Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Informing the exam room: understanding the process of translating evidence-based medical research into clinical care
(USC Thesis Other)
Informing the exam room: understanding the process of translating evidence-based medical research into clinical care
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
INFORMING THE EXAM ROOM:
UNDERSTANDING THE PROCESS OF TRANSLATING EVIDENCE-BASED
MEDICAL RESEARCH INTO CLINICAL CARE
by
Katherine Anne Elder
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(COMMUNICATION)
August 2017
Copyright 2017 Katherine Anne Elder
ii
“Wherever the art of medicine is loved, there is also a love of humanity.”
- Hippocrates
iii
DEDICATION
For Jake, Lawrence, and my parents: my human reminders of the importance of high
quality medicine.
iv
ACKNOWLEDGMENTS
It may take a village to raise a child, but an entire universe of academics, policy
makers, educators, and friends and family were essential to the completion of this
dissertation. I am endlessly grateful to the intellectual community that has encouraged me
to explore, to fail and flourish, and to work with my own interpretation of what it means
to be a scholar. Peter Clarke, Patricia Riley, Tom Goodnight, and Tom Valente have been
hugely important sources of wisdom during my time at USC. In particular, my adviser,
Lynn Miller, has been unendingly supportive and instructive. I have learned a great deal
from her, not only about statistics, but about how to design the perfect kitchen.
My former colleagues at the U.S. Department of Health and Human Services are
intrinsically motivated to do good. Their work matters, even though it often goes
unrecognized and underappreciated. I must thank Dr. Robert M. Kaplan in particular and
attribute much of the motivation behind this project to him. Dr. Kaplan has taught me that
brilliance and humility are not mutually exclusive, and the presence of people like him in
our government gives me great hope.
I also must thank Christine Yin, M.D. for her insight, wisdom, and substantial
contributions to the design of this project.
I’ve been told that writing a dissertation can be a lonely process, but mine has
been anything but. I am indebted to the support of my brilliant and talented friends: Dr.
LeeAnn Sangalang (whose dissertation I had next to mine during the entire drafting
process to ensure I was on the right track), Alex Gilinsky, Sadé Campbell, Denise
McKinney, Donna Hoffman, Leila Bighash, and Leanne Juzaitis have made me laugh,
calmed me down, and supported me in all of my life choices, good and bad. Brandon
v
Golob in particular has been my teammate throughout all of this. I am in awe of his
wisdom and humbled by his friendship and seemingly limitless tolerance of and patience
with my nonsense. I could not have done this without him, and I am fairly certain he
knows that.
Finally, I must thank my family. My bonus sister, Megan, suffers from a dreadful
affliction in which she sees the good in everything and everybody. She embraces new
experiences fearlessly, and I am endlessly amazed by her. I am lucky to have my very
own twofer in Phil: baby brother and best friend. He and his wife, other Katie, have
provided a never-ending supply of memes and pictures of cats throughout the hardest
parts of the writing process, and they only screen about 15% of my calls, for which they
should be awarded a Nobel Peace Prize. My parents are the definition of #goals.
Everything I learned about thinking, writing, educating, and just being a human being, I
learned from them. My Mom handles every major life change with grace and strength,
and my Dad takes on everything that comes his way with the conviction that he can
master anything. Without these five people, I would be living on a beach somewhere,
spending all of my time listening to Sublime records, eating burritos, and surfing. I do not
know whether to thank or admonish them for this.
vi
TABLE OF CONTENTS
Epigraph ii
Dedication iii
Acknowledgments iv
List of Tables viii
List of Figures ix
Abstract x
Chapter One: Introduction 1
Health Care in the United States 4
Comparative Effectiveness Research 6
Types of CER 8
Project and Chapter Summaries 9
Chapter Two 10
Chapter Three 10
Chapter Four 11
Chapter Five 11
Chapter Two: Background 12
Translational Research 12
Translational Research in Public Health and Health Care 13
Translation of CER 18
Influences on Physician Practice Patterns 22
Personal Characteristics 23
Institutional/Economic Dynamics 26
Information Processes 29
Chapter Three: Method 31
Sampling Procedure 31
Sample Characteristics 31
Case Study 32
Measures 34
Case 34
Personal Characteristics 36
Information Processes 38
Institutional/Economic Dynamics 38
Demographics 41
Analytic Procedure 41
vii
Chapter Four: Results 44
Preliminary Analyses 44
Research Questions 45
Personal Characteristics (RQ1) 45
Institutional/Economic Dynamics (RQ2) 48
Information Processes (RQ3, RQ4) 51
Chapter Five: Discussion 55
Personal Characteristics and Translation 56
Institutional/Economic Dynamics and Translation 57
Information Processes and Translation 58
Limitations and Future Directions 59
Concluding Thoughts 61
Bibliography 63
Appendices
Appendix A: Survey 72
viii
LIST OF TABLES
Table 2.1: Landmark Clinical Trials and Rates of Use by 2000 13
Table 2.2: Trait-Based Typology of Physicians 25
Table 4.1: Bivariate Pearson Correlations Between Likeliness to
Change Scores 44
Table 4.2: Deviations from Likeliness to Change Averages 45
Table 4.3: Bivariate Pearson Correlations Between Physician
Personality Trait and Average Likeliness to Change 47
Table 4.4: Estimated Marginal Means of Likeliness to Change Given a
Particular Source 51
Table 4.5: Bivariate Pearson Correlations Between Likeliness to
Change Given a Source and Source Credibility 54
ix
LIST OF FIGURES
Figure 1.1: Role of CER 2
Figure 2.1: Pipeline of Research Attrition 15
Figure 2.2: Stakeholder-Based Influence Model of Translation of CER 21
Figure 4.1: Visualization of Confidence Intervals of Likeliness to
Change Given a Particular Source 52
x
ABSTRACT
This project examines the processes and motivations for different stakeholders to
either facilitate or impede the translation of comparative effectiveness research into
patient care. This project proposes a stakeholder-based theoretical framework for this
process and examines one stakeholder in particular – physicians – to understand the
effects of their translation-related decision making. Cardiologists were presented with a
case and were asked a variety of questions that can loosely be sorted into three
categories: personal characteristics, institutional/economic dynamics, and information
processes. The results indicate that the source of information (including perceived
credibility of those sources) matters greatly to cardiologists when deciding whether to
make a change in practice. Results also suggest that cardiologists who feel rushed or
stressed at work are less likely to make practice-based changes, regardless of source of
information. Other economic, institutional, and personal characteristics were not
significantly correlated with the decision to change practice.
1
CHAPTER ONE: INTRODUCTION
Primum non nocere
First, do no harm
In the past decade, American politicians and policymakers began to see a need for
what is referred to in the United States as comparative effectiveness research (known in
other contexts as health technology assessment or evidence-informed policymaking)
(Chalkidou et al., 2009; Denis, Hébert, Langley, Lozeau, & Trottier, 2002; Tunis,
Benner, & McClellan, 2010). Though definitions vary, comparative effectiveness
research (CER) generally refers to an evaluation of the clinical effectiveness of two or
more medical treatments/procedures or drugs that are used to treat the same condition, the
purpose of which is to inform decision-making – starting at high policy levels and
funneling down to interpersonal communication between doctors and patients (Chalkidou
et al., 2009; Dreyer et al., 2010). CER is based on the findings of initial clinical research
and, ultimately, has the same potential to affect health outcomes and practices. Indeed,
evidence-based medical research has resulted in a variety of medical and health care
innovations (Conway & Clancy, 2009; Sussman, Valente, Rohrbach, Skara, & Pentz,
2006). Figure 1.1 summarizes this conceptualization of the role of CER.
2
Figure 1.1 Role of CER
CER is hotly contested in political circles, but is clearly an important component
of an effective and efficient health care system (Chalkidou et al., 2009; Jakicic et al.,
2015; Sox & Greenfield, 2009). Unfortunately, CER in the public sector is little
understood and utilized (Conway & Clancy, 2009; Umscheid, Williams, & Brennan,
2010). CER is based in part on the reality that many patients (and, indeed, clinicians) find
it difficult to sift through an ever-increasing body of research. CER promises to enable
effective, up-to-date decision-making in clinical settings (Gabriel & Normand, 2012).
The focus of this dissertation is to address one critical gap: understanding the
stakeholder-based factors that impact the process of translating CER into clinical care.
Converting research into practice in the U.S. health care system is not only a
matter of considerable importance, recognized by policy makers, researchers, and
practitioners alike, but also a process in which there is considerable room for
improvement. This process is known as translation. Translation research is unique and
requires special skills, as it not only involves academically-trained staff and researchers
with a background in fieldwork, but also requires that researchers communicate with
3
individuals in other roles and fields (Sussman et al., 2006). There are two types of
translational research identified in the literature: Type I, or “bench to bedside,” refers to
converting research-produced knowledge into a product for testing on human subjects,
and Type II, which refers to converting research-produced knowledge into clinical and
health care practice. Both of these translation processes are complicated by a variety of
time lags and delays (Zoë Slote Morris, Wooding, & Grant, 2011; Woolf, 2008).
Regardless of what type or phase is reflected in translational research, the consequence of
this research-practice gap is disconcerting and potentially life-threatening.
A recently accepted and deeply troubling estimate is that it takes 17 years to
translate 14% of original research into practice (E. a. Balas & Boren, 2000). This is
especially alarming considering that this lag is often applied to biomedical and health
research, with implications for patient care (E. a. Balas & Boren, 2000; L. W. Green,
Ottoson, García, & Hiatt, 2009; L. W. Green, 2009; Z. S. Morris, Wooding, & Grant,
2011; Stross & Harlan, 1979). The two disquieting figures estimated in this time lag –
that referring to the number of years it takes to publish work and that referring to how
little original research gets implemented – are separate, but related, problems. First, the
consequential nature of this hypothesized lag cannot be overstated; any gap in time
between the discovery of new health and biomedical information and its application by
practitioners deprives patients of the benefits of cutting edge medical research (Becker,
1970). Quality health care, in essence, is dependent on achievements made in original
medical research (E. a. Balas & Boren, 2000). There are troubling implications of how
little research eventually makes it into the hands of practitioners. Principal among them is
that, based on the attrition process described in the present paper, original research that is
4
eventually published may not be representative of, or applied to, different populations
and contexts (L. W. Green, 2009).
Health Care in the United States
In the United States in 2014, life expectancy at birth was 78.8 years (76.4 for
males and 81.2 for females). Between 2004 and 2014, infant mortality decreased by 14%.
In the same period, life expectancy for Black Americans increased more than that of
White Americans, narrowing the gap between the two to 3.4 years. In 2014, the ten
leading causes of death in the United States (accounting for 74% of the 2.6 million deaths
that year) were heart disease, cancer, chronic lower respiratory diseases, unintentional
injuries, stroke, Alzheimer’s disease, diabetes, influenza and pneumonia, kidney disease,
and suicide (National Center for Health Statistics, 2016).
Although reforming health care in the United States has been and continues to be
a political and policy priority, one area of general agreement is the need to address gaps
in quality and efficiency, in addition to curbing staggering costs. Some data suggest that
nearly 30% of U.S. health care spending reflects medical care of questionable value, and
the National Academy of Medicine (NAM, formerly the Institute of Medicine or IOM)
estimates that fewer than 50% of treatments delivered in the U.S. are evidence-based
(Tunis et al., 2010). In 2014, 8.2% of people living in the U.S. reported either delaying
or not receiving medical care due to cost. In the same year, 5.6% did not receive
prescription drugs for the same reason (National Center for Health Statistics, 2016).
In 2010, President Obama signed into law the Patient Protection and Affordable
Care Act (“Affordable Care Act” or “ACA”) that was designed to expand insurance
5
coverage to millions of uninsured people living in the United States. Signature
components of this legislation include (1) the creation of state-based health insurance
exchanges through which people can purchase coverage, with income-based credits
available, and (2) an expansion of Medicaid to 133% of the federal poverty level (Kaiser
Family Foundation, 2013). By the beginning of 2014, many of the ACA’s major
provisions were in effect. Between 2013 and mid-2015, the percentage of uninsured
adults aged 18-64 decreased from 20.5% to 12.7%. In 2014, the personal health care
expenditure breakdown was as follows: 33.9% by private health insurance, 22.7% by
Medicare, 17.4% by Medicaid, and 12.9% out-of-pocket. The remaining expenditures
were paid for by other types of programs (National Center for Health Statistics, 2016).
Coinciding with (and, in part, due to) these improvements in coverage is an
increase in cost. In 2014, personal health care expenditures in the U.S. increased by 5%
from 2013, totaling $2.6 trillion. Similarly, per capita health care expenditures increased
from $7,727 in 2013 to $8,054 in 2014. Prescription drug expenditures increased by
12.2% between 2013 and 2014, totaling $297.7 billion (National Center for Health
Statistics, 2016). Indications suggest that this trend will continue; in 2015, U.S. health
care expenditures reached $3.2 trillion, or $9,990 per capita. This situates health care
spending as accounting for 17.8% of the U.S. economy in 2015 (Centers for Medicare &
Medicaid Services, 2016). This proportion is uncommon; in 2013, the United States’
health care expenditure as a share of its gross domestic product (GDP) was almost 50%
more than France, the next-highest spender (17.1% and 11.6%, respectively) (Squires &
Anderson, 2015).
6
Comparative Effectiveness Research
In recent years, several countries, including the United States, have established
agencies that are tasked with evaluating health care options to inform policymaking. In an
article examining these agencies in Britain, France, Australia, and Germany, Chalkidou et
al. (2009) found that, though each of the four countries approached evidence-based
medical research in a way that would fit within its own unique health system, there were
several elements the countries’ approach had in common that predicted the successful
implementation of CER. These elements include: (1) inclusiveness of stakeholders, (2)
transparency, (3) ability to change and accommodate a shifting environment, and (4)
independence of the implementing agency from the government. The researchers also
found that all countries had in common the reality that CER was designed to meet the
needs of various stakeholders: public and private payers, patients, physicians, and
policymakers.
The inclusion of CER in the U.S. health policy apparatus has not been as smooth
a process as in other countries, and the need for CER in the United States has recently
become the source of political contention (Chalkidou et al., 2009). Despite that,
policymakers have recognized the importance of this evidence-based research for
decades, with the Office of Technology Assessment publishing a statement in 1994 on
the importance of the Agency for Healthcare Research and Quality (AHRQ) – an
operating division of the U.S. Department of Health and Human Services – whose
primary objective is to support CER. The impetus behind funding CER was recognized
again in 2003, when officials from AHRQ and the Centers for Medicare and Medicaid
Services (CMS) drew attention to issues preventing health care decision-makers and
7
stakeholders from making informed choices; namely, that the major sources of funding
for clinical research (the National Institutes of Health and the private medical products
industry) do not endeavor to ensure that research conducted will assist in clinical
decision-making. In 2007, the Congressional Budget Office testified before the House
Ways and Means Subcommittee on Health to the potential benefits of CER, both on
ensuring positive health outcomes and cost effectiveness in the health care system
(Chalkidou et al., 2009). It is important to note, however, that CER efforts undertaken by
the U.S. Government do not address the relative cost effectiveness of different treatments
or procedures, though the American College of Physicians and others have called for a
cost-effectiveness analysis to be included in CER (Weinstein & Skinner, 2010).
Policy support for CER has increased since President Obama took office. In 2009,
the American Recovery and Reinvestment Act (“ARRA” or “Recovery Act”) provided
$1.1 billion for CER. $300 million of these funds went to AHRQ, with the remaining
$800 million split evenly between DHHS and NIH (Conway & Clancy, 2009). The
Affordable Care Act expanded AHRQ’s CER efforts even further by establishing the
Patient-Centered Outcomes Research Institute (PCORI). PCORI is comprised of non-
government individuals representing varied clinical health expertise and assists in
informed and appropriate decision-making about health care by supporting evidence-
based medical and health research. CER activities have focused on a number of priority
conditions to the American public, including cancer, cardiovascular disease (e.g., stroke,
hypertension), mental health disorders, diabetes, infectious disease, obesity, and
substance abuse (PCORI, 2017). Funding for PCORI will sunset in 2019 unless
reauthorized by Congress (Jakicic et al., 2015). In addition to government-affiliated
8
groups, integrated health systems and managed care organizations (e.g., Kaiser
Permanente) have cause to support CER and establish related centers, and some have
done so due to the potential return on investment of this type of work (Umscheid et al.,
2010).
Although AHRQ sponsors and encourages CER initiatives, there is no CER body
in the U.S. Instead, the CER environment is more accurately conceptualized as research
organizations generating CER, as opposed to a central CER decision-making entity. This
should not suggest, however, that establishing such an entity has not occurred to
policymakers. In 2007 and 2008, legislation was introduced several times in both the
House and Senate that would establish a body whose primary task would be to
disseminate CER information (Chalkidou et al., 2009). That said, it is unlikely that, given
the current political climate in which many urge less government involvement in the
health care system, there will be no serious attempts to establish a public CER entity in
the U.S. anytime soon.
Types of CER
Though scholars agree on the importance of rigor in methodological approach to
this kind of research, CER has taken the form of a review of existing information (using
methodologies such as systematic reviews and meta-analyses), of nonexperimental
observational information (using methodologies such as case studies, electronic medical
records, and claims), and experiments such as randomized clinical trials (RCTs) (Dreyer
et al., 2010; Jakicic et al., 2015; Luce et al., 2009). Further, there are at least three types
of RCTs that can be applied to CER: 1) cluster RCTs (in which patient care units are
9
randomized), 2) pragmatic trials (large trials in which patients are enrolled into
treatments that are typical of clinical practice), and 3) adaptive trial design (a relatively
new approach that involves discontinuing one comparison as soon as it has been found to
be less effective than its comparator) (Jakicic et al., 2015). Researchers differ on the ideal
means of conducting CER; some suggest that RCTs are the most rigorous available
approach to generating this evidence (Luce et al., 2009), while others question claims of
generalizability from RCTs (Davis & Howden-Chapman, 1996). Others argue that
observational research must complement clinical trials in order to encompass a broad
range of patients and to generalize results (Blanco, Rafful, & Olfson, 2013; Dreyer et al.,
2010). Others yet endorse meta-analysis as a particularly valuable form of CER that can
account for a broad range of intervention effects (Conn, Ruppar, Fagin, Phillips, &
Chase, 2012).
Project and Chapter Summaries
The focus of this dissertation is to understand the stakeholder-based factors that
impact the translation of CER into clinical care. One stakeholder – physicians – is
selected for this project. This is because physicians must undoubtedly be involved in the
translation of CER, as they have the role of interfacing with patients. A CER case based
on a real-world scenario is provided to cardiologists, who are then asked to reflect on
their motivations for expediting, delaying, or blocking that research. Cardiologists were
selected not only because of the case study chosen for this project, but because there is a
paucity of research examining cardiologist motivations and practice behavior. This
project is a pilot study and is exploratory in nature; it includes a new theoretical model, a
10
CER-specific survey created for this project, and a focus on a small, infrequently – and
often difficult to assess – population.
Chapter Two
The second chapter is comprised of two components. The first provides a
definition of translational research and summarizes the research that employs this term to
explain public health and health care outcomes. The discussion then turns to prior studies
that have examined the translation of CER specifically. Incorporated in this discussion is
a framework developed for this project that hypothesizes the stakeholder-based
motivations and interactions that may explain the translation of CER into clinical care.
The second component of chapter two focuses on the influences on physician
practice patterns. Prior literature suggests that influences on physician practice patterns
may fall into at least one of the three following categories: personal characteristics,
institutional/economic dynamics, and information processes. Research questions
examining how and why physicians decide to make changes to their medical practice are
presented in this segment of the chapter.
Chapter Three
The third chapter provides a methodological overview of this study, including the
sampling procedure and sample characteristics, a description of the case study and
measures used in the survey, and a summary of analytical procedures used to address the
research questions outlined in chapter two. A survey is distributed online to a sample of
cardiologists who are asked to provide information about their practice and respond to a
cardiac-specific CER case study.
11
Chapter Four
The fourth chapter details the results of analyses of the research questions as
described in chapters two and three. A series of descriptive statistics, bivariate
correlations, and linear mixed-effects models are employed for this analysis.
Chapter Five
The final chapter contextualizes the findings of this study both in reference to the
theoretical framework hypothesized in chapter two and to our understanding of the
translation of CER results in the United States.
12
CHAPTER TWO: BACKGROUND
The gap between what we know and what we do in public health is lethal to Americans, if
not the world. – David Satcher, MD, PhD, Former U.S. Surgeon General
Translational Research
“Translational Research” first appeared in PubMed in 1993, and since then,
scholars have increasingly taken on studying the concept (Demaria, 2013). Rubio et al.
(2010) compiled working definitions for many terms associated with translational
research. Basic science (or, alternatively, basic research), according to the American
Cancer Society, refers to laboratory research that provides a foundation for clinical
research. The National Science Foundation points out that this research is undertaken
without thought of application to practical use. Clinical research, according to the NIH,
encompasses research in a variety of more practical fields (e.g., epidemiology, health
services, behavioral health, and patient-centered research).
Despite the increase in studies involving translational research in the past few
decades, there does not appear to be a unifying definition of the term (Demaria, 2013;
Rubio et al., 2010; Woolf, 2008). The general conclusion is that translation refers to
converting information from basic science to clinical settings and, ultimately, to public
health outcomes (Drolet & Lorenzi, 2011; Marincola, 2003; Pober, Neuhauser, & Pober,
2001; Rubio et al., 2010; Sussman et al., 2006). Some refer to translational research as
capturing “bench-to-bedside” activities, which generally refers to converting basic
science into new treatments, devices, and/or drugs for testing on human subjects. This is
also known as Type I translation. Type II translation, on the other hand, refers more
broadly to converting medical research into practice, decision-making, and public health
outcomes (Rubio et al., 2010; Woolf, 2008). Both types are important and are most
13
successful when combined, but Type I research tends to eclipse Type II research in the
United States (Woolf, 2008).
Translational Research in Public Health and Health Care
In 2000, Balas and Boren published a review that claims that there is a 17-year lag
in translating 14% of original research into health practice. To help support this claim, the
authors summarized landmark clinical trials and rates of use by the year 2000, presented
in Table 2.1.
Table 2.1 Landmark Clinical Trials and Rates of Use by 2000 (Balas & Boren, 2000).
Clinical Procedure
Landmark
Trial
Rate of Use
(2000)
Time Passed
(Years)
Flu vaccination 1968 55% 32
Thrombolytic therapy 1971 20% 29
Pneumococcal vaccination 1977 35.6%
23
Diabetic eye exam 1981 38.4% 19
Beta-blockers after MI 1982 61.9% 18
Mammography 1982 70.4% 18
Cholesterol screening 1984 65% 16
Fecal occult blood test 1986 17% 14
Diabetic foot care 1993 20% 7
Whether or not the gap exists is not a question: countless researchers, advocates,
and even government agencies have taken on the task of closing this gap (Balas & Boren,
2000; Berwick, 2003; Denis, Hébert, Langley, Lozeau, & Trottier, 2002; Green, Ottoson,
García, & Hiatt, 2009; Green, 2009). In fact, translation is now an integral component in
much health-related federal funding (Lenfant, 2003; Sussman et al., 2006). The value of
attributing blame does not appear to be under question, either, with researchers
14
distributing fault between stubborn practitioners set in their ways and scientists assuming
that their published results will be put into use (L. W. Green et al., 2009; L. W. Green,
2009).
Balas and Boren (2000) present a pipeline of research attrition (Figure 2.1) in
which they describe the steps involved between conducting original research and
implementing its findings. This pipeline suggests that an overwhelming majority of
original research is never translated into practice. The first step of this process involves
preparing original research for submission to a publication. Even at this early step, an
estimated 18% of research never gets submitted, often because a researcher assumes that
negative results are unpublishable. This is troubling and is not inclusive of what a
practitioner might find to be helpful or useful information (L. W. Green, 2009).
15
Figure 2.1 Pipeline of Research Attrition (Balas & Boren, 2000).
The next significant leak in the pipeline occurs between submission and
acceptance, in which an estimated 46% of studies submitted are not accepted for
publication and therefore no longer have potential for eventual implementation. It is
suggested that these studies get rejected for publication largely due to issues with
sampling, statistical power, and statistical design (L. W. Green, 2009).
There is no attrition between acceptance and publication, and the time lag is
relatively minimal. However, there is an estimated 35% attrition rate between publication
16
and indexing in bibliographic databases. A considerable portion of the estimated 17-year
time lag between research and practice occurs at the next stage, between bibliographic
indexing and inclusion in reviews, guidelines, and textbooks. Balas and Boren suggest
that this time lag ranges from 6-13 years, with only half of the indexed studies surviving
for eventual inclusion. The original research that has survived all of these stages does
eventually get used by practitioners, but not for over 9 more years (L. W. Green, 2009).
Taken as a whole, the pipeline model suggests that there are many procedural
barriers to dissemination of evidence-based knowledge and information. One
bureaucratic barrier to the quick translation of research into evidence lies in the structure
of the American policy making apparatus. The production and initial dissemination of
research is largely organized through centralized, or federal, institutions. Activities
occurring at this level include funding of research, bibliographic indexing, and
dissemination. The actual implementation of this research, however, occurs largely at a
decentralized level, through practitioners and policy makers at state and local levels. This
gap, then, may partly be due to geographic and organizational distinctions (L. W. Green
et al., 2009).
There are many other variables that are likely to impede the translation of
research into patient care. One is coverage: coverage decisions may not allow for a
presentation to patients of all available options for treatment (Shah et al., 2010). There
are other hypothesized contributors to the rate at which an innovation in health care is
adapted (borrowing from Diffusion of Innovations). One in particular is the perception of
the innovation itself and its relative advantage over alternatives (Berwick, 2003; Des
Jarlais et al., 2006). Other barriers proposed include a lack of a dissemination plan in
17
clinical trial designs, marketing programs, and problems with continuing medical
education (CME) information presentation (Avorn & Fischer, 2010). Other researchers
yet posit that communication issues between researchers in different fields, a lack of
familiarity of the research among practitioners, and even a failure in composition of
clinical guidelines contribute as barriers to translation (Sussman et al., 2006).
Approaches to studying the translation of health-related research into practice
hold some assumptions, including that adopters and adopting systems act as rational
actors (Denis et al., 2002) and that the relative advantage of the recommendations borne
from CER impact the rate and scope of its translation (Des Jarlais et al., 2006). In reality,
of course, these approaches and the assumptions that accompany them are rarely
generalizable (Denis et al., 2002). For example, Des Jarlais et al. (2006), on the basis of
two case studies (D.A.R.E. and syringe exchange), propose that there are at least three
models that explain the relationship between a public health innovation and its
translation: 1) diffusion on the basis of firm evidence of effectiveness, 2) diffusion
without evidence of effectiveness (likely to occur during a public health crisis situation),
and 3) lack of diffusion despite evidence of effectiveness.
Literature on how to close the translation gap abounds. Researchers suggest that
updating information technology and computerized systems, in addition to the growing
popularity of online e-journals, will help close the gap and call into question the integrity
of research published in traditional peer-reviewed journals (E. A. Balas, 2001; Cain &
Mittman, 2002). Others suggest that academic medical centers are the key to conducting
translational research because they include both clinicians and laboratory-based scientists
(Pober et al., 2001). Another approach is a five-phase approach to research that is
18
inclusive of translation efforts from the outset: 1) basic research, 2) methods
development, 3) efficacy trials, 4) effectiveness trials, and 5) dissemination trials. This
approach, however, fails to account for how a program or product can be applied in real
world settings from the beginning of the development of that program (Sussman et al.,
2006).
Translation of CER
Interestingly, tackling the problem of translating forms of CER into practice is not
new to the past few decades, nor even the past century. In the mid-1800s, Ignác
Semmelweis in Europe and Oliver Wendell Holmes Sr. in the United States conducted
their own form of CER related to hand washing by health professionals before delivering
babies and found that maternal death and infection were reduced when clinicians washed
their hands. Despite these findings, decades passed before clinicians incorporated these
practices into patient care (Avorn & Fischer, 2010; E. a. Balas & Boren, 2000). Even
today, there is an incorrect assumption among some that, as soon as comparative
effectiveness research is published that can improve efficiency, quality, and cost-
effectiveness of care, those findings will be immediately integrated by clinicians into
patient care (Avorn & Fischer, 2010; Macintyre, 2012).
Although CER is generally discussed in the context of generating knowledge, the
value of CER lies in the ability to translate this new knowledge into clinical practice
(Shah et al., 2010). In fact, AHRQ has funded the Eisenberg Center, which was founded
to translate CER results into short, easy-to-read summaries for consumers, clinicians, and
policy makers (although it is unclear whether these summaries have been used to
19
translate CER research into patient care) (Shah et al., 2010). There have been some
studies that focus specifically on translating the results from CER into practice and
patient care. In one pilot study, Shah et al. (2010) sought to determine whether decision
aids would be effective in providing information on diabetes medication to patients. Their
results indicated that, though patients exposed to the decision aid condition were more
aware of and knowledgeable about their treatment options, adherence to treatment, the
study’s main outcome, was not impacted by exposure to decision aids. Physician report
cards have also been recommended as a potential means to push for more evidence-based
decision-making as it relates to CER, though the effectiveness of these report cards is
inconclusive (Avorn & Fischer, 2010). Other mechanisms that may be used to translate
CER into patient care may include clinical guidelines and continuing medical education
(CME) (Shah et al., 2010).
In addition to researchers, the Institute of Medicine has advocated for the
inclusion and meaningful participation of stakeholder involvement in many aspects of
CER, including setting priorities and disseminating results (Devine et al., 2013; Sox &
Greenfield, 2009). Scholars recognize this need as well, stating that stakeholders are
likely not only to have different roles in the process (e.g., translational research has its
origins in the pharmaceutical industry), but different motivations as well (Lean, Mann,
Hoek, Elliot, & Schofield, 2008). The relationships between these stakeholders are ever-
changing, representing feedback loops between basic and applied research, public
opinion and the media, political context, and public health initiatives, to name a few
(Macintyre, 2012; Sussman et al., 2006).
20
There are two existing stakeholder-based frameworks for translating CER
relevant for discussion, both of which focus on stakeholder engagement as an optimal
outcome. The first project defines stakeholder as “an individual or group who is
responsible for or affected by health- and health care-related decisions that can be
informed by research evidence” and sorts stakeholders into seven categories: patients and
the public, providers, purchasers (e.g., employers), payers (e.g., insurers,
Medicare/Medicaid), policy makers, product makers, and principal investigators
(Concannon et al., 2012). The second, less alliterative framework for stakeholder
engagement in translation of CER defines a stakeholder as “individuals, organizations, or
communities that have a direct interest in the process and outcomes of a project, research,
or policy endeavor” and sorts stakeholders into the following categories: patients and
consumers, clinicians, health care providers (e.g., hospitals), payers and purchasers,
policy makers and regulators, life sciences industry, researchers, and research funders
(Deverka et al., 2012). These definitions are conceptually similar. What these
frameworks lack, however, is a discussion of the relative influence that different
stakeholders have on each other, in addition to the other institutional, economic, and
personal characteristics that may accelerate or impede the translation of CER. Thus,
based on a review of the literature and my own personal experience in health policy at
AHRQ and DHHS, I propose a stakeholder-based model of translating CER that
hypothesizes the relationships between stakeholders, where solid lines reflect direct
communication, and dotted lines reflect influence without direct communication (Figure
2.2).
21
Figure 2.2 Stakeholder-Based Influence Model of Translation of CER
A stakeholder-based model to explain the translation of CER into practice would
likely include a variety of individuals and organizations at all levels of U.S. political and
economic life. Inclusion of these stakeholders would facilitate both evidence-based
medicine and evidence-based health policy (Dobrow, Goel, & Upshur, 2004). These
stakeholders, depending on the nature of the case of CER, might include 1)
policy/governance (e.g., Food and Drug Administration), 2) professional organizations
(e.g., American Heart Association), 3) pharmaceutical and medical device industry
groups (e.g., PhRMA), 4) political leaders and parties, and/or 5) advocacy/non-profit
groups (e.g., Susan G. Komen for the Cure). In this framework, as opposed to the two
described in the preceding paragraph, non-government organizations are separated into
22
two categories of industry and non-profit, as they are likely to have different motivations
for the translation of CER. Two stakeholders, however, must undoubtedly be involved in
any CER-related translation model: physicians, and patients. The implementation of
research results must include the acquisition and ease of use of this reliable information
by patients and physicians, and by the partnership of the two (Sussman et al., 2006). In
2006, Sussman et al. outlined a list of research questions that might target physician
behavior in understanding translation of research into practice. These suggested research
questions include gauging physician motivations or barriers to translation and comparing
physicians by specialization, to name just two. This dissertation addresses both of these
issues. For the purposes of this project, I examine the influences on and motivations of
physicians with respect to ability and willingness to make changes in their practice based
on emerging CER results. I am specifically interested in analyzing the responses of
cardiologists to CER results specific to their field for two reasons: 1) there is a dearth of
research focusing specifically on cardiologist decision-making and information-seeking
behavior, and 2) the research that has examined physician implementation of research
results suggests that variation in specialty impacts the degree to which doctors make
changes to practice. Selecting only cardiologists reduces that variation.
Influences on Physician Practice Patterns
Understanding the unique needs, motivations, and constraints for different
stakeholders (as summarized in Figure 2.2) in the translation process is an important
component to understanding how – or whether – medical research will be translated into
patient care. Physicians bear the responsibility of direct interface with patients, and their
livelihood depends on their ability to provide effective and efficient care. Staying up-to-
23
date in their field, however, presents a substantial challenge. Balas and Boren (2000)
estimate that, despite the proliferation of health-related research in the past few decades,
health care practices are not prepared or equipped to integrate this information. As a
result, textbooks are often immediately outdated. They estimate that physicians would
have to read 6,000 articles per day to stay informed and current in their specialty.
The sheer volume of medical-related research is not solely to blame in identifying
and understanding translation gaps between research and practice. This section will
highlight several other factors that are likely to impact the translation of research into
clinical care. A review of the literature leads me to conclude that these influences can be
broadly categorized into three groups: 1) personal qualities, 2) institutional/economic
dynamics, and 3) information processes
1
.
Personal Characteristics
As with any other profession, no two physicians are, or practice, alike. The
willingness of a physician to incorporate any clinical results into patient care can depend
on a variety of factors, including source credibility, the importance of practical concerns,
and their tendency to diverge from standard practice. A psychometric instrument was
developed by Green, Gorenflo, and Wyszewianski (2002) to sort physicians into four
trait-related categories based on how they respond to new clinical information. Seekers
are likely to consider systematically gathered data the most reliable source of information
and feel confident in their ability to evaluate the data themselves. Seekers are also more
willing and likely to make changes in their practice based on evidence, even if these
1
These categories are not mutually exclusive but are presented as such for clarity.
24
changes do not reflect practice norms. Receptives, similar to seekers, place a great deal of
weight on evidence. Unlike seekers, however, they are likely to rely on the judgment of
respected colleagues, researchers, and/or leaders in the field. They are willing to make
changes to practice that depart from practice norms, but only if the evidence is
compelling.
Traditionalists, the third physician typology, view practical experience as the
most compelling and credible source of evidence. This suggests that they are likely to
consult their own experience or that of trusted colleagues when making practice-based
decisions. Traditionalists are willing to make changes to their practice if those changes
are suggested by a respected leader. They are less concerned with efficiency when
integrating a new practice.
Finally, pragmatists’ practice patterns are precisely what the name suggests: they
are focused on the day-to-day needs that arise from their practice. Rather than evaluate
sources of information, pragmatists are concerned with how a proposed change will
affect their workload and efficiency. These types of physicians are summarized in Table
2.2.
25
Table 2.2 Trait-Based Typology of Physicians (L. A. Green, Gorenflo, & Wyszewianski,
2002).
Physician Typology Sources Consulted Divergence Patterns
Seekers Journals
Very willing to change based on compelling
scientific evidence
Receptives
Journals, respected
colleagues
Somewhat willing to change based on
compelling scientific evidence
Traditionalists Practical experience
Willing to change practice based on prior
experience
Pragmatists N/A
Willing to change practice based on impact on
day-to-day flow
In order to examine the impact of physician characteristics on likelihood of
changing practice, the following research question was developed:
RQ1: Does variation in physician characteristics predict, given a hypothetical
scenario, a physician’s general willingness to change practice?
There are, of course, other personal characteristics that might predict a
physician’s ability and willingness to change his/her practice based on the latest CER
results. One obvious characteristic is the ability to understand these results in the first
place. In other words, even if doctors were aware of cutting edge research being
conducted in their field and statistics relevant to their practice, they might not understand
the results or implications of the research. Greg Gigerenzer, a prominent statistician and
director of the Harding Center for Risk Literacy in Berlin, gave a series of statistics
workshops to over a thousand practicing gynecologists. In one session, only one in five of
the doctors responded correctly to a basic statistics question, which is a poorer result than
if the doctors had selected at random. This inability to understand the nature of statistical
research – particularly regarding probability – also impacts a comprehension of survival
26
rates. Gigerenzer surveyed 412 U.S. doctors and found that nearly 75% of them
mistakenly believed that higher survival rates meant more lives were saved. Framing also
played a role in the results of this survey; “doctors would recommend a test to a patient
on the basis of a higher survival rate than they would on the basis of a lower mortality
rate” (Kremer, 2014, para. 20). Reversing the process, clinicians might be too unfamiliar
with the health research apparatus to present researchers with pressing questions and
issues (Ioannidis, 2004; Marincola, 2003). These factors were not explored in this study
but should be considered in future research.
Institutional/Economic Dynamics
Understanding the institutional (e.g., type of practice) and economic (e.g.,
reimbursement mechanisms) dynamics involved in physician practice is likely to shed
light on how – or whether – results from CER are translated into patient care.
Organizational/institutional characteristics and structures are likely to have an impact on
the rate or degree to which practicing physicians incorporate CER results into practice.
Sussman et al. (2006) outlined a variety of research questions that might target the role of
the organization in translation, including: How does the organization structure impact
translation? What are the information-seeking opportunities or approaches in an
organization? Dobrow et al. (2004) describe a decision-making context, referring to
environmental factors in which decisions are made while acknowledging that it is nearly
impossible to identify all factors. They differentiate between environmental factors in an
internal decision-making context (e.g., role of participants, processes employed to arrive
at a decision) and an external, fixed decision-making context (e.g., disease-specific
epidemiology, political factors).
27
Economic dynamics are fundamental to translational research, as much of it is
conducted by industry or industry-funded centers (Ioannidis, 2004). These dynamics,
however, are also important to consider as having a potential impact on patient care. One
example of the importance of economic dynamics in translating CER results into patient
care can be found in the 2007 Clinical Outcome Utilizing Revascularization and
Aggressive Drug Evaluation (COURAGE) CER study, which evaluated the relative
effectiveness of angioplasty versus aggressive drug treatment alone in treating blocked
coronary arteries. Although the CER trial found there to be no difference in morbidity or
mortality, there were no changes or reversals in the patterns of care. Cardiologists, in this
case emblematic of clinicians who rely on performing procedures for their livelihood,
have seemingly ignored these results and continue to perform angioplasties in lieu of less
costly and sometimes safer alternatives despite a lack of evidence of comparative
effectiveness (Avorn & Fischer, 2010).
There are some studies that suggest that reimbursement mechanism (e.g., fee-for-
service, partial or full capitation, salary) may impact physician behavior (Hickson,
Altemeier, & Perrin, 1987; Hillman et al., 1998; Khullar et al., 2015), though the degree
of impact is inconclusive.
Consistent with the relationship hypothesized by the model of CER translation
presented in Figure 2.2, gauging the influence that industry representatives have on
physicians and patient care has been examined extensively (Bennett, Casebeer, Kristofco,
& Strasser, 2004; Campbell et al., 2007; Caudill, Johnson, Rich, & McKinney, 1996;
Chew et al., 2000) and is of particular interest in this study. Concerns over the impact of
industry representatives on patient care have resulted in organizations (including for-
28
profit industry companies) to develop codes of conduct that purport to emphasize benefit
to patients and discourage giving gifts to physicians (Campbell et al., 2007). There is
some indication that physicians that rely on drug samples and receive financial or
material gifts from industry representatives may be less likely to make evidence-based
practice changes (Chew et al., 2000).
Prior research documents and hypothesizes the importance of institutional
dynamics in the translation of medical research and innovations (Berwick, 2003; Dobrow
et al., 2004; Whippen & Canellos, 1991). Physicians are overwhelmed both with respect
to their time and with respect to the myriad available sources of information. This
burnout is likely to have an impact on information-seeking behavior (Bennett et al.,
2004; Clarke et al., 2013; Dawes & Sampson, 2003; Whippen & Canellos, 1991). There
is also some evidence that type of practice may contribute to burnout (which, by
extension, contributes to information-seeking). Whippen and Canellos (1991), for
example, found that institution- or university-based oncologists were less likely to record
incidence of burnout.
The following research question was developed to examine the relationship
between practice-based indicators (institutional and economic) and likelihood of a
physician changing practice:
RQ2: Does variation in type of practice variables predict, given a hypothetical
scenario, a physician’s general willingness to change practice?
Much of the institution-related research on physician practice patterns focuses on
the impact of institutional dynamics on information-seeking behavior (and, by extension,
29
practice patterns). Acknowledging this overlap, we now turn to physician information
processes as a potential predictor of practice change.
Information Processes
Despite being overwhelmed by information volume, source credibility has been
cited as an important predictor of learning (Bennett et al., 2004; Dawes & Sampson,
2003; Lagace & Twible, 1991). There is no clear consensus, however, on what source
physicians find to be the most credible; some research suggests that physicians prefer
online means of acquiring information (Kannampallil et al., 2013), while others argue for
the importance of colleagues (Clarke et al., 2013; González-gonzález et al., 2007;
Hillman et al., 1998; Leckie, Pettigrew, & Sylvain, 1996). There is also some evidence
that physicians use information that maximizes information gain, regardless of the
cognitive effort required to process the information (Kannampallil et al., 2013). This
might suggest that medical journals are viewed as the most preferred source.
An important follow-up to this discussion examines the impact of perceived
credibility of industry representatives. Lagace and Twible (1991) found that physicians’
satisfaction with information received from industry representatives is based on whether
physicians perceive the representatives to be credible. Further, Chew et al. (2000), in a
survey of 154 practicing physicians, state that physicians often used drug samples
provided by pharmaceutical representatives because of a desire to reduce costs to the
patient. However, the same study suggests that physicians may prescribe drugs that are
not their preferred choice based on accessibility, meaning that compliance with
guidelines might decrease.
30
The following research questions were developed to gauge the relationship
between source of information and likelihood of changing practice, and to examine
potential moderators of that relationship:
RQ3: Do sources of information, given a hypothetical scenario, differently predict
patterns of physician change of practice?
RQ4: If so, what is it about these sources (e.g., credibility, nature of change
suggested) that might predict response to a given source in changing physician
practice?
In summary, the present investigation examines the conditions that may predict
how – or whether – physicians translate results from CER into practice. This study
employs frameworks and data from prior studies on translation and on physician
decision-making and practice, combined with survey items developed for this study to
examine the translation of CER specifically and designed for cardiologists as a target
audience. Independent variables examined are loosely categorized as belonging to at least
one of three categories: personal characteristics of physicians, institutional/economic
dynamics, and information processes.
31
CHAPTER THREE: METHOD
“It is difficult to get a man to understand something when his salary depends on his not
understanding it.” – Upton Sinclair
The present investigation attempts to shed light on the relative influences of
information processes, institutional/economic dynamics, and personal characteristics on
physicians in predicting change in practice based on CER results. The methodological
procedure to achieve this goal is described in this chapter.
Sampling Procedure
An online survey was administered to 42 cardiologists practicing medicine in the
United States. Olson Research Group, a market research firm with a robust database of
U.S.-based health professionals that includes over 20,000 cardiologists, was contracted
with for this project. Olson recruited forty respondents; the remaining two were recruited
through my personal network. All participants completed the survey online; the forty
recruited by Olson completed the survey on the Group’s own platform, and the two
recruited through my network completed an identical survey on Qualtrics.
Sample Characteristics
All 42 cardiologists selected for this study practice medicine and provide direct
patient care in the United States, and 83.3% of them (n = 35) graduated from medical
school in either the United States or Canada. An overwhelming majority of 95.2% (n =
40) of those surveyed were men, and only 4.2% (n = 2) were women. When asked to
indicate race/ethnicity, 28.6% (n = 12) selected Asian, 2.4% (n = 1) selected Pacific
Islander, 69% (n = 29) selected White (non-Hispanic), and 2.4% (n = 1) selected Other.
Respondents were also presented with the options of African American/Black (non-
32
Hispanic), Hispanic, and Native American/Alaska Native. The earliest recorded
graduation year from medical school was 1968, and the most recent was 2007. The
average number of years certified as a cardiologist is 21.21 (SD = 7.835), and the range is
38 years. Exactly half (21) of the cardiologists fall into the highest annual income bucket
offered on the survey ($300,000 +); 11.9% (n = 5) recorded < $100,000, 9.5% (n = 4)
recorded $100,000-$150,000, 4.8% (n = 2) recorded $150,001 - $200,000, 7.1% (n = 3)
recorded $200,001 – 250,000, and 16.7% (n = 7) recorded $250,001 – 300,000.
Case Study
The case selected for examination in this project focuses on pharmaceutical-
related CER. This case was chosen for a few reasons: 1) it demonstrates clearly the role
of CER in overturning or reversing practice, 2) it relates to a recent, real-life controversy
in cardiology, and 3) the case for overturning standard practice is very clear.
The relationship between what is recommended by policy bodies and what
eventually takes place in an examination room is complicated. A variety of organizations
endeavor to influence the way a physician prescribes a course of treatment for a patient.
These organizations likely include: (1) government regulatory bodies (e.g., Food and
Drug Administration), (2) professional organizations that produce guidelines for practice
(e.g., American Heart Association), and (3) large-scale industry lobbying organizations
seeking to advance that industry’s financial bottom line (e.g., PhRMA).
β-blockers are prescribed for a number of conditions and are most commonly used
for the management of cardiac arrhythmias and hypertension. According to the annual
survey published by AHRQ, the Medical Expenditure Panel Survey (MEPS),
33
Metroprolol, a β-blocker, was ranked fifth in the total number of people living in the
United States who received prescriptions in 2011 (totaling 11.7 million people). Given
this high volume, it is crucial that researchers and clinicians understand the uses for and
potential impacts of pharmaceuticals, not only through clinical trials, but through CER
mechanisms such as meta-analyses.
Researchers have conducted research on pharmaceutical drugs for decades. One
body of initial clinical research examining the effects of pharmaceutical treatments on
health outcomes succinctly demonstrates the research concerns described in this section
and serves as the focus of this case study. This family of trials, led by Don Poldermans, is
called the Dutch Echocardiographic Cardiac Risk Evaluation Applying Stress
Echocardiography (DECREASE). Findings from this research, beginning in 1999 with
the first conducted trial, contributed to providing the justification to prescribe
perioperative β-blockers to three classes of patients undergoing high-or immediate-risk
surgery (Bouri, Shun-Shin, Cole, Mayet, & Francis, 2014). These guidelines were
published by the European Society of Cardiology (ESC) in 2009. After combining the
data from the DECREASE trials and other trials, ESC determined there to be a neutral
effect on mortality and considerable benefit of initiation of β-blockers (Husten, 2013).
It appears that, as early as 2005, researchers began to point out the flaws in the
DECREASE trials (Bolsin & Colson, 2014). It was not until 2011, however, that
Poldermans was found to have engaged in scientific misconduct, throwing into question
the reliability of the DECREASE trials. This was complicated by the fact that Poldermans
served as the chairman for the committee that drafted the guidelines for β-blocker use
(Husten, 2013). In 2013 (revised in 2014), Bouri et al. published a CER meta-analysis of
34
the non-DECREASE RCTs of β-blockers and their impact on perioperative mortality
(among other outcomes). They analyzed nine trials with a total of over 10,000 patients
and found that the neutral effect on mortality suggested by the DECREASE trials was
inaccurate. Instead, Bouri et al. found a 27% risk increase in 30-day mortality and
concluded that guidelines based in any part on DECREASE data should be retracted
(Bouri et al., 2014). Applying this calculation to National Health Service data, Bolsin and
Colson (2014) estimate 10,000 avoidable deaths per year in the United Kingdom alone
due to published guidelines using these inaccurate data. As of 2014, the ESC and the
American College of Cardiology/American Heart Association had partially revised and
updated their guidelines based on non-DECREASE data (Bolsin & Colson, 2014).
Measures
Participants completed the following questions online. Questions included
information about their practice, opinions about their field, information-seeking opinions
and behavior, and responses to the case prompt. Appendix A contains the survey in its
entirety as it was presented to respondents.
Case
The specific case referring to the DECREASE family of trials was presented at
the conclusion of the survey. In collaboration with a practicing physician, the following
case prompt was developed:
Clinical guidelines recommend the prescription of beta-blockers in several
classes of patients, including those who will be undergoing non-cardiac surgery.
You learn that the series of trials that provide justification to prescribe
35
perioperative beta-blockers in these classes of patients is insecure; much of the
data had been lost, and that which remained was found to contain serious flaws,
including, in one case, complete fabrication of a dataset. The European Society of
Cardiology has stopped recommending the use of beta-blockers in patients
undergoing non-cardiac surgery, but the American Heart Association guidelines
based on these trials have not been retracted.
Participants were then asked, “Which of the following would best describe your
reaction had you received the information about the insecurity of the clinical trials from:”
and provided with the following nine sources: professional medical journal, non-profit
patient advocacy organization, medical-related website (e.g., WebMD), colleague, CME
class, news broadcast, industry representative (e.g., pharmaceutical, medical device),
professional association (e.g., International Association of Cardiologists; U.S. Preventive
Services Task Force), and online resource for clinical guidelines (e.g., National Guideline
Clearinghouse). Respondents were asked to select one of five reactions on a Likert scale:
Not at all likely to change practice, somewhat unlikely to change practice, neutral,
somewhat likely to change practice, or very likely to change practice.
A factor analysis suggests that source items load on two factors: professional
sources (professional medical journal, colleague, CME class, industry representative,
professional association, and online resource for clinical guidelines) (α = .836), and non-
professional sources (non-profit patient advocacy organization, medical-related website,
and news broadcast) (α = .793). However, the two factors were highly correlated (r =
.581, p < .001). Additionally, the nine items combined had higher reliability (α = .866),
and reliability analysis suggested that removing any of the items would not improve this
36
score. Given that the sample for the factor analysis was small and the reliability of the
overall measure high compared to the separate measures, physicians’ average “likeliness
to change” score was computed across all nine items. Source-specific “likeliness to
change” scores were analyzed when appropriate.
Personal Characteristics
Physicians’ practice patterns and styles of response to new information were
measured by adapting the seventeen-item psychometric instrument by Green, Gorenflo,
and Wysenwianski (2002) that separates physicians into four categories that describe
their response to new information. These categories are designed to capture the essential
information-seeking and -adapting nature of physicians as opposed to the nuance of the
situations in which they make decisions (i.e., trait, not state). The authors of this
instrument posit that there are three underlying factors of these seventeen items: evidence
versus experience (i.e., whether they view scientific evidence as superior to personal
experience/authority), nonconformity (i.e., degree of comfort with deviating from
standard practice/recommendations), and practicality (i.e., the amount of importance
assigned to issues such as workflow management and patient satisfaction). Responses
were measured on a five-point Likert scale, anchored by 1 (strongly agree) and 5
(strongly disagree).
The eighteen adapted items (separating one item into two for clarity) measuring
evidence-experience were: “randomized controlled trials are the most reliable way to
know what really works”, “evidence-based medicine makes a lot of sense to me”,
“clinical experience is the most reliable way to know what really works”, “patient care
37
should be based when possible on randomized controlled trials”, “patient care should be
based when possible on the opinions of respected authorities”, “the best practice
guidelines are based on the results of randomized controlled trials”, and “evidence-based
medicine is not very practical in real patient care”. The adapted items reflecting
nonconformity were: “I am comfortable practicing in ways different than other doctors”,
“it is best to change the way I treat a certain problem when my local colleagues are
making the same changes”, “the opinions of respected authorities should guide clinical
practice”, “I am often critical of accepted practices”, “my colleagues consider me to be
someone who marches to my own drummer”, and “it is not prudent to practice out of step
with physicians in my area”. Finally, the adapted items reflecting practicality factor: “I
don’t have the time to read up on every practice decision”, “I follow practice guidelines if
they are not too much hassle”, “I am too busy taking care of patients to keep up with the
recent literature”, “I am uncomfortable doing things differently from the way I was
trained”, and “I follow practice guidelines as long as they don’t interfere too much with
the flow of patients”.
With the data collected for this study, after reverse coding when necessary (items
5, 7, 9, 13, 16, and 18), subscales were devised based on the prior literature. Factor
analysis with such a small sample size, with low average correlations among items, was
inappropriate (MacCallum, Widaman, Zhang, & Hong, 1999). The reliability for the
evidence-experience subscale was α = .424. Reliability for the nonconformity scale was α
= .616, and reliability for the practicality scale was α = .565. Because of these low
reliability figures, rather than construct new subscales based on a factor analysis of this
small sample, each item was treated independently.
38
Information Processes
Four questions were adapted from the National Survey on Medical
Professionalism at the Institute for Health Policy at Harvard Medical School (2003) that
relate to information processes. These include: “In the last 3 years, have you served as a
reviewer for a professional journal?”, “How often do you communicate with other
clinicians via email?”, “How often do you access professional journals online?”, and “On
average, how many professional journals do you read monthly?”.
A variety of credibility items were created for this study. Respondents were asked
to respond to a 5-point Likert scale anchored by 1 (not at all credible) and 5 (very
credible) to reflect their perceived credibility of the following sources: peer-reviewed
journal articles, non-profit patient advocacy organizations (e.g., Susan G. Komen for the
Cure), medical-related websites (e.g., WebMD), colleagues, CME classes, news
organizations (e.g., NBC, CBS), industry representatives (e.g., pharmaceutical, medical
device), and professional associations (e.g., American Heart Association).
Institutional/Economic Dynamics
Survey questions designed to capture the nature of a physician’s practice – both
with respect to reimbursement and to organizational structure – were adapted from the
National Survey on Medical Professionalism at the Institute for Health Policy at Harvard
Medical School (2003). To address institutional dynamics, the following items were
included: “How would you best characterize the organization of your medical practice?”,
“Do you currently hold a faculty appointment at a medical school?”, and “In a typical
work week, how much time do you spend in direct patient care services?”. An additional
39
question was created for this survey: “In a typical work week, how many patients do you
see?”.
Respondents were then asked the following yes/no questions in response to the
prompt “In the last 3 years, has either your main practice or hospital undertaken formal
initiatives in any of the following areas?”: error reduction, patient confidentiality, quality
improvement, access for the poor or underserved, and cultural competence. Respondents
were also asked the following yes/no questions in response to the prompt “In the last 3
years, have you”: participated in a formal medical error reduction initiative in your office,
clinic, hospital, or other health care setting; undergone competency assessment by a
provider organization or health plan; participated in the development of formal clinical
practice guidelines; reviewed another physician’s medical records for quality
improvement reasons; encouraged one or more patients to enroll in a clinical trial;
provided care, with no anticipation of reimbursement, in a setting serving poor and
underserved patients; provided health-related expertise to local community organizations
(e.g., school boards, parent-teacher organizations, athletic teams, local media); and
encouraged your professional society to address a public health issue or policy issue that
is not primarily concerned with physician welfare.
To gauge the impact of economic dynamics on a physician’s likeliness to make
changes in practice, the following questions were adapted from the National Survey on
Medical Professionalism at the Institute for Health Policy at Harvard Medical School
(2003): “Approximately what percentage of your patients is covered as follows:
Medicaid, Medicare, Uninsured and unable to pay”, “In your main practice, please check
the predominant reimbursement mechanism by which you are compensated for the care
40
you provide to patients”, and “Please estimate your personal total before tax
compensation for all clinical activities and services in 2016”. Respondents were also
asked to rate the extent to which they agreed with the following statements (using a 5-
point Likert scale anchored by 1 - strongly agree and 5 - strongly disagree): “physicians
should provide necessary care regardless of the patient’s ability to pay”, “physicians
should put the patient’s welfare above the physician’s financial interests”, and
“physicians should know the overall cost of care they provide”.
To address the influence of the “Industry” stakeholder on patient care
hypothesized in Chapter 2, the following questions were adapted from the National
Survey on Medical Professionalism at the Institute for Health Policy at Harvard Medical
School (2003): “In an average month, how many times do you meet with representatives
from drug, device, or other medically-related companies?”, “which of the following have
you received in the last year from drug, device, or other medically-related companies?”,
and “excluding any food, beverages, and drug samples you may have received in your
workplace, please estimate the total value of all goods and services you received in the
last year from drug, device, or other medically-related companies”. Respondents were
also asked to provide their opinion using a 5-point Likert scale anchored by 1 (strongly
agree) and 5 (strongly disagree) on the role of pharmaceuticals in patient care using the
following items: “patients tend to misuse drug samples”, “drug samples permit quicker
initiation of therapy”, “drug samples deprive the patient of pharmacists’ screening for
drug interaction”, “drug samples fulfill an educational role through demonstration”, “drug
samples are a source of medication for patients who cannot afford them”, and “supplies
of drug samples are inconsistent” (Chew et al., 2000). A composite score for “drug
41
opinion” was created by reverse coding some items so that the higher end of the scale for
responses (5) reflects a positive opinion about drug samples and use, and the lower end of
the scale for responses (1) reflects a negative opinion about drug samples and use.
Finally, after the case study was presented to respondents, the survey asked
physicians to rate the influence of following institutional/economic factors in their
decision-making: “ease of integration of change (e.g., can the practice handle this change
as is? Does it have to be referred out? Is there a need to hire or develop new expertise? Is
there a need to incorporate new technology?)”, “change in immediate cash flow”, “impact
of change on patients (i.e., does the new information promise to save lives, or does it
simply recommend an alternative treatment with similar efficacy)”, “anticipated practice-
related consequences of deviation from existing guidelines (e.g., fines)”, and “degree of
risk for patients”. These items were measured on a 5-point Likert scale anchored by 1
(not at all influential) and 5 (very influential) and were developed for the purposes of this
project.
Demographics
Respondents were asked a variety of demographic questions, including
race/ethnicity, gender, year graduated from medical school, place of graduation from
medical school, and number of years certified as a cardiologist.
Analytic Procedure
Due to the sample size and the exploratory nature of this study, analyses were
conducted using primarily descriptive and inferential statistics. In general and unless
otherwise noted below, all correlations used, as a dependent variable, a computation of
42
the mean of all nine “likeliness to change” items, as all nine of them loaded on one factor
(α = .866).
Addressing the first research question (does variation in physician characteristics
predict, given a hypothetical scenario, a physician’s general willingness to change
practice?) was frustrated by a low reliability of the subscales used in prior literature.
Because of concerns about sample size, rather than categorizing physicians as a “type” by
collapsing items into three factors, each item was examined individually. Pearson
correlation coefficients were calculated to examine the impact of these items on a general
willingness to change and on willingness to change given a particular source.
The second research question (does variation in type of practice variables predict,
given a hypothetical scenario, a physician’s general willingness to change practice?) was
examined using correlation coefficients (Pearson r or Kendall tau-b, depending on the
nature of the variables and whether or not they are normally distributed). The variables
that were correlated with the general likeliness to change score were whether the
respondent holds a faculty appointment; number of patients seen in a week; type of
organization of practice; percentage of patients covered by Medicaid, Medicare, and
uninsured/unable to pay; formal initiatives undertaken by main practice or hospital;
feelings of being rushed or stressed; opinion of drug sample/use; and the impact of the
following variables on decision-making: change in immediate cash flow, impact of
change on patients/degree of risk for patient, and ease of integration of change; a count of
the number of material or financial goods received by industry representatives; and
finally, opinions about whether physicians should put the patient’s welfare over the
physician’s financial interests.
43
For the third research question (do sources of information, given a hypothetical
scenario, differently predict patterns of physician change of practice?), I used a linear
mixed-effects model analysis to determine the interaction between type of source and
likeliness of changing practice. Finally, for the fourth question (if these sources do
differently predict patterns of physician change of practice, what is it about these sources
that might have this effect?), I used Pearson r to examine the relationship between a
respondent’s likeliness to change score for a source and perceived credibility of that
source.
44
CHAPTER FOUR: RESULTS
“To the individual who devotes his or her life to science, nothing can give more
happiness than when results immediately find practical application. There are not two
sciences. There is science and the application of science, and these two are linked as the
fruit is to the tree.”
– Louis Pasteur
Preliminary Analyses
Before examining the research questions, a bivariate correlation matrix was
produced to explore the degree to which the nine dependent variables (likeliness to
change after hearing the case from the nine potential sources) correlate. This correlation
matrix is presented in Table 4.1.
Table 4.1 Bivariate Pearson Correlations Between Likeliness to Change Source
Journal
Non-
profit
Website Colleague
CME
class
News
Industry
rep
Professional
assoc.
Online
clinical
guidelines
Journal 1.00 .247 .470** .587** .504** .264 .386* .453** .428**
Non-profit 1.00 .710** .380* .248 .480** .285 .301 .316*
Website 1.00 .410** .496** .508** .442** .550** .410**
Colleague 1.00 .613** .335* .366* .426** .327*
CME class 1.00 .318* .479** .535* .324*
News 1.00 .420* .300 .334*
Industry rep 1.00 .498** .502**
Professiona
l assoc. 1.00 .582**
Online
clinical
guidelines 1.00
Note: * p < .05 ** p < .01
45
The sources for the likeliness to change scores are highly correlated and have
positive coefficients, suggesting that a respondent was either generally likely or generally
unlikely to change practice across the board. Scores were then computed to reflect the
extent to or tendency with which a physician’s likeliness to change scores deviated from
the average score across physicians. The average deviations for all nine variables, in
addition to the composite variable combining and averaging all likeliness to change
scores per physician, were quite low. However, the range figures were generally high (at
least 3.0 in all cases but one). This indicates that physicians deviate from each other quite
a bit. These numbers are summarized in Table 4.2.
Table 4.2 Deviations from Likeliness to Change Averages
Mean SD Range
Journal Deviation .0038 .841 3.00
Non-profit Deviation .0005 .917 4.00
Website Deviation -.0043 1.00 4.00
Colleague Deviation .0029 .759 3.00
CME class Deviation .0033 .786 3.00
News Deviation .0038 1.041 4.00
Industry rep Deviation .0014 .973 4.00
Professional assoc. Deviation -.0010 .889 3.00
Online clinical guidelines Deviation .0048 .989 3.00
Combined average Deviation .0017 .636 2.56
Research Questions
Personal characteristics (RQ1)
Because of the low reliability of the predicted subscales, I analyzed all eighteen
physician trait items and their relationship with a physician’s average likeliness to change
independently of each other. As a reminder, the scores for the trait items range from 1
46
(strongly agree) to 5 (strongly disagree), and the scores for average likeliness to change
range from 1 (very unlikely) to 5 (very likely). The Pearson correlation coefficients are
reported in Table 4.3.
47
Table 4.3 Bivariate Pearson Correlations Between Physician Personality Trait and
Average Likeliness to Change
Item Hypothesized Factor r
Randomized controlled trials are the most reliable way to
know what really works.
Evidence-experience -.080
I am comfortable practicing in ways different than other
doctors.
Nonconformity .155
Evidence-based medicine makes a lot of sense to me. Evidence-experience -.222
I don't have the time to read up on every practice decision.
Practicality .038
It is best to change the way I treat a certain problem when
my local colleagues are making the same changes.
Nonconformity -.143
I follow practice guidelines if they are not too much hassle. Practicality -.021
The opinions of respected authorities should guide clinical
practice.
Nonconformity -.083
I am too busy taking care of patients to keep up with the
recent literature.
Practicality -.036
Clinical experience is the most reliable way to know what
really works.
Evidence-experience -.086
I am uncomfortable doing things differently from the way I
was trained.
Practicality .059
I am often critical of accepted practices. Nonconformity .334*
Patient care should be based when possible on randomized
controlled trials.
Evidence-experience .268
Patient care should be based when possible on the opinions
of respected authorities.
Evidence-experience
-
.333*
My colleagues consider me to be someone who marches to
my own drummer.
Nonconformity .106
I follow practice guidelines as long as they don't interfere
too much with the flow of patients.
Practicality -.051
It is not prudent to practice out of step with other
physicians in my area.
Nonconformity -.014
The best practice guidelines are based on the results of
randomized controlled trials.
Evidence-experience -.255
Evidence-based medicine is not very practical in real
patient care.
Evidence-experience .204
Note: * p < .05
48
Based on this analysis and with respect to this sample, physician trait is not a
compelling category of items in determining whether cardiologists are likely to make
CER evidence-based changes to practice or in determining what sources cardiologists
view as credible. Only two items (“I am often critical of accepted practices” and “Patient
care should be based when possible on the opinions of respected authorities”)
significantly correlate with average likeliness to change. This suggests that source of
information, as opposed to personal characteristics, may be an important predictor of
likeliness to change practice.
Institutional/Economic Dynamics (RQ2)
Testing the second research question involved a variety of variables that either fall
under the heading of institutional dynamics or economic dynamics. The results are
presented below with that distinction.
Institutional Dynamics. I first examined whether there is a correlation between
whether a physician holds a faculty appointment at a university and that physician’s
average likeliness to change practice. This relationship is not statistically significant (rτ =
-.090, p = .495). Neither is there a statistically significant relationship between the
number of patients seen in a typical work week and average likeliness to change practice
(r = -.055, p = .730). Finally, there is no statistically significant relationship between any
of the formal initiatives questions and average likeliness to change practice. This was true
for whether a physician’s practice had undertaken formal initiatives in error reduction (rτ
= -.026, p = .846), patient confidentiality (rτ = -.039, p = .769), quality improvement (rτ =
49
-.018, p = .892), access for the poor or underserved (rτ = -.041, p = .758), and cultural
competence (rτ = .117, p = .375).
Physicians were then asked to rank the degree to which they feel rushed and
stressed at work. Interestingly, there appears to be some correlation between these items
and a physician’s average likeliness to change practice. There is a significant relationship
between feeling rushed when consulting with a patient and average likeliness to change (r
= .307, p < .048). The relationship between feeling stressed at work and likeliness to
change practice approaches conventional statistical significance (r = .296, p = .057).
Along these lines, there is a significant relationship between a physician’s average
likeliness to change practice and that physician’s concern over the ease of integration of
the proposed change (r = .354, p = .021).
Economic Dynamics. I first examined whether reimbursement mechanism have
an effect on average likeliness to change practice. Of the four options available to
respondents, 31 said that the predominant reimbursement mechanism by which they are
compensated for patient care was fee-for-service; 1 respondent said partial capitation, one
said full capitation, and 9 said salary. As a result, I analyzed cardiologists who are
reimbursed by fee-for-service compared with the 11 who are not using a Kendall tau-b
coefficient used for correlating variables that are not normally distributed. This
relationship is insignificant (rτ = .026, p = .841). The Pearson correlations between the
percentage of patients covered by Medicaid or Medicare and average likeliness to change
are also insignificant (r = .042, p = .792; and r = .120, p = .449, respectively). The
relationship between percentage of a physician’s patients who are uninsured/unable to
pay and that physician’s average likeliness to change is also insignificant, though it
50
approaches significance (r = .263, p = .093). Physicians’ average likeliness to change was
not impacted by concerns of change in immediate cash flow (r = .153, p = .334) or the
degree of risk for patients (r = .071, p = .653), though the relationship between the impact
of change on patients and a physician’s likeliness to change approaches significance (r =
.268, p = .086).
I then examined a variety of variables that might explain the impact of industry
opinions and involvement on both a physician’s likeliness to change practice and a
physician’s likeliness to change practice given information from an industry
representative. I first looked at whether count data reflecting material or financial goods
received from industry representatives correlates with average likeliness to change. This
relationship is insignificant (r = .093, p = .557), as is the relationship between goods
received and likeliness to change given that the source of information was an industry
representative (r = -.098, p = .537). Next, I correlated the variables of the drug opinion
composite (where a higher score reflects a more positive attitude toward drug samples
and use) and average likeliness to change. This relationship is significant (r = .356, p =
.021). The relationship between the drug opinion composite and likeliness to change
given an industry representative as a source of information is also significant (r = .395, p
= .01).
Finally, when gauging physicians’ opinions on how physicians should behave
when weighing the needs of their patients with their own financial needs, there was no
significant relationship between average likeliness to change and responses to the
statement “physicians should provide necessary care regardless of the patient’s ability to
pay” (r = -.059, p = .712), nor with the statement “physicians should know the overall
51
cost of the care they provide” (r = -.194, p = .219). That said, the relationship between
average likeliness to change and responses to the statement “physicians should put the
patient’s welfare above the physician’s financial interests” approached significance (r = -
.268
2
, p = .086).
Information Processes (RQ3, RQ4)
To test the third research question (do sources of information, given a
hypothetical scenario, differently predict patterns of physician change of practice?), I
used a linear mixed-effects model analysis. The results suggest a significant difference in
changing practice based on source of information [F(1,8) = 22.35, p < .001]. The
estimated marginal means of the likeliness of changing practice given a particular source
are reported in Table 4.4. These results indicate that source of information does predict a
physician’s likeliness to change practice.
Table 4.4 Estimated Marginal Means of Likeliness to Change Given a Particular Source
95% Confidence Interval
Est. Marginal Mean Standard Error Lower Bound Upper Bound
Journal 3.976 .130 3.714 4.238
Non-profit 2.810 .141 2.524 3.095
Website 3.214 .154 2.902 3.526
Colleague 3.357 .117 3.121 3.594
CME class 3.667 .121 3.422 3.912
News 2.476 .161 2.152 2.801
Industry rep 2.929 .150 2.625 3.232
Professional
assoc. 3.881 .137 3.604 4.158
Guidelines 3.595 .153 3.287 3.903
2
Lower scores for the opinion item reflect “strongly agree,” which explains the negative coefficient.
52
Using these numbers, I examined the overlap (or lack thereof) of 95% confidence
intervals to determine whether the means of likeliness to change practice given a
particular source were statistically different from each other. These confidence intervals,
using the numbers presented in Table 4.4, are visualized in Figure 4.1.
Figure 4.1. Visualization of Confidence Intervals of Likeliness to Change Given a
Particular Source
As we can see from these numbers, the estimated marginal mean for likeliness to
change based on information from a medical journal (the source most likely to prompt
change) is significantly higher (p < .05) than that for colleagues, medical-related
53
websites, industry representatives, non-profit organizations, and news broadcasts as a
source. The estimated marginal mean for the source least likely to prompt change – a
news broadcast – is significantly different from every source except for industry
representatives and non-profits. This information suggests that source of information
does matter in a physician’s decision-making process. This leads us to the fourth research
question: what is it about a source that may predict a physician’s response to that source?
To address the fourth research question, I used Pearson correlation coefficients to
examine whether there is a relationship between a respondent’s likeliness to change score
for a source and perceived credibility of that source. In the case of every source except
medical journals and non-profit organizations (the latter of which approaches
significance), the relationship between likeliness to change for a source and perceived
credibility of that source is statistically significant (p < .05)
3
. These coefficients are
summarized in Table 4.5.
3
An online resource for clinical guidelines (e.g., National Guideline Clearinghouse) was not used in this
part of the analysis, as credibility of these guidelines is not in question because the guidelines represent
fact.
54
Table 4.5 Bivariate Pearson Correlations Between Likeliness to Change Given a Source
and Source Credibility
r
Journal .200
Non-profit .279
Website .394*
Colleague .346*
CME class .426**
News .408**
Industry rep. .501**
Professional assoc. .376*
Note: * p < .05 ** p < .01
55
CHAPTER FIVE: DISCUSSION
“You cannot solve problems by continuing to use the same solutions that created the
problems in the first place.” – Albert Einstein
The purpose of this project was to understand the stakeholder-based factors that
influence the process of translating comparative effectiveness research into clinical care.
There were some underlying assumptions guiding this project: 1) that translation is likely
a case-specific process that might look different depending on the nature of the CER (i.e.,
the medical case presented to a group of stakeholders); 2) that a framework for translation
of CER must be stakeholder-based; 3) that motivations to expedite, delay, or block the
translation process vary among and within key stakeholders; and 4) that those
motivations are complicated and case-specific, and reflect a combination of
informational, institutional, economic, and personal considerations and characteristics. A
theoretical, stakeholder-based framework was created to hypothesize the translation of
CER into clinical care. Based on the above assumptions and proposed framework, a case
was developed that reflects a real-world example of CER suggesting a change in practice
that would represent a deviation from medical guidelines. This was done acknowledging
the limitations of using only one case and the resultant lack of generalizability; the
project was undertaken as an exploratory endeavor to begin with a specialized case and
group to start understanding motivational and influential processes in translation. I
elected to examine one stakeholder – physicians providing direct patient care – for
analysis, as physicians are, ultimately, on the front lines of converting research into
patient care. Given the cardiac focus of the case, I recruited cardiologists to complete a
survey that examined their personal, economic, institutional, and informational
motivations that would explain their decision to expedite, delay, or block the translation
56
of CER into practice. The results indicate that many of the above assumptions could hold
true: the process of translation could very likely be case-specific, translational models
must be inclusive of stakeholders and of the influence they exert over each other, and that
motivations to expedite or delay translation are likely multi-faceted and complicated. In
this chapter, I discuss a variety of possible factors that might predict physician behavior
as it relates to incorporating CER results.
Personal Characteristics and Translation
Based on my analysis, personal characteristics did not appear to be a compelling
predictor of physician behavior. This may be due to the poor reliability of measurement,
or it could be that personal characteristics, or “type,” are not strong predictors of
willingness or likeliness to change practice. To differentiate these possibilities, we need
better measurement of these characteristics.
That said, these data indicate a couple of interesting points. First, physicians were
not categorizable; their concerns over practicality, tendency towards nonconformity, and
relative emphasis placed on the value of evidence versus experience were fluid and might
be case-specific. Personality-based models to explain physician behavior do not take into
account the importance not only of institutional/economic dynamics, but of the nuances
of the behavior physicians are being encouraged to adopt.
Second, the two items that were significantly correlated with likeliness to change
practice could shed some light on motivations to change. Responses to “I am often
critical of accepted practices” were significantly correlated with likeliness to change, and
the coefficient suggests that respondents who strongly disagreed with this statement were
57
more likely to change practice. This suggests an interesting idea: doctors may not be
critical of accepted practices but are nevertheless willing to make changes in their own
practice. This again suggests that personality trait as hypothesized in prior literature may
not have a compelling impact on physician behavior. Responses to the statement “Patient
care should be based when possible on the opinions of respected authorities” were also
significantly correlated with likeliness to change. This coefficient suggests that doctors
who agree with that statement are likely to make changes in their own practice. Source
credibility, then, appears to have a compelling impact on this sample of physicians, which
I will discuss later in this chapter.
Institutional/Economic Dynamics and Translation
In general, institutional characteristics including type of practice, formal
initiatives undertaken by practice, and patient load did not have much of an impact on the
likeliness to change practice. However, physician stress did significantly correlate with
willingness to change. The coefficients suggest that physicians who do not feel rushed
when consulting with a patient or stressed at work are more likely to change practice.
Along these lines, physicians were concerned with the ease of integration of the change
proposed by the CER with respect to their decision of whether or not to make the change.
All of this could suggest that physicians feel that it is incumbent on them to play an
evaluative role in the change suggested, and that those who feel rushed and stressed at
work may not perceive there to be enough time to critically assess new information,
regardless of the source from which it originates. This is consistent with Kannampalil et
al.’s (2013) finding that physicians use information that maximizes information gain,
regardless of the cognitive effort necessary to process the information.
58
Interestingly and contrary not only to the literature but to my own intuition,
physicians were not strongly motivated by financial concerns. For this sample of
physicians, reimbursement mechanism, patient coverage, concerns over immediate cash
flow, or business relationships with industry representatives did not appear predictive of
willingness to change practice. This is particularly noteworthy given that every physician
indicated having received at least one of the ten possible material or monetary goods
from drug or device companies in the past year. Only 42.9% (n = 18) of this sample
indicated having received $0 worth of goods and services in the last year (indicating that
they received food, beverages, and/or drug samples at a minimum). This could suggest
that physicians are primarily concerned with patient wellbeing and safety. As an
extension to this, physicians who had a positive attitude towards drug samples and use
were significantly more likely to make a change in their practice that would discontinue
the initiation of drug therapy in some patients. Cardiologists sampled may believe in the
power of drugs in treating patients while acknowledging the paramount importance of
prescribing the drugs responsibly.
Information Processes and Translation
Results from this dissertation confirm the belief stated in chapter two that CER
translation models must be inclusive not only of different stakeholders, but of the
influence those stakeholders have over each other. The first conclusion to be drawn from
analyses of information processes is simple: the source matters. Physicians were
significantly more likely to make changes in practice when provided with CER
information in a medical journal or via a professional association; industry
representatives, non-profit organizations, and news broadcasts were not compelling
59
enough to move a physician to change practice. In fact, likeliness to change practice
given a news broadcast as a source falls closer to “unlikely to change” than to a neutral
response.
In examining what aspect of these sources may have prompted physicians to
change practice given a source, it seems clear that source credibility is extremely
important. With the exception of only two sources – journals and non-profit organizations
(which reflect both sides of the likeliness to change spectrum) – source credibility and
likeliness to change given a source were significantly correlated. There is a lesson to be
taken from stakeholders (e.g., industry representatives, non-profit organizations, research
institutions): when working with cardiologists (and, presumably, physicians representing
other specialties), it is extremely important to understand their information needs and
how to move them to view one’s organization as credible.
Limitations and Future Directions
The primary limitation of this study relates to the small sample size, which has
been discussed elsewhere in this chapter. A number of analyses that could have gauged
the moderating and mediating influences of different variables on predictors of physician
behavior were simply not possible with this small sample and inadequate power for such
analyses. Future examinations into stakeholders in the CER translation process must have
as large a sample as possible in order to generalize findings.
The choice of case presented to respondents might also be a limitation for a few
reasons. First, the case is based on a real-life example; it is possible that some (if not all)
participants had heard of this research scandal, and they were not asked this in the survey.
60
Second, the change recommended in this case is less likely to prompt strong financial
concerns than other case studies. While it is true that there is a link between industry
payments to physicians and physicians prescribing brand-name drugs (Yeh, Franklin,
Avorn, Landon, & Kesselheim, 2016), physicians’ financial interests may play more of a
role in deciding whether to accept CER results that suggest that patients benefit more
from diet and exercise than they do from an expensive surgery, for example. This is the
case with percutaneous coronary interventions, also known as heart stents. Despite the
CER publication of these results ten years ago (Boden et al., 2007), there is little evidence
that the number of heart stent procedures has substantially declined. Finally, the CER
case selected for this study strongly suggests that physicians are in immediate and serious
danger if physicians do not make the recommended change in practice. This knowledge is
perhaps more likely to move physicians to act than a case where minimal harm is
anticipated. Future iterations of this study should take on an experimental form,
manipulating these aspects of the case studies with which respondents are presented.
In this dissertation, four exploratory questions were posed: 1) Does variation in
physician characteristics predict, given a hypothetical scenario, a physician’s general
willingness to change practice? 2) Does variation in type of practice variables predict,
given a hypothetical scenario, a physician’s general willingness to change practice? 3) Do
sources of information, given a hypothetical scenario, differently predict patterns of
physician change of practice? and 4) If so, what is it about these sources that might
predict response to a given source in changing physician practice? The most compelling
set of variables that predict patterns of physician change in practice are the source of
information and the sense of feeling rushed and/or stressed at work. These findings
61
provide support for the theoretical model proposed in chapter two in that they touch on
the influence of other stakeholders in the translation process and on the varying
motivations within and between stakeholders to accelerate or impede translation.
Scholars interested in understanding motivations of physicians may conduct systematic
interviews to capture individual differences not previously assessed. Researchers
interested in other stakeholders may adapt parts of the survey constructed for this study to
gauge whether the importance of other stakeholders is similarly impactful outside of the
“physician” category, or whether other motivations (e.g., financial) become more
prominent instead.
Concluding Thoughts
While researchers, policy makers, and practitioners bemoan the slow pace with
which translation occurs (Lenfant, 2003), it seems widely acknowledged in projects
examining translational research that the desire for speedy translation may be misplaced
(Zoë Slote Morris et al., 2011; Sussman et al., 2006). The motivation behind this claim is
understandable: there is a concern that a medical innovation has not been sufficiently
tested by the time initial results are published (Sussman et al., 2006) and that the time lag
that characterizes translation may ultimately help protect patients. While I concede that
this may be the case with new research, I believe that we should endeavor to facilitate
translation of CER as quickly as possible. Comparative effectiveness research does not
provide support for new treatment; it is evaluative of existing treatments and has the
potential to reverse or correct harmful practice based on insecure research. The
importance of AHRQ and efforts to support CER cannot be overstated. Comparative
62
effectiveness research has tremendous potential to save lives, minimize costs, and
streamline inefficiencies in our complicated and overburdened health care system.
63
BIBLIOGRAPHY
Avorn, J., & Fischer, M. (2010). “Bench to behavior”: Translating comparative
effectiveness research into improved clinical practice. Health Affairs, 29(10), 1891–
1900. doi:10.1377/hlthaff.2010.0696
Balas, E. A. (2001). Information Systems Can Prevent Errors and Improve Quality.
Journal of the American Medical Informatics Association, 8, 398–399.
doi:10.1136/jamia.2001.0080398
Balas, E. a., & Boren, S. a. (2000). Managing Clinical Knowledge for Health Care
Improvement. Yearbook of Medical Informatics. doi:10.2196/ijmr.2811
Bennett, N. L., Casebeer, L. L., Kristofco, R. E., & Strasser, S. M. (2004). Physicians’
Internet information-seeking behaviors. Journal of Continuing Education in the
Health Professions, 24(1), 31–38 8p. doi:10.1002/chp.1340240106
Berwick, D. M. (2003). Disseminating innovations in health care. JAMA : The Journal of
the American Medical Association, 289(15), 1969–1975.
doi:10.1001/jama.289.15.1969
Boden, W. E., O’Rourke, R. A., Teo, K. K., Hartigan, P. M., Maron, D. J., Kostuk, W. J.,
… Weintraub, W. S. (2007). Optimal Medical Therapy with or without PCI for
Stable Coronary Disease. New England Journal of Medicine, 356(15), 1503–1516.
doi:10.1056/NEJMoa070829
Bolsin, S. N., & Colson, M. (2014). Perioperative beta-blockade: guidelines do not reflect
the problems with the evidence from the DECREASE trials. BMJ: British Journal of
Medicine, 349.
Bouri, S., Shun-Shin, M. J., Cole, G. D., Mayet, J., & Francis, D. P. (2014). Meta-
64
analysis of secure randomised controlled trials of β-blockade to prevent
perioperative death in non-cardiac surgery. Heart (British Cardiac Society), 100(6),
456–64. doi:10.1136/heartjnl-2013-304262
Cain, M., & Mittman, R. (2002). Diffusion of Innovation in Health Care.
Campbell, E. G., Gruen, R. L., Mountford, J., Miller, L. G., Cleary, P. D., & Blumenthal,
D. (2007). A national survey of physician-industry relationships. New England
Journal of Medicine, 356, 1742–1750.
Caudill, T. S., Johnson, M. S., Rich, E. C., & McKinney, W. P. (1996). Physicians,
pharmaceutical sales representatives, and the cost of prescribing. Archives of Family
Medicine, 5(4), 201–206.
Centers for Medicare & Medicaid Services. (2016). National Health Expenditures 2015
Highlights. Retrieved from https://www.cms.gov/Research-Statistics-Data-and-
Systems/Statistics-Trends-and-
Reports/NationalHealthExpendData/Downloads/highlights.pdf
Chalkidou, K., Tunis, S., Lopert, R., Rochaix, L., Sawicki, P. T., Nasser, M., & Xerri, B.
(2009). Comparative effectiveness research and evidence-based health policy:
experience from four countries. The Milbank Quarterly, 87(2), 339–367.
doi:10.1111/j.1468-0009.2009.00560.x
Chew, L. D., O’Young, T. S., Hazlet, T. K., Bradley, K. A., Maynard, C., & Lessler, D.
S. (2000). A physician survey of the effect of drug sample availability on
physicians’ behavior. Journal of General Internal Medicine, 15(7), 478–483.
Clarke, M. A., Belden, J. L., Koopman, R. J., Steege, L. M., Moore, J. L., Canfield, S.
M., & Kim, M. S. (2013). Information needs and information-seeking behaviour
65
analysis of primary care physicians and nurses: A literature review. Health
Information and Libraries Journal, 30(3), 178–190. doi:10.1111/hir.12036
Concannon, T. W., Meissner, P., Grunbaum, J. A., McElwee, N., Guise, J. M., Santa, J.,
… Leslie, L. K. (2012). A new taxonomy for stakeholder engagement in patient-
centered outcomes research. Journal of General Internal Medicine, 27(8), 985–991.
Conway, P., & Clancy, C. (2009). Comparative-effectiveness research—implications of
the Federal Coordinating Council’s report. New England Journal of Medicine,
361(4), 328–330. doi:10.1056/NEJMp0905631
Davis, P., & Howden-Chapman, P. (1996). Translating research findings into health
policy. Soc Sci Med, 43(5), 865–872. doi:x
Dawes, M., & Sampson, U. (2003). Knowledge management in clinical practice: A
systematic review of information seeking behavior in physicians. International
Journal of Medical Informatics, 71(1), 9–15. doi:10.1016/S1386-5056(03)00023-6
Demaria, A. N. (2013). Translational research? Journal of the American College of
Cardiology, 62(24), 2342–2343. doi:10.1016/j.jacc.2013.11.008
Denis, J.-L., Hébert, Y., Langley, A., Lozeau, D., & Trottier, L.-H. (2002). Explaining
diffusion patterns for complex health care innovations. Health Care Management
Review, 27(3), 60–73. doi:10.1097/00004010-200207000-00007
Des Jarlais, D. C., Sloboda, Z., Friedman, S. R., Tempalski, B., McKnight, C., & Braine,
N. (2006). Diffusion of the D.A.R.E and syringe exchange programs. American
Journal of Public Health, 96(8), 1354–1358. doi:10.2105/AJPH.2004.060152
Deverka, P. A., Lavallee, D. C., Desai, P. J., Esmail, L. C., Ramsey, S. D., Veenstra, D.
L., & Tunis, S. R. (2012). Stakeholder participation in comparative effectiveness
66
research: defining a framework for effective engagement. Journal of Comparative
Effectiveness Research, 1(2), 181–194.
Devine, E. B., Alfonso-Cristancho, R., Devlin, A., Edwards, T. C., Farrokhi, E. T.,
Kessler, L., … Flum, D. R. (2013). A model for incorporating patient and
stakeholder voices in a learning health care network: Washington State’s
comparative effectiveness research translation network. Journal of Clinical
Epidemiology, 66(8 SUPPL.8), 1–14. doi:10.1016/j.jclinepi.2013.04.007
Dobrow, M. J., Goel, V., & Upshur, R. E. G. (2004). Evidence-based health policy:
Context and utilisation. Social Science and Medicine, 58, 207–217.
doi:10.1016/S0277-9536(03)00166-7
Dreyer, N. A., Tunis, S. R., Berger, M., Ollendorf, D., Mattox, P., & Gliklich, R. (2010).
Why observational studies should be among the tools used in comparative
effectiveness research. Health Affairs, 29, 1818–1825.
doi:10.1377/hlthaff.2010.0666
Drolet, B. C., & Lorenzi, N. M. (2011). Translational research: Understanding the
continuum from bench to bedside. Translational Research, 157(1), 1–5.
doi:10.1016/j.trsl.2010.10.002
Gabriel, S. E., & Normand, S.-L. T. (2012). Getting the Methods Right - The Foundation
of Patient-Centered Outcomes Research. New England Journal of Medicine, 367(9),
787–790. doi:10.1056/NEJMp1209051
González-gonzález, A. I., Dawes, M., Sanchez-Mateos, J., Riesgo-Fuertes, R., Escortell-
Mayor, E., Sanz-Cuesta, T., & Hernandez-Fernandez, T. (2007). Information Needs
and Information-Seeking Behavior of Primary Care Physicians. Annals of Family
67
Medicine, 5(4), 345–353. doi:10.1370/afm.681.Jos
Green, L. A., Gorenflo, D. W., & Wyszewianski, L. (2002). Validating an instrument for
selecting interventions to change physician practice patterns. Journal of Family
Practice, 51(11), 938–942.
Green, L. W. (2009). Making research relevant: If it is an evidence-based practice,
where’s the practice-based evidence? Family Practice, 25(May), 20–24.
doi:10.1093/fampra/cmn055
Green, L. W., Ottoson, J. M., García, C., & Hiatt, R. a. (2009). Diffusion theory and
knowledge dissemination, utilization, and integration in public health. Annual
Review of Public Health, 30, 151–174.
doi:10.1146/annurev.publhealth.031308.100049
Hickson, G. B., Altemeier, W. A., & Perrin, J. M. (1987). Physician reimbursement by
salary or fee-for-service: effect on physician practice behavior in a randomized
prospective study. Pediatrics, 80(3).
Hillman, A. L., Ripley, K., Goldfarb, N., Nuamah, I., Weiner, J., & Lusk, E. (1998).
Physician financial incentives and feedback failure to increase cancer screening in
Medicaid managed care. American Journal of Public Health, 88(11), 1699–1701.
Husten, L. (2013). How Heart Guidelines Based on Disgraced Research May Have
Caused Thousands of Deaths. Retrieved from
https://www.forbes.com/sites/larryhusten/2013/07/31/european-heart-guidelines-
based-on-disgraced-research-may-have-caused-thousands-of-deaths/#7017160c2e33
Ioannidis, J. P. (2004). Materializing research promises: opportunities, priorities and
conflicts in translational medicine. Journal of Translational Medicine, 2, 5.
68
doi:10.1186/1479-5876-2-5
Jakicic, J. M., Sox, H., Blair, S. N., Bensink, M., Johnson, W. G., King, A. C., …
Ainsworth, B. E. (2015). Comparative effectiveness research: A roadmap for
physical activity and lifestyle. Medicine and Science in Sports and Exercise, 47(8),
1747–1754. doi:10.1249/MSS.0000000000000590
Kaiser Family Foundation. (2013). Summary of the Affordable Care Act. Retrieved from
http://kff.org/health-reform/fact-sheet/summary-of-the-affordable-care-act/
Kannampallil, T. G., Franklin, A., Mishra, R., Almoosa, K. F., Cohen, T., & Patel, V. L.
(2013). Understanding the nature of information seeking behavior in critical care:
Implications for the design of health information technology. Artificial Intelligence
in Medicine, 57(1), 21–29. doi:10.1016/j.artmed.2012.10.002
Khullar, D., Chokshi, D. A., Kocher, R., Reddy, A., Basu, K., Conway, P. H., &
Rajkumar, R. (2015). Behavioral economics and physician compensation - promises
and challenges. New England Journal of Medicine, 372(24), 2281–2283.
Kremer, W. (2014, July 7). Do doctors understand test results? BBC World Service.
Retrieved from http://www.bbc.com/news/magazine-28166019
Lagace, R. R., & Twible, J. L. (1991). The impact of source credibility characteristics on
physicians’ satisfaction with pharmaceutical salespeople. Journal of Ambulatory
Care Marketing, 4(2), 53–63.
Lean, M. E. J., Mann, J. I., Hoek, J. A., Elliot, R. M., & Schofield, G. (2008).
Translational research. British Medical Journal, 337(7672), 705–706.
doi:10.1136/bmj.a863
Leckie, G. J., Pettigrew, K. E., & Sylvain, C. (1996). Modeling the information seeking
69
of professionals: a general model derived from research on engineers, health care
professionals, and lawyers. Library Quarterly, 66(2), 161–193.
Lenfant, C. (2003). Clinical Research to Clinical Practice — Lost in Translation? New
England Journal of Medicine, 349, 868–874.
Luce, B. R., Kramer, J. M., Goodman, S. N., Connor, J. T., Tunis, S., Whicher, D., &
Schwartz, J. S. (2009). Medicine and Public Issues Annals of Internal Medicine
Rethinking Randomized Clinical Trials for Comparative Effectiveness. Annals of
Internal Medicine, 131(3), 206–209.
MacCallum, R. C., Widaman, K. F., Zhang, S., & Hong, S. (1999). Sample Size in Factor
Analysis. Psychological Methods, 4(1), 84–99.
Macintyre, S. (2012). Evidence in the development of health policy. Public Health,
126(3), 217–219. doi:10.1016/j.puhe.2012.01.026
Marincola, F. M. (2003). Translational Medicine: A two-way road. Journal of
Translational Medicine, 1, 1. doi:10.1186/1479-5876-1-1
Morris, Z. S., Wooding, S., & Grant, J. (2011). The answer is 17 years, what is the
question: understanding time lags in translational research. Jrsm, 104, 510–520.
doi:10.1258/jrsm.2011.110180
Morris, Z. S., Wooding, S., & Grant, J. (2011). The answer is 17 years, what is the
question: understanding time lags in translational research. Journal of the Royal
Society of Medicine, 104, 510–20. doi:10.1258/jrsm.2011.110180
National Center for Health Statistics. (2016). Health, United States, 2015: With Special
Feature on Racial and Ethnic Health Disparities. Health, United States, 2015.
Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/27308685
70
PCORI. (2017). PCORI. Retrieved from http://www.pcori.org/
Pober, J. M. J. S. M. S., Neuhauser, C. S., & Pober, J. M. J. S. M. S. (2001). Obstacles
facing translational research in academic medical centers. Faseb J, 15, 2303–2313.
doi:10.1096/fj.01-0540lsf 15/13/2303 [pii]
Rubio, D. M., Schoenbaum, E. E., Lee, L. S., Schteingart, D. E., Marantz, P. R.,
Anderson, K. E., … Esposito, K. (2010). Defining translational research:
Implications for training. Academic Medicine, 85(3), 470–475.
doi:10.1097/ACM.0b013e3181ccd618.Defining
Shah, N. D., Mullan, R. J., Breslin, M., Yawn, B. P., Ting, H. H., & Montori, V. M.
(2010). Translating comparative effectiveness into practice: the case of diabetes
medications. Medical Care, 48(6 Suppl), S153–S158.
doi:10.1097/MLR.0b013e3181d5956c
Sox, H. C., & Greenfield, S. (2009). Comparative Effectiveness Research : A Report
From the Institute. Annals of Internal Medicine, (June), 203–205.
Squires, D., & Anderson, C. (2015). U.S. Health Care from a Global Perspective.
Retrieved from http://www.commonwealthfund.org/publications/issue-
briefs/2015/oct/us-health-care-from-a-global-perspective
Stross, J. K., & Harlan, W. R. (1979). The dissemination of new medical information.
JAMA : The Journal of the American Medical Association, 241, 2622–2624.
doi:10.1001/jama.1979.03290500030017
Sussman, S., Valente, T. W., Rohrbach, L. a, Skara, S., & Pentz, M. A. (2006).
Translation in the health professions: converting science into action. Evaluation &
the Health Professions, 29(1), 7–32. doi:10.1177/0163278705284441
71
Tunis, S. R., Benner, J., & McClellan, M. (2010). Comparative effectiveness research:
Policy context, methods development and research infrastructure. Statistics in
Medicine, 29(19), 1963–1976. doi:10.1002/sim.3818
Umscheid, C. A., Williams, K., & Brennan, P. J. (2010). Hospital-based comparative
effectiveness centers: Translating research into practice to improve the quality,
safety and value of patient care. Journal of General Internal Medicine, 25(12),
1352–1355. doi:10.1007/s11606-010-1476-9
Whippen, D. A., & Canellos, G. P. (1991). Burnout syndrome in the practice of
oncology: results of a random survey of 1,000 oncologists. Journal of Clinical
Oncology, 9(10), 1916–1920.
Woolf, S. H. (2008). The Meaning of Translational Research and Why it M atters. Jama,
299, 211–213. doi:10.1001/jama.2007.26
Yeh, J. S., Franklin, J. M., Avorn, J., Landon, J., & Kesselheim, A. S. (2016).
Association of Industry Payments to Physicians With the Prescribing of Brand-name
Statins in Massachusetts. JAMA Internal Medicine, 176(6), 763–768.
72
APPENDIX A: SURVEY
Thank you for participating in this survey. We appreciate your feedback. The purpose of
this study is to gain a better understanding of how cardiologists stay up-to-date on
information in their field. Although you may not receive any benefits from participating
in this study, the information that you and other participants provide may help inform the
strategies of future health communications and health research. This survey will take
approximately fifteen minutes to complete.
This study is anonymous. You will not be asked to provide any identifying personal
information such as your name or address. When the results of the research are published
or discussed at conferences, no information will be included that would reveal your
identity.
Your participation is voluntary. Your refusal to participate will involve no penalty. You
may withdraw your consent at any time and discontinue participation without penalty.
You are not waiving any legal claims, rights, or remedies because of your participation in
this research study.
Investigator contact information:
Katherine Elder, PhD(c), MA, MPA
Annenberg School for Communication and Journalism
University of Southern California
kelder@usc.edu
IRB contact information:
University Park IRB, Office of the Vice Provost for Research Advancement, Stonier
Hall, Room 224a, Los Angeles, CA 90089-1146, (213) 821-5272 or upirb@usc.edu
[Page Break]
1. In what year did you graduate from medical school? [open-ended]
2. Did you graduate from medical school in either the United States or Canada?
a. Yes
b. No
3. Please indicate your gender:
a. Male
b. Female
c. Other gender identity
d. Prefer not to state
4. Please indicate your race/ethnicity
a. African-American/Black (non-Hispanic)
b. Asian
c. Hispanic
d. Native American/Alaska Native
73
e. Pacific Islander
f. White (non-Hispanic)
g. Other
5. For how many years have you been certified as a cardiologist? [open-ended]
6. Do you currently hold a faculty appointment at a medical school?
a. Yes, full-time
b. Yes, part-time
c. Yes, voluntarily
d. No
7. How often do you communicate with other clinicians via email?
a. Frequently
b. Rarely
c. Never
8. On average, how many different professional journals do you read monthly?
[open-ended]
[Page Break]
9. Do you currently provide direct patient care?
a. Yes
b. No
10. In a typical work week, how many patients do you see? [open-ended]
11. Approximately what percentage of your patients is covered as follows? [sliding
scale, 0-100%]
a. Medicaid
b. Medicare
c. Uninsured and unable to pay
12. Please rate the degree to which you agree/disagree with the following statements:
Strongly
Agree
Agree Neutral Disagree Strongly
Disagree
I often feel
rushed when
consulting
with a
patient
I often find
myself
feeling
stressed at
work
I often feel
that there is
74
not enough
time to give
patients
adequate
care
13. In your main practice, please check the predominant reimbursement mechanism
by which you are compensated for care you provide to patients:
a. Fee-for-service
b. Partial capitation
c. Full capitation
d. Salary
14. How would you best characterize the organization of your medical practice?
a. Solo or two-person practice
b. Single specialty group
c. Multispecialty group
d. Staff/group model HMO
e. University
f. Hospital
g. Other
15. In an average month, how many times did you meet with representatives from
drug, device, or other medically-related companies? [open-ended]
16. Which of the following have you received in the last year from drug, device, or
other medically-related companies? Check all that apply.
a. Food and/or beverages in your workplace
b. Free drug samples
c. Honoraria for speaking
d. Payment for consulting services
e. Payment for services on a scientific advisory board or board of directors
f. Payment in excess of costs for enrolling patients in industry-sponsored
trials
g. Costs of travel, time, meals, lodging, or other personal expenses for
attending meetings
h. Gifts that you receive as a result of prescribing practices
i. Free tickets to cultural and/or sporting events
j. Free or subsidized admission to meetings or conferences for which CME
credits are awarded
17. Please estimate your personal total before tax compensation for all clinical
activities and services in 2016.
a. <$100,000
b. 100,000 – 150,000
c. 150,001-200,000
75
d. 200,001-250,000
e. 250,001-300,000
f. $300,000 +
18. Excluding any food, beverages, and drug samples you may have received in your
workplace, please estimate the total value of all goods and services you received
in the last year from drug, device, or other medically-related companies.
a. None
b. $1-100
c. 101-500
d. 501-1,000
e. 1,001-5,000
f. 5,001-10,000
g. 10,001-25,000
h. $25,001 +
[Page Break]
19. In the last 3 years, has either your main practice or hospital undertaken formal
initiatives in any of the following areas?
Yes No
Error reduction
Patient confidentiality
Quality improvement
Access for the poor or
underserved
Cultural competence
20. In the last 3 years, have you:
Yes No
Participated in a formal medical error reduction initiative in your office,
clinic, hospital, or other health care setting?
Undergone competency assessment by a provider organization or health
plan?
Served as a reviewer for a professional journal?
Participated in the development of formal clinical practice guidelines?
76
Reviewed another physician’s medical records for quality improvement
reasons?
Encouraged one or more patients to enroll in a clinical trial?
Provided care, with no anticipation of reimbursement, in a setting serving
poor and underserved patients?
Provided health-related expertise to local community organizations (e.g.,
school boards, parent-teacher organizations, athletic teams, local media)?
Encouraged your professional society to address a public health issue or
policy issue that is not primarily concerned with physician welfare?
[Page Break]
21. Please rate the degree to which you agree/disagree with the following statements:
Strongly
Agree
Agree Neutral Disagree Strongly
Disagree
Randomized controlled trials are the
most reliable way to know what
really works.
I am comfortable practicing in ways
different than other doctors.
Evidence-based medicine makes a lot
of sense to me.
I don’t have the time to read up on
every practice decision.
It is best to change the way I treat a
certain problem when my local
colleagues are making the same
changes,
I follow practice guidelines if they
are not too much hassle.
The opinions of respected authorities
should guide clinical practice.
I am too busy taking care of patients
to keep up with the recent literature.
77
Clinical experience is the most
reliable way to know what really
works,
I am uncomfortable doing things
differently from the way I was
trained.
I am often critical of accepted
practices,
Patient care should be based when
possible on randomized controlled
trials.
Patient care should be based when
possible on the opinions of respected
authorities.
My colleagues consider me to be
someone who marches to the beat of
my own drum.
I follow practice guidelines as long
as they don’t interfere too much with
the flow of patients.
It is not prudent to practice out of
step with other physicians in my
area.
The best practice guidelines are
based on the results of the
randomized controlled trials.
Evidence-based medicine is not very
practical in patient care.
22. Please rate the degree to which you agree/disagree with the following statements:
Strongly
Agree
Agree Neutral Disagree Strongly
Disagree
Patients tend to misuse drug samples.
Drug samples permit quicker
initiation of therapy.
78
Drug samples deprive the patient of
pharmacists’ screening for drug
interaction.
Drug samples fulfill an educational
role through demonstration.
Drug samples are a source of
medication for patients who cannot
afford them.
Supplies of drug samples are
inconsistent.
Physicians should provide necessary
care regardless of the patient’s ability
to pay.
Physicians should put the patient’s
welfare above the physician’s
financial interests.
Physicians should know the overall
cost of the care they provide.
[Page Break]
23. Please rate the following sources in terms of credibility:
Not at all
credible
Somewhat
not credible
Neutral/No
opinion
Somewhat
credible
Very
credible
Non-profit patient
advocacy organizations
(e.g., Susan G. Komen
for the Cure)
Medical-related websites
(e.g., WebMD)
Your colleagues
CME classes
News organizations (e.g.,
NBC, CBS)
79
Industry representatives
(e.g., pharmaceutical,
medical device)
Professional associations
(e.g., American Heart
Association)
Peer-reviewed journal
articles
[Page Break]
Please read the following case and respond to the questions that follow.
Clinical guidelines recommend the prescription of beta-blockers in several classes of
patients, including those who will be undergoing non-cardiac surgery. You learn that the
series of trials that provide justification to prescribe perioperatve beta-blockers in these
classes of patients is insecure; much of the data had been lost, and that which remained
was found to contain serious flaws, including, in one case, complete fabrication of a
dataset The European Society of Cardiology has stopped recommending the use of beta-
blockers in patients undergoing non-cardiac surgery, but the American Heart
Association guidelines based on these trials have not been retracted.
24. Which of the following would best describe your reaction had you received the
information about the insecurity of the clinical trials from:
Not at all
likely to
change
practice
Somewhat
unlikely to
change
practice
Neutral Somewhat
likely to
change
practice
Very
likely to
change
practice
A professional medical
journal
A non-profit patient
advocacy organization
A medical-related website
(e.g., WebMD)
A colleague
A CME class
80
A news broadcast
An industry representative
(e.g., pharmaceutical,
medical device)
A professional association
(e.g., International
Association of
Cardiologists; U.S.
Preventive Services Task
Force)
An online resource for
clinical guidelines (e.g.,
National Guideline
Clearinghouse)
25. Please rate the following factors with respect to importance in your decision
making:
Not at all
influential
A little
influential
Neutral Influential Very
influential
Ease of integration change
(e.g., can the practice handle
this change as is? Does it
have to be referred out? Is
there a need to hire or
develop a new expertise? Is
there a need to incorporate
new technology?)
Change in immediate cash
flow
Impact of change on patients
(i.e., does the new
information promise to save
lives, or does it simply
recommend an alternative
treatment with similar
efficacy?)
Anticipated practice-related
consequences of deviation
from existing guidelines
81
(e.g., fines)
Degree of risk for patients
Abstract (if available)
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Communicating organizational knowledge in a sociomaterial network: the influences of communication load, legitimacy, and credibility on health care best-practice communication
PDF
Value based purchasing: decision-making processes underlying hospital acquisitions of orthopedic devices
PDF
Competing logics and organizational failure: the structuring of California prison health care
PDF
Developing and testing a heuristic-systematic model of health decision making: the role of affect, trust, confidence and media influence
PDF
Social support in Cambodia: the role of peer educators in behavior change
PDF
Health-related quality of life correlates of the treatment decision making process of newly diagnosed prostate cancer patients
PDF
Translating research into practice: a community-based medication management intervention
PDF
Application of professional learning outcomes into the classroom: an evaluation study
PDF
Relationship formation and information sharing to promote risky health behavior on social media
PDF
Building a unicorn: management of innovation, collaboration, and change in a Silicon Valley start-up
PDF
The evolution of decision-making quality over the life cycle: evidence from behavioral and neuroeconomic experiments with different age groups
PDF
Understanding the decision making process of California urban schools superintendents through Bolman and Deal's four leadership frames
PDF
Traditional media, social media or word-of-mouth? Examining the information sources that influence Chinese students' decision-making processes when applying to U.S. graduate schools
PDF
Rediscovering the communication engine of neighborhood effects: how the interaction of residents and community institutions impacts health literacy and how it can be leveraged to improve health c...
PDF
Leadership corruption and the implications to the corporate brand
PDF
The effects of framing and actuarial risk probabilties on involuntary civil commitment decisions
PDF
Getting to the Core: an examination into the resources, strategies and skills superintendents employed as they implemented the Common Core state standards and the politics in play
PDF
The evolution of knowledge creation online: what is driving the dynamics of Wikipedia networks
PDF
Staying ahead of the digital tsunami: strategy, innovation and change in public media organizations
PDF
The new news: vision, structure, and the digital myth in online journalism
Asset Metadata
Creator
Elder, Katherine Anne
(author)
Core Title
Informing the exam room: understanding the process of translating evidence-based medical research into clinical care
School
Annenberg School for Communication
Degree
Doctor of Philosophy
Degree Program
Communication
Publication Date
07/24/2017
Defense Date
06/01/2017
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
cardiologists,comparative effectiveness,health communication,OAI-PMH Harvest,physician decision-making,translation
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Miller, Lynn (
committee chair
), Riley, Patricia (
committee member
), Valente, Thomas (
committee member
)
Creator Email
katherine.elder@gmail.com,kelder@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c40-415219
Unique identifier
UC11214734
Identifier
etd-ElderKathe-5635.pdf (filename),usctheses-c40-415219 (legacy record id)
Legacy Identifier
etd-ElderKathe-5635.pdf
Dmrecord
415219
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Elder, Katherine Anne
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
Tags
cardiologists
comparative effectiveness
health communication
physician decision-making
translation