Download Using health services data to explore the links between cancer care

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts
no text concepts found
Transcript
Defining a role for hospitals
and health care systems in
addressing cancer disparities
KIM F. RHOADS, MD, MS, MPH, FACS
CALIFORNIA CANCER REGISTRARS ASSOCIATION
43RD ANNUAL EDUCATION CONFERENCE
NOVEMBER 3, 2016
Disclosures
•
NO COMMERCIAL OR FINANCIAL
CONFLICTS
Motivation for the work: Jeannette
Barnes (1935-1993)
58 years old
• Presented with a locally
advanced breast mass
(stage IIIB)
•
Breast cancer care in a safety net hospital

No planned surgical intervention



No breast fellowship trained surgeon
No radiation therapy

?No radiation therapist

?No radiation oncologist

?No palliative care service
Given antibiotics and chemotherapy

Died during the hospitalization
Trends in Cancer Disparities
California Breast Cancer Trends
1980-2004
California Colorectal Cancer Trends
1980-2004
NCI website: http://statecancerprofiles.cancer.gov last access 7/17/12
Traditional explanations for disparities
in cancer survival
Patient
Characteristics
Chronic disease/comorbid states
 Late stage at diagnosis
Insurance Status
Lack of access to care contributes to poor
outcomes
Genetic/epigenetic explanations
 Minorities have more aggressive tumors
Expanding the explanations for cancer
survival disparities
•
•
Disparities in Treatment
• Lower quality
treatment may drive
worse outcomes
Surgical Volume &
Outcomes
• Use of low volume
hospitals results in
worse outcomes
•
Racial Disparities in Late Survival after Rectal Cancer
• African Americans more likely to be treated for
rectal cancer by low volume hospitals and fail to
receive adjuvant therapy.
•
•
Morris AM, Wei Y, Birkmeyer NJ, et al. J Am Coll Surg. 2006
Race and Surgical Mortality in the United States
• Black patients undergoing cancer operations in low
percent black hospitals have better cancer survival.
•
Lucas FL, Stukel TA, Morris AM, et al. Annals of Surg 2006
Does Hospital Context and Quality Drive
Disparities?
•
•
What is known:
•
Racial and ethnic disparities in cancer exist
•
Patient characteristics are associated with poor cancer outcomes
What is not well described:
•
What is the role of hospital characteristics on these disparities?
•
How are hospital characteristics associated with cancer outcomes
and cancer disparities?
The IOM on hospitals serving high percentages
of patients with Medicaid
“Because of Medicaid’s low
reimbursement rates for
doctors and hospitals, poor,
disproportionately minority
beneficiaries are subject to
largely separate, often
segregated systems of
hospitals and neighborhood
clinics. These systems often
adopt their own norms of
medical practice, shaped by
tight resource constraints.”
Unequal Treatment, Confronting Racial and Ethnic Disparities in HealthCare , IOM 2002
Quality of Colon Cancer
Outcomes in Hospitals with a High
Percentage of Medicaid Patients
RHOADS KF, ACKERSON LK, JHA AK, DUDLEY RA. JACS 2008
Methods
•
Data Sources
•
•
California Cancer Registry linked to Office of Statewide Health Planning
and Development Discharge data
•
ICD-9CM coding of diagnosis, procedures and co-morbidities
•
Individual level/patient characteristics + what happens in hospitals
Added hospital financial characteristics
•
•
Defined High Medicaid hospitals based on Medicaid Utilization Rate
Analytic approach
•
Hierarchical modeling
High Medicaid Hospitals serve higher proportions
of minorities and uninsured patients
Patient Characteristics
(n=18,000)
High Medicaid
Hospital (%)
Non-High
Medicaid Hospital
(%)
38.6
12.5
24.9
24.0
76.5
6.2
9.9
7.4
Race/Ethnicity
White (non-Hispanic)
Black (non-Hispanic)
Hispanic
Asian Pacific Islander
Insurance Status
Private Insurance
Medicaid
No Insurance
Medicare
Unknown
25.5
16.6
10.7
37.9
9.3
49.2
2.5
1.4
45.1
1.8
Rhoads KF, Dudley RA. J Am Coll Surg 2008
High Medicaid hospitals are associated with
higher colon cancer mortality
4
3.5
3
*p <0.001
2.5
High Medicaid Hospital
2
Non-High Medicaid Hospital
1.5
*p = 0.04
1
0.5
0
30days
1 year
% Mortality at 30 days and 1 year for patients with colon cancer
(California 1998-99)
Rhoads KF, Dudley RA. J Am Coll Surg 2008
High Medicaid hospitals are associated with
higher colon cancer mortality
4.5
4
3.5
3
p = 0.18
2.5
Private Insurance
2
1.5
Medicaid
p = 0.07
1
0.5
0
30 day
1 year
% Mortality at 30 days and 1 year for patients using High Medicaid
Hospitals (Colon Cancer, California 1998-99)
Rhoads KF, Dudley RA. J Am Coll Surg 2008
Understanding disparities in cancer survival:
place matters
Where you go for
treatment…can influence
survival
How location of care can impact outcomes of care:
Donabedian’s healthcare quality triad
Structural
Quality
•Surgical Volume
•Surgical Specialists
•Med/Rad Oncologists
•Rad Onc Facilities
• PET/CT scanners
Donabedian. Millbank Quarterly 2005
Process
Quality
Outcome
Quality
•Tumor Board Meetings
•Preoperative Evaluation
•Receipt of (Neo)/Adjuvant
•Number of Lymph Nodes
sampled
Do hospitals serving a high percentage of
Medicaid patients perform poorly on
evidence-based care for colon cancer ?
RHOADS KF, NGO JV, WELTON ML, DUDLEY RA. JHCPU, 2013
Defining evidence based care for colon
cancer (NCCN guidelines)
1. Resections for colon cancer follow
oncologic principles.
2. At least 12 Lymph node should be
examined after resection to determine
disease stage.
Stage I-II – 0 positive lymph
nodes
Stage III – Any positive lymph
nodes
3. Disease stage determines therapy
Stage I-II – No chemotherapy
Stage III – Chemotherapy
Stage
III
Available at nccn.org
High Medicaid hospitals lag behind others for
uptake and rate of 12 LN examination
Trends in Receipt of adequate LN examination
Proportion of Patients
0.8000
0.7000
0.6000
0.5000
*
*
0.4000
0.3000
HMH40
*
Teach
HVH
0.2000
0.1000
0.0000
Consensus panel recommendation
Colon Cancer, California, 1996-2006; from Rhoads et al. JHCPU 2013
Patients treated in HMH settings received chemotherapy
for stage III disease at lower rates
Trends in Receipt of Appropriate Chemo
(Stage III | Colon Cancer)
0.8500
Proportion of Patients
0.8000
0.7500
HMH40
0.7000
Teach
0.6500
HVH
0.6000
0.5500
0.5000
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
(Colon Cancer, California, 1996-2006; from Rhoads et al. JHCPU 2013
Understanding disparities in cancer survival:
place matters
Where you go for treatment
determines the quality of
care you get… and can
influence survival
How location of care can impact outcomes of care:
Donabedian’s healthcare quality triad
Structural
Quality
•Surgical Volume
•Surgical Specialists
•Med/Rad Oncologists
•Rad Onc Facilities
• PET/CT scanners
Donabedian. Millbank Quarterly 2005
Process
Quality
Outcome
Quality
•Tumor Board Meetings
•Preoperative Evaluation
•Receipt of (Neo)/Adjuvant
•Number of Lymph Nodes
sampled
Disparities in cancer mortality vary by tumor type
Tumor type
5-year survival rate
Absolute
difference
Black
White
Prostate
97.5%
99.9%
2.4%
Pancreas
4.6%
4.7%
0.1%
Liver
6.5%
9.1%
2.6%
Lung
12.2%
15.2%
3%
Esophagus
10.5%
16.8%
6.3%
Colorectal
55.5%
65.6%
10.1%
Breast
76.6%
89.8%
13.2%
Bladder
64.8%
83.2%
18.4%
Head & Neck
40.5%
62.1%
21.6%
Uterine
61.8%
86.4%
24.6%
Morris and Rhoads JACS 2010.
Source: http://seer.cancer.gov/canques/survival.html last accessed 7/08
How do integrated systems
address disparities in colon
cancer?
RHOADS KF, PATEL MI, MA Y, AND SCHMIDT L. JCO, MARCH 2015
Methods
•
Data Source
•
•
Predictor Variables
•
•
Hospital characteristics—fully integrated (Berkeley Forum definition) versus all other health
care settings
Outcomes
•
•
•
•
CCR + OSHPD + California hospital financial data (2001-06)
Delivery of evidence based care (NCCN guidelines)
5- year survival
Racial/ethnic disparities
Analytic Approach
•
Propensity Score Matching
•
with clinical, demographic AND social characteristics
Rates of NCCN guideline compliance were higher*
in Integrated versus other settings
* There was no different between settings for 12 LN
examination; rates ranged from 43.3% to 49.1%
Rhoads KF, Patel MI, Ma Y, and Schmidt L. JCO, March 2015
Propensity score matched survival shows integrated
settings associated with survival advantage
Rhoads KF, Patel MI, JCO, March 2015
IHS are associated with smaller gaps in racial/ethnic
survival
California, Colon Cancer, 2001-2006; from: Rhoads KF, Patel MI, JCO, March 2015
Mortality disparities in integrated versus other
settings; quality of care makes the difference
Baseline Model
Race/ethnicity
White
Black
Hispanic
API
Integrated System
HR (95%CI)
1.0 (referent)
0.96 (0.72-1.04)
0.92 (0.77-1.10)
0.83 (0.67-1.04)
P
NS
NS
NS
All Other Systems
HR (95%CI)
1.0 (referent)
1.15 (1.04-1.27)
0.90 (0.83-0.90)
0.79 (0.68-0.81)
P
<0.001
0.009
<0.001
-----------------------------------Baseline + Evidence Based Care----------------------------------White
Black
Hispanic
Asian
1.0 (referent)
0.79 (0.65-0.96)
0.86 (0.72-1.02)
0.79 (0.64-1.00)
0.02
0.09
0.05
1.0 (referent)
1.09 (0.98-1.21)
0.87 (0.90-0.95)
0.73 (0.67-0.81)
NS
0.001
0.001
Models adjusted for age; gender; Charlson comorbidity score; stage of disease. Abbreviations: HR-hazard ratio; 95%
CI-95% confidence interval; API-Asian/Pacific Islander
Interaction terms crossing race * location of care no qualitative difference (not shown)
Conclusion: improving quality of care & cancer
care equity can address cancer disparities
•
California’s largest integrated system has higher adherence to
evidence based care guidelines than other systems in the state
•
Propensity score matched comparisons showed lower mortality after
care within an integrated system
•
There were smaller disparity gaps in treatment in the integrated system
and no detectable racial differences in survival
•
Most importantly, adjusting survival models for delivery of evidence
based care eliminated survival disparities in all other settings
•
Similar findings in Acute Myeloid Leukemia
•
Patel MI & Rhoads KF, CEBP March 2015
Why is the role of hospitals and
health care systems so important
in addressing cancer disparities?
How location of care can impact outcomes of care:
Donabedian’s healthcare quality triad
Structural
Quality
•Surgical Volume
•Surgical Specialists
•Med/Rad Oncologists
•Rad Onc Facilities
• PET/CT scanners
Donabedian. Millbank Quarterly 2005
Process
Quality
Outcome
Quality
•Tumor Board Meetings
•Preoperative Evaluation
•Receipt of (Neo)/Adjuvant
•Number of Lymph Nodes
sampled
Minorities under utilize facilities that
deliver high quality cancer care
HUANG LC, MA Y, NGO JV, RHOADS KF. CANCER 2014
HUANG LC, MA Y, NGO JV, RHOADS KF. DIS COL AND RECTUM 2015
Higher proportions of minorities in California live
close to an NCI center
30%
% of patients treated for CRC
25%
20%
15%
Nearby
10%
10%
12%
14%
26%
White
Black
Hispanic
API
5%
0%
Percentage of each racial group living nearby (within 5 miles of) an NCI hospital (Colorectal
Cancer , California 1996-2006. (N=79,231))
Huang LC, Rhoads KF. Cancer 2014
Insurance status does not explain minority
under-utilization of NCI settings
Baseline Model
Odds ratio
P-value
Insurance Model
Odds ratio
P-value
Race
White (ref)
1.0
1.0
Black
0.83
0.008
0.81
<0.001
Hispanic
0.72
<0.001
0.70
<0.001
API
1.40
<0.001
1.39
<0.001
Insurance
Private (ref)
1.0
Medicaid
1.85
<0.001
Medicare
2.10
<0.001
No insurance
0.86
0.311
Missing data
1.93
<0.001
Huang LC, Rhoads KF. Cancer 2014
Neighborhood demographics are more important
predictor of NCI use than travel distance
Lower odds of utilization
Higher odds of utilization
Huang LC, Rhoads KF. Cancer 2014
Where do we go from here?
Implications for increasing equity and
quality in cancer care
A conceptual framework for understanding disparities in
cancer survival
California Cancer Registry Data:
Current State

CCR data is collected and curated under California state
law whose language focuses on reporting of cancer
incidence (epidemiology); and detection of geographic
high incident areas

There is no language (currently) in the law to support the
use of data for quality reporting or quality improvement
Leveraging the CCR to measure & improve
quality of care and outcomes

Section 103875 of the Health and Safety Code recommended amendments:
c) The director shall analyze data collected under the program to assess , measure
and publicly report on the quality of cancer care in the state. In assessing and
measuring the quality of cancer care, the director shall define and identify oncology
providers. In publicly reporting the quality of cancer care in the state, the director shall
identify oncology providers but not any cancer patients….

Section 103885 of the Health and Safety Code recommended amendments:
a) “The director shall also identify and include in the statewide system cancer care
quality measures for use in public reporting.
Fighting cancer with data: Enabling the California cancer registry to Measure and Improve care.
California Health Care Foundation, November 2014 (Available online at:
http://www.chcf.org/~/media/MEDIA%20LIBRARY%20Files/PDF/PDF%20F/PDF%20FightingCancerDataRegistry.pdf)
Registry data can play a critical role in helping
hospitals & healthcare systems address disparities
•
Patient Level: Increasing patient access to cancer care quality information
•
•
•
Use data to identify high and low performing health systems and hospitals
Hospital level: Monitoring hospital performance
•
Use the data to identify high and low performing hospital (by name or characteristics)
•
Designing and deploying large scale interventions to improve quality (at the hospital level)
Health Care System level: Inform CMS reimbursement policy for Oncology Care
Model sites
•
NCCN guidelines define high quality care for OCM
•
Use registry data to document compliance with guidelines (over time)
•
Use long term data to rationally risk adjust reimbursement policies (Pay for Performance programs), to
avoid unintended consequences of financial penalties in low resource settings
Where are we trying to go: Equality versus Equity
Everyone gets the same
resources
Everyone gets the resources they
need
IN CLOSING…
Defining a role for hospitals
and health care systems in
addressing cancer
disparities
[email protected]
THANK YOU FOR
WHAT YOU DO!