Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 159
7
Applying Evidence for Patient-Centered
Care: Standards and Expectations
INTRODUCTION
Communicating evidence to patients is a critical challenge in transition-
ing to a health system centered on patients’ individual values and prefer-
ences. The public is bombarded with inconsistent health messages and has
little background or training in how to evaluate the information presented
in television newscasts, daily papers, and on the Internet. Moreover, many
messaging campaigns may contribute to this confusion because of over-
simplification, emotional appeals, and conflicting advice.
Value and science-based care are concepts that will require rethinking
how information is shared with patients. Committing to communicating
evidence in all its complexity while ensuring it is understandable and per-
tinent to individual patient circumstances will be a challenging task. Com-
munication strategies need to be evaluated so that shortfalls can be made
clear, and effort should be given to developing new approaches to teach
patients about evidence-based medicine.
The papers in this chapter address how to apply standards for evidence
in the context of individual patient-centered care. They point to the im-
portance of generating evidence applicable to individual patient outcomes,
preferences, and values. This type of evidence is necessary to provide care
that is more effective for the individual patient and more efficiently deliv-
ered. Additionally, the papers take up the nature of difficulties in commu-
nicating this evidence to patients and examine strategies that have proven
valuable in informed decision making.
159
OCR for page 160
160 PATIENTS CHARTING THE COURSE
In the first paper, Dale Collins Vidal of the Dartmouth Institute for
Health Policy and Clinical Practice points out that the current informed
consent process fails to help patients understand the trade-offs—or spe-
cific risks and benefits—in comparing treatment options. When competing
treatment options are “preference sensitive,” decision making about treat-
ment should incorporate a patient’s values and preferences. To ensure that
patients have the tools they need to make an informed choice, providers
must adequately communicate the risks, benefits, alternatives, experience,
and cost.
In the second paper, Clifford Goodman of The Lewin Group addresses
the limitations of evidence hierarchies that have been used for decades.
He highlights the limitations of what until now has been considered “best
evidence,” for example, meta-analyses and randomized controlled trials
(RCTs). He suggests moving away from RCTs as best evidence for a num-
ber of reasons, including their focus on population-based care, the time lag
in obtaining scientific results, high costs, and the lack of applicability to
individual patients. He points to several methods that can better capture
evidence applicable to personalized medicine, which is becoming increas-
ingly important with advances in genomic data.
Fran M. Visco of the National Breast Cancer Coalition addresses trans-
lating and communicating evidence when the recommendations for care
are uncertain. She reviews the barriers to understanding science-driven
care, including the adoption of practices that have become standard even
though evidence to back them is limited; the oversimplification of messag-
ing that misleads the public; and the promulgation of guidelines that are
self-serving. She highlights the pressing need to evaluate how to deliver
complex messages about interventions so they can be made pertinent to
individual patients.
THE ROLE OF EVIDENCE IN PATIENT-CENTERED CARE
Dale Collins Vidal, M.D., M.S.1
Dartmouth Institute for Health Policy and Clinical Practice
In Crossing the Quality Chasm, the Institute of Medicine (IOM) defines
patient-centeredness as “providing care that is respectful of and responsive
to individual patient preferences, needs, and values, ensuring that patient
values guide all clinical decisions” (IOM, 2001). However, studies reveal
that the current informed consent process falls far short of this goal and
frequently fails to help patients understand the specific risks and benefits of
1 The author would like to acknowledge Allison J. Hawke, Sue Burg, and Sherry Thornburg
for their contributions in preparing this manuscript.
OCR for page 161
161
APPLYING EVIDENCE FOR PATIENT-CENTERED CARE
treatment options (Holmboe et al., 2000). This is true even when the deci-
sion involves “effective care”—care that is supported by strong evidence
and usually depends less on an individual’s personal values and preferences
(O’Connor et al., 2007). When treatments are not supported by adequate
evidence or when they involve trade-offs that could impact a patient’s
quality of life, the traditional informed consent process is inadequate for
helping patients make informed treatment decisions. In these situations, the
process should more appropriately be framed as informed choice, incorpo-
rating a discussion of treatment alternatives, the evidence associated with
them, and the patient’s personal values.
To achieve the IOM’s goals around patient-centered decision making,
a framework is needed that will allow patients, families, and providers to
engage in successful, mutual healthcare interactions, allowing patient and
family values to guide all healthcare decisions. This paper presents elements
of such a framework and describes tools used by the Dartmouth Hitchcock
Medical Center to support it.
Categories of Care
Effective Care
John Wennberg is credited with defining three categories of care: effec-
tive, preference-sensitive, and supply-sensitive (Table 7-1). Effective care is
that which is supported by high-quality evidence demonstrating that the
benefits of a proposed treatment or intervention are large compared with
the potential harms. Clinicians, and most patients, agree on the appropriate
course of action for effective care. When care is deemed effective, the clini-
cian typically makes a recommendation, and the goal is patient compliance
or increased uptake by a population (Wennberg, 2002). Examples include
the use of antibiotics for treatment of bacterial pneumonia, screening with
pap smears, and preventive flu vaccinations for healthcare workers.
Preference-Sensitive Care
Because of conflicting, uncertain, or insufficient information as well as
differing personal preferences, the quality of evidence often does not allow
for a clear “right” choice. A quality decision for these “preference-sensitive”
medical decisions requires that the patient be knowledgeable about the
options and that his or her personal values inform the choice (Sepucha et
al., 2004). Examples include elective surgeries such as LASIK eye surgery;
use of screening tests, such as prostate-specific antigen for prostate cancer;
and prevention of cervical cancer with human papillomavirus vaccinations.
An effective care recommendation may become a preference-sensitive deci-
OCR for page 162
162 PATIENTS CHARTING THE COURSE
TABLE 7-1 Summary and Description of Categories of Care
Category of Typically Characterized
Care by: Goal Examples
Effective • Proven clinical • Increase • Treatment—antibiotics
data on treatment uptake for community-acquired
effectiveness pneumonia
• Benefits large • Screening—pap smear
compared with harms • Prevention—flu shot
Preference- • Multiple treatment • Patient • Treatment—LASIK
Sensitive options participation surgery
• Lack of evidence- • High decision • Screening—prostate-
based treatment quality specific antigen (PSA)
options • Prevent • Prevention—human
• Benefits and risks are overuse papillomavirus (HPV)
uncertain of options vaccination
• Significant trade-offs patients do
for the patient not value
Supply- • Lack of evidence • Varies across • Care for patients with
Sensitive on comparative the country progressive chronic
effectiveness of based illness (e.g., lung
treatments primarily on disease, cancer, diabetes,
• Few guidelines capacity heart failure)
regarding delivery of • Amplified by
care fee-for-service
• Assumption that more model
care is better
SOURCE: Wennberg et al., 2002.
sion if a patient does not readily accept the clinician’s recommendation.
Preference-sensitive care avoids the overuse of options patients do not
value. The essential feature common to all preference-sensitive conditions
is that choice of treatment is up to the patient (Wennberg, 2002).
Supply-Sensitive Care
Wennberg and colleagues have demonstrated that capacity dictates
how healthcare resources are used (Wennberg et al., 2002). Supply-sensitive
care exists when the number of available hospital beds drives how often
patients are hospitalized rather than cared for in the outpatient setting, or
the frequency of referrals to specialists. The Dartmouth Atlas Project2 has
2 See http://www.dartmouthatlas.org/downloads/reports/supply_sensitive.pdf (accessed Oc-
tober 14, 2010).
OCR for page 163
163
APPLYING EVIDENCE FOR PATIENT-CENTERED CARE
demonstrated that supply-sensitive care varies widely across the United
States. This variation may result in underuse of effective care and improper
use of preference-sensitive care. The latter may occur when patients are not
provided accurate information about their options and are not encouraged
to incorporate their own values and preferences into care decisions.
Healthcare Variation
The Dartmouth Atlas has used Medicare data to map geographic varia-
tions in care for more than 20 years. The project reports on resource and
medical care use among beneficiaries living in 3,436 hospital service areas—
aggregated into 306 hospital referral regions—examining unwarranted
variations in the above three categories of clinical care. It has revealed the
presence of significant geographic variations in healthcare prices, practices,
and providers; patient characteristics; and patient preferences. The project
also has found that regions that spend more on health care and have higher
rates of utilization do not experience better health outcomes compared with
regions spending less on health and using healthcare services less frequently
(Fisher et al., 2003).
Although higher rates of effective treatments are preferred, it is difficult
to define the “right rate” for preference-sensitive treatments. The question
raised by Wennberg and colleagues is, given the variation in the rates of pro-
cedures across the United States, which rate is right (Weinstein et al., 1998)?
To illustrate the issue, Figure 7-1 shows the variations in spinal surgery across
hospital referral regions in the United States. The number of spinal surgeries
ranges from fewer than 2 to almost 11 per 1,000 Medicare enrollees. Even
among high-performing academic medical centers such as Intermountain,
Geisinger, Mayo Clinic, and Dartmouth-Hitchcock Medical Center, there is
more than a twofold variation in the rates of spinal surgery. Why is this the
case? In this area of preference-sensitive care, the indications for surgery are
not universally agreed upon, and the evidence to support the decision-making
process is imperfect for both the clinician and the patient. In these cases, it
is impossible to know which rate is right, and ideally, well-informed patients
should decide whether the intervention is right for them. Thus it is incumbent
on the healthcare system to provide patients with accurate, balanced informa-
tion and encourage them to participate in the decision.
Helping Patients Make Treatment Decisions
Decision Quality
When there is no clear answer as to which treatment is best, the pa-
tient should decide. High-quality decisions depend on adequate knowledge,
OCR for page 164
164 PATIENTS CHARTING THE COURSE
11.0
Back Surgery per 1,000 Enrollees
9.0
Rate of
Academic Center Surgery
7.0
Salt Lake City, UT 5.64
(Intermountain)
Danville, PA 5.24
5.0 (Geisinger)
United States average 4.28
Rochester, MN 3.63
3.0 (Mayo)
Lebanon, NH 2.26
(Dartmouth)
1.0
FIGURE 7-1 Turnip plot showing variation in rates of spinal surgeries.
SOURCE: The Dartmouth Atlas of Health Care. Copyright Trustees of Dartmouth
Figure 7-1.eps
College.
clarification of values, and resulting treatment choices that are consistent
with the patient’s values (Sepucha et al., 2004). There are, however, real
challenges to providing patients with optimal support.
Creating systems and processes to inform patients about the benefits
and risks of treatment alternatives, then assessing whether they have the
information they need to make high-quality decisions, is an important
requirement of today’s healthcare system. This requirement, however, may
appear to be in direct conflict with the demand for increased clinical ef-
ficiency and documentation that has resulted in ever shorter interactions
with patients.
Risks of the Typical Informed Consent Process
In cases in which effective care is indicated, a recommendation from
the physician may be appropriate, along with a discussion of the potential
benefits and harms with the decision maker. This approach is in keeping
with the traditional model of informed consent. Yet studies reveal that the
typical informed consent process usually does not help patients understand
the specific risks and benefits of treatment options (Holmboe et al., 2000).
OCR for page 165
165
APPLYING EVIDENCE FOR PATIENT-CENTERED CARE
Benefits of Shared Decision Making and Patient-Centered Care
Often the choice between competing treatment options is not so clear
and requires preference-sensitive decisions. In these situations, the treat-
ment choice should take into account an individual’s values and preferences
regarding the potential outcomes. Shared decision making is a process
that incorporates evidence-based medicine and requires both patients and
physicians to contribute information and participate. Providing decision
support and participating in patient-centered care requires skills and core
competencies. Several evidence-based practices foster communication and
shared decision making at the decision point, including involvement of both
the patient and the physician, information sharing by both parties, and
reaching an agreement about the treatment options (Charles et al., 1997).
Recent trials of decision-support systems designed to help patients
understand their treatment options reveal that informed patient choice
results in different patterns of practice from those found with patients who
experience usual care. However, medical opinion rather than patient prefer-
ence tends to dominate the treatment choice (Wennberg, 2002).
Replacing standard informed consent practices with a patient-centered
model of informed choice involves incorporating standardized communica-
tion of evidence, including risks, benefits, alternatives, experience, and cost.
These changes may lead to differences in healthcare utilization and can
empower patients to make informed decisions about their care (Krumholz,
2010). Providing patient-centered care depends on the comprehensive train-
ing of healthcare providers, increased consumer health literacy, and the
successful identification of implementation models (Edwards and Elwyn,
2009).
Use of Information Technology to Feed Forward the
Right Patient Information at the Right Time
Dartmouth-Hitchcock Medical Center uses health information technol-
ogy resources, such as video decision aids and electronic surveys, to collect
information from patients about their health history, behaviors, and health
preferences in order to provide them with feedback and decision support
and facilitate clear communication at the point of care. For example, all
patients with breast cancer are electronically screened for psychosocial,
financial, and emotional problems (such as anxiety and depression). Using
validated surveys such as the Patient Health Questionnaire 9, the com-
puter system can immediately score the survey and identify patients who
reach preset clinical thresholds. When a patient reaches that threshold, the
computer automatically e-mails Dartmouth-Hitchcock’s social workers to
intervene with patients identified as having a need.
OCR for page 166
166 PATIENTS CHARTING THE COURSE
In addition, our computer software systems allow us to collect infor-
mation about patients’ treatment preferences to share with their physician.
When patients come in for their appointments, their clinicians are armed
with information focused on what is meaningful to the patients. This facili-
tates discussion between clinician and patient and allows for the creation
of a personalized and tailored treatment plan.
Impact of Decision Aids on Patients’ Choices
A number of RCTs have shown the benefits of decision aids in support-
ing specific decisions. However, few trials have shown the impact on treat-
ment choice. At Dartmouth-Hitchcock, the use of decision aids has resulted
in approximately 30 percent of patients changing their initial treatment
preference. In an RCT, Deyo and colleagues studied the impact of a video
decision aid designed to inform patients about treatment options for back
pain. The overall rate of surgery was 22 percent lower in the group view-
ing the video than in controls (26 percent vs. 33 percent). The researchers
concluded that the video decision aid appeared to facilitate decision making
and could help ensure informed consent (Deyo et al., 2000).
Summary
Patients and providers should each have the benefit of making decisions
with the best available evidence. When treatments are not supported by
adequate evidence or when they involve trade-offs that can variably impact
a patient’s quality of life, the decision-making process should be structured
in a way that supports informed patient choice, by incorporating a discus-
sion of treatment alternatives, the best evidence available, and the patient’s
personal values.
At Dartmouth-Hitchcock Medical Center, a patient’s self-reported
health information is collected by computer systems. This information
is synthesized, and reports are created instantaneously to feed informa-
tion back to the patient about his/her health behaviors and conditions.
This same software can support the provision of sophisticated decision
aids when patients are facing preference-sensitive treatment choices. Ad-
ditionally, information on patients’ treatment preferences—including their
understanding of the key facts about the treatment and how they would
value the different possible outcomes of care—can be collected and fed
forward to their treating clinician at the point of care. This integration of
technology, patient information, and evidence provides a framework for
patient-centered care and informed choice.
OCR for page 167
167
APPLYING EVIDENCE FOR PATIENT-CENTERED CARE
EVIDENCE STANDARDS AND APPLICATION:
RIGHT CARE, RIGHT PATIENT, RIGHT TIME
Clifford Goodman, Ph.D.
The Lewin Group
Standards of evidence for what works in health care are generally ap-
preciated as instrumental to the widely shared goal of getting the right care
to the right patient at the right time. Such standards are used to support
clinical practice guidelines and other best practices, formulary decisions,
coverage policies, and more. These evidence standards are oriented largely
toward the tenets of experimental methods and often presented in frame-
works or hierarchies of study design. These hierarchies have not been static.
They have evolved over the past few decades to better address the particular
circumstances of healthcare decisions. However, persistent shortcomings in
the practical utility of current approaches for appraising evidence, as well
as certain mismatches between evidence hierarchies and the questions they
are intended to help answer, suggest that these approaches and hierarchies
need to be revisited and reconsidered in a broader context ranging from
identification of research priorities bringing research to practice.
Evidence Hierarchies and Their Application
Evidence standards can inform and improve healthcare decisions and
policies when they are used to appraise the quality of available evidence
on the impacts of healthcare interventions. Although evidence hierarchies
have evolved over the last two decades, they are oriented largely toward
the relative strength of evidence regarding the causal effect of an interven-
tion on particular health outcomes or other endpoints. In these hierarchies,
internal validity is associated with study design, and the resulting evidence
hierarchy of primary studies—i.e., those that collect original data—typically
places RCTs at the top as the “gold standard” for primary evidence. RCTs
are followed, typically, by nonrandomized controlled trials, various pro-
spective and retrospective observational (nonexperimental) studies with
concurrent or historical control groups, case series, single cases, and finally
(though not a form of evidence per se) expert opinion. In some hierarchies,
systematic reviews and meta-analyses of RCTs, where available, are placed
above RCTs (Sackett, 1989; USPSTF, 2008).
Notable variations or adaptations of this basic framework recognize
that implementation of a study design, not just the nature of the study
design itself, affects the quality of the evidence yielded. Accordingly, these
variations account for whether a study was “well designed” by employing
an acceptable means of randomization, by eliminating or minimizing other
OCR for page 168
168 PATIENTS CHARTING THE COURSE
potential biases or threats to internal validity, and by employing adequate
statistical power to detect potential treatment effects; what kind of com-
parison is involved (e.g., “head-to-head” trials of an intervention vs. a
standard of care rather than indirect comparison of the two or comparison
with placebo); whether the emphasis is on clinically relevant, rather than
surrogate, outcomes; and other attributes (Atkins et al., 2005; Harbour and
Miller, 2001; McAlister et al., 1999; USPSTF, 2008).
Limitations of Evidence Hierarchies
Despite improvements over the years, characteristics inherent in evi-
dence hierarchies can limit the development and appraisal of evidence nec-
essary for informing decisions about the right care for the right patient at
the right time. That there are more than 60 published evidence hierarchies
signals a lack of satisfaction or consensus in the field. Apart from incon-
sistencies across hierarchies in the definitions and categorization of study
designs (e.g., multiple and overlapping definitions of “cohort” and “quasi-
experimental” studies) and in rankings of study designs (e.g., whether RCTs
or systematic reviews reside at the top), most existing hierarchies are limited
by the mismatch between their original use and current application and the
associated overreliance on the RCT.
Evidence hierarchies are based largely on methodological principles
for assessing pharmacological models of therapy, including emphasizing
experimental control; placebo control groups where possible; randomized
assignment; narrowly defined patient groups; blinding of patients, clini-
cians, and investigators; and other attributes intended to enhance internal
validity. However, this approach can jeopardize external validity—that is,
the generalizability of findings to patients, settings, and other circumstances
of real-world care. For example, RCTs may provide clear results when as-
sessing the effect of a drug compared with a placebo in a narrowly defined
patient population with a specific health problem. However, such RCTs
may not provide information relevant for those making choices among
alternative therapies used in practice or for those with multiple comorbidi-
ties along with the specific health problem. Further, while RCTs and other
experimental study designs are intended to address the internal validity of
the causal effect of an intervention on outcomes, they have been misapplied
to other types of research questions pertaining to healthcare technologies,
such as the accuracy of a screening or diagnostic test, the prognostic ac-
curacy of a test, or rare or delayed adverse effects of a therapy.
As is recognized by some evidence hierarchies, relying on study type
alone for assessing the quality of evidence can be misleading. For example,
poor design or implementation of a high-ranking study type may yield
findings that are less valid than those from study types lower on the hier-
OCR for page 169
169
APPLYING EVIDENCE FOR PATIENT-CENTERED CARE
archy that are well designed and rigorously implemented. Further, evidence
hierarchies often cannot adequately accommodate or combine results from
multiple study design types, even when no single study design can answer
some evidence questions.
Limitations of Randomized Controlled Trials
RCTs are usually important, and sometimes essential, components of
a rigorous evidence base for demonstrating the causal effects of healthcare
interventions on outcomes. However, RCTs have important limitations,
including some circumstances in which underlying assumptions about them
are not valid (Rawlins, 2008; Walach et al., 2006). For example, while
undertaking to randomize patients to one intervention or another assumes
equipoise between the two, patients and clinicians involved in RCTs often
bring preferences for one or the other. Whereas RCTs assume a lack of
knowledge about the merits of the interventions in questions, there may
indeed be relevant evidence from other sources (e.g., phase II drug trials)
that undermines the utility of the null hypothesis in experimentation. RCTs
also have a preference for specificity—i.e., that only the specific effects
attributable to an intervention are therapeutically valid. Another assump-
tion underlying RCTs that often is not valid relates to the context indepen-
dence of the effect—i.e., that there is some “true” magnitude of efficacy
or a stable effect size independent of the context in which an intervention
is used.
Another major weakness concerns incorrectly assuming that the find-
ings about a therapeutic effect from an RCT are externally valid—that is,
readily transferable into clinical practice—if the exclusion and inclusion
criteria of the trial match the characteristics of a given patient. Although
this shortcoming of RCTs is generally well recognized, its extent and sig-
nificance for patient care are not well understood.
Other methodological problems with RCTs can include relying on
intermediate endpoints rather than health outcomes; having inadequate
statistical power or duration for assessing benefits and harms, especially
those that are rare or delayed; and being unsuccessfully blinded to patients,
clinicians, or other investigators. Probability theories may pose problems
(especially with frequentist approaches) in the form of arbitrary selection
of p-values and the difficulty of applying them to everyday patient care;
the multiplicity of endpoints compared, stopping rules, and analysis of
subgroups; and resistance to Bayesian approaches.
Furthermore, there are various circumstances in which RCTs can be
inappropriate. For example, they may raise bioethical and legal concerns;
they may be difficult or impossible to conduct for rare diseases; and they
may be unnecessary for very large treatment effects or to establish causa-
OCR for page 176
176 PATIENTS CHARTING THE COURSE
aids) readily accessible at the place and time of clinician–patient decisions
(Pearson et al., 2009).
The field of comparative effectiveness research (CER) emphasizes the
need to rely on multiple evolving methods, including more advanced data
infrastructure and linkages among claims data, electronic health records,
registries, and other sources. Further, consistent with the intent of Congress
and national priority-setting reports for CER, the field should focus on
subgroup analyses and priority populations, which will expand the base
of well-founded evidence for patient decisions (FCC, 2009; IOM, 2009).
Indeed, this focus on CER will help ensure that it not only generates
population-based evidence but also supports progress toward personalized
medicine—including patients’ genomes; their health states; and the behav-
ioral, environmental, socioeconomic, cultural, and other personal determi-
nants of their response to healthcare interventions (Goodman, 2009).
Several other trends reflect this sharper focus on meeting the needs
of decision makers. Patient input has been advanced as central to discus-
sions of CER priority setting, study design, and identification of key out-
comes, with citizen councils and forums in Europe, Canada, and the United
States offering various models for engaging the public. As noted above
with respect to pharmacogenomic testing, evidence appraisals are increas-
ingly attuned to intervention types, applications, and settings. Regulators
and payers are more explicit about evidence expectations, including study
designs and designation of health outcomes and other study endpoints.
Bayesian and adaptive clinical trial designs, the focus of increased interest,
offer flexible variations on primary data collection that optimize the use of
accumulated findings to derive evidence more efficiently for responsive and
nonresponsive patient groups (Orloff et al., 2009).
Finally, much greater and earlier interaction among innovators, regu-
lators, payers, and health technology assessment functions is occurring on
evidence expectations or requirements well in advance of regulatory deci-
sions or coverage decisions by payers. Of importance, these discussions are
focused on anticipating evidence needs throughout the life cycle of a tech-
nology and often include clinician and patient input at the outset. Rather
than retrospective reaction to fill evidence gaps or make determinations
based on the available evidence, this approach allows for coordination of
evidence development that takes advantage of the right methods for the
right questions and builds toward a totality of evidence on what works best
for individual patients.
OCR for page 177
177
APPLYING EVIDENCE FOR PATIENT-CENTERED CARE
TRANSLATION AND COMMUNICATION NEEDS FOR
CARE IN THE FACE OF UNCERTAIN EVIDENCE
Fran M. Visco, J.D.
National Breast Cancer Coalition
Approaches to translation and communication needs for health care
must take into account the broader context within which the public receives
health information. The public receives healthcare information from a vari-
ety of sources including healthcare providers, traditional media, the Internet,
family and friends, and patient and medical advocacy organizations. For
many reasons, the temptation is to provide messages that are short and
simple. But when evidence is uncertain, healthcare messages are not simple,
and there needs to be a system that is honest about this uncertainty. Health
messages must not be presented as absolute or simple when evidence is not.
We need to educate the media, the public, and the medical community about
evidence, and we need to institute oversight mechanisms to ensure that
health communications are based on the best available evidence.
The public must trust that the healthcare system works for their benefit.
The public’s expectation is that care is communicated and delivered in a
system in which
• patient needs are paramount,
• care is based on evidence,
• risks and benefits will be explained and understood, and
• interventions will change with new knowledge.
The Context in Which the Public Receives
Information About Health Care
The public uses various sources for health information. Survey data
indicate that these sources vary depending on age and education. For ex-
ample, a 2009 survey conducted by the Pew Internet and American Life
Project found that when asked, “Now thinking about all the sources you
turn to when you need information or assistance in dealing with health or
medical issues, please tell me if you use any of the following sources”:
• 86 percent of all adults ask a health professional, such as a physician;
• 68 percent ask a friend or family member;
• 57 percent use the Internet;
• 54 percent use books or other printed reference material;
• 33 percent contact their insurance provider; and
• 5 percent use another source not mentioned above.
OCR for page 178
178 PATIENTS CHARTING THE COURSE
Although the Pew report found that Internet use did not replace other,
traditional sources of health information, significant differences in health
information sources did emerge among age groups. For example, adults
aged 18–29 are significantly less likely than older adults to consult a health
professional (79 percent, compared with 88 percent of those aged 30–49,
89 percent of those aged 50–64, and 89 percent of those aged 65+). Younger
adults are more likely than older adults to consult a friend or family member.
Seventy-eight percent of adults aged 18–29 and 72 percent of those aged
30–49 consult a friend or family member, compared with 58 percent of
those aged 50–64 and 59 percent of those aged 65+ (Fox and Jones, 2009).
The Pew survey also found significant differences in health information
sources by education level, with 94 percent of college graduates consult-
ing a health professional, compared with only 83 percent of high school
graduates. These findings are similar to those reported in a 2008 Center
for Studying Health System Change report. According to that report, health
information–seeking behavior differed by education level, with 72 percent
of people with a graduate degree seeking information from any source,
compared with 42 percent of those without a high school diploma (Tu and
Cohen, 2008).
In a survey commissioned by the National Breast Cancer Coalition
(NBCC) in 2009, a slightly different pattern of health information sources
was found among women seeking information on breast cancer. Among
women actively looking for information about breast cancer or its treat-
ments, the four most common sources were the Internet, talking with a friend
or relative, talking with a doctor or other medical professional, and magazine
articles. Overall, the Internet was by far the most common source, consulted
by 71 percent of women as compared with 53 percent who spoke with
friends and relatives, 45 percent who spoke with a doctor, and 43 percent
who found information in magazines. This pattern was seen in all age groups.
In addition, all age groups were more likely to consult a friend or relative
than a physician for breast cancer information. Women aged 25–34 were
more likely to seek breast cancer information in a magazine (48 percent) than
to speak with a physician (40 percent), while women aged 50+ were slightly
more likely to speak with a friend or relative (49 percent) or a breast cancer
survivor (47 percent) than with a physician (43 percent).
When NBCC survey respondents were asked what they had heard or
read about breast cancer and where, television programs were the most
likely response (57 percent). The next most common responses were the
Internet and advertisements, both of which had increased in usage over
the prior two years (from 29 percent to 38 percent, and 15 percent to
34 percent, respectively).
Given these findings on sources of information, no communication
strategy can focus simply on physicians and patients; it is also essential to
OCR for page 179
179
APPLYING EVIDENCE FOR PATIENT-CENTERED CARE
take into account messages received through print, television, and online
media. Also, historical communications about an intervention must be
considered in crafting any new messages.
Several recent examples illustrate confusion in healthcare communities
and among the public about recommendations for health interventions when
• the intervention became standard with limited or no evidence
(e.g., breast self-examination, hormone replacement therapy, use
of erythropoiesis-stimulating agents as supportive therapy, manage-
ment of ductal carcinoma in situ);
• uncertainty exists among healthcare professionals; and
• self-interests of healthcare professions, advocacy groups, and the
media appear to trump the evidence.
A Case History: The U.S. Preventive Services Task Force
The recent uproar over the USPSTF recommendations for breast cancer
screening captures many of the difficulties of communicating health infor-
mation today. In 2007, the USPSTF outlined refined methods for devel-
oping recommendations that included an outline of communication and
dissemination strategies (Guirguis-Blake et al., 2007). Unfortunately, these
strategies focused on dissemination to professionals via medical journals
and to federal agencies, professional societies, and quality improvement
organizations via public meetings. As the recent breast cancer screening
example illustrates, these are not the only important audiences. Strategies
for disseminating information to the media, policy makers, and the public
are also crucial components of any communication plan.
On November 16, 2009, the USPSTF revised its recommendations on
screening for breast cancer in the general population (USPSTF, 2009). In
summary:
• The USPSTF does not recommend that women automatically begin
mammography screening at the age of 40. Instead, it recommends
that the decision to start regular, biennial mammography screening
before age 50 should be an individual one and take into account
patient context, including the patient’s values regarding specific
benefits and harms.
• The USPSTF recommends mammography screening every other
year for women aged 50–74.
• The USPSTF concludes that evidence is insufficient to determine the
harms and benefits of mammography screening in women over 74.
• The USPSTF recommends against healthcare providers teaching
breast self-examination.
OCR for page 180
180 PATIENTS CHARTING THE COURSE
• Evidence was insufficient for the USPSTF to make a recommen-
dation on clinical breast examination, digital mammography, or
magnetic resonance imaging.
These recommendations were not significantly different from those
issued in 2002.
The release of these updated recommendations was communicated in
an article in the Annals of Internal Medicine, a biweekly medical journal.
This typical communication strategy for releasing USPSTF recommenda-
tions had not led to mass confusion and hysteria in the previous year when
the prostate cancer screening recommendations were updated. But the reac-
tion to the revised breast cancer screening recommendations from the pub-
lic, policy makers, the media, and the healthcare community was far from
typical (Goldberg, 2009).
The timing of the release of the new recommendations was unfortu-
nate, coinciding with a congressional vote on a healthcare reform bill after
months of a particularly contentious debate. As a result, the revised screen-
ing recommendations were cast as one more example of “big government
rationing health care,” which added to the hysteria around the recommen-
dations’ release. Moreover, there have for years been simplistic messages
about breast cancer screening—for example, that “early detection saves
lives”—from the American Cancer Society, the National Cancer Institute,
patient groups, trade associations, and the media. However, the evidence
behind these statements and campaigns was not part of any discussion
with the public, and the public health experts and primary care physi-
cians with expertise in the area were on the periphery of these messaging
campaigns. To further complicate the situation, Congress weighed in and
demanded that the National Cancer Institute provide a clear message on
screening, despite the lack of strong evidence of overall benefit. Indeed, in
1997 the Senate held hearings and passed a nonbinding resolution in sup-
port of mammograms for women under age 50 by a vote of 98-0 (Kassirer,
1997).
Years of communications about screening resulted in an ad campaign’s
being converted into absolute truth. In fact, any evidence of the limitations
of mammography screening and questioning of the evidence on which
these simplistic messages were based was ignored or vilified. For example,
a research article reported data indicating that “the natural course of some
screen-detected invasive breast cancers is to spontaneously regress” (Zahl
et al., 2008). Rather than taking the opportunity to explain the evidence to
the public, the American Cancer Society’s director of cancer screening was
quoted in USA TODAY as saying, “It’s important that people not wonder
if women lost their breasts for no reason. That’s reprehensible conjecture”
(Szabo, 2008).
OCR for page 181
181
APPLYING EVIDENCE FOR PATIENT-CENTERED CARE
Against this background, when the USPSTF issued its revised recom-
mendations in the Annals of Internal Medicine, many in the media, policy
makers, medical trade associations, and healthcare providers attacked the
Task Force and the revised guidelines. What did the public hear?
• The American College of Radiology, a medical trade association
representing radiologists and the field of medical imaging, stated:
“Countless American women will die needlessly from breast cancer
each year” (American College of Radiology, 2009).
• The American Cancer Society responded: “With its new recom-
mendations, the USPSTF is essentially telling women that mam-
mography at age 40 to 49 saves lives; just not enough of them”
(American Cancer Society, 2009).
• ABC News misrepresented the guidelines with the headline “Stop
Annual Mammograms, Govt. Panel Tells Women Under 50” and
implied that cost savings, not evidence, motivated the change.
“Anecdotally, most people in the United States can think of a
woman they know who caught breast cancer through a routine
mammogram long before she turned 50. Many patient advocates
wonder if money fueled the decision” (Cox, 2009).
• A Fox News Sunday interview included Bernadine Healy, M.D.,
the former director of the National Institutes of Health, strongly
urging women to ignore the USPSTF screening recommendations
because they will result in more women dying of breast cancer (Fox
News, 2009).
• Secretary of Health and Human Services Sebelius undermined the
credibility of the Task Force and its guidelines by stating: “[The
U.S. Preventive Services Task Force] has presented some new evi-
dence for consideration, but our policies remain unchanged. My
message to women is simple. Mammograms have always been
an important life-saving tool in the fight against breast cancer,
and they still are today. Keep doing what you have been doing
for years—talk to your doctor about your individual history, ask
questions, and make the decision that is right for you” (Goldberg,
2009).
• Daniel Kopans, M.D., a radiologist and director of breast imaging
at Massachusetts General Hospital, is often quoted in the media
along the lines of his letter to the editors of the Annals of Internal
Medicine, which stated: “Suggesting that these guidelines are based
on clear evidence is not supported by the facts. . . . I believe that
some of the advisors to the USPSTF have major, undisclosed, career
interests in the guidelines. They have received funding for what I
believe are nihilistic approaches, constituting more insidious con-
OCR for page 182
182 PATIENTS CHARTING THE COURSE
flicts of interest than the obvious conflicts of radiologists, such as
myself. . . . Task Force members had no expertise in mammography
screening or even breast cancer care. . . . In sum, the new USPSTF
guidelines are unscientific, endanger women through false analyses,
and should be withdrawn” (Kopans, 2010).
Unfortunately, these attacks dominated the media and the public per-
ception of the updated guidelines, despite the fact that public health and
primary care experts supported the recommendations:
• The American College of Physicians, a professional organization
for physicians specializing in internal medicine, had issued clinical
practice guidelines in 2007 for mammography screening among
women aged 40–49 that encouraged physicians to carefully assess
individual women’s risks for breast cancer and discuss with them
the potential benefits and harms of mammography screening so they
can make informed individual decisions about screening (Qaseem et
al., 2007).
• The Cochrane Collaboration published a systematic review of
mammography screening in 2006 and concluded that while such
screening likely reduces breast cancer mortality, the magnitude of
the effect is uncertain, and the screening also results in some women
receiving a cancer diagnosis even though their cancer would not
have led to death or sickness (Gøtzsche and Nielsen, 2006).
• The American Academy of Family Physicians recommended that
mammography screening begin in average-risk women at age 50,
and that all women aged 40–49 be counseled about the risks and
benefits of mammography before making a decision to undergo
screening (AAFP, 2003).
• The Annals of Internal Medicine conducted a readers’ survey and
found that among clinician respondents, 67 percent reported that
they will stop offering routine mammograms to women in their 40s
(Annals of Internal Medicine, 2010).
While this is but one case study, it received a significant amount of
attention. An atmosphere was created that undermined the trust between
the American people and public health officials.
Lessons Learned
We need to be honest. We would all prefer that the correct message
be simple and certain. However, it most often is not. We need better poli-
cies to ensure that the public, the media, healthcare providers, and policy
OCR for page 183
183
APPLYING EVIDENCE FOR PATIENT-CENTERED CARE
makers all have the tools they need to understand and explain uncertainty,
understand evidence, and keep the needs of patients paramount.
Professional associations need stronger oversight because many clini-
cal practice guidelines are issued by professional societies. Often, the de-
velopers of guidelines have a financial conflict of interest in the use of the
interventions highlighted in the guidelines. For example, the American
College of Radiology issues statements on breast cancer screening, although
its members are not public health experts and have a financial interest in
the outcome of such recommendations. The American Society for Clinical
Oncology still recommends and has actively lobbied the federal govern-
ment on coverage issues for the use of erythropoiesis-stimulating agents as
supportive therapy for cancer patients, despite evidence that they actually
stimulate tumor progression (FDA, 2010; Rizzo et al., 2008).
The public can receive particularly confusing messages when profes-
sional societies differ publicly in their assessment of the evidence, as the
American College of Radiology and the American College of Physicians did
on breast cancer screening for women aged 40–49. What ethical responsi-
bilities do these trade associations have in making these pronouncements
public? How do we communicate to the public who the experts are?
Campaigns to better educate the public, policy makers, and the media
about the importance of evidence are crucial. We should not underestimate
the public’s ability to understand and accept evidence. In the 2009 NBCC
consumer survey, for example, consumers identified comparative effective-
ness research as more likely than other healthcare reforms to improve qual-
ity of care for breast cancer patients.
Projects such as NBCC’s Project LEAD® (Leadership, Education, and
Advocacy Development) training courses are important. Such courses on
critically evaluating research and evidence need to be made more broadly
available to the general public and journalists. Much work on this front
is also being done by others, including Gary Schwitzer with his popular
HealthNewsReview blog3; the Dartmouth Institute for Health Policy and
Clinical Practice’s Center for Medicine and the Media4; and the Foundation
for Informed Medical Decision Making.5
3 For more information see http://www.healthnewsreview.org/blog/ (accessed October 15,
2010).
4 For more information see http://tdi.dartmouth.edu/centers/medicine-and-the-media/ (ac-
cessed October 15, 2010).
5 For more information see http://www.informedmedicaldecisions.org/index.html (accessed
October 15, 2010).
OCR for page 184
184 PATIENTS CHARTING THE COURSE
REFERENCES
AAFP (American Academy of Family Physicians). 2003. Summary of policy recommendations
for periodic health examinations, rev 5.4. Leawood, KS: American Academy of Family
Physicians.
American Cancer Society. 2009. American Cancer Society responds to changes in USPSTF mam-
mography guidelines. http://www.cancer.org/docroot/med/content/med_2_1x_american_
cancer_society_responds_to_changes_to_uspstf_mammography_guidelines.asp (accessed
October 15, 2010).
American College of Radiology. 2009. USPSTF mammography recommendations will re-
sult in countless unnecessary breast cancer deaths each year. http://www.acr.org/
MainMenuCategories/media_room/FeaturedCategories/PressReleases/USPSTFMammoRecs.
aspx (accessed October 14, 2010).
Annals of Internal Medicine. 2010. Editorial: When evidence collides with anecdote, politics,
and emotion: Breast cancer screening. Annals of Internal Medicine 152(8):531-532.
Atkins, D., P. Briss, M. Eccles, S. Flottorp, G. Guyatt, R. Harbour, S. Hill, R. Jaeschke, A.
Liberati, N. Magrini, J. Mason, D. O’Connell, A. Oxman, B. Phillips, H. Schunemann,
T. Edejer, G. Vist, J. Williams, and T. G. W. Group. 2005. Systems for grading the quality
of evidence and the strength of recommendations II: Pilot study of a new system. BMC
Health Services Research 5(1):25.
Charles, C., A. Gafni, and T. Whelan. 1997. Shared decision-making in the medical encoun-
ter: What does it mean? (or it takes at least two to tango). Social Science & Medicine
44(5):681-692.
Cox, L. 2009. Stop annual mammograms, govt. panel tells women under 50. ABC Medical
News Unit.
Deyo, R., D. Cherkin, J. Weinstein, J. Howe, M. Ciol, and A. J. Mulle. 2000. Involving
patients in clinical decisions: Impact of an interactive video program on use of back
surgery. Medical Care 38(9):959-969.
Ebell, M. H., J. Siwek, B. D. Weiss, S. H. Woolf, J. Susman, B. Ewigman, and M. Bowman.
2004. Strength of Recommendation Taxonomy (SORT): A patient-centered approach to
grading evidence in the medical literature. American Family Physician 69(3):548-556.
Edwards, A., and G. Elwyn. 2009. Shared decision making in health care: Achieving evidence
based patient choice. Oxford: Oxford University Press.
FCC (Federal Coordinating Council for Comparative Effectiveness Research). 2009. Report
to the President and Congress. http://www.hhs.gov/recovery/programs/cer/cerannualrpt.
pdf (accessed October 11, 2010).
FDA (Food and Drug Administration). 2010. Information on erythropoiesis-stimulating agents
(ESA) epoetin alfa (marketed as Procrit, Epogen) darbepoetin alfa (marketed as Aranesp).
Safety announcement. http://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafety
InformationforPatientsandProviders/ucm109375.htm (accessed October 15, 2010).
Fisher, E. S., D. E. Wennberg, T. A. Stukel, D. J. Gottlieb, F. L. Lucas, and E. L. Pinder. 2003.
The implications of regional variations in Medicare spending. Part 1: The content,
quality, and accessibility of care. Annals of Internal Medicine 138(4):273-287.
Fox News. 2009. Transcript: Dr. Bernadine Healy on “FNS.” http://www.foxnews.com/
story/0,2933,576291,00.html (accessed October 15, 2010).
Fox, S., and S. Jones. 2009. The social life of health information. http://www.pewinternet.
org/Reports/2009/8-the-social-life-of-health-information (accessed October 14, 2010).
Goldberg, P. 2009. HHS secretary rebukes task force guidelines on breast cancer screening.
The Cancer Letter 35(43).
OCR for page 185
185
APPLYING EVIDENCE FOR PATIENT-CENTERED CARE
Goodman, C. 2009. Comparative effectiveness research and personalized medicine: From
contradiction to synergy. http://www.lewin.com/content/publications/Lewin_CER-PM.
pdf (accessed October 14, 2010).
Gøtzsche, P. C., and M. Nielsen. 2006. Screening for breast cancer with mammography.
Cochrane Database of Systematic Reviews 4.
Guirguis-Blake, J., E. Calonge, T. Miller, A. Siu, S. Teutsch, and E. Whitlock. 2007. Current
processes of the U.S. Preventive Services Task Force: Refining evidence-based recommen-
dation development. Annals of Internal Medicine 147(2):117-122.
Harbour, R., and J. Miller. 2001. A new system for grading recommendations in evidence
based guidelines. BMJ 323(7308):334-336.
Holmboe, E., D. Fiellin, E. Cusanelli, M. Remetz, and H. Krumholz. 2000. Perceptions
of benefit and risk of patients undergoing first-time elective percutaneous coronary
revascularization. Journal of General Internal Medicine 15(9):632-637.
Howick, J., P. Glasziou, and J. K. Aronson. 2009. The evolution of evidence hierarchies: What
can Bradford Hill’s ‘guidelines for causation’ contribute? Journal of the Royal Society of
Medicine 102(5):186-194.
IOM (Institute of Medicine). 2001. Crossing the quality chasm: A new health system for the
21st century. Washington, DC: National Academy Press.
———. 2009. Initial national priorities for comparative effectiveness research. Washington,
DC: The National Academies Press.
Kassirer, J. 1997. Practicing medicine without a license—the new instructions by Congress.
New England Journal of Medicine 336:1747.
Kopans, D. 2010. The Annals is ignoring and denigrating fundamental scientific issues that
make the USPSTF guidelines on mammography screening, scientifically, unsupportable.
http://www.annals.org/content/early/2010/02/12/0003-4819-152-8-201004200-00210.
short/reply (accessed June 20, 2011).
Krumholz, H. M. 2010. Informed consent to promote patient-centered care. Journal of the
American Medical Association 303(12):1190-1191.
McAlister, F. A., A. Laupacis, G. A. Wells, D. L. Sackett, and for the Evidence-Based Medicine
Working Group. 1999. Users’ guides to the medical literature. XIX. Applying clinical
trial results. B. Guidelines for determining whether a drug is exerting (more than) a class
effect. Journal of the American Medical Association 282(14):1371-1377.
O’Connor, A. M., J. E. Wennberg, F. Legare, H. A. Llewellyn-Thomas, B. W. Moulton, K. R.
Sepucha, A. G. Sodano, and J. S. King. 2007. Toward the “tipping point”: Decision aids
and informed patient choice. Health Affairs 26(3):716-725.
Orloff, J., F. Douglas, J. Pinheiro, S. Levinson, M. Branson, P. Chaturvedi, E. Ette, P. Gallo, G.
Hirsch, C. Mehta, N. Patel, S. Sabir, S. Springs, D. Stanski, M. R. Evers, E. Fleming, N.
Singh, T. Tramontin, and H. Golub. 2009. The future of drug development: Advancing
clinical trial design. Nature Review of Drug Discovery 8(12):949-957.
Pearson, S.-A., A. Moxey, J. Robertson, I. Hains, M. Williamson, J. Reeve, and D. Newby.
2009. Do computerised clinical decision support systems for prescribing change practice?
A systematic review of the literature (1990-2007). BMC Health Services Research
9(1):154.
Qaseem, A., V. Snow, K. Sherif, et al. 2007. Screening mammography for women 40 to 49
years of age: A clinical practice guideline from the American College of Physicians. An-
nals of Internal Medicine 146:511-515.
Rawlins, M. D. 2008. De testimonio: On the evidence for decisions about the use of therapeu-
tic interventions. The Harveian Oration delivered before the Fellows of the Royal College
of Physicians of London on 16 October 2008. London: Royal College of Physicians.
OCR for page 186
186 PATIENTS CHARTING THE COURSE
Rizzo, J., M. Somerfield, and K. Hagerty. 2008. Use of Epoetin and Darbepoetin in patients
with cancer: 2007 American Society of Clinical Oncology/American Society of Hema-
tology clinical practice guideline update. Journal of Clinical Oncology 26(1):132-149.
Sackett, D. L. 1989. Rules of evidence and clinical recommendations on the use of anti-
thrombotic agents. Chest 95(2 suppl):2S-4S.
Sepucha, K. R., F. J. Fowler, Jr., and A. G. Mulley, Jr. 2004. Policy support for patient-centered
care: The need for measurable improvements in decision quality. Health Affairs Suppl
Web Exclusives:VAR54-62.
Szabo, L. 2008. Study: Do breast tumors go away on their own. USA Today, http://www.
usatoday.com/news/health/2008-11-24-breast-cancer_N.htm (accessed June 20, 2011).
Teutsch, S. M., L. A. Bradley, G. E. Palomaki, J. E. Haddow, M. Piper, N. Calonge, W. D.
Dotson, M. P. Douglas, and A. O. Berg. 2009. The Evaluation of Genomic Applications
in Practice and Prevention (EGAPP) initiative: Methods of the EGAPP working group.
Genetics in Medicine 11(1):3-14
Tu, H. T., and G. R. Cohen. 2008. Striking jump in consumers seeking health care information.
http://www.hschange.com/CONTENT/1006/# (accessed October 14, 2010).
USPSTF (U.S. Preventive Services Task Force). 2008. Procedure manual. http://www.
uspreventiveservicestaskforce.org/uspstf08/methods/procmanual.htm (accessed Octo -
ber 14, 2010).
———. 2009. Screening for breast cancer: US Preventive Services Task Force recommendation
statement. Annals of Internal Medicine 151(10):716-726, W-236.
Walach, H., T. Falkenberg, V. Fonnebo, G. Lewith, and W. B. Jonas. 2006. Circular instead
of hierarchical: Methodological principles for the evaluation of complex interventions.
BMC Medical Research Methodology 6:29.
Weinstein, J., D. Goodman, and J. Wennberg. 1998. The orthopaedic workforce: Which rate
is right? Journal of Bone and Joint Surgery 80(3):327-330.
Wennberg, J. E. 2002. Unwarranted variations in healthcare delivery: Implications for aca-
demic medical centres. BMJ 325(7370):961-964.
Zahl, P., J. Maehlen, and H. Welch. 2008. The natural history of invasive breast cancers de-
tected by screening mammography. Archives of Internal Medicine 168(21):2311-2316.
