Mr. Chairman, thank you for giving me the opportunity to address the Committee on the very
timely and important issue of health care quality. AHCPR's mission is to provide science-based
information that will improve decision making at all levels -- from patients, to clinicians, to
health care system leaders, to public and private policymakers. AHCPR's goal is to ensure in an
increasingly market-based health care system that unbiased, state-of-the-science information
drives informed decision making.
Today, I would like to provide you with my perspective on quality, not only as the
Administrator of Agency for Health Care Policy and Research (AHCPR), but as someone who
has spent his entire career in clinical medicine. My perspective on the quality of health care is
also shaped by my experiences as the Chair of Medicine at Georgetown University, as the Chair
of the Physician Payment Review Commission (PPRC), and as a professor of medicine.
What is Quality?
Most health professionals see quality as having three dimensions: structure, process, and
outcome. Structure represents the basic characteristics of physicians, hospitals, other
professionals and other facilities. It describes whether there are well-educated health
professionals, appropriate hospitals, nursing homes, and clinics, as well as well-maintained
medical records and good mechanisms for communication between clinicians. For example: Is
the mammography equipment up to date and maintained properly? Are the cardiologists
well-trained and board certified?
Structure is the framework in which we practice, and although the education of professionals
and the facilities in which we practice is among the best in the world, let us never take them for
If the structure is solid we can concern ourselves with the process of medical care. Concern
for process suggests that quality is determined not just by having the right people and facilities
available, but also means the right things must get done in the right way. Process includes
questions like: Was the mammogram done for a woman at risk for breast
cancer? Was the heart attack treated in the most up-to-date manner?
The third dimension, outcome, reflects the end result of care. Did people get better? Was
disease or disability reduced? Was it reduced as much as it could have been, given what we know
is scientifically possible? This is an area of increasing interest, but one in which we don't know is
striking. We need to be able to measure the outcomes of care so that we know which types of
care really help patients and so that we can look to instances of poor opportunities for
improvement. For example, outcomes tells us whether breast cancer was detected early enough
to treat effectively? Did the patient survive the heart attack with the highest level of functioning?
I have felt for years that we need to ensure that we are protecting the quality of the health care
provided to this Nation's citizens by developing science-based, reliable quality measurement and
Patients' Concerns About Quality
The context of our discussion about quality is that our country has a market-oriented health
care system. Whether publicly or privately financed, a basic element of market-oriented health
care is the opportunity for informed choices by purchasers, by parents and by those acting on
their behalf. An essential part of health care quality improvement is to empower with
information the ultimate consumer of health care - the patient. And we know that Americans are
concerned about the quality of care they receive.
A recent Kaiser Family Foundation poll found that a majority of Americans rate their health
plans a "'B" or higher. However, 55 percent of respondents in managed care plans and 34
percent with traditional insurance responded "yes" to question about whether their plans "would
be more concerned about saving money than about what is the best medical treatment."
We need an infrastructure in place that will provide consumers with information on health
care quality. This information should include outcomes of treatments, patient assessments, and
other quality indicators.
Clinicians' Need For Information on Quality
Today I would like to concentrate on the clinical aspects of quality once someone has
gained access to that care.
Clinicians want to do the right thing. Physicians and other health care professionals need
better scientific information on which treatments are most appropriate for which patients and at
what point during the course of their care. I can not overstate the importance of building the
evidence base for clinical practice. It gives health care professionals the unbiased information
they need to make effective, timely diagnoses and to provide appropriate treatment. Physicians
need to know what works to provide quality health care, and their patients deserve no less.
The ever-changing, ever-growing medical literature is making it difficult for busy physicians
and other health care professionals to keep up with the latest scientific evidence. For example, it
is estimated that if a physician read two peer-reviewed journal articles each night, at the end of
the year, he would be 800 articles behind in his reading. While it is good to have a large body of
information, we need to provide this information in a useful format.
Having a readily accessible evidence base for treatment also will help improve the
communication between patients and their doctors. Together physicians and their patients can
use this information to find the most effective, appropriate, and least burdensome treatment.
This sharing of information and communication is the foundation of a good doctor-patient
In my role as the head of a federal health agency, I see the importance for providing a
sound evidence base for clinical practice in order to assure and improve its quality. This is
invaluable to a health care system that is faced with the critical issue of what care is appropriate.
Medical practice varies widely in this country. The issue of variation is not new to you. Dr.
John Wennberg's work has shown that medical practice varies widely in this country. AHCPR
has sponsored a substantial portion of Dr. Wennberg's work in the area of prostate disease. His
research team found that the rate of radical prostatectomy for Medicare patients in Fort Worth,
Texas is twice the rate in Dallas, Texas (3. 1 per 1000 Medicare enrollees versus 1. 9 per 1000).
These variations can vary region to region State to State, or within States. For example, the rate
for radical prostatectomy for Medicare patients in Baltimore, Maryland is approximately three
times the rate in Salisbury, Maryland.
Variation provides us an opportunity to study what care is appropriate, how much is enough,
and what is fair. Let me qualify that variation is not inherently bad. In some cases, variation is
caused by geographical, epidemiological, or cultural preferences. For example, we expect to
have a higher rate of skin cancer in the South, and therefore more treatment for skin cancer.
In other cases, variation may point to areas of uncertainty in medical practice. This
variation has demonstrated that there are inconsistencies in how health care is delivered in this
country. What those inconsistencies mean is a subject for further research and data collection,
and point to the need for better information on what works, when and for whom. We need more
research on and knowledge about health care outcomes to understand whether variation in a
medical practice should be celebrated or eliminated.
An important component of variation is personal preference and values. Patients and their
clinicians must weigh personal preferences when making health care decisions. For example,
research has indicated that some patients with laryngeal cancer would rather risk living fewer
years than undergo a procedure that would cause them to lose their voices. For them, the issue is
not the length of their lives but the quality of their lives. That is a decision about outcomes that
patients make reflecting their own priorities, their own values, their own choices.
Although individual preferences are important, we know that there are some essential issues
of quality that are common to all patient encounters. We know that certain drugs and certain
immunizations should be given in certain clinical circumstances. Our challenge is to provide
consumers with information on quality that will help them make decisions about the care they
receive according to their individual needs and desires.
These decisions about the clinical services that will serve patients' needs represent one level
of choice that requires valid information. When we decide on a health plan and a clinician or
hospital we deserve the kind of information we can get when choosing a car, a home, and most
other products and services that we use our limited resources to buy.
A survey cosponsored by AHCPR and The Kaiser Family Foundation found that a large
majority of Americans (nearly 90% in every case) felt that quality information --such as how a
plan cares for its members who have health problems, ease of getting care, and success in
treating or managing disease -- was "very important" when choosing a health plan.
With the growing complexity of the marketplace, the demand for this kind of information is
growing. We cannot leave the other health care stakeholders out of the mix. We must also
ensure that health system leaders and policymakers, also have the information they need to make
Quality Measurement and Improvement
Certainly a key element in quality health care is to assure that the three elements are in place
-- that the structure of the system is strong and that clinicians are able to know what to do and are
able to do it with skill and expertise. We also need to measure how well we do in the structure
and process of care, to measure the outcomes we achieve, and to identify areas where there is an
opportunity for improvement. We need to develop more methods, and better methods,
for measuring quality. We also need to communicate the results in useful, understandable
formats that help improve health care decision making. And, it is critically important that this
information is available in the public domain.
We have been successful in doing this in the automobile industry. For instance, if I am
buying a car, I know that I can find data on the safety, efficiency, and reliability of different car
models. This data is based on accepted measurements, such as crash tests, service records, and
Like the automobile industry, we must make it the goal for our health care system to provide
similar information on the quality of health care services. To that end, we must strive to develop
accepted measures and instruments used to gauge and improve the quality of health care services.
Last Friday, the Secretary of Health and Human Services announced the release of AHCPR's
Consumer Assessments of Health Plans Survey (CAHPS), a series of questionnaires designed to
be used by public- and private-sector health plans, employers, and other organizations to survey
their members and employees. The information from CAHPS questionnaires, presented in the
CAHPS report formats can help consumers and group purchasers compare health plans and make
more informed choices based on quality. Both the questionnaires and report formats have been
tested widely in demonstration sites around the country.
We know that consumers select health plans based on the recommendations of their families,
friends and colleagues at work. While this is a good start, CAHPS provides information on the
experiences of hundreds of people who are already are in a particular plan. This provides a view
of a health plan that is more representative and more reliable that what can be captured by the
opinions of a few individuals. Since CAHPS allows for comparisons among similar and across
different types of plans (managed care vs. fee for service), consumers will be able to get a
complete picture of what each option provides and how it stacks up to what else is available.
CAHPS already is being used by a wide range of private sector organizations, including Ford
Motor Company's Department of Health Care Quality which is testing in two markets, and five
large health plans, including NylCare and United Health Care. The surveys have been used by
more than 20 states, including California, Maryland, New Jersey, Washington, Texas, and
The Secretary also announced last Friday that HCFA will begin fielding a CAHPS survey,
developed in partnership by AHCPR and HCFA, to assess the quality of care in Medicare
managed care plans.
In the near future, you will get an opportunity to use CAHPS to pick your health care
coverage for you and your family. The Office of Personnel Management will be using CAHPS
to help federal employees select health plans based on information on quality.
I am not suggesting that all providers and plans in every clinical setting and every region in
this country be evaluated using the exact same measures. Measures and instruments should not
be one-size-fits-all, but should reflect the diversity of needs and uses. What I am advocating is a
"department store" of accepted quality measures, all based on science and validated for reliability
and usefulness, where users of measures can pick the set that fits their need, whether that need is
to compare health plans or providers, or to conduct a hospital quality improvement project.
The Public Role
As a physician who has had experience providing care in a Veterans Affairs hospital and has
cared for Medicare beneficiaries, I have had the opportunity to gain insight into the public role in
providing and paying for care. Our responsibility for quality includes government's role as
purchaser and provider but goes beyond this level to serve to help the health care market work as
effectively as possible for all Americans.
In my view, it is the responsibility of Government, in partnership with the private sector to
ensure that the science of performance measurement matures in a way that promotes effective,
efficient, and reliable measurement and reporting. Government's contribution in this partnership
plays out in four critical areas.
One, the Government supports and conducts the basic research underpinning the
science of quality measurement and quality improvement. Resting on the strong foundation of
outstanding biologic research by our colleagues such as those at the National Institutes of
Health, the Agency for Health Care Policy and Research (AHCPR) supports health services
research about the effectiveness and outcomes of medical care that serves as an essential building
block for quality measurement.
As you well know, AHCPR is not a regulatory or enforcement agency, but a scientific
research agency that sponsors, conducts and translates research. We follow the same rigorous
evaluation and peer review standards for awarding research grants as does the National Institutes
of Health. Three-quarters of AHCPR's research funds are used to support researchers throughout
the country. This research provides the evidence needed about what works and
doesn't work in health care practice, and hence what can be measured and improved.
Two, the Government can put science into practice by developing measurement tools and
instruments and testing them on an ongoing basis to ensure their reliability, validity, and
usefulness in improving the quality of health care services.
A third, and unique, contribution of the Government is that the research, measures, and tools
developed by us and our partners are in the public domain available for all to use. There were
many times during my years of practice and as Chief of Medicine at Georgetown that I wished
that we had better access to a toolbox of quality measures that would have enabled us to measure
quality and patient outcomes better.
Last, but by no means least; the fourth major role of Government is the implementation of
quality measures within Government health programs. The Government is the largest
purchaser of health care in the Nation, accounting for more than 43 percent of health care dollars
spent at the local level, and is entrusted with the care of many of this nation's most vulnerable
The Government has an interest in ensuring quality, partly because the Government is a
purchaser and provider of medical services, but also because it has a general responsibility to
help make the health care market work as effectively as possible for all Americans. How do we
hope to achieve these goals? By developing, testing, and using science-based measures and using
the results to improve quality either through consumer choke strategies or quality improvement
Mr. Chairman, while measuring quality of care is difficult now, we are making -progress in
that direction. We know that consumers, physicians, and the health care system as a whole need
information on what works and what doesn't work in health care. This information is critical,
and as I have seen through my career, essential to improving the quality of care provided in this