|
P E R S P E C T I V E S
V A R I A T I O N S
16 November 2005
Variation In Medical Care:
Time For Action
Unlocking the potential of variations
research.
By Leonard D. Schaeffer and
Dana E. McMurtry
ABSTRACT:
Variations
research has broken new ground by moving analysis of Medicare data from the
community to the hospital level. The California study of hospitals further illustrates
the power of comparing performance to demonstrate the levels of efficiency that
can be achieved. WellPoint, replicating the Dartmouth approach with its commercial
data, is partnering with medical specialty societies to make data “actionable”
and move beyond benchmarking to help providers alter practice patterns. The
shared goal is to achieve a material reduction in practice variation throughout
health care. Key elements of their partnership include establishing national
databases, reducing the administrative burden for pay-for-performance programs
and using evidence-based metrics.
Since 2004 we have mentioned Jack Wennberg and Elliott Fisher’s landmark
studies on treatment and spending variation among the nation’s most renowned
hospitals in all of our public presentations. The response is the same, regardless
of the audience’s sophistication: astonishment that more care and more
cost do not result in better outcomes. In fact, when they evaluated costs statewide,
Fisher and colleagues found a negative correlation between the cost of all medical
services that Medicare beneficiaries receive with the overall quality of care.1
The Dartmouth finding that more care and higher spending do not produce better
medical outcomes is completely contrary to deeply ingrained beliefs about the
relationship between spending and quality found in other sectors of our economy.
Audiences are shocked with the Dartmouth conclusion that up to a third of Medicare
spending is potentially wasted on unnecessary (or even harmful) care. Even many
physicians are surprised by the wide variation in use rates among the most often
performed surgical procedures and in the adoption of clinical interventions
that save lives. These expenses and interventions could be redirected to correct
the underuse of proven, effective care or to help expand coverage to the uninsured.
The current Wennberg paper, which focuses on California, continues the more
recent Dartmouth work to unlock the potential of variations research to change
care delivery.2 In older studies, individual hospital
and physician care patterns were obscured by community rates, which allowed
hospitals to claim that other hospitals were responsible for higher rates of
inefficiency or that the variation was driven by geographic destiny.3
The California study, using a benchmark approach, found that if Los Angeles
hospitals had provided the same pattern of care as in Sacramento over a five-year
period, Medicare would have saved $1.7 billion. But, more importantly, the researchers
were able to identify which hospitals were driving the higher community rate.
Variations research at the hospital-specific level creates important new opportunities
to reduce variation in care and increase efficiency in the commercial, Medicare,
and Medicaid markets. Inspired by the Dartmouth approach, some health plans
are replicating the Dartmouth Medicare-based research using their own commercial
data. The objective is to better understand the wide divergence in practice
patterns within their own hospital networks for those conditions that account
for sizable health care spending and drive plan-specific medical trends.
Making Data ‘Actionable’
WellPoint has conducted comparative analyses of utilization and costs similar
to Dartmouth’s for the most common inpatient procedures, prescription
drug use, advanced imaging, and other health data across providers. Our goal
is to make findings “actionable”—that is, to achieve a material
reduction in the variation in care. For example, our analyses have determined
that extreme variation exists in the use and cost of coronary artery bypass
graft (CABG) across our geographies. WellPoint data show CABG admissions ranging
from 0.3 per 1,000 in one southeastern city to 1.7 per 1,000 in another centrally
located city. The use of advanced imaging technology also varies widely, with
forty-five magnetic resonance imaging (MRI) scans per 1,000 performed in northern
Ohio and 105 per 1,000 in southern New Hampshire.
In contrast to Medicare, commercial health plans negotiate contracted rates
for health care services, and the unit cost for certain surgical services can
vary severalfold. Interestingly, our data do not demonstrate a strong correlation
between cost and quality outcomes; nor is there a positive correlation between
the volume of this surgical procedure and mortality among the hospitals in our
sample for this particular procedure, which challenges previous findings that
doing more procedures results in better outcomes.
It is also interesting to note that what payers might view as a “dream”
hospital—high quality, high efficiency, lower cost—often does not
match public perception. In fact, some hospitals that are low quality and high
cost for coronary services enjoy strong reputations as the “best”
in their communities. Wennberg and colleagues emphasize that strategies to encourage
Medicare patients to use more efficient hospitals will not reduce overuse of
supply-driven care because these resources will still be used by some beneficiaries
and the private sector. Overcoming the power of historical reputations, patient
“loyalty” to high-cost providers and the intervention of regulators
in the provider-contracting process will challenge both the private and public
sectors.
Using Relationships To Drive Systemic Change
Historically, the vulnerability of variations research has been that documenting
relative differences in use cannot discern the “right” amount of
care that should be provided. Wennberg and colleagues suggest that in the absence
of knowing the “proper processes of care,” establishing benchmarks
of efficiency is the best approach to evaluating performance. The problem, however,
is that benchmarking is not persuasive enough to change providers’ behavior
throughout health care. In addition, a focus on comparing performance alone
risks making the discussion more about waste than about the pursuit of higher-quality
care and better outcomes for more patients.
Achieving large-scale change in physicians’ behavior requires strategies
to identify and increase the use of scientifically based clinical care. State-of-the-art
knowledge coupled with endorsement of proven approaches to improve clinical
outcomes by medical specialty societies are essential to reducing unwarranted
variation, improving quality and efficiency, and avoiding unintended consequences.
For this reason, WellPoint has developed an approach for partnering with medical
specialty societies and academic medical centers to identify evidence-based
practices and apply them to care delivery. Driving systemwide change requires
establishing national databases, minimizing the administrative burden for pay-for-performance
programs, and using evidence-based quality metrics.
A New Partnership Model
WellPoint’s model is based on an initially successful partnership with
the American College of Cardiology (ACC), using its CathPCI Registry, which
represents more than 600 institutions and more than 2.5 million patient records.
The partnership creates a single point of data entry through a national registry.
Hospitals participating in WellPoint’s coronary quality program receive
reports from the ACC and WellPoint.
In addition to reducing the administrative burden on hospitals, WellPoint and
the ACC also collaborated to establish quality outcome metrics for collection
and reporting. Since the goal is to improve care delivered to all patients (not
just WellPoint members), reporting reflects all hospital patients. An established
relationship ensures that these metrics can change with the support of the physician
and hospital community as the science base evolves.
Importantly, a growing number of our network hospitals receive a portion of
WellPoint’s overall reimbursement based on performance on an expanding
set of clinical measures that have been endorsed by specialty societies, the
Joint Commission on Accreditation of Healthcare Organizations (JCAHO), and the
National Quality Forum (NQF). WellPoint also creates financial incentives for
hospitals to report data to the ACC CathPCI database.
WellPoint is also establishing partnerships with the Society of Thoracic Surgeons
and the American College of Radiology (ACR). The spiraling costs of radiology
are a particular concern to both the Centers for Medicare and Medicaid Services
(CMS) (given the role of imaging costs in Part B premium increases) and to the
private sector. Advanced imaging costs have been increasing 15–25 percent
annually. In addition, primary care physicians and specialists are acquiring
their own imaging centers to which they can refer patients. WellPoint has worked
collaboratively with the ACR to apply its standards for advanced imaging, which
will help ensure more appropriate use of these important advanced imaging technologies.
Moving Beyond The Carrot And The Stick
Some argue that variation is a fixture of U.S. health care—there is no
need to worry, they say, because today’s variation will go away only to
be replaced by variation associated with tomorrow’s new treatments. The
marketplace, however, is becoming less tolerant of variation that contributes
to higher care costs but does not improve quality. Public and private payers
are demanding increases in value to justify increases in costs.
Many purchasers, armed with information technology, will continue to implement
strategies to reduce variation and move to evidence-based care and reimbursement.
These initiatives—whether draconian sticks (such as the decision by the
California Public Employees Retirement System, or CalPERS, to drop expensive
hospitals) or economic carrots (such as quality incentive programs)—will
alter patterns of care in many institutions and systems. WellPoint’s partnerships
with specialty societies, however, reflect our belief that engaging medical
leadership is necessary to inspire better performance and drive transformative
change throughout the health care system for the benefit of all patients independent
of geography or payer.
The authors thank Sam Nussbaum, executive vice president and chief medical
officer, WellPoint Inc., for his insight, thoughtful comments, and data.
NOTES
1. E.S. Fisher et al., “The Implications of Regional Variations
in Medicare Spending. Part 1: The Content, Quality, and Accessibility of Care,”
Annals of Internal Medicine 138, no. 4 (2003): 273–287.
2. J.E. Wennberg et al., “Evaluating the Efficiency of
California Providers in Caring for Patients with Chronic Illnesses,” Health
Affairs, 16 November 2005, content.healthaffairs.org/cgi/content/abstract/hlthaff.w5.526.
3. J.R. Knickman, “The Dartmouth Data: Moving from Analysis
to Action,” Health Affairs, 7 October 2004, content.healthaffairs.org/cgi/content/abstract/hlthaff.var.121
(21 October 2005).
Leonard Schaeffer is chairman
of WellPoint Inc., headquartered in Indianapolis, Indiana. Dana McMurtry (dana.mcmurtry{at}wellpoint.com)
is WellPoint's vice president for public policy.
Read related articles
by John
Wennberg and colleagues, Max
Baucus, Thomas
Priselac, Uwe
Reinhardt, and Barry
Straube.
DOI: 10.1377/hlthaff.w5.552
©2005 Project HOPE–The People-to-People Health
Foundation, Inc.
|
