Transcript Addressing Disparities Through Organizational Quality

Addressing Disparities Through
Organizational Quality Improvement
Efforts
David R. Nerenz, Ph.D.
Center for Health Services Research
Henry Ford Health System
October 21, 2005
Overview
• Health Care Disparities
• Reasons for Disparities
• Hospitals, Health Plans, and Quality
Improvement
• Three Challenges:
– Identifying Significant Disparities
– Measuring Effects of QI Initiatives
– Setting Priorities
IOM Report, 2002: Assessing the Quality
of Minority Health Care
“Disparities in the health care
delivered to racial and ethnic
minorities are real and are
associated with worse outcomes
in many cases, which is
unacceptable.”
- Alan Nelson, retired physician, former
president of the American Medical
Association and chair of the committee
that wrote the Institute of Medicine
report, Unequal Treatment: Confronting
Racial and Disparities in Health Care
Ratio of Minority-White Death Rates,
1994-1996
Death rates of minority Americans compared
to those of white Americans at various ages
3.0
African American
2.0
American Indian/Alaska Native
White,
non-Latino
Latino
1.0
Asian/Pacific Islander
0.0
0-14
15-24
25-44
Age Groups
Ratios are based on deaths per 100,000 resident population
SOURCE: DHHS Health, Unites States, 1998
45-64
65+
Screening: Percent with Early Stage Cancer* Among
Women with Breast Cancer,
1978-1987 (Detroit)
29%
30%
24%
20%
19%
20%
16%
15%
White
Black
10%
0%
1978-1981
1982-1984
*Tumors <2cm and no auxiliary lymph involvement at diagnosis
SOURCE: Swanson, M et al., 1990
1985-1987
Indicators of Children’s Access
to Care, 1987
Adjusted Odds Ratios
Minority Children vs. White Children
3.0
2.5
2.0
2.44
Ratios > 1.0
indicate
minority more
likely than
white children
2.12
1.77
1.54
1.45
1.5
1.0
0.56
0.5
0.0
Has a usual
Does not see
Source of Care
a specific
physician
DATA: 1987 NMES
SOURCE: Newacheck et at, 1996
No Afterhours
Emergency
Care
Travel time
of 30 min. +
Wait time of 60
min. +
Physician
Not Seen for
Selected
Symptoms
Heart Procedure Rates for
Blacks and Whites, 1980 vs.1993
(Ratio of Black/White procedure rates) *
1.0
Equal procedure rate
Ratios
<1.0
indicates
Blacks
less likely
than
Whites to
undergo
procedure
0.0
1980
1993
Cardiac
Catheterization
•Rates were age-adjusted
•SOURCE: Gillum R.F., et al., 1997
1980
1993
Angioplasty
1980
1993
Bypass
Surgery
Coronary Artery Surgery Rates by Race and
Disease Severity, 1984-1992
Percent Receiving Bypass Surgery
80%
Whites
African Americans
60%
45%
40%
35%
31%
25%
20%
0%
Mild Disease
Source: Peterson, et al., 1997.
Severe Disease
Disparities in Cardiac
Revascularization




5,000 Medicare
beneficiaries in 5 states
– 1991 and 1992
RAND
appropriateness
criteria
Some gender
disparities noted as
well
Epstein et al, Medical
Care, 2003
100
80
60
Black
White
40
20
0
CABG or
PTCA
Clinically
Appropriate
Clinically
Inappropriate
Evidence of racial/ethnic
differences in cardiac care
1984-2001
11 studies find no
racial/ethnic
difference in care
(14%)
68 studies find
a racial/ethnic
difference in
care
(84%)
Source: Kaiser Family
Foundation
2 studies find racial/ethnic
minority group more likely
than whites to receive
appropriate
care (2%)
Total= 81 studies
Minorities are Less Satisfied with The
Quality of Care They Receive
Percent adults 19-64 privately insured
60
Total
White
Black
41*
36*
40
31
36
Hispanic
38
34
29
24*
20
0
Better Care With Different Health Plan
Rate Care from Doctor as Excellent
*Significantly different from whites at p<.05 or less
Source: The Commonwealth Fund Biennial Health Insurance Survey (2003)
THE
COMMONWEALTH
FUND
Isn’t It All About Poverty and
Lack of Insurance?
Rates of Hospitalization for Coronary Artery Bypass
Surgery Among Medicare Beneficiaries, 1993
6
Coronary Artery Bypass Surgery
Procedures per 1000 beneficiaries per
year*
4.8
4.9
4.8
White
4.6
Black
4
2.2
2
0
2.1
2.2
1.8
<$13,001
$13,001$16,300
$16,301$20,500
Annual Income
*Rates were adjusted for age and sex to the total Medicare population.
SOURCE: Gomick, ME et al., 1996
>$20,500
Infant Mortality Rates: Mothers 20+ Years by
Education and Race/Ethnicity, 1995
College +
Latino
African American
White
5.0
12
4.7
High
School
Less than
High School
5.9
14.8
6.4
6.0
17.3
7.6
Infant deaths per 1,000 live births
DATA: CDC National Center for Health Statistics
SOURCE: DHHS. Health, United States, 1998
16
Across Income Groups, African Americans Are Most Likely to
Go without Needed Care Because of Cost
Percent adults 19-64 privately insured going without needed care
Total
White
60
40
African American
Hispanic
52*
37
36
36*
31
24
23
29
20
0
Income below 200% FPL
Income 200% or above FPL
^Did not fill prescription, did not get specialist care, or skipped recommended test because of cost.
*Significantly different from whites at p<.05 or better
Source: The Commonwealth Fund Biennial Health Insurance Survey (2003)
THE
COMMONWEALTH
FUND
Across Income Groups – Hispanics Are Most At
Risk for Forgoing Preventive Care
Percent adults 19-64 privately insured with blood pressure check in past year
Total
100
84
White
87
African American
90
86
74*
90
Hispanic
95
84*
50
0
Income below 200% FPL
Income 200% or above FPL
*Significantly different from whites at p<.05 or better
Source: The Commonwealth Fund Biennial Health Insurance Survey (2003)
THE
COMMONWEALTH
FUND
Evidence on Disparities
“Racial and ethnic minorities tend to receive a
lower quality of health care than non-minorities,
even when access-related factors, such as patient’s
insurance status and income, are controlled. ” (my
emphasis)
IOM Report, Unequal Treatment:
Confronting Racial and Ethnic Disparities
in Health Care, 2002
Some of these studies are pretty old –
haven’t things changed since people
started studying this?
Reperfusion Therapy in Medicare
Beneficiaries with Acute MI
Group
% Eligible
receiving reperfusion
White men
59%
White women
56%
Black men
50%
Black women
44%
Canto JG, Allison JJ, Kiefe CI, Fincher CI, Farmer R, Sekar P,
Person S, Weissman NW.
Relation of race and sex to the use of reperfusion therapy in Medicare
beneficiaries with acute myocardial infarction. N Engl J Med 2000 Apr
13;342(15):1094-100
Disparities in Management of AMI –
Changes Over Time (1994-2002)
• Data Source – NRMI
• 600,000 Patients
• Significant disparities
in several measures;
no change over time
• Some disparities
became not significant
in adjusted analyses
•
Vaccarino et al, NEJM,
August 18, 2005
Receipt of Major Surgical Procedures –
Medicare Data 1992-2001
• Focus on 9 surgical
procedures
• Analysis by hospital referral
regions for three
procedures
• No evidence of change in
disparity over 10-year
period
• Disparity reduction in
22/158 regions, but no
elimination of disparity in
any region
•
Jha et al, NEJM, August 18,
2005
Disparities in Medicare Managed Care
(HEDIS) Measures Over Time
• Standard, widely-used
quality measures
• Trends from 1997 or
1999 to present
• Improvements in
quality overall,
reduction in
disparities in some
HEDIS measures,
but not all
•
Trivedi et al, NEJM, August
18, 2005
Disparities in Medicare Managed Care
(HEDIS) Measures Over Time (Cont.)
• Additional
HEDIS
measures
• Change in
disparity from
first to most
recent year
•
Trivedi et al, NEJM,
August 18, 2005
All of these studies involve large
national samples – what about
disparities within single health care
organizations?
Comprehensive Diabetes Care:
Foot Exam Performed
Rate
100
80
60
40
20
0
White
African American
Hispanic
White vs. African American (p<0.001), White vs. Hispanic (p<0.001) and
White vs. Asian (p<0.001).
Asian
Overall
Asthma: Outpatient Follow-up
After Acute Episodes
• Core concept:
Outpatient follow-up
after either ER visit or
admission
• Children 5-17 years old
• Standard based on
national expert panel
guidelines
70
60
50
40
Caucasian
30
AfricanAmerican
20
10
0
Follow-up Rate
Comparison of non-Hispanic/Hispanic Breast Cancer
Screening by Commercial, Medicare Risk, and Medicaid
Products in a Single Health Plan, 2000
Rate
100
80
60
40
20
0
Medicaid
Medicare Risk
Commercial
non-Hispanic
P=.001 non-Hispanic population
Hispanic
Possible Explanations for Disparities
• Environmental Factors
–
–
–
–
Income/Poverty
Insurance coverage
Geographic access
Poor-quality facilities and
providers in minority
neighborhoods
– Cultural competence of
providers and systems
– Language barriers
– “Institutional racism”
• Individual Factors
– Cultural beliefs and
preferences
– Trust in providers and
organizations (lack of)
– Literacy
– Biased clinical decision-making
– Some possible biological
differences
Conceptual Model of
Health Care Disparities
Patient Factors
Health Care
System Factors
Treatments
Environmental
Factors
Provider Factors
Outcomes
Conceptual Model of
The Operation of a Car
Engine Factors
Chassis and
Body Factors
Movement of
Car
Environmental
Factors
Driver Factors
Arrival at
Destination
Understanding the Provider Contribution to Disparities
Class
Pt Race/ethnicity
(Michelle VanRyn)
Provider
cognition and affect
regarding patient
Provider
Interpretation of
Symptoms
Provider
Decision-making
(Diagnosis,Treatment)
Pt Behavior in
Encounter
Provider Behavior in
Encounter
(e.g., question-asking selfdisclosure, assertiveness)
(e.g., participatory style, warmth,
content, information giving,
question-asking)
Treatment
or
Service Pt
Received
Encounter characteristics
Pt Satisfaction
Pt Behavior
Pt Cognitive & Affective Factors
Culture
(e.g., acceptance of advice, attitude, self-efficacy, intention, feelings of
competence, attitude toward med care, trust)
(e.g. self management,
information-seeking, utilization)
CHRONOLOGY
DIAGNOSIS 
BIRTH 
TREATMENT
No chemotherapy,
Age
INNATE
SCLC / advanced NSCLC
Sex / Gender
No surgery,
localized NSCLC
Race / ethnicity (B vs. W)
No Rx, any
Socioeconomic status
EXPOSURES /
ACQUIRED FACTORS
Marital status (spouseless)
Smoking status (current)
Illicit drug (use vs. not)
Comorbidities (adverse)
CANCERRELATED
Symptoms (adverse)
Histology NOS
Stage (higher / advanced)
SURVIVAL
DIAGNOSTIC
INTENSITY
Unstaged cancer
Conceptual Model for Contribution of
Race/Ethnicity and SES to
Treatment and Outcome Disparities
Education
Financial
Access
Job Class
Benefits,
Schedule
Flexibility
Historical Bias/
Discrimination
Comprehension,
Trust
Language
Race/Ethnicity
Health Beliefs,
Trust in Providers
Culture
Biological
Differences
Tumor Histology,
Comorbidities
Bias/Discrimination in
Medical Settings
Disparities in Outcomes
Income
Disparities in Treatment
SES
Literacy
Disparities in Cardiac Surgery –
Steps in a Process
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Notice and interpret symptoms
Get primary care doctor appointment
Get tests
Deal with insurance and payment issues
Get referral to Cardiology
Get Cardiology appointment
Get additional tests
Discuss treatment options and ask questions
Get referral to Cardiac Surgeon
Get appointment with surgeon
Get additional tests
Discuss treatment options and ask questions
Get surgery scheduled
Arrange time off work and family support
Deal with comorbid conditions
Get surgery
Disparities in Breast Cancer
Treatments and Outcomes
Bynum et al
Mammography
Lipscombe et al
Comorbidities,
Obesity
Bibb, Mandelblatt,
Others
Yood et al, Others
Stage at Diagnosis
Treatment by
Stage
Stark et al
Histology
Chlebowski et al
Tammemagi et al,
Griggs et al
Survival
Hershman et al,
Dignam
Grann et al
Understanding Underlying Factors –
Role of Comorbidities
• Comorbidities as mediators of treatment
choices, or of treatment effectiveness
• Comorbidities as predictors of survival or other
health outcomes, independent of treatment for
primary condition being studied
• Higher prevalence of comorbidities (e.g.,
hypertension, diabetes) among minority patients.
Survival Disparities in Breast Cancer –
Role of Comorbidities
•
Comorbidity and Survival
Disparities Among Black and
White Patients With Breast
Cancer
•
C. Martin Tammemagi, PhD; David
Nerenz, PhD; Christine NeslundDudas, MA; Carolyn Feldkamp, PhD;
David Nathanson, MD
JAMA. 2005;294:1765-1772.
•
Approximately 900 Black and White
Women with Breast Cancer, 10-Year
Follow-up
Comorbidities and Breast Cancer (Cont.)
• African American
women more likely than
White women to have
comorbidities
• Comorbidities associated
with survival
• Comorbidities explained
most of the disparity in
all-cause survival, but not
in cancer-specific
survival.
Disparities in Lung Cancer Outcomes
HR (Black vs. White) = 1.21 (95% CI 1.05, 1.38; p = 0.008)
Median survival Blacks = 8.5 months; Whites = 11.2 months
Kaplan-Meier survival estimates, by race
1.00
Survival proportion
0.75
0.50
0.25
White
0.00
Black
0
1
2
3
Follow-up in years
4
5
Kaplan Meier survival plots for 1154 LCA patients,
stratified by
Adverse symptoms
1.00
1.00
0.75
0.75
.
Survival proportion
.
Survival proportion
Adverse comorbidity
0.50
No comorbidity
0.25
0.50
Absent
0.25
Comorbidity
0.00
0
1
2
3
Follow-up in years
Present
0.00
4
5
0
1
2
3
Follow-up in years
4
5
Predictors of LCA survival – hazard ratios & distributions (OR) by race/ethnicity
Prognostic Factors
Innate
Age (per 10 year increase)
Race/ethnicity (Black vs. White)
Acquired
SES * (BGMHI $10,000)
Marital status (spouseless vs. not)
Univariate HR
Multivariate HR
Black
White
1.16 (p <0.001)
1.21 (p = 0.008)
1.24 (p <0.001)
67.5 yr
-
67.1 yr
-
$19,913
50.3%
$38,822
37.1%
0.37 (p < 0.001)
1.72 (p < 0.001)
0.92 (p < 0.001)
1.27 (p = 0.001)
Odds RatioB vs. W
Smoking status (current smk vs. not)
Illicit drug use (user vs. not)
Adverse comorbidity (≥1 vs. 0)
Diagnostic intensity
Unstaged (vs. stage I)
Histology (unspecified vs. SqCCA)
Cancer-related factors
1.29 (p < 0.001)
2.17 (p < 0.001)
1.45 (p < 0.001)
1.32 (p < 0.001)
1.99 (p = 0.004)
1.42 (p < 0.001)
53.4%
4.3%
65.7%
44.2%
0.3%
59.0%
1.45 (p = 0.003)
15.67 (p < 0.001)
1.33 (p = 0.02)
4.61 (p < 0.001)
1.80 (p < 0.001)
2.71 (p < 0.001)
1.26 (p = 0.03)
7.8%
28.0%
4.8%
20.9%
1.69 (p = 0.03)
1.47 (p = 0.006)
Adverse symptoms (≥1 vs. 0)
Stage I
Stage II
Stage III
Stage IV *
Treatment
Surgery in localized NSCLC *
Chemotherapy (SCLC, III/IV NSCLC)
2.20 (p < 0.001)
referent group
2.16 (p < 0.001)
3.49 (p < 0.001)
7.09 (p < 0.001)
1.65 (p < 0.001)
referent group
2.11 (p = 0.001)
3.37 (p < 0.001)
6.90 (p < 0.001)
80.7%
20.2%
6.6%
33.2%
40.0%
70.5%
26.8%
5.1%
33.5%
34.5%
1.75 (p < 0.001)
0.23 (p < 0.001)
0.43 (p < 0.001)
55.2%
45.9%
70.1%
53.5%
0.53 (p = 0.01)
0.74 (p = 0.03)
Any treatment (treated vs. not)
Surgery
Chemotherapy
Radiation therapy
0.34 (p < 0.001)
70.6%
19.9%
41.3%
42.2%
80.8%
30.4%
45.5%
43.1%
0.57 (p < 0.001)
0.51 (p < 0.001)
0.48 (p < 0.001)
0.89 (p = 0.15)
OR (adv vs. local) =
1.33 (p = 0.03)
Why Think About Disparities in Terms
of Quality of Care?
• Relatively strong science base of published
literature and evidence-based guidelines
conceptual and moral clarity
• Build on existing staff, data collection
infrastructure, and organizational relationships
• Build on existing QI concepts, models, and
approaches
Disparities in Standard Hospital
Measures of Quality of Care
• JCAHO/CMS Standard Measure Set
– CHF
– AMI
– Pneumonia
• Commonwealth Fund – HRET Project
• RWJF Initiative – “Expecting Success”
CMS/JCAHO Measures
for CHF and AMI
100
*
90
80
Percent
70
60
Black
White
50
40
30
20
*
10
0
Disch Inst
LVEF
ACE @
Disch
Smoking
Advice
Asprin Arrival
Aspirin Disch
ACEI for
LVSD
Smoking
Advice
B-Block at B-Block @
Disch
Arriv
Mortality
Analysis of Disparities –
Basic Requirements
• Well-defined, accepted measures of quality,
access, satisfaction, clinical outcomes
• Data on race/ethnicity, SES, primary language
Methods of Data Collection –
Direct from Members/Patients
• Pros:
– Most Flexible
– Generally Preferred for
Accuracy
• Cons:
– Can be Expensive
– Can Raise Concerns
about Health Plan’s Real
Objectives
Methods of Data Collection –
Geocoding
• Pros:
– Relatively Easy, Fast, and
Inexpensive
– Requires only
Information you Already
Have
• Cons:
– Won’t Work in
Residentially Integrated
Areas
– Won’t Identify Small,
Dispersed Groups
Methods of Data Collection –
Surname Recognition
• Pros:
– Relatively Easy, Fast, and
Inexpensive
– Can be Combined with
Geocoding
• Cons:
– Only Works for Groups
with Distinct Names
– May not Work in All
Market Areas
Health Plans as Catalysts for
Quality Improvement
As agents of purchasers, health plans:
• Organize Quality Improvement Projects and Programs
• Define Important Quality Domains and Develop Measures
• Disseminate Practice Guidelines
• Identify High-Priority Target Populations
• Identify High-Priority Clinical Conditions
• Develop Incentive Systems
• Direct Communications to Members
Comparison of Caucasian and African American
HbA1c Testing in a Single Plan
Rate
100
80
60
40
20
0
1998
1999
Caucasian
2000
African American
Multiple Disparities in HEDIS Measures
in Single Health Plan
(Six-State Medicaid Project)
70
60
Percent
50
40
Caucasian
African American
Hispanic
30
20
10
0
HbA1c
Testing
Appropriate Prenatal
Good
Care
Asthma
Glycemic
Meds
Control
Source: Single Health Plan analysis of HEDIS data – 2003, unpublished
Asthma Medication Management
Reporting Year 2003
African-American
Caucasian
Numerator Denominator Rate
Numerator
Denominator Rate
All Co’s
411
600
69%
698
921
76%
A
189
272
69%
174
218
80%
B
153
213
72%
375
499
75%
C
69
115
60%
149
204
73%
Breast Cancer Screening
Reporting Year 2003
African-American
Caucasian
Numerator Denominator Rate
Numerator
Denominator Rate
All Co’s
1116
1468
76%
2581
3168
81%
A
390
519
75%
536
650
82%
B
435
561
78%
1415
1719
82%
C
291
388
75%
630
799
79%
Appropriate Testing for Children with Pharyngitis
Reporting Year 2003
African-American
Caucasian
Numerator Denominator Rate
Numerator
Denominator Rate
All Co’s
211
304
69%
689
837
82%
A
94
142
66%
161
204
79%
B
85
116
73%
349
429
81%
C
32
46
70%
179
204
88%
Quality Improvement Interventions
in Single Health Plans
• Patient Reminders
• Provider Reminders
• Culturally-Sensitive Member Education
Materials
• Disease Management Programs
• Partnerships with Community Groups to Raise
Awareness of Prevention
Improvements in Quality of Care for
African American Health Plan Members
with Diabetes
HbA1c Testing
LDL-C Testing
100
100
80
80
Percent
60
60
2003
40
2004
20
2003
Percent
2004
40
20
0
African American Members
0
African American Members
Another Approach to Evaluating
QI Program Success
100
90
80
70
Percent
• Asthma severity
definition involving
ER visits and
admissions
• Focus on AfricanAmerican members
with asthma
• Used shift in
distribution of
severity categories
as measure of
program success
• Statistically
significant using
Chi-square test
60
Pre-Intervention
Post-Intervention
50
40
30
20
10
0
Mild
Moderate
Severe
Childhood Immunization –
Combo I – (HEDIS 1999 Definition)
25
Percent
20
15
Hispanic
Total
10
5
0
2002
2003
Rolling
20032004
First Q
2004
Example of Quality of Care Disparities
Found in HEDIS Data
(Six-State Medicaid Project)
70
60
Percent
50
40
Caucasian
African American
Hispanic
30
20
10
0
HbA1c
Testing
Appropriate Prenatal
Good
Care
Asthma
Glycemic
Meds
Control
Source: Single Health Plan analysis of HEDIS data – 2003, unpublished
Which Disparity to Work On?
•
•
•
•
•
•
Largest absolute difference?
Statistical significance of difference?
Size of denominator population?
Likelihood of making a difference?
Cost-effectiveness of intervention(s)?
Something else?
– Purchaser incentives/preferences
– Community preferences
Commonwealth Fund Simulation
Modeling Project - Basic Premise
• Limited budgets and other resources to invest in
disparity reduction initiatives
• A reasonable evidence base exists for modeling
effects of disparity reductions on several
measures of health
– Mortality
– Quality of life
– Attendance at work or school
HEDIS Effectiveness of Care Measures - 2004
•
•
•
•
•
•
•
•
•
•
•
Childhood Immunization Status
Adolescent Immunization Status
Appropriate Treatment for Children
With Upper Respiratory Infection
Appropriate Testing for Children With
Pharyngitis
Colorectal Cancer Screening
Breast Cancer Screening
Cervical Cancer Screening
Chlamydia Screening in Women
Osteoporosis Management in Women
Who Had a Fracture
Controlling High Blood Pressure
Beta-Blocker Treatment After a Heart
Attack
•
•
•
•
•
•
•
•
•
•
•
Cholesterol Management After Acute
Cardiovascular Event
Comprehensive Diabetes Care
Use of Appropriate Medications for
People With Asthma
Follow-Up After Hospitalization for
Mental Illness
Antidepressant Medication
Management
Medical Assistance With Smoking
Cessation
Flu Shots for Adults Age 50–64
Flu Shots for Older Adults
Pneumonia Vaccination Status for
Older Adults
Medicare Health Outcomes Survey
Management of Urinary Incontinence
in Older Adults
Simulation Models –
Basic Features
• Generally use existing published data, although
it’s possible to collect and use primary data.
• Define key clinical/biological events and assign
probabilities (and utility values) to those events.
• Create model structure that matches essential
features of key clinical or biological processes.
Example of State Transitions
When Cancer Found (Stage III)
Sensitivity Analysis - Diabetes
M
am
m
og
r
ap
hy
a
im
ist
ic
"
-"
Ba
se
"
As
th
m
ist
ic
"
"B
as
e"
pt
im
-"
O
pt
og
r
ap
hy
m
"O
Bl
oc
ke
r-
Al
l
H
bA
1c
ia
be
te
s-
ia
be
te
s-
Bl
oc
ke
r-
M
am
Be
ta
Be
ta
D
D
QALY Gain
Per 1,000
QALY Gains for Alternative
Disparity Reduction Initiatives
50
40
30
20
10
0
Simulation Models Conclusions / Next Steps
•
•
•
•
•
Under basic sets of assumptions, health benefits related to
eliminating disparities in four existing HEDIS measures are
relatively modest.
There is a significant range – approximately a hundred-fold
variation - of potential benefits across all measures examined.
Health benefits associated with other quality of care disparities
in these clinical populations may be more significant (e.g., other
dimensions of diabetes care, breast cancer treatment vs. breast
cancer screening)
We would like to expand the modeling effort to include
attention to these other quality of care disparities.
As available data allow, these methods can be applied to other
HEDIS measures and other specific groups.
Overall Conclusions
• Plenty of documentation of disparities – need to know
more now about underlying reasons and potential
solutions.
• Policy changes are important, but clinical change
happens at the local, single organization level.
• Quality improvement concepts and structures are a
useful way to address disparities.
• We’ve made some progress in handling data collection
and measurement challenges, but we have much yet to
do.