Quality: Where Have We Been?
Where Are We Headed?
Elizabeth A. McGlynn, Ph.D.
Presentation to the University of Texas System Clinical Effectiveness & Safety Conference
September 26, 2013
Outline of Talk
• A Brief History of Quality Measurement and Reporting
• An Examination of Policy Solutions: Old and New
• An Illustration of What is Possible When All the Parts
Are Working Together
2
In 1853, Florence Nightingale Identifies
Potentially Preventable Deaths
160 years later we face the same issues
In 1917, Ernest Codman Calls on
Hospitals to Release Surgical Outcomes
So I am called eccentric for saying in
public that hospitals, if they wish to
be sure of improvement:
• Must find out what their results are
• Must analyze their results
• Must compare their results with
other hospitals
• Must welcome publicity not only
for their successes, but for their
errors
Such opinions will not be eccentric a
few years hence
4
Early Comprehensive Studies Find
Substantial Quality Deficits (1960s-70s)
• Morehead (1964)
– Concerns about rising health care costs
– 57% of hospital care for Teamsters in New York City meets
optimal care standards
• Payne (1970)
– 41% of post-discharge care in Hawaii met standards
• Rhee (1976)
– Physician performance varies by condition, ranging from 45%
for stroke to 91% for cesarean section
Variations in Service Delivery:
First Published in 1973
Inpatient Knee Replacement Per 1000 Medicare Enrollees, 2007
6
Appropriateness Posited As One
Explanation for Observed Variations (1980s)
7
RAND Health Insurance Experiment Demonstrates
Relationship between Finances and Utilization (1980s)
• Co-pays lead to lower
utilization
• Lower utilization has no
effect on health of
average person
• Reductions in utilization
occur for both
appropriate &
inappropriate services
8
Medical Outcomes Study Established Utility of
Patient Reported Outcomes (1990s)
• Short form measures of
patient reported
outcomes are valid
• Impact of chronic
conditions on
functioning established
• Led to changes in data
collected in clinical trials
• Interest has accelerated
in recent years
9
IOM Raises Public Awareness About
Medical Errors (1999)
• 44,000 – 98,000
preventable deaths
annually
• Subsequent study:
1.12M events in 1.07M
hospital stays
• Government policy in
this decade focuses on
nonpayment for errors
10
IOM Highlights Quality Problems in
Report (2001)
• Summary of literature
establishes likely causes
of quality deficits
• Identifies current
practices that impede
quality care,
• Explores how systems
approaches can be used to
implement change
• Recommends a
fundamental design of the
US health care system
11
RAND Study: American Adults Receive
55% of Recommended Care (2003)
Care that
meets
quality
standards
McGlynn et al, 2003
Children Receive Less than Half of
Recommended Ambulatory Care (2007)
Care that
meets
standards
Mangione-Smith et al., 2007
Quality of Care for Cardiopulmonary
Problems Varies Widely
Coronary artery disease
Hypertension
Heart failure
Stroke
Chronic lung disease
Asthma
High cholesterol
Pneumonia
Atrial fibrillation
0
McGlynn et al., 2003
20
40
60
80
% of recommended care delivered
100
Significant Variation Exists in Management
of Adults’ General Medical Problems
Low back pain
Depression
Osteoarthritis
BPH
Headache
Diabetes
Ulcers
Alcohol dependence
0
McGlynn et al., 2003
20
40
60
80
% of recommended care delivered
100
How Does Quality Vary by Market?
SEATTLE
•
•
SYRACUSE
•
BOSTON
LANSING
•
CLEVELAND
•
•
NEWARK
INDIANAPOLIS
•
ORANGE
COUNTY
•
•
PHOENIX
LITTLE
ROCK
•
GREENVILLE
•
•
MIAMI
We Found You Aren’t Safe Anywhere…
Boston
Overall
Greenville
Indianapolis
Little Rock
Newark
Orange Co
Syracuse
30
Kerr et al., 2004
40
50
60
70
80
90
100
% of recommended care delivered
We Found You Aren’t Safe Anywhere…
Boston
Overall
Preventive
Greenville
Indianapolis
Little Rock
Newark
Orange Co
Syracuse
30
Kerr et al., 2004
40
50
60
70
80
90
100
% of recommended care delivered
We Found You Aren’t Safe Anywhere…
Boston
Overall
Preventive
Greenville
Indianapolis
Acute
Little Rock
Newark
Orange Co
Syracuse
30
Kerr et al., 2004
40
50
60
70
80
90
100
% of recommended care delivered
We Found You Aren’t Safe Anywhere…
Boston
Overall
Preventive
Greenville
Indianapolis
Acute
Chronic
Little Rock
Newark
Orange Co
Syracuse
30
Kerr et al., 2004
40
50
60
70
80
90
100
% of recommended care delivered
No One Is Immune From Quality Deficits
Gender
Male
Female
White
Black
Hispanic
Other
Race
18-30
31-64
65+
Age
0
Asch et al, 2006
20
40
60
% of recommended care delivered
80
100
Greatest Differences Found in Metrics &
Conditions Included in VA System
VA Metrics &
Incentives
Different
Metrics for VA
Conditions
US
VA
No VA
Measure
0
20
40
60
80
% of recommended care delivered
Asch et al., 2004
100
Outline of Talk
• A Brief History of Quality Measurement and Reporting
• An Examination of Policy Solutions: Old and New
• An Illustration of What is Possible When All the Parts
Are Working Together
23
OK, Now What?
• Transparency (Public Reporting)
– Quality
– Price
• Financial incentives (Pay for Performance)
– Hospitals, physicians
– Patients
• Health information technology (Meaningful Use)
– Electronic medical records
– Decision support tools
• Organizational change (ACOs, Medical Home)
– Putting the pieces together
What Do We Know About the Effect of
Public Reporting?
Entity
Health Plans
Hospitals
Physicians
Effects on
Consumer
Choices
Quality
Improvement
Efforts
Clinical
Outcomes
Mixed
No evidence
No evidence
No effect
Positive
Mixed
Mixed
No evidence
Positive
So, Is It Worth It?
• Public reporting is the right thing to do
• But it may not have large effects by itself
– Other factors (financial incentives) still
dominate
– Biggest impact likely on providers rather than
consumers
– Continuing work on the implementation of these
programs is necessary
If Just Reporting Doesn’t Work, What
About Incentives?
• Pay-for-Performance (P4P) programs use financial
incentives to motivate hospitals and doctors to
increase adherence to best practices
• Providers receive differential payments based on
performance on a set of specified measures:
–
–
–
–
Clinical quality
Resource use (efficiency)
Patient experience
Information technology use or capabilities
P4P Is Not a New Idea
“If a physician makes a large incision with an operating
knife and cures it, or if he opens a tumor (over the eye)
with an operating knife, and saves the eye, he shall
receive ten shekels in money.”
“If a physician makes a large incision with the
operating knife, and kills him, or opens a tumor with
the operating knife, and cuts out the eye, his hands
shall be cut off.”
-- Code of Hammurabi, 1750 B.C.
P4P Generated Slightly Greater
Improvements than Public Reporting
Comparison of Performance on Composite of 10
Measures: Q4 2003-Q3 2005
90
88
86
Premier P4P hospitals
Performance
84
rate (%)
82
80
CMS P4R hospitals
∆= 2.8% points after adjusting for hospital differences*
78
76
Q1
Q2
*Lindenaur et al., 2007 (NEJM)
Q3
Q4
Q5
Q6
Q7
Q8
What Do We Know about Pay-forPerformance?
• Few evaluations of pay for performance
(P4P) have been conducted
• The published studies show modest
positive results
• P4P program design related to program
impact (devil is in the details)
• P4P alone is unlikely to solve quality and
cost problems, but may be useful when
combined with other policy levers
Health Information Technology: Using
Modern Tools for Improvement
• All modern industry uses information technology to
manage knowledge, processes
• Complexity of medicine has increased exponentially
• Many-to-many matching problem better handled by
computers than human brains
What Do We Know About the Effect of
HIT on Quality?
• Systematic review (Chaudry et al, 2007)
– 257 studies met inclusion criteria
– 25% of studies from 4 academic centers
– Benefits include increased adherence to
evidence-based medicine, enhanced
surveillance and monitoring, decreased
medication errors
– Questions about generalizability beyond the
benchmark institutions
What Do We Know (cont.)?
• Impact on quality improvement (Jones et al,
2010)
– Basic systems provide small improvement in
performance for 3 conditions
– Upgrades, improvements slow improvement
• Heterogeneity of experience
– Hard to assess the technology & its use on a
broad scale
– Led U.S. to develop standards for “meaningful
use” (as basis for incentives)
What About New Care Models?
• ACA opened the door for substantial experimentation
with new models of care delivery and payment
– ACOs, Medical Homes, and More!
• Lots of activity happening
– No one wants to be left behind
• Not much is known yet
• The types of changes we are experimenting with are
really hard to evaluate
• So, they may become the new normal – whether or
not they deliver the goods
34
Outline of Talk
• A Brief History of Quality Measurement and Reporting
• An Examination of Policy Solutions: Old and New
• An Illustration of What is Possible When All the Parts
Are Working Together
35
It Takes A System…and Maybe a
Village
• Quality improvement is a
team sport
• It requires tools,
opportunities, and (good)
attitudes
• Integrated delivery systems
are well positioned to
deliver on the promise
• How far non-integrated or
virtually integrated systems
can go remains to be tested
36
An Example: A System At Work
BMI
37
Core Capabilities for Kaiser’s
Performance Improvement System
•
•
•
•
•
•
Leadership sets priorities and accountability
Systems approach to improvement
Measurement capability
Learning organization
Improvement capability
Culture
Cervical Cancer Screening Rate Up
and Unnecessary Testing Down
100%
HEDIS Pap - IQR
75%
82.0%
85.6%
PAPA Overutilization Rates - Overall
86.6%
86.1%
50%
33.9%
21.6%
25%
12.6%
8.8%
0%
2006
39
2007
2008
2009
Approaching 100% On Many Joint
Commission Core Measures
100%
90%
80%
70%
HF Bundle
SCIP Composite
CAP Composite
AMI Bundle
60%
2Q07
3Q07
4Q07
1Q08
2Q08
3Q08
4Q08
40
Source: Quality and Risk Management/TJCCore Measures, May 2011
1Q09
2Q09
3Q09
4Q09
1Q10
2Q10
3Q10
4Q10
Better Than National Average
Hospital Mortality Rates
0.8
US Medicare Overall
KPSC
0.7 (most recent)
0.7
0.6
0.52
0.5
1Q07
2Q07
3Q07
4Q07
1Q08
2Q08
41
Source: Quality and Risk Management/KP Insight May 2011
3Q08
4Q08
1Q09
2Q09
3Q09
4Q09
1Q10
2Q10
3Q10
Kaiser Permanente Outperforms the
Nation in Blood Pressure Control
% With Hypertension Whose
Blood Pressure is Controlled
100
80
60
40
20
0
US
Kaiser Permanente
Summary
• The quest for quality has
been long & arduous
• No magic solutions have
been found
• Islands of excellence can
be found, but all too rarely
reproduced
• But the search
continues…and we remain
hopeful!
43
Questions?
Questions?