing results. The most suited DOE technique for this study was the one-way multivariate
analysis of variance (MANOVA).
Along with an extensive literature review, the research question “Does addressing HRO principles,
Patient Safety factors, and Care Coordination activities combined in one package as a system
element positively impact the performance of a hospital?” was tested using the MANOVA
technique as well as Discriminant Analysis, Reliability Test, Factor Analysis, and individual
Analysis of Variance (ANOVA) to create a more comprehensive investigation and to deliver an
insightful conclusion.
6

The statistical software used was Statistical Package for the Social Sciences (SPSS), which is
officially called IBM SPSS Statistics. This is software that is very commonly utilized and is very
popular for healthcare related studies and Health Science.
The results of these analyses revealed that indeed this proposed package of system elements has
improved the performance of Keck Hospital and the experience of care by the patients. The null
hypothesis of this study was defined as “Applying HRO principles, embedding Patient Safety
factors, and implementing Care Coordination activities together as one package of a system
element does not improve hospitals’ performance” and the result of the MANOVA test rejected
this null hypothesis and the remaining set of analyses confirmed the same result.
The following chapters describe and explain the endeavors of this study to evaluate the proposed
model; assess the combination of the HRO principles, Patient Safety factors, and Care
Coordination activities; analyze the methods selected to perform this research; and examine the
impact of this package of system elements on the performance of Keck Hospital and the
enhancement of the patients’ experience to put this study’s hypothesis to test.

7

II. Significance, Novelty, and Impact of this Research
The latest report, “Better Health Care and Lower Costs”, from the President’s Council of Advisors
on Science and Technology published in May 2014 heavily emphasized improvement through
systems engineering stating that systems engineering approaches have reduced waste and
increased reliability in many industries and that remarkable positive results have emerged in the
few healthcare organizations that have implemented such concepts (The President’s Council of
Advisors on Science and Technology, 2014).
These efforts have transformed health care at much larger scales, such as
coordinating operations across an entire hospital system or across a community.
Systems tools and methods, moreover, can be used to ensure that care is reliably
safe, to eliminate inefficient processes that do not improve care quality or people’s
health, and to ensure that health care is centered on patients and their families. (The
President’s Council of Advisors on Science and Technology, 2014)

Applying systems approaches can help healthcare organizations to design their operations with
safety measures, to reduce redundancies, standardize work processes, and account for human
factors (Kaplan et al., 2013). It helps with identifying the infrastructure needed to measure and
monitor the health outcomes and the relationships among different subsystem elements (Kaplan et
al., 2013). In addition, a systems approach fosters continuous improvement tools to help with
handling healthcare’s multifaceted processes (Kaplan et al., 2013).
Following the systems engineering approaches, this study proposes a package of system elements
including HRO principles, Patient Safety factors, and Care Coordination activities combined
together and aims to analyze and assess the impact of this set of system elements on the
performance of healthcare delivery systems. This investigation is to determine whether this
fundamental set can help make improvements in hospitals’ performance in delivering care.
8

HRO principles are a set of five strategies used to achieve and sustain an exceptional safety record
in organizations that deal with hypercomplex and tightly coupled processes where a failure leads
to a catastrophe

and endangers the lives of human beings. Patient Safety measures aim to reduce
medical errors, improve the quality of care that patients receive, and prevent safety breaches. They
are the first step to accomplish efficiency and effectiveness in the system. Care Coordination
activities are a set of actions that involve all the care providers who deliver a service to a patient
in his/her path of receiving medical attention and sharing important clinical information with them
to minimize the danger of care fragmentation. These three elements will further be elaborated on
in detail in the following chapter when the literature review unfolds these concepts.
It is understood that “performance” may be interpreted widely and with many definitions, so with
the help of subject matter experts in the medical field a set of performance measures that can best
describe and represent the performance of a hospital and most other healthcare delivery systems
was identified to be included in this research.
Distinctive features of this study include:
- Evaluating the combination of HRO principles, Patient Safety measures and best practices,
and Care Coordination activities.
- Assessing the effectiveness of merging the three identified elements into one package of
system components.
- Identifying a set of measures that comprehensively characterizes the performance of a
hospital.
- Introducing a methodology to convert the data collected individually for each measure with
specific scales and scopes into a single rating system.
9

- Determining the measures that have most significantly been impacted, positively, as a
result of the proposed system elements’ package.
- Examining the final effect of the proposed package of system elements on the performance
of the hospital.
- Analyzing the obstacles in embedding this package of system elements into the healthcare
delivery system and the opportunities it will bring about.
- Investigating the efforts needed to achieve High Reliability in healthcare settings.
- Identifying the level of endeavors necessary to instill the patient safety measures and best
practices in healthcare delivery system as well as incorporating the concept of serious
reportable events.
- Recognizing the importance of care coordination activities in the final outcomes of care.
Innovation has never been the easiest solution to humans’ problems, but it has been the most
effective. This is an unprecedented and innovative initiative that is introduced in this study, and it
is known that it may be complicated to fully implement and achieve final results that can be
sustained. However, both literature studies and the result of the investigation at the studied
hospital, which will be discussed in the following chapters of this dissertation, lead to the
conclusion that the performance of the healthcare delivery system in many aspects individually
and in the hospital’s overall conduct will improve significantly, the quality of care will elevate,
and many opportunities for future enhancements will flourish as a result of this proposed package
of High Reliability, Patient Safety, and Care Coordination. In addition, the patients’ experiences
and their care outcomes will be revolutionized as a result of much safer, reliable, and highly
coordinated care.

10

III. Literature Review
System approaches have applicability for a variety of issues facing the health and
health care system, including improving patient safety; preventing disease with a
community-based approach; enhancing coordination and communication between
care team members; managing the growing complexity of biomedical evidence and
diagnostic and treatment options; and continually improving the quality, value, and
outcomes of care. (Kaplan et al., 2013)

HRO principles, Patient Safety measures, and Care Coordination activities, packaged together to
form a system element, symbolize a three pillar groundwork for care delivery. The three towers
provide a comprehensive patient centered care system where patients receive reliable and safe care
within and among all providers.

Figure 4: HRO, Patient Safety, and Care Coordination Package

Figure 3: HRO, Patient Safety, and Care Coordination Three Pillar Groundwork

HRO Principles
Care
Coordination
Patient Safety
HRO
Care Delivery System
11

A. High Reliability Organization Principles
High reliability organizations are organizations that are running very complex and tightly coupled
processes (Wilson, Burke, Priest, & Salas, 2005). Failures in their systems can become
catastrophic and endanger the lives of human beings. They need to operate without any error and
require a certain set of guidelines to be implemented in their systems to place safety at the top of
their priorities. Specifically, HROs internal relationships among different departments and entities
and their thoughtful and mindful interrelations of actions separate them from other organizations
(Wilson et al., 2005). They need to develop the capability to handle disasters and put the
resolutions in place while performing in a very dynamic environment. Teams in HROs are
considered High Reliability Teams (HRTs), which have different responsibilities and sets of
authorities compared to teams in ordinary organizations. Honesty, sympathy, dignity, humility,
duty, and trust should be the main characteristics of HRT members.
HROs perform in “hypercomplex” environments where multiple teams need to cooperate together
to ensure safety (Hines, Luna, & Lofthus, 2008). These organizations comprise of “tightly
coupled” teams whose performance affect each other’s (Hines et al., 2008). HROs have each role
and responsibility clearly identified with “hierarchical differentiation” to allow the coordination
of tasks to run smoothly and cohesively while the decision making in the time of a crisis is given
to the most knowledgeable teammate (Hines et al., 2008). There are “multiple levels of decision
making” all connected together to deliver a task in HROs with a need for a clear communication
path to avoid any conflicts (Hines et al., 2008). HROs have a “high degree of accountability” when
an error happens that leads to a catastrophe (Hines et al., 2008). HRT members need to receive
continuous and immediate feedback from each other so they make adjustments as necessary to
prevent faults in their interconnected system (Hines et al., 2008). HROs run under “time
12

constrains” and it certainly affects how staff can complete their tasks, when they need to ask for
help, and when they need to stop and move to other procedures (Hines et al., 2008).
Healthcare organizations are considered to be high reliability as they deal with human life and they
provide services to patients who seek help to achieve a better health status or to remain healthy.
Hence, performing the correct task at the right time with the right team of professionals and
available technology is a necessity not a luxury in healthcare organizations. From delivering minor
preventive care to offering the most complicated and exotic surgeries and treatment options,
healthcare organizations need to put the safety of their patients and their personnel at the top of
their priorities and minimize the probability of harm to them. From an Institute of Medicine (IOM)
report “the failure of a planned action to be completed as intended or the use of a wrong plan to
achieve an aim” is the definition of a Medical Error (Christianson, Sutcliffe, Miller, & Iwashyna,
2011). Sadly, medical errors are quite frequent.
Since industries, such as commercial airlines and nuclear power plants, have achieved much higher
safety standards and were able to sustain their outstanding results, we think there is merit in
learning from them about high reliability rules and implementing their roadmaps in achieving a
high performance by following certain conducts. They have learned to very effectively cope with
hazards, and we can certainly learn from them to become closer to having perfect safety records
in the Healthcare industry. All HROs share the same condition that their environments are very
dynamic and in continuous change. Therefore, they need to continually seek perfection and thrive.
These industries have learned that implementing effective teamwork and practical team behaviors
is a must to achieve safe performance and that sustaining such success requires effective leadership
with a profound understanding of safety culture (Leonard & Frankel, 2011). Effective teamwork,
which is the combination of structured communication, effective critical language, psychological
13

safety, situational awareness, and effective leadership behaviors, forms the infrastructure to ensure
that personnel have the information, resources, tools, and staff necessary to work effectively
together and provide safe care (Leonard & Frankel, 2011).
There are five guidelines described as HRO principles to accomplish outstanding safety levels and
to sustain such records as follows (Chassin & Loeb, 2013; Christianson et al., 2011; Hines et al.,
2008):
- Problem Detection (Christianson et al., 2011; Hines et al., 2008):
 Preoccupation with Failure: using failure and near misses to achieve a better
understanding from strengths and weaknesses of the system (Christianson et al., 2011).
A patient’s perspective can help us improve the system. Error reporting should be
encouraged with a mentality that accepts the idea that human error is inevitable. The
system should be continuously observant of signs hinting of safety hazards (Chassin &
Loeb, 2013).

HRTs should be trained to provide and receive constructive feedback to
quickly learn from mistakes and self-correct (Wilson et al., 2005). They need to
monitor each other’s behaviors and review their performance to fix the flaws and
shortfalls of the teams and system (Hines et al., 2008; Wilson et al., 2005).
 Reluctance to Simplify: avoiding the tendency to minimize the problems (Christianson
et al., 2011). Safety breaches come in various forms and sometimes are hidden in the
complexity of the procedures (Chassin & Loeb, 2013). Therefore, there is a major
difference in the outcome between identifying the error in the early stages vs. detecting
it when it is out of control (Chassin & Loeb, 2013). Due to the complexity of such
organizations, to prevent simplifying the safety issues we should avoid approaches that
only look at problems from one lens and perspective, and try to consider multifaceted
14

tactics. Staff should be encouraged to dig deep down to the roots of safety and reliability
issues and to see if from the patients’ standpoint they received the service that was most
important and vital to them or not. HRTs will collect all relevant information from the
field to adjust their efforts and reallocate their recourses, which requires flexibility that
should exist in HROs. They ensure that all members know their responsibilities and
roles to set appropriate goals and to enhance their performance, but also to stay flexible
enough to adopt to unexpected events and to be able to take on other roles as necessary.
 Sensitivity to Operations: keep the holistic perspective of the system to recognize how
different modules of the system work together and why problems are contagious. Most
of the time failures do not appear suddenly, although they look unexpected. They start
with minor changes in the organization’s operations. Hence, any nonconformity or
abnormality in the system, no matter how small and insignificant, should be reported
and team members should not only be allowed to point out the threats but also feel a
responsibility and obligation to speak up (Chassin & Loeb, 2013). A gap analysis
between work planned and work done will tremendously help with early identification
of potential errors (Conway, 2014). Failures are often the result of multiple errors
occurring at the same time or on a continuous basis. HRT members should transfer
transparent, clear, complete, and accurate information to each other in a closed loop
manner to create a shared mental model of their environment to undertake the best
strategies for dealing with the failures, and in addition for learning purposes for future
situations (Wilson et al., 2005). Information is the critical part of any system, and
especially in healthcare the success of the system and people to achieve a perfect model
depends on it. Hence, poor communication can often result in a higher risk of failures.
15

- Problem management (Christianson et al., 2011; Hines et al., 2008):
 Resilience: developing the capability to deal with sudden adverse events (Christianson
et al., 2011; Hines et al., 2008). Despite the fact that HROs work towards an error-free
environment, safety breaches may occur and it is vital that they do not disable the
system (Chassin & Loeb, 2013). The organization should develop the capability to
identify and contain the error before it spreads or turns into a more dramatic hazard
(Chassin & Loeb, 2013). At the time of a failure, staff will involve patients to perform
a root cause analysis to understand what went wrong, why that happened, and how to
avoid its reoccurrence. HRTs should back each other up and help others in performing
their tasks with both emotional support and physical and technical help in cases of
adverse events. They should continuously provide positive feedback while monitoring
teammates’ performance towards the completion of each task pertaining to one’s role
(Wilson et al., 2005). HRTs are encouraged to ask for help when they are overloaded
because there is nothing more important than the safety of their patients, so nothing
should compromise this goal.
 Deference to Expertise: identify the expertise of the front line staff and leave the
decision making authority to handle adverse events to them no matter where in the
hierarchy of the organizational chart they stand (Christianson et al., 2011; Hines et al.,
2008). HROs should develop the capability to spot the individuals who can best handle
the newly identified threat based on their expertise and let them handle the situation
and the decision making (Chassin & Loeb, 2013). It is crucial to empower front line
personnel and to allow them to make critical decisions at the time of the adverse event,
but also the patients and their families should be engaged in creating a resolution since
16

there is a chance that at the time of the chaos the patients know more about their clinical
history than the staff. HRT members are promoted to speak up and give their opinions,
trigger actions, and propose practical fixes (Wilson et al., 2005; Hines et al., 2008).
IOM published a report in 2004 discussing and proposing the empowerment of nurses,
who are mainly the front line staff, to make decisions in regards to patient care
(Pronovost et al., 2006).
High reliability creates a culture of “high expectations” in running a system and not to accept that
human errors are avoidable. This is partly to recognize unsafe practices and create customized
solutions for each setting. This is a culture of no-tolerance for risky conditions, aggressive
manners, and bullying teammates. The high reliability of a system lies in the empowerment of all
staff regardless of their hierarchical positions.
“Collective Mindfulness”, use of “Robust Process Improvement” tools, and “Culture of Safety”
are the building blocks of high reliability and the great achievements of the above mentioned
principles (Chassin & Loeb, 2011).
Collective mindfulness is the main feature of HROs, which is accomplished through implementing
HRO guidelines
22
. HRT members are aware of the severe consequences that any mistake can bring
about and the deadly implications it may have, and therefore they are extra cautious in regards to
following safety procedures. They continuously look for breaches of safety to fix them in the early
stages when they are very easy to correct. When there are multiple failure modes with small
chances of happening, they still need to be considered and for HRTs to look at the system as a
whole, because those failures can happen at the same time together and fail the system at once.
Such organizations pay serious attention to the lessons learned from failures to analyze the series
of events that happened prior to any adverse event and to detect the flaws in the system that can
17

jeopardize the safety as well as measures and arrangements necessary to prevent future
catastrophes (Chassin & Loeb, 2013).
HROs use Robust Process Improvement (RPI) tools to eliminate safety deficiencies by dissecting
multifaceted safety problems. The Joint Commission Center for Transforming Healthcare
promotes using RPI tools, including Lean, SixSigma, and Change Management processes, which
are systematic approaches to solve complex healthcare problems and to improve the quality and
safety in healthcare (Robust Process Improvement, n.d.). RPI tools can reliably measure the extent
of the problem, identify the root causes, propose solutions for most important causes, prove the
success of the solutions, and ensure sustainable improvements over time (Chassin & Loeb, 2011).
Lean is a tool that focuses on customers (patients) and aims to eliminate waste to create a flow
throughout the value stream (care process). It is generally a low-cost method to enhance the
processes and endorse error-proof systems. Change Management is the ideology to make the
change happen in any organization in the most efficient and successful manner by creating a shared
vision among employees who will be affected by that and to lessen the resistance via effective
communication strategies. Lean-Six Sigma is a blended approach of a mix of Lean and SixSigma
that creates a philosophy whose emphasis is on customers, increasing value, and improving
quality, safety and productivity that has the strengths of both methods (Robust Process
Improvement, n.d.).
Based on Dr. James Reason’s description of the culture of safety organizations, they must collect,
analyze, and publish safety information. Individuals must feel free and comfortable to report safety
issues without fear of blame knowing that management will take actions to fix that concern. So,
the culture of safety resides on trust, reporting safety concerns, and continuous improvement.
Effective interdisciplinary collaboration has a substantial role in improving patient health
18

outcomes in healthcare organizations through empowering the culture of safety (Garman et al.,
2006). This is the result of having trust among team members, mutual respect between staff and
management, having clear communications and strong coordination, considering shared
responsibilities, and transparency. In addition, research has revealed that effective teams with
coordinated functions improve patient safety (Garman et al., 2006).
It is important to mention that implementing the five principles of high reliability by themselves
does not ensure patient safety. They collectively create a state of mindfulness that is needed for
reliability, and this consequently becomes the prerequisite for safety (Hines et al., 2008). Figure 5
is a graphical presentation that demonstrates this relationship (Hines et al., 2008).

Figure 5: Five specific concepts that help create state of mindfulness needed
for reliability, which in turn is prerequisite for safety (Hines et al., 2008)

Although implementing HRO is resource consuming it is evident that the enhanced organization
will benefit from having a higher patient satisfaction ratio, higher employee satisfaction ratio,
better patient health outcomes, and a safer environment; not to mention, that eliminating and
preventing patient safety risks and catastrophes is a considerable achievement.
19

There are certain strategies that HROs can undertake to better improve the performance of their
teams:
- Cross Training: to create a shared mental model among members of different functional
departments so they know what each department does at a high level (Wilson et al., 2005).
- Team Coordination Training: to demonstrate effective team coordination in times of
disaster and to encourage providing back-up support to other team members (Wilson et al.,
2005).
- Team Self-Correction Training: to teach how to provide and accept constructive feedback
and to monitor, identify, and correct each other’s deficiencies (Wilson et al., 2005).
- Scenario-Based Training: to provide the opportunity for HRT members to learn from
embedded trigger events that stimulate certain behaviors and allow them to gain skills on
how to deal with failures (Wilson et al., 2005).
We need to keep in mind that HRTs consist of people from different educational backgrounds and
skills with different levels of understanding. Their experiences shape their view of their
professions and how they interact with others, which impact organizational behaviors (Garman et
al., 2006).
Physicians have chosen a highly competitive career path with long years of study and fellowships
in different hospitals and clinical settings. They are highly goal oriented and focused on their
personal achievements. Their main responsibility is to make a correct and accurate diagnosis and
deliver clinical care, for which they get compensated very well both financially (based on the
volume of patients they have and procedures they perform) and with high prestige comparing to
other clinical staff (Garman et al., 2006). The culture of medical schools generally focuses on
individual contributions and success, and values individualistic competencies rather than team
20

work and cooperation (Garman et al., 2006). Physicians are trained to stay focused and keep
themselves away from patients’ stress and anxiety to be able to perform their tasks precisely, but
this causes emotional distance from patients at times (Garman et al., 2006).
“Dr. Bernard Lown, Nobel Peace Prize Winner: “In my view the lost art of listening
is a quintessential failure of our health care system. I think that you cannot heal the
health care system without restoring the art of listening and of compassion. You
cannot ignore the patient as a human being. A doctor must be a good listener. A
doctor must be cultured in order to understand where the patient lives, why he lives
like that, and also realize that the leading cause of disease in the world is poverty.”
(Brink, 2014)
Nurses generally go through less competitive and less intensive education series compared to
physicians. Their main responsibility is to care for patients and follow physicians’ orders, for
which they mainly get paid on an hourly basis and of course a part of their reward is the admiration
and positive feedback from their patients and other teammates. Their financial compensation was
historically low, which is now getting fixed in many hospitals/clinics; however, since there is a
link between patient safety and working 12+ hour shifts, so it is not recommended to increase their
salary by allowing them to work more, rather than by increasing their hourly rate (Garman et al.,
2006). They are in contact with patients much more than physicians as they need to provide both
clinical and non-clinical care to patients (Garman et al., 2006). Their observations and suggestions
have saved many doctors from making fatal mistakes while they were providing care to their
patients (Altman, 2014).
Considering these differences and with the knowledge that there exist an intimidating behavior
mainly expressed by physicians and received by nurses, a high reliability culture encourages an
environment that opens up lines of communication and empowers all levels of care providers to
express their observations without the fear of becoming the target of disrespectful attitudes. High
reliability promotes a culture in which all care providers aim to avoid risky and unsafe practices
21

with a clear understanding that their jobs will be untouched and even rewarded if they prevent
harm to patients.
Effective care delivery starts with forming powerful teamwork and implementing reliable care
processes and it results in a safe and low risk service. Creating an environment that inspires
effective teamwork needs a cultural shift to allow highly accomplished and talented groups of
people to learn how to collaboratively work towards a goal together in a very dynamic and complex
healthcare system (Leonard & Frankel, 2011).
HRO principles empower the organization to undertake an approach so that they can redirect their
efforts towards potential failures to predict unforeseen adverse events and allocate the necessary
resources to resolve the issues when they occur. Implementing HRO principles helps in responding
to uncertainties that exist and risks that are associated with complex and interconnected systems.
Although applying HRO principles does not eliminate the treats to the system, it certainly helps in
carrying out a positive and effective response to such adverse events. In addition, implementing
HRO principles creates and embeds a culture of safety in the heart and foundation of the
organization. This is of especially high importance in healthcare organizations since there are less
regulatory audits in regards to safety and reliability in healthcare and generally healthcare
organizations lack care coordination in the current decentralized system, not to mention the
continuous change of staff due to shifts and temporary jobs (Christiansin et al., 2011). Healthcare
staff also suffer from having to make decisions with either no information about the patients or
sometimes too much information to rely on in emergency cases (Christiansin et al., 2011).
Leadership skills in both categories of clinical leadership, as well as administration leadership,
although of tremendous importance in achieving high performance in an organization, is not taught
in healthcare. However the need for such teamwork practices is evident. The leaders of healthcare
22

organizations have to be the role model and put effective teamwork, quality, and safety on top of
their priorities and convey the same message to front line staff. In addition, setting organization’s
standards in team behaviors, such as organizational fairness, can help create a culture that promotes
healthy teamwork and, furthermore, reduces the hierarchical distance between personnel to
encourage them to speak up freely and make the leaders more approachable (Leonard & Frankel,
2011).
In healthcare organizations, executives’ leadership styles in response to disasters, how they
allocate priorities to different tasks, where patients rank among their organizational goals and
objectives, the culture of safety, and their learning and continuous improvement methodologies,
can distinguish the system’s efficiency in responding to adverse events. Not having the capability
of acting on time, quickly, and in an effective manner can put the organization in a very devastating
status. For healthcare organizations to better achieve HRO goals there is a need for leadership
commitment to zero quality failures, implementation of effective process improvement methods
in the system, and the enactment of the culture of safety (Chassin, 2012). To integrate the principles
of high reliability we need to develop the right culture, hire and retain the right people, put in place
the right processes, and provide personnel with the right tools and equipment (Gauthier, Davis, &
Schoenbaum, 2006).
Among all the efforts required towards assimilating HRO principles in the heart of any healthcare
organization one should consider that a culture that represents mutual respect plays a vital role.
People should feel safe to raise their concerns when they catch a safety breach knowing that it will
be taken seriously and that they will not be under any resentment from other coworkers. This is
how truly effective teamwork can become feasible (Leonard & Frankel, 2011).
23

In spite of having great knowledge about the concept of high reliability, theoretically and as
explained above, and despite all the efforts towards safer care in the 18 years after the publication
of “The Err is Human”, still millions of Americans are harmed in our healthcare delivery system
and suffer from hospital acquired infections, medication errors, and many more injuries while they
receive care (Aspden, Wolcott, Bootman, & Cronenwett, 2007; Klevans et al., 2007; Chassin &
Loeb, 2013).
Even a transition among different healthcare settings can cause danger and about 1 in 5 patients
suffer from a type of adverse event while transitioning from hospital to their home (Bodenheimer,
2008; Forster, Murff, Peterson, Gandhi, & Bates, 2003). Still, in this era, wrong side surgeries and
surgeries on wrong patients are as common as fifty times each week in the U.S. (estimated from:
Minnesota Department of Health 2013; Chassin & Loeb, 2013). Even scenarios of fire breaking
out in surgery suites has been observed nearly 600 times in a year with the major side effects and
injuries to the patients (ECRI Institute, 2013; Chassin & Loeb, 2013).
This lead to the fact that High Reliability is not a state to be achieved at once, rather this is a form
of maturity in the organizational context that needs to be gradually reached and that “one-size-fits-
all” solutions need to be customized. It is not as simple for hospitals to suddenly adopt HRO
principles and let the new system take over, but first the capacity to make certain changes needs
to be made available. May be the first step is to recognize where the hospital is standing with
respect to the reliability status and then to plan for milestones to be achieved (Chassin & Loeb,
2013).
The below triangle illustrates the building blocks to achieving high reliability based on The Joint
Commission guidelines identifying the three domains of Leadership, Safety Culture, and Robust
24

Process Improvement as components of change (Joint Commission Resources Quality & Safety
Network Resource Guide, 2011):

Figure 6: Building Blocks of High Reliability
(Joint Commission Resources Quality & Safety Network Resource Guide, 2011)

The three domains of change start with the leadership’s commitment to zero harm. This is the
initial and most critical step, without which the remaining changes will not follow successfully.
Everyone involved in the formal and informal leadership of a hospital should be united for a “Zero
Harm to Patient” goal and should not be satisfied with anything below that level (Chassin & Loeb,
2013).
The second change is to implement a practical and purposeful safety culture throughout the
organization that supports high reliability and prevents any harm to the patients, where there is
nothing above the safety of patients and personnel (Chassin & Loeb, 2013).
Lastly the final change comes with the extensive incorporation of process improvement tools with
lean, six sigma, and change management, as described earlier, to help healthcare systems form
25

safety processes that do not fail and to additionally assist with the root cause analysis at the time
of adverse events or close calls (Chassin & Loeb, 2013).
Based on the same body of knowledge, Dr. Chassin and Dr. Loeb have developed a framework for
improvements to help hospitals make changes step-by-step and move towards the maturity model
with High Reliability (Chassin & Loeb, 2013). This practical framework represents the same
triangle with the fourteen identified healthcare components that are linked to a better quality
performance, as based on the literature, for the four phases of Beginning, Developing, Advancing,
and Approaching a high reliability status for each of the three domains of change. The fourteen
healthcare components comprise Board, CEO/Management, Physicians, Quality Strategy, Quality
Measures, Information Technology, Trust, Accountability, Identifying Unsafe Conditions,
Strengthening Systems, Assessment, Methods, Training, and Spread. The three domains of change,
as mentioned above, consist of Leadership, Culture of Safety, and Process Improvement
(Goeschel, Wachter, & Pronovost, 2010; Jha & Epstein, 2010; Weiner, Shortell, & Alexander,
1997; Chassin & Loeb, 2013).
The following table (Table 1) represents these four stages of maturity for each of the identified
components that would define progress toward high reliability (Chassin & Loeb, 2013). On the
left side of the table, the Harm Reduction Measures of this study are included. They will further
be elaborated on in the next chapter. Due to the size of the table it has been broken into multiple
pages but they all form one table conceptually.
26

Table 1: High Reliability Health Care Maturity Mode

High Reliability Health Care Maturity Model: Practical Framework for Improvement
(Extracted from Chassin and Loeb, 2013)
Organizational Maturity Components

Leadership

Board CEO/Management
Harm Reduction
Measures
Maturity Stages Maturity Stages

Beginning Developing Advancing Approaching Beginning Developing Advancing Approaching
Harm Reduction Measures
Safety
Postop Sepsis
Board’s
quality
focus is
nearly
exclusively
on
regulatory
compliance
.
Full
board’s
involvemen
t in
quality is
limited to
hearing
reports
from its
quality
committee.
Full board
is engaged
in the
developme
nt of
quality
goals and
approval of
a
quality plan
and
regularly
reviews
adverse
events and
progress on
quality
goals.
Board
commits to
the goal of
high
reliability
(i.e., zero
patient
harm) for all
clinical
services.
CEO/mana
gement’s
quality
focus is
nearly
exclusively
on
regulatory
compliance
.
CEO
acknowled
ges need
for
plan to
improve
quality and
delegates
the
developme
nt
and
implementa
tion of a
plan to a
subordinate
.
CEO leads
the
developme
nt
and
implementa
tion of a
proactive
quality
agenda.
Management
aims for zero
patient
harm for all
vital clinical
processes;
some
demonstrate
zero or near-
zero rates of
harm.
Falls Rate
Effectiveness
and Mortality
Mortality
Efficiency and Patient
Centeredness
7 Day Readmission Rate
Rate Hospital
Communication with Doctors
Communication About
Medicines
Care Transition

27

Physicians Quality Strategy

Beginning Developing Advancing Approaching Beginning Developing Advancing Approaching
Harm Reduction Measures
Safety
Postop Sepsis
Physicians
rarely lead
quality
improveme
nt
activities;
overall
participatio
n by
physicians
in these
activities
is low.
Physicians
champion
some
quality
improveme
nt
activities;
physicians
participate
in these
activities
in some
areas but
not widely.
Physicians
often lead
quality
improveme
nt
activities;
physicians
participate
in
these
activities in
most
areas, but
some
important
gaps
remain.
Physicians
routinely lead
clinical
quality
improvement
activities
and accept
the leadership
of
other
appropriate
clinicians;
physicians’
participation
in
these
activities is
uniform
throughout
the
organization.
Quality is
not
identified
as a
central
strategic
imperative.
Quality is
one of
many
competing
strategic
priorities.
Quality is
one of the
organizatio
n’s top
three or
four
strategic
priorities.
Quality is the
organization’
s
highest-
priority
strategic goal.
Falls Rate
Effectiveness
and Mortality
Mortality
Efficiency and Patient
Centeredness
7 Day Readmission Rate
Rate Hospital
Communication with Doctors
Communication About
Medicines
Care Transition

28

Quality Measures Information Technology

Beginning Developing Advancing Approaching Beginning Developing Advancing Approaching
Harm Reduction Measures
Safety
Postop Sepsis
Quality
measures
are not
prominentl
y displayed
or
reported
internally
or
publicly;
the only
measures
used are
those
required by
outside
entities and
are not
part of
reward
systems.
Few quality
measures
are
reported
internally;
few or
none are
reported
publicly
and are not
part of
reward
systems.
Routine
internal
reporting of
quality
measures
begins,
with the
first
measures
reported
publicly
and the
first quality
metrics
introduced
into staff
reward
systems.
Key quality
measures are
routinely
displayed
internally and
reported
publicly;
reward
systems for
staff
prominently
reflect the
accomplishm
ent of
quality goals.
IT provides
little or no
support
for quality
improveme
nt.
IT supports
some
improveme
nt
activities,
but
principles
of safe
adoption
are not
often
followed.
IT
solutions
support
many
quality
initiatives;
the
organizatio
n commits
to
principles
and the
practice
of safe
adoption.
Safely
adopted IT
solutions are
integral to
sustaining
improved
quality.
Falls Rate
Effectiveness
and Mortality
Mortality
Efficiency and Patient
Centeredness
7 Day Readmission Rate
Rate Hospital
Communication with Doctors
Communication About
Medicines
Care Transition

29

Safety Culture

Trust Accountability

Beginning Developing Advancing Approaching Beginning Developing Advancing Approaching
Harm Reduction Measures
Safety
Postop Sepsis
Trust or
intimidatin
g
behavior is
not
assessed.
First codes
of behavior
are
adopted in
some
clinical
department
s.
CEO and
clinical
leaders
establish a
trusting
environme
nt for all
staff by
modeling
appropriate
behaviors
and
championin
g
efforts to
eradicate
intimidatin
g
behaviors.
High levels of
(measured)
trust
exist in all
clinical areas;
self-policing
of codes of
behavior is in
place.
Emphasis
is on
blame;
discipline
is not
applied
equitably
or with
transparent
standards;
no
process
exists for
distinguishi
ng
“blameless
”
from
“blamewort
hy” acts.
The
importance
of equitable
disciplinary
procedures
is
recognized,
and some
clinical
department
s adopt
these
procedures.
Managers
at all levels
accord
high
priority to
establishin
g all
elements of
safety
culture;
adoption of
uniform
equitable
and
transparent
disciplinary
procedures
begins
across
the
organizatio
n.
All staff
recognize and
act on
their personal
accountability
for
maintaining a
culture of
safety;
equitable and
transparent
disciplinary
procedures
are fully
adopted
across the
organization.
Falls Rate
Effectiveness
and Mortality
Mortality
Efficiency and Patient
Centeredness
7 Day Readmission Rate
Rate Hospital
Communication with Doctors
Communication About
Medicines
Care Transition

30

Identifying Unsafe Conditions Strengthening Systems

Beginning Developing Advancing Approaching Beginning Developing Advancing Approaching
Harm Reduction Measures
Safety
Postop Sepsis
Root cause
analysis is
limited
to adverse
events;
close calls
(“early
warnings”)
are not
recognized
or
evaluated.
Pilot “close
call”
reporting
programs
begin in
few
areas; some
examples
of
early
interventio
n to
prevent
harm can
be found.
Staff in
many areas
begin to
recognize
and report
unsafe
conditions
and
practices
before they
harm
patients.
Close calls
and unsafe
conditions
are routinely
reported,
leading
to early
problem
resolution
before
patients are
harmed;
results are
routinely
communicate
d.
Limited or
no efforts
exist to
assess
system
defenses
against
quality
failures and
to remedy
weaknesses
.
RCAs
begin to
identify the
same
weaknesses
in system
defenses in
many
clinical
areas, but
systematic
efforts
to
strengthen
them are
lacking.
System
weaknesses
are
cataloged
and
prioritized
for
improveme
nt.
System
defenses are
proactively
assessed, and
weaknesses
are
proactively
repaired.
Falls Rate
Effectiveness
and Mortality
Mortality
Efficiency and Patient
Centeredness
7 Day Readmission Rate
Rate Hospital
Communication with Doctors
Communication About
Medicines
Care Transition

31

Assessment

Beginning Developing Advancing Approaching

Harm Reduction Measures
Safety
Postop Sepsis
No
measures
of safety
culture
exist.
Some
measures
of safety
culture are
undertaken
but
are not
widespread
; little if
any attempt
is made to
strengthen
safety
culture.
Measures
of safety
culture are
adopted
and
deployed
across
the
organizatio
n; efforts to
improve
safety
culture are
beginning.
Safety culture
measures are
part of
the strategic
metrics
reported to
the board;
systematic
improvement
initiatives are
under way to
achieve a
fully
functioning
safety culture.

Falls Rate

Effectiveness
and Mortality
Mortality

Efficiency and Patient
Centeredness
7 Day Readmission Rate

Rate Hospital

Communication with Doctors

Communication About
Medicines

Care Transition

32

Performance Improvement

Methods Training

Beginning Developing Advancing Approaching Beginning Developing Advancing Approaching
Harm Reduction Measures
Safety
Postop Sepsis
Organizatio
n has not
adopted a
formal
approach to
quality
manageme
nt.
Exploration
of modern
process
improveme
nt
tools
begins.
Organizatio
n commits
to adopt the
full suite
of Robust
Process
Improveme
nt (RPI)
tools.
Adoption of
RPI tools is
accepted fully
throughout
the
organization.
Training is
limited to
compliance
personnel
or
to the
quality
department.
Training in
performanc
e
improveme
nt tools
outside the
quality
department
is
recognized
as critical
to success.
Training of
selected
staff
in RPI is
under way,
and a plan
is in place
to broaden
training.
Training in
RPI is
mandatory
for all staff,
as appropriate
to
their jobs.
Falls Rate
Effectiveness
and Mortality
Mortality
Efficiency and Patient
Centeredness
7 Day Readmission Rate
Rate Hospital
Communication with Doctors
Communication About
Medicines
Care Transition

33

Spread

Beginning Developing Advancing Approaching

Harm Reduction Measures
Safety
Postop Sepsis
No
commitme
nt to
widespread
adoption of
improveme
nt methods
exists.
Pilot
projects
using
some new
tools are
conducted
in a few
areas.
RPI is used
in many
areas to
improve
business
processes
as
well as
clinical
quality
and safety;
a positive
ROI is
achieved.
RPI tools are
used
throughout
the
organization
for all
improvement
work;
patients
are engaged
in
redesigning
care
processes,
and RPI
proficiency is
required for
career
advancement.

Falls Rate

Effectiveness
and Mortality
Mortality

Efficiency and Patient
Centeredness
7 Day Readmission Rate

Rate Hospital
Communication with Doctors
Communication About
Medicines

Care Transition

34

The potential application and contribution of this table is that healthcare delivery system
administrators can use it to assess their system and determine where they stand as far as the
maturity levels of high reliability. The addition of the harm reduction measures, which can vary
for different hospitals to some extent in the same context with HRO maturity phases allows the
decision makers to envision how these factors can interact and influence each other and enable
executives to qualitatively identify the interdependencies of harm reduction measures and maturity
phases.
There does not exist a hospital yet that has achieved a high reliability full maturity status for all
their activities (Chassin & Loeb, 2013). This framework was developed as a result of a very
comprehensive study and the efforts of many experts in the field of high reliability as well as the
healthcare industry and many lessons learned from other HRO industries. It has been tested in
hospital settings and has been greatly studied for accuracy and completeness. The Joint
Commission has offered a tool to hospitals called ORO
TM
2.0 as a resource for High Reliability
Assessment based on the same framework as shown in the table above.

B. Patient Safety
There has been a great devotion to reducing medical errors and improving patient safety,
particularly in the last 18 years after “The Err is Human” report published by the IOM, and
therefore this subject has become the main topic of care improvement endeavors. There has been
lots of efforts to make the underlying causes of safety breaches apparent, which is a great strategy
for increasing public awareness, but still due to lack of systematic approaches and other
fundamental changes that need to be resolved in advance, such as professional organizational
35

attitudes in the medical field, it is not an easy task to implement the many great available safety
interventions and to embed them into the foundation of the U.S. health system.
Patient safety must become the major priority in healthcare delivery and the root of accomplishing
efficiency and effectiveness (Hines et al., 2008). If it would be the other way around, then there is
a high chance that we create a very efficient and effective system that consists of highly unsafe
processes (Hines et al., 2008). Therefore, only when safety becomes the principle will the
relationship among safety, efficiency, and effectiveness results in a positive outcome (Hines et al.,
2008).
The National Quality Forum has indicated that based on the statistics they obtained, approximately
two million healthcare associated infections occur every year, which causes about 90,000 deaths
and more than $4.5-5.7 billion in hospital healthcare costs. The IOM has reported that about
400,000 preventable adverse drug events occur annually in US hospitals, another 800,000 in long-
term care, and more than 500,000 among Medicare patients in outpatient settings. Lengthy hospital
stays that are unnecessary cause additional illness, pain, and emotional distress for patients and
their families. Healthcare associated infections and preventable medication errors are just the two
categories of patient safety breaches. Addressing these issues will lead to an affordable, effective,
and equitable care delivery process. According to this data and much more information acquired
for other preventable causes of death in the U.S., the National Quality Strategy identified the
following three roadmaps to redirect the priorities to create safer care (National Quality Forum
Patient Safety, n.d.; National Quality Forum Patient Safety 2015):
1. Reduce preventable hospital admissions and readmissions.
2. Reduce the incidence of adverse healthcare-associated conditions.
3. Reduce harm from inappropriate or unnecessary care. (National Quality
Forum Patient Safety, n.d.; National Quality Forum Patient Safety 2015)
36

The National Quality Forum in a collaboration with the Agency for Healthcare Research and
Quality (AHRQ, 2005) has selected 34 Safe Practices and 29 Serious Reportable Events. They
collectively along with the National Quality Forum ‘Patient Safety Measures’ help to improve
patient safety performance in the US by reducing adverse events and medical errors (National
Quality Forum Patient Safety, 2015).
These practices and measures have been selected based on their generalizability, effectiveness in
reducing harm, usability, and their benefits to patient safety. Additionally The Joint Commission
in their 2013 annual report on quality and safety “Improving America’s Hospitals” stated that their
achievement in improving quality of care and patient safety has been the result of hospitals’
mindfulness and sharing of common measures, goals, and solutions. Moreover, they stated that
hospitals strive to always do the right thing at the right time for patients and try to constantly
improve the quality and safety of healthcare (The Joint Commission, 2013). This is indeed a great
accomplishment to collect and track the data and provide a comparison tool for both patients and
physicians as well as creating a healthy competitive environment in terms of providing safer
settings for all entities involved.
Still in this era, many routine safety practices fail, like hand hygiene, medication administration,
or communication in care transition, which are completely preventable (Adverse Events after
Hospital Discharge, n.d.; Chassin, 2012). There is a lack of a unified definition on what should be
considered “harm” to patients, and that has caused unclear estimates of how many patients have
been hurt due to avoidable mistakes through the care processes and this impedes the improvement
efforts. Safety investigations in the past decade have shown that 20 to 30% of hospitalized patients
experience harm during their stay (Kaplan et al., 2013).
37

In addition to the lack of standardization, the care process is a complex task that increases the
probability of human errors since there are many participants simultaneously providing multiple
levels of care. New technologies, such as Electronic Health Records and advanced medical
equipment, on the one hand have reduced human errors by minimizing the complexity of the care
process and introducing automation to healthcare. On the other hand, they need precise
implementation and execution alongside several layers of training for clinical personnel to be
successful in enhancing patient safety measures.
To be able to create a safe environment for patients and personnel, three main factors need to be
considered (Wilson et al., 2005):
- Organization Factors that are the basis of having a deep understanding and appreciation to
reach high reliability status (Wilson et al., 2005). They create the infrastructure of high
reliability.
- Team Factors to train and encourage individuals to act as high reliability teams (Wilson et
al., 2005). They promote team work approaches and transparency in communication.
- Developmental Strategies that help to convert the organization into a high reliability entity
smoothly with minimum disruption (Wilson et al., 2005). Wilson et al. (2005) in their
“Promoting healthcare safety” paper have very eloquently put together Figure 7 that shows
a structure for promoting safety in the workplace.
38

Figure 7: Workplace Safety (Wilson et al., 2005)

Looking into human factors can identify safety, quality, and reliability challenges by
understanding how humans interact with technologies and processes (Kaplan et al., 2013). In
typical healthcare organizations safety is defined and measured more by its absence than its
presence. Safety defenses protect the healthcare system and at the same time they can cause
catastrophic failures if they are not being implemented correctly (Reason, 2000). Many healthcare
organizations try to minimize the inconsistency in their personnel’s performances to reduce the
human error, but this strategy tightens their hands when they need to very quickly and flexibly
respond to an unexpected event and make necessary changes to maintain the safety of patients in
the healthcare system’s dynamic environment. Sometimes there is a belief in absolute safety that
is not practical but a preoccupation with failure (as mentioned in the HRO principles) can lead to
higher reliability of actions.
39

A noteworthy point to consider is that physicians generally become very task and goal oriented.
Consequently, unless a task is directly impacting their professional pathway (monetary rewards,
fame, and receiving acknowledgements by peers) they do not see it as necessary or important
enough to act upon. Patient safety interventions are not yet being seen as important to many
physicians as they should be. Until it becomes a matter of life and death for patients, at which time
it is sadly too late, some physicians do not take any actions. Furthermore, the medical atmosphere
does not commonly promote or reward the reporting of unsafe practices and, hence, many,
especially in lower hierarchical levels (nurses, interns, and residents), would rather drop the issue.
The National Quality Forum has proposed the following categories of safe practices to enhance
patient safety (AQRH, 2005; NQF Safe Practices Consensus Committee, 2010). The validity of
these points has been proven in other safety literature both in healthcare and other high reliability
organizations.
- Creating Safety Culture.
- Matching healthcare needs with service delivery capabilities.
- Enabling information transfer and Promoting clear communication.
- Implementing and embracing safe practices.
- Increasing safe medication use. (AQRH, 2005)

The comprehensive list of the National Quality Forum top 30 safe practices is as follows (AQRH,
2005; NQF Safe Practices Consensus Committee, 2010):
1. Create a healthcare culture of safety.
2. For designated high-risk, elective surgical procedures or other specified care,
patients should be clearly informed of the likely reduced risk of an adverse outcome
at treatment facilities that have demonstrated superior outcomes and should be
referred to such facilities in accordance with the patient’s stated preference.
3. Specify an explicit protocol to be used to ensure an adequate level of nursing
based on the institution’s usual patient mix and the experience and training of its
nursing staff.
40

4. All patients in general intensive care units (both adult and pediatric) should be
managed by physicians having specific training and certification in critical care
medicine ("critical care certified").
5. Pharmacists should actively participate in the medication-use process, including,
at a minimum, being available for consultation with prescribers on medication
ordering, interpretation and review of medication orders, preparation of
medications, dispensing of medications, and administration and monitoring of
medications.
6. Verbal orders should be recorded whenever possible and immediately read back
to the prescriber.
7. Use only standardized abbreviations and dose designations.
8. Patient care summaries or other similar records should not be prepared from
memory.
9. Ensure that care information, especially changes in orders and new diagnostic
information, is transmitted in a timely and clearly understandable form to all of the
patient’s current healthcare providers who need that information to provide care.
10. Ask each patient or legal surrogate to recount what he or she has been told
during the informed consent discussion.
11. Ensure that written documentation of the patient's preference for life-sustaining
treatments is prominently displayed in his or her chart.
12. Implement a computerized prescriber order entry system.
13. Implement a standardized protocol to prevent the mislabeling of radiographs.
14. Implement standardized protocols to prevent the occurrence of wrong-site
procedures or wrong-patient procedures.
15. Evaluate each patient undergoing elective surgery for risk of an acute ischemic
cardiac event during surgery, and provide prophylactic treatment of high-risk
patients with beta blockers.
16. Evaluate each patient upon admission, and regularly thereafter, for the risk of
developing pressure ulcers. This evaluation should be repeated at regular intervals
during care. Clinically appropriate preventive methods should be implemented
consequent to the evaluation.
17. Evaluate each patient upon admission, and regularly thereafter, for the risk of
developing deep vein thrombosis (DVT)/venous thromboembolism (VTE). Utilize
clinically appropriate methods to prevent DVT/VTE.
18. Utilize dedicated anti-thrombotic (anti-coagulation) services that facilitate
coordinated care management.
19. Upon admission, and regularly thereafter, evaluate each patient for the risk of
aspiration.
20. Adhere to effective methods of preventing central venous catheter-associated
blood stream infections.
41

21. Evaluate each pre-operative patient in light of his or her planned surgical
procedure for the risk of surgical site infection, and implement appropriate
antibiotic prophylaxis and other preventive measures based on that evaluation.
22. Utilize validated protocols to evaluate patients who are at risk for contrast
media-induced renal failure, and utilize a clinically appropriate method for reducing
risk of renal injury based on the patient’s kidney function evaluation.
23. Evaluate each patient upon admission, and regularly thereafter, for risk of
malnutrition. Employ clinically appropriate strategies to prevent malnutrition.
24. Whenever a pneumatic tourniquet is used, evaluate the patient for the risk of an
ischemic and/or thrombotic complication, and utilize appropriate prophylactic
measures.
25. Decontaminate hands with either a hygienic hand rub or by washing with a
disinfectant soap prior to and after direct contact with the patient or objects
immediately around the patient.
26. Vaccinate healthcare workers against influenza to protect both them and
patients from influenza.
27. Keep workspaces where medications are prepared clean, orderly, well lit, and
free of clutter, distraction, and noise.
28. Standardize the methods for labeling, packaging, and storing medications.
29. Identify all "high alert" drugs (e.g., intravenous adrenergic agonists and
antagonists, chemotherapy agents, anticoagulants and anti-thrombotics,
concentrated parenteral electrolytes, general anesthetics, neuromuscular blockers,
insulin and oral hypoglycemics, narcotics and opiates).
30. Dispense medications in unit-dose or, when appropriate, unit-of-use form,
whenever possible. (AQRH, 2005)

The National Quality Forum leads a “Patient Safety” project that addresses the following areas
and more (National Quality Forum Patient Safety 2015):
- Fall screening and risk management.
- Medication reconciliation.
- Measures from applicable settings, such as skilled nursing facilities and inpatient
rehabilitation facilities.
42

- Unplanned admission-related measures from other settings (i.e., hospitalization for patients
on dialysis).
- All-cause and condition specific admission measures.
- Condition-specific readmissions measures.
- Measures examining length of stay.
Michael Leonard, a physician leader for patient safety at Kaiser Permanente in Colorado, offers
the following definition of a culture of safety: “No one is ever hesitant to speak up regarding the
well-being of a patient, and everyone has a high degree of confidence that their concern will be
heard respectfully and be acted upon” (Conway, Federico, Stewart, & Campbell, 2010).
Healthcare organizations with a robust safety culture are much better at handling catastrophic
events effectively. The principles of the culture of safety allow all team members to help in
preventing and eliminating adverse events in the first place and acting upon them very quickly
when they happen (Singer et al., 2003). A safety culture also helps the communication route to
pass the information speedily from patients to different levels of the organization since
transparency is highly promoted. In a health system with a strong safety culture patients and their
involved family members receive the support they need. In such cases the resolutions will occur
when both patients and clinical team members are treated respectfully without putting the blame
on one’s shoulder.
There are many elements, including patients, providers, tasks, technology and tools, team,
environment, and organizational factors, that form the overall culture of safety (Singer et al., 2003).
The safety culture influences not only the organization’s response to catastrophes but also how the
organization avoids errors in the first place. In brief, to successfully implement the culture of safety
in health organizations the following few points at a minimum need to be considered:
43

- To introduce a systems perspective.
- To obtain leadership’s commitment to zero quality failure.
- To set achievable milestones and stick to the plan.
- To provide continuous training to remind staff about the importance of patient safety and
how to use available tools and techniques.
- To promote a “team approach” towards the patient safety subject and encouraging
communication among all different levels of personnel.
- To inspire safety initiatives, such as “Reporting System”, “Patient Safety Survey”, and
“Safety Audits”, besides all safety measures in place.
- To allocate time and financial resources to “Safety Culture Training”.
- To keep the communication lines open and reassure all staff to point out non-compliance
cases.
- To engage and empower patients and their families in the care process and decision
making.
- To identify the Patient Safety Responsibility hierarchy that may not be one to one
corresponding to the organizational ladder.
- To remove barriers, such as shortage of personnel, tools, equipment, etc.
- To incentivize those staff members who report safety issues, mitigate potential safety
failures, invent new methodologies to better carry out the care process to reduce unsafe
practices, and those who improve the care process to avoid safety breaches.
We need to be able to measure patient safety defects to monitor the reliability of our system and
that means that we should be able to measure the number of patient safety failure cases with the
exact number of patients who were harmed to the total population at risk and of course the severity
44

is also important (Pronovost et al., 2006). If we do not know what has gone wrong precisely, we
will not be able to fix it. This does not mean that we should not consider the non-rate-base measures
since they also impact patient safety. Pronovost et al. (2006) have introduced a framework that
beautifully distinguishes the differences between these categories of Rate-Based and Non-Rate-
Based measures:
- Rate-Based Measures (Pronovost et al., 2006):
 Outcome Measure: The result achieved; how often do we harm patients?
 Process Measure: What we do; how often do we use evidence-based medicine?
- Non-Rate-Based Measures (Pronovost et al., 2006):
 Structural Measures: How care is organized; how do we know we learned from our
mistakes?
 Context Measures: The context in which care is delivered; how well have we
created the culture of safety?
Non-rate-based measures are not related to a specific discipline while rate-based measures are
particularly designed for each sector of healthcare (Pronovost et al., 2006).

C. Care Coordination
The U.S. healthcare system has been battling fragmentation for a long time. This issue has forced
serious risks into the Patient-Centered care delivery system. Care coordination is a set of activities
that minimizes the dangers of fragmentation by ensuring that all providers involved in patient’s
care share important clinical information and have clear shared expectations about their roles. Care
coordination is a pathway to improve the population’s health, enhance patient experience, and
reduce the rate of increase in the cost of care (Safety Net Medical Home Initiative, 2013).
45

An example of a succesfull care coordination process can be found in the CareOregon project.
CareOregon is a Portland Oregon-based nonprofit Medicaid health plan. They developed two
innovative programs to help optimize care for their enrollees: a patient-centered medical home
initiative in safety-net clinics and a multidisciplinary case management program for members at
high risk of poor health outcomes. To implement these programs, the health plan emphasized the
use of learning communities through which independent providers could acquire, share, and
practice techniques to improve population health and enhance the patients’ experience.
Meanwhile, they could decrease the cost also. They started in 2004 by providing a centralized case
management system and care coordination service to patients with a highest risk for poor health
outcomes. Multidisciplinary care management teams operating from CareOregon’s Portland
headquarters helped these patients find critical community-based resources, resolve difficult
behavioral issues and self-management problems, and improve their ability to follow a treatment
plan (Klein and McCarthy, 2010).
In the CareOregon intervention, the CareSupport program yielded savings of $400 per member
per month (or $5,000 per year) in the year following a member’s enrollment, while maintaining or
slightly improving the patients’ quality of life (Klein and McCarthy, 2010). In addition to the
implementation of patient-centered medical homes in safety-net clinics, pilot sites have been
associated with greater continuity of care and improved health screening and chronic care
management. They observed a 7% increase in the proportion of patients with controlled blood
pressure and diabetes over one year, with the best-performing clinics exceeding national
benchmarks (Klein and McCarthy, 2010).
CareOregon was able to achieve the following results with their care coordination model (Klein
and McCarthy, 2010):
46

- 10.8% increase in the proportion of diabetic patients receiving HbA1c testing to measure
their blood sugar control.
- 7.6% increase in the proportion of diabetic patients with blood sugar under control (HbA1c
<8).
- 7.6% increase in the proportion of hypertensive patients with blood pressure under control
(<140/90).
- 3.4% increase in the proportion of female patients screened for cervical cancer (pap test
within three years) among five clinics.
- 12.2% increase in the proportion of young children who were up-to-date on immunizations
at one clinic.
- More than threefold increase in the proportion of patients screened for depression within
one year.
CareOregon surveyed their patients in the medical homes to assess their experience of care. The
mean score for healthcare teams, on a scale of 0 to 10 (with 10 being the best care-team possible),
was 8.41 (Klein and McCarthy, 2010). When asked whether they “usually or always” got all
aspects of patient-centered care, 80.2% of patients responded yes, while 19.8% responded no
(Klein and McCarthy, 2010).
The case of CareOregon care coordination model is a good example from the perspective that they
had to deal with sicker patients: many patients had complex health problems. Almost 66% of adult
members suffered from at least one of 12 common chronic health conditions, such as diabetes,
depression, or chronic heart failure. Almost 30% of adults suffered from three or more of these
chronic conditions. Many of these conditions are exacerbated by psychosocial difficulties, such as
unsafe housing, emotional distress, and language barriers. 33% of members do not speak English
47

as a first language. These complex needs often manifest in frequent use of the hospital emergency
department. 8% of CareOregon’s Medicaid members made four or more visits to the ER in the last
12 months prior to the study (Gindi, Cohen, & Kirzinger, 2012). Caring for this population often
requires providers to devote time and resources to understanding the barriers this population face
in adhering to treatment plans and to address social needs that fall outside of the traditional scope
of medical care. Without such efforts, vulnerable patients may suffer adverse health outcomes
(Ignagni, 2010).
Following CareOregon, the Southcentral Foundation started using a similar care coordination
model for Alaskan patients and the results were extremely promising (Klein and McCarthy, 2010):
- Urgent care and emergency department visits dropped by more than 40%.
- Daytime urgent care visits dropped from 82 visits per 1,000 to 27. The visits declined by
40% in the first two years, which Douglas Eby, M.D., M.P.H., SCF’s vice president of
medical services, attributed to the implementation of same-day access and a well-designed
primary care platform.
- Primary care visits to the clinic have decreased by 20% while phone and e-mail clinical
interactions have increased significantly.
- Hospital days and admissions decreased by 25% within the first three years of the program.
- Visits to specialists dropped by 60%.
- A survey of patients in each of the recent years found that 91% of patients rated their overall
care favorably continuously.
- By implementing same-day access, the backlog of patients waiting for behavioral health
services dropped from 1,300 to nearly zero in 2007.
48

- More than three-quarters of the quality measures were at or above the 75th percentile on
Healthcare Effectiveness Data and Information Set (HEDIS) results reported by the
National Committee for Quality Assurance.
The basis of care coordination activities follow the model presented below to develop a sense of
community and collegiality among health providers and participants to help spread the best
practices (Safety Net medical Home Initiative, 2013):

Figure 8: Elements of Successful Care Coordination (Safety Net Medical Home Initiative, 2013)

The result of this model of care coordination includes, but is not limited to, having reduced the
length of stay in hospitals, lowered operation costs, deceased adverse events, and improved
patients’ health outcomes.
Figure 9: Care Coordinated Model (Safety Net Medical Home Initiative, 2013)

Develop
Relationship
Establish
Connectivity
Assume
Accountability
Support the
Patient
49

IV. Methodology
A. Research Design
The literature suggested that HRO principles, Patient Safety factors, and Care Coordination
activities would in fact enhance and improve the performance of the care delivery systems
specifically in hospital settings, which are the main point of providing care. To also statistically
test for the validity of the suggested outcomes from the literature review, this package of system
elements was tested in a real life situation.
The research question of this study is:
Does addressing HRO principles, Patient Safety factors, and Care Coordination activities
combined in one package as a system element positively impact the performance of a
hospital?
This study is to investigate how this proposed package of system elements would impact the
performance of hospitals and the outcome of care, and to evaluate the effectiveness of the
integration of HRO, Patient Safety, and Care Coordination. The Null Hypothesis is defined with
the following statement and the goal is to assess if the null hypothesis can be rejected or not:
H
0
: Addressing HRO principles, Patient Safety factors, and Care Coordination activities
combined in one package as a system element does not improve hospital’s performance.
To test the research question, the Design of Experiment (DOE) method was carefully chosen. DOE
is a systematic technique to identify the causal relationship between variables that impact a process
and the outgoing results.
A critical advantage of using DOE is that “Lack of Success” is not the same as
“Failure.” One of the greatest benefits of DOE is the ability to terminate unfruitful
lines of investigation using the objective evidence of the validated models
50

generated to scientifically justify this decision. This occurs when the model
predictions have been verified (validating the model) and when those predictions
show the impossibility of meeting all necessary goals simultaneously. (Figard,
2009)

DOE offers a variety of techniques and the first step is to identify the method that best fits the
study. To do so, independent variables and dependent variables needed to be clearly defined to
present the layout of the DOE.
Independent Variable: The proposed package of system elements (HRO, Patient Safety, and Care
Coordination) is considered to be the independent variable in this work. This is a categorical type
(Yes 1 – No 0), and since the three elements have been integrated together and presented as one
package, there is only one independent variable under investigation. The aim was to assess the
impact of this one independent variable on the performance of the hospital in this study.
Dependent Variables: 20 harm reduction measures that indicate the performance of the hospital
were selected and eight were nominated for the final analysis. These measures are considered the
dependent variables of this study. They are quantitative (numbers from 0 to 100) and the goal was
to evaluate the impact of the proposed package of system elements on this set of performance
measures.
Data collection and how the eight measures were selected among the identified 20 items will be
explained in detail in the next part.
According to Dr. Abedi (pers. com.), an expert in the field of experimental design, Multivariate
Analysis of Variance (MANOVA) is the most suitable method to run this analysis (Mertler and
Vannatta, 2002). The following table summarizes the different techniques that are used in
statistical analysis with DOE (Mertler and Vannatta, 2002) and illustrate the pathway to choose
One-Way MANOVA for this study.
51

Table 2: Design of Experiment Techniques

MANOVA can be used when there are several quantitative dependent variables and there is a need
to run an overall statistical test on the set of variables. In addition, when there is a need to identify
how independent variables influence the responses of the dependent variables, MANOVA is a
perfect choice (French, Macedo, Poulsen, Waterson, & Yu, 2008). In MANOVA there are usually
one or more nominal independent variables and several continuous dependent variables (French et
al., 2008). This is a method that is extremely helpful in assessing and analyzing the main effects
of independent variables, the importance of the dependent variables, the strength of association
between dependent variables, and the effects of variance (French et al., 2008).
MANOVA forms linear combinations of the dependent variables while simultaneously controlling
for variates, which is very helpful in this design. “Testing the multiple dependent variables is
accomplished by creating new dependent variables that maximize group differences. These
artificial dependent variables are linear combinations of the measured dependent variables”
(French et al., 2008).
If the multivariate test is significant it can be concluded that the proposed package of systems
elements (HRO, Patient Safety, and Care Coordination) has been significant in improving the
performance of the hospital.
52

There are several test statistics that can be utilized for MANOVA. Hotelling’s Trace is the ratio of
effect variance to error variance. The Pillai-Bartlett criterion is the effect variances. Roy’s Largest
Root represents the largest eigenvalues. However, the most commonly used test for MANOVA is
Wilk’s Lambda, which is equal to the ratio of error variances to the effect variance plus error
variance. When there are only two levels for an effect all of the tests would be identical.
Wilks' lambda is a test statistic that is very commonly used in MANOVA to test whether there are
differences between the means of identified groups of subjects to a combination of dependent
variables. Wilks's lambda distribution is a probability distribution used in multivariate hypothesis
testing and is the multivariate generalization of the univariate F-distribution. So Wilks' lambda test
in MANOVA is the equivalent of the F-test in ANOVA, but in a multivariate setting with a
combination of dependent variables. Table 3 contains a summary comparison of the available test
statistics (Han, n.d.).
Table 3: MANOVA Test Statistics
Test Characteristics Formula
Wilks' Lambda - Mid conservative test
- Widely used for MANCOVA
- Determines whether the groups are different without worrying
about linear combinations of dependent variables
- Good balance between power and assumptions. Use if
assumptions appear to be met.

Roy’s Largest Root - Varies as far as conservatism
- Measures the differences only on the first canonical root
- Appropriate and very powerful when the dependent variables are
strongly interrelated on a single dimension
- Most likely to be affected by violations of assumptions

HE
-1

Hotelling’s Trace - Very liberal test
- Best for designs with manipulated variables
tr (HE
-1
)
Pillai-Bartlett - Very conservative test
- Best for designs with many problems, such as “unbalanced
design”
- Most robust when assumptions are not met

Seffect = H Serror = E trace = sum of the diagonal elements
53

The Total Sum of Squares is the total of the Sum of Squares of between groups and the Sum of
Squares of within groups (French et al., 2008; Han, n.d.; Todorov & Filzmoser, 2010):
SS Total = SS bg + SS wg
This is equal to:
∑
∑ = n ∑ + ∑
∑

GM is the grand mean matrix calculated with one value for each dependent variable averaged
across all individuals in the matrix.

⋮

Since there is only one independent variable in this study there is no need for the SSbg to be
partitioned. The following column matrix (vector) shows the n values of dependent variables
(French et al., 2008; Han, n.d.; Todorov & Filzmoser, 2010):
,…

… … … .

A multivariate F value, which is the Wilks' λ, can be obtained based on a comparison of the error
covariance matrix and the effect of the covariance matrix (French et al., 2008; Han, n.d.; Todorov
& Filzmoser, 2010):
Λ
|
|

54

An estimate of F can be calculated through the following equations (French et al., 2008; Han, n.d.;
Todorov & Filzmoser, 2010):

,
1

Λ

4

5

p = number of DVs

1
2

2
2

Since Wilks’ λ is equal to the variance not accounted for by the combined dependent variables,
then (1 – λ) is the variance that is accounted for by the best linear combination of dependent
variables (French et al., 2008; Han, n.d.; Todorov & Filzmoser, 2010). Eta Squared, which is the
measure of the effect size, indicates the strength of the relationship between variables. The formula
is as follows for both Eta Squared and Partial Eta Squared:
1 Λ
1 Λ

The following section will explain in detail the measures studied, the data collected, and the
analysis to test the hypothesis of this work.

55

B. Data Collection and Analysis
The data used in this study was collected from Keck Hospital of USC. After completion of the
documentations required by the Institutional Review Board (IRB) to obtain approval and receiving
the authorization that this investigation can proceed, Keck Hospital of USC provided their data
collected monthly for the time frames of 2013 and 2015.
Data collected from 2013 represent the “before state” when Keck Hospital was performing their
operations based on their basic fundamentals of providing care. The 2015 data characterize the
“after state” when new policies and strategies in regards to enhancing their performance were in
place.
In 2014 the executive management decided to step by step incorporate new measures to monitor
and policies to embed in their daily operations, with the aim of providing a higher quality of care
to their patients and achieving superior performance. The deliberated changes were implemented
with new strategies in Keck Hospital starting in 2014, and flourishing health outcomes were visible
and evident starting in 2015. That was the reason for the two years of 2013 and 2015, as Before
and After states of Keck Hospital’s performance, were selected for this study.
The measures listed in table 4 are the set of 20 items that Keck Hospital provided their monthly
data for. This group of measures are the dependent variables in the MANOVA analysis and will
be further elaborated on in the next chapter.

56

Table 4: List of Dependent Variable Measures
Measure Scale
Mortality Percentage of Death Observed
Length of Stay Days
Falls Per 1,000 Patient Days
Pressure Ulcer
Reportable HAPUs per 1,000 Patient
Days
Postop Sepsis Observed Rate Per 1000 Cases
Postop Respiratory Failure Observed Rate Per 1000 Cases
Accidental Puncture-Laceration Observed Rate Per 1000 Cases
Readmission 30 Percentage 30 Day Readmit
Readmission 7 Percentage 7 Day Readmit
Rate Hospital HCAHPS Top Box
Communication with Nurses HCAHPS Top Box
Responsiveness of Hospital Staff HCAHPS Top Box
Communication with Doctors HCAHPS Top Box
Hospital Environment Cleanliness HCAHPS Top Box
Hospital Environment Quietness HCAHPS Top Box
Pain Management HCAHPS Top Box
Communication About Medicine HCAHPS Top Box
Discharge Information HCAHPS Top Box
Care Transition HCAHPS Top Box
Recommend the Hospital HCAHPS Top Box

Each of these measures have their own scales.
HCAHPS Top Box: The Hospital Consumer Assessment of Healthcare Providers and Systems
(HCAHPS) survey is the first national, standardized, publicly reported survey of patients'
perspectives of hospital care. Beginning in 2002, the Center for Medicare and Medicaid Services
(CMS) partnered with the AHRQ to develop and test the HCAHPS survey. CMS implemented the
57

HCAHPS survey in October 2006, and the first public reporting of HCAHPS results occurred in
March 2008 (HospitalHCAHPS, n.d.; CAHPS Hospital Survey, n.d.).
The “top-box” is the most positive response to HCAHPS Survey items. The “top-box” response is
“Always” for five HCAHPS composites (Communication with Nurses, Communication with
Doctors, Responsiveness of Hospital Staff, Pain Management, and Communication about
Medicines) and two individual items (Cleanliness of Hospital Environment and Quietness of
Hospital Environment), “Yes” for the Discharge Information composite, “‘9’ or ‘10’ (high)” for
the Overall Hospital Rating item, “Definitely yes” for the Recommend the Hospital item, and
“Strongly agree” for the Care Transition composite (HospitalHCAHPS, n.d.; CAHPS Hospital
Survey, n.d.).
Patient Day: A unit of time during which the service of a health facility is used by a patient. For
example 20 patients in a hospital for one day would be 20 patient days.
Harm from Falls per 1,000 Patient Days: Number of inpatient falls with injuries on the unit
divided by the number of inpatient days on the unit, multiplied by 1,000.
Reportable HAPUs per 1,000 patient days: Hospital-acquired pressure ulcers per 1,000 patient
days.

C. Harmonizing the Scale
Since these measures have been collected based on their own respective scales, to be able to assess
the impact of the recommended package of system elements (HRO, Patient Safety, and Care
Coordination) on the performance of Keck Hospital, a strategy to harmonize the scales and
normalize the data was developed. Each of the measures represents an aspect of performance of
58

the hospital and collectively they shape a pattern of movement demonstrating how Keck hospital
has been progressing towards providing a higher quality of care. This group of measures came
from very comprehensive perspectives of both patients (from HCAHPS data) and the hospital
itself.
To harmonize the data, the actual evidence was converted to a true representation of the data in a
0 to 100 scale as an indication of Keck Hospital’s performance (0 being worse and 100 being the
highest performance). For each measure publicly available data was used to determine the proper
ranges and levels. With curve fitting the actual data has been converted to a normalized set
representing Keck Hospital’s performance. From this point of the study forward all the analyses
have been conducted based on this new set of data.

59

V. Results
Before proceeding with the statistical analysis of the data several comparison charts were created
to visually construct the trend of Keck Hospital’s performance from 2013 to 2015. The following
figures indicate the effect of the proposed package of system elements (HRO, Patient Safety, and
Care Coordination) on Keck Hospital’s performance for each of the 20 harm reduction measures.
As shown in the Figure 10, the majority of the red bars that indicate the performance of Keck
Hospital in 2015 stand above the blue bars that represent the 2013 performance.

Figure 10: Performance Comparison Chart

Figure 11 also signifies the performance of Keck Hospital in a monthly illustration, and it is evident
that the red line is above the blue one demonstrating that even with all the seasonal changes the
performance in 2015 of Keck Hospital surpassed that of 2013. Another noteworthy point is that
60

the trend lines express much less turbulence in the data in 2015 when comparing to 2013 and that
by itself is considered an improvement.

Figure 11: Monthly Performance with Tend Lines

In addition to the performance charts, a box plot has been created. A box plot is a standardized
way of displaying the distribution of data based on the five number summary: minimum, first
quartile, median, third quartile, and maximum. Figure 12 demonstrates the distribution of the data
and indicates smaller standard deviations in 2015, which means that the data points are closer to
the expected value in 2015 than in 2013.

Trend 2013 Trend 2015
61

Figure 12: Distribution of Data

In the next step, the set of data was prepared for the analysis and exported into the IBM Statistical
Package for the Social Sciences (SPSS) software. SPSS is one of the most popular statistical
packages that can perform highly complex data manipulation and analysis. As previously
described, this study consists of 20 quantitative dependent variables (performance measures) and
one categorical independent variable (proposed package of system elements). That makes
MANOVA the best model to test this experiment and to examine the differences in the two vectors
of means between the groups of 2013 data vs. the 2015 data.
62

To be able to run MANOVA, however, first the normality and homogeneity of the data has to be
tested. These are the two most important assumptions of multivariate analysis. The dependent
variables should be normally distributed within groups, but if the non-normality is caused by
skewness rather than by outliers the F test is still robust to non-normality. In addition, homogeneity
of variances assumes that the dependent variables exhibit equal levels of variance across the range
of predictor variables (French et al., 2008). The following graphs and tables emerged from the
SPSS analysis in phase one when testing for normality and homogeneity of data, and they
expressed insightful findings.
The frequency histogram option in SPSS indicates whether the data follows normal distribution
(Figure 13). Among all twenty measures “Pressure Ulcer”, “Postop Respiratory Failure”, “Postop
Sepsis”, and “Recommend the Hospital” did not pass the normality assumption.

63

Figure 13: SPSS Normality Test Results

In addition, Levene's test is used to test if the data have equal variances and verifies the
homogeneity of the data. This option tests the null hypothesis that the error variance of the
dependent variable is equal across the groups. The desired value to insure homogeneity of the data
is p-value sig > = 0.05. Table 5 has the SPSS result of Leven’s test that indicates that “Pressure
64

Ulcer”, “Post Op Respiratory Failure”, “Communication with Nurse”, “Responsiveness of
Hospital Staff”, “Cleanliness”, “Quietness”, “Pain Management”, “Discharge Information”, “Care
Transition”, and “Recommend the Hospital” are the measures that did not pass the homogeneity
assumption.
Table 5: Levene’s Test of Homogeneity

65

Table 6 illustrates a summary of the measures and whether they have passed normality and
homogeneity assumptions of MANOVA. All the measures highlighted in green could potentially
be included in the final analysis. Measures highlighted in purple miss one of the assumptions, but
still they could be used for the analyses in this study. The three measures highlighted in gray violate
both MANOVA assumptions and hence need to be eliminated from the analysis.
Table 6: Summary of Normality and Homogeneity Test Results

At this point of the study, out of the 20 measures, three were eliminated due to violating MANOVA
assumptions. With the remaining 17 measures there exists another restriction. The data was
collected for 24 months (12 for each year of 2013 and 2015). This was a limited number of
observations. According to Dr. Abedi (pers. com.), an expert in the field of experimental design,
Measures Normality Homogeneity
Length of Stay Yes Yes
Fall Yes Yes
Pressure Ulcer No No
Post Op Sepsis No Yes
Post Op Respiratory Failure No No
Accidental Puncture or Laceration Yes Yes
30 Day Readmit Yes Yes
7 Day Readmit Yes Yes
Rate Hospital Yes Yes
Communication with Nurse Yes No
Responsiveness of Hospital Staff Yes No
Communication with Doctors Yes Yes
Cleanliness Yes No
Quietness Yes No
Pain Management Yes No
Communication About Medicine Yes Yes
Discharge Information Yes No
Care Transition Yes No
Recommend the Hospital No No
Mortality Yes Yes
66

the maximum number of dependent variables that could be included in this study to keep the
robustness of the MANOVA intact with 24 data points was eight.
In the next step, eight of the 17 measures needed to be nominated for the remaining analysis. To
make the final selection the opinions of subject matter experts (Dr. Stephanie L. Hall, Chief
Medical Officer of Keck Hospital of USC and USC Norris Cancer Hospital and Dr. Philip D.
Lumb Chair and Professor of the Department of Anesthesiology of Keck Hospital of USC) were
sought. The goal was to select those eight measures that were 1) most desired to be included in
this study from a medical perspective, 2) closely linked and related to HRO, Patient Safety, and
Care Coordination, and 3) a comprehensive representation and characterization of the performance
of Keck Hospital.
Table 7 contains the finalized list of dependent variables that were included in this work. From
this point forward the data for this list of eight measures has been inputted in SPSS for the
remaining analysis.
Table 7: Final Selection of Dependent Variables
Selection of Dependent Variables
Measures Relation to Package Normality Homogeneity
Mortality Combo Yes Yes
Fall Care Coordination Yes Yes
Care Transition Care Coordination Yes No
7 Day Readmission HRO Yes Yes
Rate Hospital HRO Yes Yes
Communication with Doctors HRO Yes Yes
Communication About Medicine HRO Yes Yes
Post Op Sepsis Patient Safety No Yes

67

Based on the format of the collected data and the structure of this study, and in accordance with
the recommendations of Dr. Abedi (pers. com.), an expert in the field of experimental design, it
was decided that in addition to MANOVA the following extra analyses should be conducted:
- MANOVA
 To run an overall statistical test on the set of variables and identify how the
independent variables influence the responses of the dependent variables.
- Discriminant Analysis
 To identify the measures that most significantly impact the performance of Keck
Hospital with the proposed package of system elements.
- Reliability Test
 To assess the internal consistency among the identified measures.
- Factor Analysis
 To develop a strategy for data reduction.
- Individual ANOVAs
 To test each identified measure against H0.

Although MANOVA was sufficient to test the null hypothesis and investigate the research
question, to achieve a complete and all-inclusive conclusion, undertaking a more comprehensive
set of analyses was required. The following parts describe the results of the aforementioned
analyses when using SPSS.
Note: For all the results generated by SPSS, Group 1 indicates Keck Hospital’s performance in
2013, which is also titled the Before state. Group 2 refers to the 2015 performance data when new
strategies and policies symbolizing the proposed package of HRO, Patient Safety, and Care
68

Coordination were in place. The 2015 data is the After state in the layout of the DOE in this study.
For ease of use in SPSS the code name of the package is HROPSCC. The code names of all
measures are available in the appendix A.

A. Manova
After completion of the MANOVA analysis in SPSS the following results emerged:
1) Descriptive Statistics in Table 8 compare the mean and standard deviation of each of the
identified measures among the two groups. With the exception of “Communication with Doctors”
and “Mortality” measures, the remaining items show improvements in performance from 2013 to
2015 when considering the mean and standard deviation.
Table 8: Descriptive Statistics
Descriptive Statistics

HROPSCC Mean Std. Deviation N
FALL 1 77.1328 10.78876 12
2 84.8306 5.77213 12
Total 80.9817 9.33063 24
SEPS 1 78.6169 18.21262 12
2 92.1468 9.94930 12
Total 85.3819 15.92907 24
READMT7 1 77.1948 7.01799 12
2 84.4940 6.45818 12
Total 80.8444 7.57638 24
RATEHOSP 1 84.2414 7.02407 12
2 89.2036 3.49437 12
Total 86.7225 5.98828 24
COMWDOC 1 83.6878 6.21003 12
2 82.7120 3.48445 12
Total 83.1999 4.94965 24
COMABMED 1 76.1526 5.22078 12
2 80.0449 3.93366 12
69

Total 78.0988 4.93846 24
CARETRANS 1 84.7365 3.69518 12
2 86.4103 1.62206 12
Total 85.5734 2.91883 24
MORTAL 1 89.2194 4.51950 12
2 89.0343 3.92263 12
Total 89.1269 4.13967 24

2) Box’s Test (Table 9) was used to test for the equality of the covariance between the groups.
This is the equivalent of a multivariate homogeneity of variance designed to insure the robustness
of the MANOVA. A P-Value (sig) greater than 0.05 demonstrates the equality of the covariance
between the groups, which is desired when using MANOVA.
Table 9: Box’s Test

Box's Test of
Equality of
Covariance
Matrices
a

Box's M 59.912
F .993
df1 36
df2 1628.589
Sig. .482
Tests the null
hypothesis that the
observed covariance
matrices of the
dependent variables are
equal across groups.
a. Design: Intercept +
HROPSCC

3) The results of the Levene’s Test of Equality of Error Variances are illustrated in Table 10. This
test examined the null hypothesis to check whether the error variances of the dependent variables
70

are equal across groups. This is basically to test for the homogeneity of the variance. It was
observed that except for the “Care Transition” measure, the remaining items have a P-value (sig)
greater than 0.05. That means the observed differences in the sample variances are based on
random sampling from a population with equal variances for all measures except “Care
Transition”.
Table 10: Levene’s Test of Equality of Error Variances
Levene's Test of Equality of Error Variances
a

F df1 df2 Sig.
FALL 1.922 1 22 .180
SEPS 2.894 1 22 .103
READMT7 .404 1 22 .531
RATEHOSP 3.805 1 22 .064
COMWDOC 4.209 1 22 .052
COMABMED .626 1 22 .437
CARETRANS 7.005 1 22 .015
MORTAL .441 1 22 .514
Tests the null hypothesis that the error variance of the dependent
variable is equal across groups.
a. Design: Intercept + HROPSCC

4) There are four major statistics used for MANOVA: Pillai's Trace, Hotelling's Trace, Roy's
Largest Root, and Wilks’ λ. All four have been included in Table 11, although the focus was on
Wilks’ λ. P-Values (sig) less than or equal to 0.05 reject the null hypothesis. It was observed that
in all four statistics the P-value sig is equal to 0.00, and therefore it can be interpreted that “group
2, has a higher performance comparing to group 1 as a result of the proposed package of system
elements”. Partial Eta squared indicates the variance explained by the independent variable, which
is the effect size of the MANOVA. In this model the Partial Eta squared is 0.843, and that is
significantly high, meaning the changes in the performance from group 1 to 2 have been the result
71

of the HRO, Patient Safety, and Care Coordination package. In addition, the Observed Power of
the test is the absolute highest indicated as 1.000 that demonstrates this has been indeed a very
powerful test.
Table 11: Multivariate Tests of MANOVA
Multivariate Tests
a

Effect Value F
Hypothesis
df Error df Sig.
Partial
Eta
Squared
Noncent.
Parameter
Observed
Power
c

Intercept Pillai's Trace 1.000 18487.857
b
8.000 15.000 .000 1.000 147902.859 1.000
Wilks' Lambda .000 18487.857
b
8.000 15.000 .000 1.000 147902.859 1.000
Hotelling's Trace 9860.191 18487.857
b
8.000 15.000 .000 1.000 147902.859 1.000
Roy's Largest Root 9860.191 18487.857
b
8.000 15.000 .000 1.000 147902.859 1.000
HROPSCC Pillai's Trace .843 10.096
b
8.000 15.000 .000 .843 80.772 1.000
Wilks' Lambda .157 10.096
b
8.000 15.000 .000 .843 80.772 1.000
Hotelling's Trace 5.385 10.096
b
8.000 15.000 .000 .843 80.772 1.000
Roy's Largest Root 5.385 10.096
b
8.000 15.000 .000 .843 80.772 1.000
a. Design: Intercept + HROPSCC
b. Exact statistic
c. Computed using alpha = .05

5) The Pearson Correlation demonstrate strong connections among dependent variables. Table 12
indicates that there are significant correlations between “Rate the Hospital” and “Communication
with Doctors”, “Communication about Medicine”, “Care Transition”, and “Mortality”.
With a slightly less strong, but still notable, connection between “Communication with Doctors”
and “Care Transition”.

72

Table 12: Pearson Correlations
Correlations
FALL SEPS READMT7 RATEHOSP COMWDOC COMABMED CARETRANS MORTAL
FALL Pearson Correlation 1 -.180 .263 .049 -.001 .252 -.215 .214
Sig. (2-tailed)

.401 .214 .820 .997 .235 .314 .316
N 24 24 24 24 24 24 24 24
SEPS Pearson Correlation -.180 1 .267 .229 -.146 .206 -.020 -.376
Sig. (2-tailed) .401

.207 .281 .496 .333 .928 .070
N 24 24 24 24 24 24 24 24
READMT7 Pearson Correlation .263 .267 1 .328 .025 .286 .123 -.267
Sig. (2-tailed) .214 .207

.117 .909 .176 .568 .206
N 24 24 24 24 24 24 24 24
RATEHOSP Pearson Correlation .049 .229 .328 1 .584
**
.534
**
.630
**
-.546
**

Sig. (2-tailed) .820 .281 .117

.003 .007 .001 .006
N 24 24 24 24 24 24 24 24
COMWDOC Pearson Correlation -.001 -.146 .025 .584
**
1 .336 .480
*
-.204
Sig. (2-tailed) .997 .496 .909 .003

.109 .018 .339
N 24 24 24 24 24 24 24 24
COMABMED Pearson Correlation .252 .206 .286 .534
**
.336 1 .258 -.265
Sig. (2-tailed) .235 .333 .176 .007 .109

.223 .211
N 24 24 24 24 24 24 24 24
CARETRANS Pearson Correlation -.215 -.020 .123 .630
**
.480
*
.258 1 -.392
Sig. (2-tailed) .314 .928 .568 .001 .018 .223

.058
N 24 24 24 24 24 24 24 24
MORTAL Pearson Correlation .214 -.376 -.267 -.546
**
-.204 -.265 -.392 1
Sig. (2-tailed) .316 .070 .206 .006 .339 .211 .058

N 24 24 24 24 24 24 24 24
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).

B. Discriminant Analysis
In the next step, Discriminant Analysis was conducted in SPSS to identify the measures that most
significantly impact the performance of Keck Hospital as a result of the proposed package of HRO,
73

Patient Safety, and Care Coordination among the selected measures. The following are the results
associated with this test:
1) The summary of the Canonical Discriminant Functions indicates that the Canonical Correlation
is 0.981. This specifies the correlation between the discriminant scores and the levels of the
dependent variable. A high correlation, considering that 1 is perfect, indicates a function that
discriminates well. Canonical Correlation to the power of 2 (0.981
2
= 0.962) represents the effect
size and indicates the magnitude of the actual effect of the predictors on the outcome.
In addition, with the Wilk’s Lambda test’s smaller lambda it shows that the group means differ
and with a P-value (sig) of 0.000 the difference is significant from group 1 to 2. This is a
representation of the proportion of the total variance in the discriminant scores that is not explained
by differences between the groups and that is the reason why the smaller lambda is preferred. Our
result of P-value sig = 0.000 confirms that this group of predictor variables predicts the outcome
at a statistically significant level and that this is a very strong model.
Table 13: Discriminant Scores
Eigenvalues
Function Eigenvalue % of Variance Cumulative %
Canonical
Correlation
1 5.385
a
100.0 100.0 .918
a. First 1 canonical discriminant functions were used in the analysis.

Wilks' Lambda
Test of Function(s) Wilks' Lambda Chi-square df Sig.
1 .157 33.370 8 .000

74

2) Table 14 is the Structure Matrix that illustrates the canonical loading (also titled discriminant
loading) of the discriminant functions. This is to demonstrate which of the measures impacts Keck
Hospital’s performance the most, in descending order. The results indicated that “Readmission in
7 days”, “Post Op Sepsis”, and “Rate the Hospital” are the top three measures with the highest
impact on Keck Hospital’s performance as a result of the proposed package of HRO, Patient
Safety, and Care Coordination.
Table 14: Structure Matrix
Structure Matrix

Function
1
READMT7 .244
SEPS .207
RATEHOSP .201
FALL .200
COMABMED .190
CARETRANS .132
COMWDOC -.044
MORTAL -.010
Pooled within-groups correlations between
discriminating variables and standardized
canonical discriminant functions
Variables ordered by absolute size of correlation
within function.

3) The Separate-Group graphs (Figure 14) visually display how well the discriminant function
works. If the two plots overlap too much that is an indication that the discriminant function is poor.
In this case they do not overlap at all and this is the representation of a decent discriminant
function.
75

Figure 14: Separate-Group Graphs
4) The Classification Results (Table 15) determines how well the discriminant function works and
if it works equally well for each group of the dependent variables. In this case, 100.0% of the
original grouped cases and 95.8% of the cross-validated grouped cases were correctly classified.
This indicates a robust discriminant function that was also expressed in the plot above.
Table 15: Classification Results

Classification Results
a,c

HROPSCC
Predicted Group Membership
Total

1 2
Original Count 1 12 0 12
2 0 12 12
% 1 100.0 .0 100.0
2 .0 100.0 100.0
Cross-validated
b
Count 1 12 0 12
2 1 11 12
% 1 100.0 .0 100.0
2 8.3 91.7 100.0
a. 100.0% of original grouped cases correctly classified.
b. Cross validation is done only for those cases in the analysis. In cross validation, each
case is classified by the functions derived from all cases other than that case.
c. 95.8% of cross-validated grouped cases correctly classified.

76

C. Reliability Test
Although it was not necessary for this study to conduct a reliability test, it was decided to perform
one to determine how Cronbach’s alpha behaves. Cronbach’s alpha is a measure of internal
consistency and it is calculated from 0 to 1. In most cases the desired value is more than or equal
to 0.7, but in this study the lower alpha demonstrates that the selected measures have minimal
overlap, hence they are a good set to characterize a comprehensive model of the performance of
Keck Hospital. In this study the Cronbach’s alpha is 0.317. This is an indication that the selected
measures can demonstrate Keck Hospital’s performance in a comprehensive fashion.
Table 16: Reliability Test

Reliability Statistics
Cronbach's
Alpha N of Items
.317 8

D. Factor Analyis
To complete this work it was decided to perform a factor analysis as a method of data reduction if
such requirements occur in future, and the following outcome emerged:
1) KMO and Bartlett's Test is a measure of sampling adequacy to check if the data set is a good fit
for factor analysis. The result indicates a P-Value (sig) of 0.004. Since it is less than the 0.05
Bartlett’s test is significant and factor analysis is appropriate, but the KMO value is 0.662 (a value
greater than 0.5 is acceptable). It is understood that the number of samples may not be perfect but
it is acceptable. It was known in the beginning that only 24 data points (12 months for each state)
were available to perform the analysis.
77

Table 17: KMO and Bartlett’s Test
KMO and Bartlett's Test
Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .662
Bartlett's Test of Sphericity Approx. Chi-Square 51.699
df 28
Sig. .004

2) Table 18 explains the actual factors that were extracted. In this case three factors with
eigenvalues greater than 1.0 met the criteria for extraction. Under the “Rotation Sums of Squared
Loadings” section it is indicated that three selected factors accounted for a cumulative of 72.164%
of the variability in all eight measures.
Table 18: Factor Analysis - Total Variance Explained
Total Variance Explained
Component
Initial Eigenvalues
Extraction Sums of Squared
Loadings Rotation Sums of Squared Loadings
Total
% of
Variance
Cumulative
% Total
% of
Variance
Cumulative
% Total
% of
Variance
Cumulative
%
1 2.920 36.504 36.504 2.920 36.504 36.504 2.489 31.115 31.115
2 1.444 18.052 54.556 1.444 18.052 54.556 1.736 21.702 52.817
3 1.409 17.608 72.164 1.409 17.608 72.164 1.548 19.347 72.164
4 .701 8.761 80.925

5 .486 6.074 86.998

6 .456 5.695 92.693

7 .386 4.830 97.523

8 .198 2.477 100.000

Extraction Method: Principal Component Analysis.

3) A Scree-Plot graph illustrates the loadings of the selected measures visually. This graph
indicates which components or factors explain most of the variability in the data and how many
factors should remain in the analysis if and when data reduction is required. In this case three
78

factors explained most of the variability. This is specified in Figure 15 where the line flattens out
and plateaus after the third factor.

Figure 15: Scree-Plot

4) The Rotated Component Matrix demonstrates the factor loadings for each variable. Table 19
explains how the variables are weighted for each factor and also determines the correlation
between the variables and the factors. In this study the analysis can be reduced into three factors.
- “Rate Hospital”, “Communication with Doctors”, and “Care Transition” load on factor 1.
- “Post Op Sepsis” and “Mortality” load on factor 2.
- “Fall”, “7 Day Readmission”, and “Communication About Medicine” load on factor 3.

79

Table 19: Rotated Component Matrix
Rotated Component Matrix
a

Component
1 2 3
FALL -.129 -.310 .850
SEPS -.106 .866 .041
READMT7 .067 .468 .621
RATEHOSP .813 .330 .270
COMWDOC .828 -.210 .080
COMABMED .462 .218 .580
CARETRANS .826 .113 -.138
MORTAL -.464 -.677 .060
Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization.
a

a. Rotation converged in 4 iterations.

E. Individual Anova
The finale of this series of statistical analyses was to perform individual ANOVAs and t-Tests.
The following results emerged from the SPSS analysis.
1) “Communication with Doctors”, “Care Transition”, and “Mortality” have a P-values greater
than 0.05 and hence cannot reject the null hypothesis. This is an indication that the proposed
package of HRO, Patient Safety, and Care Coordination did not improve the performance of these
three measures from 2013 to 2015.

80

Table 20: ANOVA and t-Test
Independent Samples Test

Levene's Test for Equality of
Variances t-test for Equality of Means
F Sig. t df
Sig. (2-
tailed)
Mean
Difference
Std. Error
Difference
95% Confidence Interval
of the Difference
Lower Upper
FALL Equal variances assumed 1.922 .180 -2.179 22 .040 -7.69777 3.53217 -15.02305 -.37250
Equal variances not assumed

-2.179 16.820 .044 -7.69777 3.53217 -15.15607 -.23948
SEPS Equal variances assumed 2.894 .103 -2.258 22 .034 -13.52986 5.99088 -25.95419 -1.10553
Equal variances not assumed

-2.258 17.029 .037 -13.52986 5.99088 -26.16790 -.89181
READMT7 Equal variances assumed .404 .531 -2.651 22 .015 -7.29921 2.75318 -13.00897 -1.58945
Equal variances not assumed

-2.651 21.850 .015 -7.29921 2.75318 -13.01124 -1.58718
RATEHOSP Equal variances assumed 3.805 .064 -2.191 22 .039 -4.96214 2.26473 -9.65891 -.26537
Equal variances not assumed

-2.191 16.131 .043 -4.96214 2.26473 -9.76000 -.16427
COMWDOC Equal variances assumed 4.209 .052 .475 22 .640 .97585 2.05560 -3.28720 5.23891
Equal variances not assumed

.475 17.302 .641 .97585 2.05560 -3.35533 5.30704
COMABMED Equal variances assumed .626 .437 -2.063 22 .051 -3.89231 1.88702 -7.80576 .02113
Equal variances not assumed

-2.063 20.445 .052 -3.89231 1.88702 -7.82309 .03846
CARETRANS Equal variances assumed 7.005 .015 -1.437 22 .165 -1.67382 1.16495 -4.08979 .74215
Equal variances not assumed

-1.437 15.087 .171 -1.67382 1.16495 -4.15561 .80797
MORTAL Equal variances assumed .441 .514 .107 22 .916 .18509 1.72754 -3.39761 3.76780
Equal variances not assumed

.107 21.573 .916 .18509 1.72754 -3.40173 3.77191

2) The results of the individual ANOVAs revealed that the majority of the measures rejected the
null hypothesis and presented significant improvements from 2013 to 2015 as a result of the
proposed package of HRO, Patient Safety, and Care Coordination.

81

Table 21: Individual ANOVAs
ANOVA
FALL
Sum of Squares df Mean Square F Sig.
Between Groups 355.534 1 355.534 4.749 .040
Within Groups 1646.862 22 74.857

Total 2002.397 23

ANOVA
SEPS
Sum of Squares df Mean Square F Sig.
Between Groups 1098.342 1 1098.342 5.100 .034
Within Groups 4737.569 22 215.344

Total 5835.912 23

ANOVA
READMT7
Sum of Squares df Mean Square F Sig.
Between Groups 319.671 1 319.671 7.029 .015
Within Groups 1000.563 22 45.480

Total 1320.234 23

ANOVA
RATEHOSP
Sum of Squares df Mean Square F Sig.
Between Groups 147.737 1 147.737 4.801 .039
Within Groups 677.030 22 30.774

Total 824.767 23

ANOVA
COMWDOC
Sum of Squares df Mean Square F Sig.
Between Groups 5.714 1 5.714 .225 .640
Within Groups 557.765 22 25.353

Total 563.478 23

82

ANOVA
COMABMED
Sum of Squares df Mean Square F Sig.
Between Groups 90.901 1 90.901 4.255 .051
Within Groups 470.033 22 21.365

Total 560.934 23

ANOVA
CARETRANS
Sum of Squares df Mean Square F Sig.
Between Groups 16.810 1 16.810 2.064 .165
Within Groups 179.140 22 8.143

Total 195.950 23

ANOVA
MORTAL
Sum of Squares df Mean Square F Sig.
Between Groups .206 1 .206 .011 .916
Within Groups 393.942 22 17.906

Total 394.147 23

83

VI. Discussion
The aim of this study was to investigate the performance of Keck Hospital at USC. The goal was
to evaluate the effects of the combination of HRO principles, Patient Safety factors, and Care
Coordination activities on the hospital’s performance. The research question was designed to help
illustrate how the integrated triple element model impacts the performance of Keck Hospital. The
Design of the Experiment was then carefully chosen to conduct the study.
In 2013 Keck Hospital was in its original setting, delivering care based on their basic fundamentals
of operations. With the aim of providing a higher quality of care, in 2014 several new policies and
strategies were introduced and modifications directed their system towards a superior
performance. The outcomes of the new approaches started to become visible in 2015.
To pursue this assessment a set of 20 measures representing the performance of Keck Hospital
were identified. The data was collected in their respective scales for all 20 measures, but for the
purpose of the final analysis it was converted to a 0 to 100 performance rating.
The data was collected on a monthly basis for two years of 2013 and 2015 providing a total of 24
data points. Since only 24 data points for each measure were available, with the help and
supervision of experts in the medical field (Dr. Stephanie L. Hall, Chief Medical Officer of Keck
Hospital of USC and USC Norris Cancer Hospital, and Dr. Philip D. Lumb Chair and Professor of
the Department of Anesthesiology of Keck Hospital of USC) eight of those 20 measures that met
the MANOVA assumptions were selected to keep the robustness of the analysis intact.
With the identification of independent and dependent variables for this research it became evident
that a one-way MANOVA would be the most appropriate technique to pursue. To answer the
research question performing only the MANOVA was adequate, however to create a
84

comprehensive outlook for the performance of Keck Hospital it was decided that further analyses
should be performed. This was so that complementary results could highlight the path to a more
insightful understanding in respect to the effectiveness of the proposed package. The IBM SPSS
software was used to conduct a series of statistical analyses including MANOVA, Discriminant
Analysis, Reliability Test, Factor Analysis, and individual ANOVAs, and profound outcomes
emerged.

A. Statistical Analyses
The MANOVA test resulted in a p-value sig equal to 0.000, which rejects the null hypothesis. This
can be interpreted as “Keck Hospital’s performance in 2015 surpassed their 2013 performance as
a result of the proposed package of HRO, Patient Safety, and Care Coordination”. With a
significantly high partial Eta squared of 0.843, it is evident that the higher performance is the result
of the proposed package since this item indicates the variance explained by the independent
variable and expresses the effect size of MANOVA. In addition the Observed Power of the test is
the absolute highest expressed as 1.000, demonstrating that this has been indeed a very powerful
test. The Pearson Correlation determines that there are significant correlations between “Rate the
Hospital” measure and “Communication with Doctors”, “Communication about Medicine”, “Care
Transition”, and “Mortality”. A slightly less strong, but still notable connection is observed
between “Communication with Doctors” measure and “Care Transition”.
In the Discriminant Analysis Wilk’s Lambda sig was equal to 0.000 and confirms that this group
of predictor variables predicted the outcome at a statistically significant level, and indicates that
this is a very strong model. In addition, the difference between the performances from 2013 to
2015 is significant. The Structure Matrix explains that “Readmission in 7 days”, “Post Op Sepsis”,
85

and “Rate the Hospital” are the top three measures with the highest impact on Keck Hospital’s
performance as a result of the proposed package. The Classification Result specifies that the
discriminant function worked very well, and equally well for both groups of dependent variables
with 100.0% of the original grouped cases and 95.8% of the cross-validated grouped cases being
correctly classified. This is an indication of a robust discriminant function.
The Reliability test generated a Cronbach’s alpha that was equal to 0.317. This is a measure of
internal consistency. In this study the lower alpha explains that the selected measures have minimal
overlap and hence they are a very comprehensive set that truly represents the performance of Keck
Hospital.
The Factor Analysis stated that for this study the analysis can be reduced to three factors that can
explain most of the variability. The three factors with eigenvalues larger than 1.0 were identified
and their loadings were demonstrated. “Rate Hospital”, “Communication with Doctors”, and “Care
Transition” loaded on factor 1. “Post Op Sepsis” and “Mortality” loaded on factor 2. “Fall”, “7
Day Readmission”, and “Communication About Medicine” loaded on factor 3.
Lastly Individual ANOVAs were conducted and the result pointed to the fact that
“Communication with Doctors”, “Care Transition”, and “Mortality” are the three measures with
p-values greater than 0.05 and hence they cannot reject the null hypothesis, individually. This
means that the proposed package in this study could not help to improve the performance of these
measures from 2013 to 2015.
The collective outcomes resulting from MANOVA, Discriminant analysis, Reliability test, Factor
analysis, and individual ANOVAs are the manifestation that the addition of the proposed package
of HRO, Patient Safety factors, and Care Coordination activities in a health system is truly
positively impactful.
86

B. Triple Element Model
The proposed package of High Reliability, Patient Safety, and Care Coordination represents a three
pillar groundwork as a foundation of the care delivery system. They complement each other to
synchronize health systems with high effectiveness. They share similar subset elements that
reinforce the impact of the package on the health systems’ performance while they have their own
unique features. They collectively provide comprehensive patient centered care where patients
receive reliable and safe care within and among all providers. The three system elements work as
three silos that can be integrated together very successfully.

Figure 16: Pictograph of the Triple Element Method

This integrated triple element model collectively provides a solid foundation for an effective and
efficient healthcare system where patients receive safe and reliable care at the right time with the
right set of professionals. This combination offers major benefits to the system including but not
limited to:
87

- Improve performance of care delivery systems.
- Enhance patients’ health outcomes.
- Reduce number and severity of adverse events.
- Promote effective and powerful teamwork and practical team behaviors.
- Empower staff and patients.
- Form a “Collective Mindfulness”.
- Develop a purposeful Safety Culture.
- Encourage continuous improvement.
- Embrace systematic change.
- Encourage health screening and preventive measures.
- Reduce the fragmentation of care.
- Enhance patient experience.

C. Limitation of this Research
HRO principles, Patient Safety factors, and Care Coordination activities were combined into one
package of a system element and the study investigated the impact of this combined package on
the performance of Keck Hospital. Therefore, the effectiveness of each individual element, as
HRO, Patient Safety, and Care Coordination, in the final outcome could not be evaluated.
The number of data points were a total of 24, 12 months in each year for 2013 and 2015. Expanding
the number of data points could potentially allow for all 20 measures to be included in the analysis
and create a more comprehensive results.

88

D. Future Studies
The natural step ahead for this research and an opportunity for future study is to implement the
proposed package in a non-teaching hospital and compare the results to Keck Hospital, which is a
teaching hospital.
Another opportunity in the future would be to design an experiment for each individual element
of HRO, Patient Safety, and Care Coordination to be added to a health system and their impacts
on the performance of a hospital to be evaluated independently.

89

VII. Conclusion
This study aimed to assess the effects of HRO, Patient Safety, and Care Coordination combined
as one package of system elements on the performance of a health system. The results of the
evaluation were found to be remarkably effective in enhancing the performance of Keck Hospital
of USC.
The factors selected to characterize the performance of the hospital comprised “Rate the Hospital”,
“Communication with Doctors”, “Communication about Medicine”, “Care Transition”, “Fall”,
“Readmission in 7 days”, “Post Op Sepsis”, and “Mortality”. They were carefully nominated from
a pool of 20 measures based on whether they conform to MANOVA assumptions, as well as expert
opinions.
The set of statistical analyses consisting of MANOVA, Discriminant analysis, Reliability test,
Factor analysis, and individual ANOVAs were determined to be performed. The result collectively
indicated that the performance of Keck Hospital in 2015 outshone the outcomes in 2013.
Furthermore, based on a detailed assessment it was revealed that the majority of measures rejected
the null hypothesis individually and their performance had improved from 2013 to 2015 as a result
of the proposed package of HRO, Patient Safety, and Care Coordination. Additionally, it was
indicated that “Readmission in 7 days”, “Post Op Sepsis”, and “Rate the Hospital” are the top three
measures with the highest impact on Keck Hospital’s performance.
The result of analyzing the integrated triple element method was found to be significantly effective
in enhancing the performance of Keck Hospital, and therefore addressing HRO principles, Patient
Safety factors, and Care Coordination combined in one package as a system element is highly
recommended to health systems. The outcomes of this investigation can help executives in health
90

organizations and hospitals to identify the measures that most significantly impact the performance
of their organizations. They can consequently focus their resources on interventions that target
those measures specifically. The results of this study are also helpful in decision making processes
for those health organizations who are debating whether to pursue HRO principles, knowing that
it is a substantial investment of resources. The outcome of this study illustrated that they can
combine HRO with Patient Safety and Care Coordination interventions to maximize the benefits,
and, in addition, it revealed that there are major benefits to their health systems with the addition
of this proposed package that can be very motivating and encouraging to help them drive this path.

91

VIII. Bibliography

Adverse Events after Hospital Discharge. (n.d.). Patient Safety Network AHRQ. Retrieved from
http://psnet.ahrq.gov/primer.aspx?primerID=11

AHRQ. (2005). 30 safe practices for better health care. Pub No. 05-P007. Retrieved from
https://archive.ahrq.gov/research/findings/factsheets/errors-safety/30safe/30-safe-practices.pdf

Altman, L. K. (2014, February). A patient’s eye view of nurses. New York Times. Retrieved from
http://well.blogs.nytimes.com/2014/02/10/a-patients-eye-view-of-
nurses/?_php=true&_type=blogs&_r=0

Aspden, P., Wolcott, J. A., Bootman, L., & Cronenwett, L. R. (2007). Preventing medication
errors. Washington, DC: National Academies Press.

Batalden, P., & Davidoff, F. (2007). What is "quality improvement" and how can it transform
healthcare? Quality and Safety in Health Care, 16(1), pp.2-3.

Bodenheimer, T. (2008). Coordinating care - A perilous journey through the health care system.
New England Journal of Medicine, 358(10), 1064–71.
92

Brink, S. (2014, January 7). Can hospitals avoid waste and prevent overtreatment? US News.
Retrieved from
http://health.usnews.com/health-news/hospital-of-tomorrow/articles/2014/01/07/can-hospitals-
avoid-waste-and-prevent-overtreatment

CAHPS Hospital Survey. (n.d.). Retrieved from
http://www.hcahpsonline.org/SummaryAnalyses.aspx

Chassin, M. R. (2012). Quality of care: how good is good enough? Israel Journal of Health Policy
Research, 1(4).

Chassin, M. R, & Loeb, J. M. (2011). The ongoing quality improvement journey: Next stop, high
reliability. Health Affairs, 30(4), 559-568.

Chassin, M. R., & Loeb, J. M. (2013). High-reliability health care: Getting there from
here. Milbank Quarterly, 91(3), 459-490. doi:10.1111/1468-0009.12023

Christianson, M. K., Sutcliffe, K. M., Miller, M. A., & Iwashyna, T. I. (2011). Becoming a high
reliability organization. Critical Care Volume, 15(6).

Committee on Human-System Design Support for Changing Technology. (2007). Human-system
integration in the system development process: A new look. The National Academies Press.
93

Conway, J. (2014). High reliability: What’s the role of the patient, family, public, and community?
Institute for Patient- and Family-Centered Care.

Conway, J., Federico, F., Stewart, K., & Campbell, M. J. (2010). Respectful management of serious
clinical adverse events. Institute for Healthcare Improvement.

ECRI Institute. (2013). Surgical fire prevention. Retrieved from
https://www.ecri.org/surgical_fires

Figard, S. (2009). The basics of experimental design for multivariate analysis. SAS global Forum.

Forster, A. J., Murff, H. J., Peterson, J. F., Gandhi, T. K, & Bates, D. W. (2003). The incidence
and severity of adverse events affecting patients after discharge from the hospital. Annals of
Internal Medicine, 138(3), 161–67.

French, A., Macedo, M., Poulsen, J., Waterson, T. & Yu, A. (2008). Multivariate Analysis of
Variance (MANOVA). Retrieved from

Garman, A. N., Leach, D. C., & Spector, N. (2006). Worldviews in collision: Conflict and
collaboration across professional lines. Journal of Organizational Behavior, 27(7), 829-849.

94

Gauthier, A. K., Davis, K., & Schoenbaum, S. C. (2006). Achieving a high-performance health
system: High reliability organizations within a broader agenda. Health Services Research,
41(4p2):1710-1720.

Gindi, R. M., Cohen, R. A., & Kirzinger, W. K. (2012). Emergency room use among adults aged
18–64: Early release of estimates from the National Health Interview Survey, January-June 2011.
CDC Report Division of Health Interview Statistics, National Center for Health Statistics.

Goeschel, C. A., Wachter, R. M., & Pronovost, P. J. (2010). Responsibility for quality
improvement and patient safety: Hospital board and medical staff leadership challenges. Chest,
138(1), 171–78.

Han, K. (n.d.). Multivariate Analysis of Variance (MANOVA) – Between group design. Central
Michigan University, College of Humanities and Social and Behavioral Sciences.

Hines, S. Luna, K., & Lofthus, J. (2008). Becoming a high reliability organization: Operational
advice for hospital leaders. Agency for Healthcare Research and Quality publication No. 08-0022.

HospitalHCAHPS. (n.d.). Retrieved from https://www.cms.gov/medicare/quality-initiatives-
patient-assessment-instruments/hospitalqualityinits/hospitalhcahps.html

Ignagni, K. (2010). Testimony for Senate Committee on Health, Education, Labor and Pensions;
Understanding health insurance premiums and the need for system-wide cost containment.
95

Jha, A., & Epstein, A. (2010). Hospital governance and the quality of care. Health Affairs, 29(1),
182–87.

Joint Commission Resources Quality & Safety Network Resource Guide; Changes for 2012:
Standards and Survey Process. (2011). Joint Commission Resources Quality and Safety Network.

Kaplan, G., Bo-Linn, G., Carayon, P., Pronovost, P., Rouse, W., Reid, P., & Saunders, R. (2013).
Bringing a systems approach to health. Institute of Medicine and National Academy of
Engineering.

Klein, S., & McCarthy, D. (2010). CareOregon: Transforming the role of a Medicaid health plan
from payer to partner. Issues Research Inc., The Commonwealth Fund.

Klevans, R. M., Edwards, J. R., Richards, C. L., Horan, T. C., Gaynes, R. P., Pollock, D. A., &
Cardo, D. M. (2007). Estimating health care associated infections and deaths in U.S. hospitals.
Public Health Reports, 122(2), 160–66.

Leonard, M. W., & Frankel, A. S. (2011). Role of effective teamwork and communication in
delivering safe, high-quality care. Mount Sinai Journal of Medicine: A Journal of Translational
and Personalized Medicine 78(6), 820-26.

Mertler, C. A., & Vannatta, R. A. (2002). Advanced and multivariate statistical methods; Practical
application and interpretation. Pyrczak Publishing.
96

Minnesota Department of Health. (2013). Adverse health events in Minnesota: Ninth annual public
report. Retrieved from http://www.health.state.mn.us/patientsafety/ae/2013ahereport.pdf

National Quality Forum, Patient Safety. (n.d.). Retrieved from
https://www.qualityforum.org/Topics/Patient_Safety.aspx

National Quality Forum, Patient Safety. (2015). Retrieved from
http://www.qualityforum.org/ProjectDescription.aspx?projectID=77836

Nelson, E. C., Godfrey, M. M., Batalden, P. B., Berry, S. A., Bothe, Jr., A. E., McKinley, K. E.,
Melin, C. N., Muething, S. E., Moore, L. G., Wasson, J. H., & Nolan, T. W. (2008). Clinical
microsystems, Part 1. The building blocks of health systems. The Joint Commission Journal on
Quality and Patient Safety.

NQF Safe Practices Consensus Committee. (2010). Safe practices for better healthcare. The
National Quality Forum.

Pronovost, P. J., Berenholtz, S. M., Goeschel, C. A., Needham, D. M., Sexton, J. B., Thompson,
D. A., Lubomski, L.H., Marsteller, J. A., Makary, M. A., & Hunt, E. (2006). Creating high
reliability in health care organizations. Health Services Research, 41(4 Pt 2):1599-1617.

Radley, D. C., McCarthy, D., Lippa, J. A., Hayes, S. L., & Schoen, C. (2014). Aiming higher:
results from a scorecard on state health system performance. The Commonwealth Fund.
97

Reason, J. (2000). Safety paradoxes and safety culture. Injury Control and Safety Promotion, 7(1),
3-14.

Robust Process Improvement. (n.d.). Joint Commission Center for Transforming Healthcare.

Safety Net Medical Home Initiative. (2013). Care coordination, reducing care fragmentation in
primary care. Qualis Health, The Commonwealth Fund, Group Health Research Institute McColl
Center for Healthcare Innovation.

Singer, S. J., Gaba, D. M., Geppert, J. J., Sinaiko, A. D., Howard, S. K., & Park, C. K. (2003). The
culture of safety: Results of an organization-wide survey in 15 California hospitals. Qual Saf
Health Care, 12, 112–118.

The Joint Commission. (2013). Improving America’s hospitals: The Joint Commission’s annual
report on quality and safety; Top performers on quality measures. The Joint Commission.

The President’s Council of Advisors on Science and Technology (PCAST). (2014). Report to the
President: Better healthcare and lower costs; Accelerating improvements through system
engineering. Executive Office of the President President’s Council of Advisors on Science and
Technology.

Todorov, V., & Filzmoser, P. (2010). Robust statistic for the One-way MANOVA. Computational
Statistics & Data Analysis, 54(1), 37-48.
98

Weiner, B. J., Shortell, S. M., & Alexander, J. (1997). Promoting clinical involvement in hospital
quality improvement efforts: The effects of top management, board, and physician leadership.
Health Services Research, 32(4), 491–510.

Wilson, K. A., Burke, C. S., Priest, H. A., & Salas, E. (2005). Promoting health care safety through
training high reliability teams, Qual Saf Health Care, 14, 303–309.

99

IX. Appendices

Appendix A: Dependent Variables Code Names in SPSS

Dependent Variables Names Dependent Variables Code Names in SPSS
Length of Stay LOS
Fall Fall
Pressure Ulcer PU
Post Op Sepsis SEPS
Post Op Respiratory Failure RESP
Accidental Puncture or Laceration ACCPULA
30 Day Readmission READMT30
7 Day Readmission READMT7
Rate Hospital RATEHOSP
Communication with Nurse COMWNUR
Responsiveness of Hospital Staff RESPHOSPSTAF
Communication with Doctors COMWDOC
Cleanliness CLEAN
Quietness QUIET
Pain Management PAINMNGT
Communication About Medicine COMABMED
Discharge Information DISCH
Care Transition CARETRANS
Recommend the Hospital RECOMHOSP
Mortality MORTAL

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Designing health care provider payment systems to reduce potentially preventable medical needs and patient harm: a simulation study

PDF

Human and organizational factors of PTC integration in railroad system and developing HRO-centric methodology for aligning technological and organizational change

PDF

Using a human factors engineering perspective to design and evaluate communication and information technology tools to support depression care and physical activity behavior change among low-inco...

PDF

A series of longitudinal analyses of patient reported outcomes to further the understanding of care-management of comorbid diabetes and depression in a safety-net healthcare system

PDF

Total systems engineering evaluation of invasive pediatric medical therapies conducted in non-clinical environments

PDF

Developing an agent-based simulation model to evaluate competition in private health care markets with an assessment of accountable care organizations

PDF

A risk analysis methodology to address human and organizational factors in offshore drilling safety: with an emphasis on negative pressure test

PDF

Delivering better care for children with special health care needs: analyses of patient-centered medical home and types of insurance

PDF

Investigating racial and ethnic disparities in patient experiences with care and health services use following colorectal cancer diagnosis among older adults with comorbid chronic conditions

PDF

Simulation modeling to evaluate cost-benefit of multi-level screening strategies involving behavioral components to improve compliance: the example of diabetic retinopathy

PDF

Evaluation of factors influencing Los Angeles Tiered-Dispatch System’s improvement on bystander CPR rate and inter reliability between electronic patient care report (ePCR) and 911 call review on...

PDF

A theoretical framework and mixed-methods investigation of document status as a social determinant of emergency department utilization…

PDF

Supporting a high value maternity system of care: prioritizing resilience of and relationships with mothers to improve maternal and child health

PDF

Towards health-conscious spaces: building for human well-being and performance

Asset Metadata

Creator Massoumi, Sanaz (author)

Core Title Investigation of health system performance: effects of integrated triple element method of high reliability, patient safety, and care coordination

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Industrial and Systems Engineering

Publication Date 11/14/2017

Defense Date 09/22/2017

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag care coordination,healthcare systems,high reliability,OAI-PMH Harvest,patient safety,performance improvement,quality of care

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Meshkati, Najmedin (committee chair), Capron, Alexander (committee member), Lumb, Philip D. (committee member), Rahimi, Mansour (committee member)

Creator Email sanazmas@usc.edu,suny.massoumi@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-454076

Unique identifier UC11264987

Identifier etd-MassoumiSa-5908.pdf (filename),usctheses-c40-454076 (legacy record id)

Legacy Identifier etd-MassoumiSa-5908.pdf

Dmrecord 454076

Document Type Dissertation

Rights Massoumi, Sanaz

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA