Transcript Lecture - University of Windsor

Breast Cancer among Women
Living in Poverty: Better Care in
Canada than in the United States
Historical Cohort Support of a
Health Insurance Explanation
Presenter Disclosures
No relationships to disclose
Funding sources:
Canadian Institutes of Health Research
Grant no. 67161-2
Manuscript status:
Social Work Research (in press)
Abstract
We studied breast cancer care among women living in
poverty in California & Ontario between 1996 & 2011.
Women in Canada were diagnosed earlier, enjoyed better
access to breast conserving surgery, radiation (RT)
and hormone therapies and survived longer. They
even experienced shorter waits for surgery and RT.
By observing the historical protective effects of Canada’s
universally accessible, single-payer health care
system we estimated the Affordable Care Act’s (ACA)
likely protections as well as its likely risks.
Background
Longstanding Rhetorical Context
Anecdotes about Canadian health care
failures abound
• Long waits for care
• Care denials—“death lists”
“Do you want government in your medicine
cabinet?”
Does systematic, empirical evidence tend to
support or refute such rhetoric ?
Clinical Context
Why study breast cancer?
•
•
•
•
•
Relatively common over the life course
Effective screens exist
Effective treatment regimes exist
Timely diagnosis & best treatment matters
Excellent prognoses can be expected:
Long survival & high quality of life
It is a sentinel health care quality indicator.
Historical-Theoretical Context
• Large Canada-US studies, exemplified by a General
Accounting Office study (1994) found nil to null
differences on breast cancer survival
• Country-by-income interaction discovered (1997)
• Canadian breast cancer survival better in
low-income (LI) neighborhoods only (RR = 1.30)
• 78 Canada-US cohorts synthesized (2009)
• Canadian women advantaged in LI neighborhoods
only (pooled RR = 1.14); < 65 yoa (RR = 1.21)
Knowledge gaps:
• Breast cancer care processes in extremely poor
places have not yet been studied.
Geo-Economic-Policy Context:
During Great Recession, 2007-11
United States
• Prevalence of people living in poverty increased
25% (12% to 15%) to 46.2 million people
• Uninsured population rose to more than 50 million
• Inadequately insured population rise to 100 million
[uncovered costs by multi- public & private payers]
Canada
• Prevalence of people living in poverty was nearly
constant at approximately 10%
• The entire population was insured for medically
necessary care by a single, public payer.
Research Question: Hypotheses
Prior to the enactment of the Affordable Care Act,
Did Canadian women living in high poverty
neighborhoods with breast cancer experience
better care than did their American counterparts?
Hypotheses:
1. Canadian women experienced better care and
survival.
2.
The care and survival of Canadian women was
even better when compared to that of uninsured
or underinsured American women.
Methods
Comparison of Historical Cohorts:
High Poverty Neighborhoods in
Ontario and California,
Women with Breast Cancer Diagnosed
Between 1996 & 2000 Followed to 2011
Sampling High Poverty Cohorts
Enhanced Ontario and California cancer registries
• Comprehensive, reliable and valid
• Diverse places well represented
Random samples stratified by urbanity: megalopolises,
small cities & rural places
• Respectively, 300 & 1,950 women (multi-”controls”)
Comparably poor places defined by Census Bureaus
• CT household poverty prevalence of 30-40+% (US)
• Poorest CTs on Stats Can’s low-income criterion
• Mdn incomes, purchasing power-adjusted in USD:
$23,175 (California) & $23,800 (Ontario)
Practical Statistical Analyses
Early diagnosis, treatment & survival rates
• Directly age-adjusted (other confounds)
• Study sample was the internal standard
• Rates per 100 participants or percentages
Standardized rate ratio (RR) comparisons with (95% CIs)
Mathematical models adjusted for multiple predictive
and potentially confounding factors
• Logistic regressions (diagnosis & treatment)
• Cox hazards regression (survival)
Notes. Key study variables had less than 3% missing data which was not
confounding. Covariates: disease stage at diagnosis, tumor grade,
tumor size and hormone receptor status.
Results
Node Negative Breast Cancer at Diagnosis?
Places (People)
Adjusted Rates (%)
Ontario (All)
65.0
California (All)
61.5
RR = 1.06 (0.96, 1.17)
Ontario (All)
65.0
CA (Uninsured or Publicly)
57.9
RR = 1.12 (1.01, 1.24)
Received Cancer-Directed Surgery?
Places (People)
Adjusted Rates (%)
Ontario (All)
96.6
California (All)
94.3
RR = 1.02 (0.99, 1.05)
Ontario (All)
96.6
CA (Uninsured or Medicaid)
93.2
RR = 1.04 (1.00, 1.08)
Received Breast Conserving Surgery?
Places (People)†
Adjusted Rates (%)
Ontario (All)
73.5
California (All)
49.6
RR = 1.48 (1.31, 1.68)
†
Among women with node negative disease.
Received Radiation Therapy?
Places (People)†
Adjusted Rates (%)
Ontario (All)
70.6
California (All)
66.4
RR = 1.06 (0.98, 1.14)
Ontario (All)
70.6
CA (Uninsured or Publicly)
60.6
RR = 1.17 (1.01, 1.35)
†
Among women with node negative disease who had
breast conserving surgery.
Experienced Long Waits for Care?
Places (People)†
Adjusted Rates (%)
Waited > 2 Months for Surgery
Ontario (All)
7.2
CA (Uninsured or Medicaid)
12.4
RR = 0.58 (0.36, 0.93)
Waited > 6 Months for Radiation Therapy
Ontario (All)
6.2
CA (Uninsured or Medicaid)
14.2
RR = 0.44 (90% CI: 0.20, 0.96)
†
Among women with non-metastasized disease.
Received “Optimum”† Care?
Places (People)‡
Adjusted Rates (%)
Ontario (All)
64.0
California (All)
47.4
RR = 1.35 (90% CI: 1.03, 1.77)
Ontario (All)
64.0
CA (Uninsured or Publicly)
43.1
RR = 1.48 (1.13, 1.94)
(CA Uninsured) RR = 1.89 (1.31, 2.72)
†
Received breast conserving surgery within 2 months of diagnosis and
received adjuvant radiation therapy within 4 months of surgery.
‡ Women with node negative disease & low to intermediate grade tumors.
Received Hormone Therapy?
Places (People)†
Adjusted Rates (%)
Ontario (All)
68.2
California (All)
41.2
RR = 1.65 (1.44, 1.89)
Ontario (All)
68.2
CA (Uninsured or Publicly)
38.3
RR = 1.78 (1.53, 2.07)
†
Women with hormone receptor positive tumors.
Survived?
We ran a number of age, stage and grade
adjusted regressions on 3- to 10-year survival.
Canadian women were advantaged in each:
Pooled RR = 1.60 (1.26, 2.02)
In each instance, when health insurance entered
the model country was no longer significant:
Pooled RR = 1.07 (0.94, 1.19)
Discussion
Summary
Women living with breast cancer in high
poverty neighborhoods received better care
in Ontario than in California.
Such access mattered in terms of their better
short- and long-term survival chances.
In the pre-Obamacare era that included a
number of treatment innovations and
increasingly effective breast cancer care,
extremely poor women in Ontario gained
access to them much more readily than did
their counterparts in California.
Interpretations
Estimated inequities among women with breast cancer
living in poverty in America over the past generation:
• 53,000 late diagnoses
• 158,000 sub-optimum treatments
• 172,000 premature deaths
Breast cancer accounts < 2% of US’s disease burden
US Women Compared to Canadian women:
• Uninsured most consistently vulnerable
• Medicaid insured nearly as vulnerable
• Significant vulnerabilities among Medicare insured
• Even certain privately insured were vulnerable
Interpretations
Language of American health care:
• Platinum, gold, silver and bronze private coverages
• “Medigap” coverages are needed for Medicare covered
• Cost-sharing among the poor, covered by Medicaid?
Seems an acceptance, not only of a multi-payer system, but of a
multi-quality and multi-outcome system that destines many to
relatively low quality care with its attendant greater risks of
suffering & early death.
ACA/Obamacare Changes:
• Tens of millions more Americans will be insured
• Majority new private plans bronze or silver , high deductibles
• Many states have not yet expanded Medicaid and considering
various, out-of-pocket, cost-sharing measures
• Many previously uninsured may become underinsured
Conclusions
The ACA will probably substantially reduce
such observed inequities.
But single-payer reform would probably further
reduce, if not completely eliminate them.
To the extent that such is not politically
feasible, advocates ought to work to ensure
that the ACA is enacted across all 50 states
in ways that are consistent with its federal
legislative intent, that is, that high quality
health care be truly available to all.
Potential Limitations
1. Race/Ethnicity Alternative Explanation
• Findings replicated among the subsample of non-Hispanic
white women in California vs. the entire ethnically diverse
Ontario sample
2.
Income Differences (US Poor are Poorer on average than
Canadian Poor)
• Findings replicated among California-Ontario subsamples
with nearly identically low incomes
• Even granting this: It is instructive to know that women
who live in Canada’s poorest neighborhoods are so much
better insured than women who live in America’s poorest
neighborhoods.
Co-Investigators
Investigator
Kevin Gorey
Nancy Richter
Madhan Balagurusamy
Affiliation
__________
School of Social Work
University of Windsor
Isaac Luginaah
GuangYong Zou
Caroline Hamm†
Department of Geography
Dept. of Epidemiol & Biostats
Department of Oncology
University of Western Ontario
Eric Holowaty
School of Public Health
University of Toronto
†&
Medical Oncology Department, Windsor Regional Cancer Center
Acknowledged Administrative,
Logistical or Research Support
Supporter
Affiliation
__________
Kurt Snipes
Janet Bates
Gretchen Agha
Cancer Surveillance and
Research Branch, California
Department of Public Health
Mark Allen
Allyn Fernandez-Ami
Arti Parikh-Patel
California Cancer Registry
Sundus Haji-Jama
School of Social Work
University of Windsor
Charles Sagoe
Cancer Care Ontario
Disclaimer
Other Agencies Involved in Data Management:
National Cancer Institute (United States),
Cancer Prevention and Public Health
Institutes of California, Centers for Disease
Control and Prevention, University of
Southern California, and the Canadian
Institute for Health Information
The ideas and opinions expressed herein are
those of the presenters and endorsement by
any affiliated or data-supportive agencies or
their contractors and subcontractors are not
intended nor should they be inferred.
Principal Investigator
Kevin Gorey
For more information about our research see my
academic website at:
www.uwindsor.ca/gorey
For any additional information, including reprint
requests, feel free to contact me at:
[email protected]