Financing coverage for Oregon

1
Financing coverage for Oregons uninsured

• John McConnell, PhD
• Oregon Health Science University

2
Objectives of this talk

• Review of health care costs
• Why is health care in the U.S. so expensive?
• Why do health care costs go up?
• Uncompensated care in Oregon
• Variations in care
• Chronic illness
• Evidence on markets
• The cost of covering the uninsured
• Review of Health Policy Commission model and
  estimates

3
Why is health care in the U.S. so expensive?
4
Why is health care in the U.S. so expensive?

• U.S. per capita spending 2.5 times greater than
  median Organization for Economic Cooperation and
  Development (OECD) country
• 50 higher than the second highest (Switzerland)
• Why so much higher than other countries?

5
Its the prices, stupid.
Anderson et al, Health Affairs
2003

• Expenses Price Quantity
• Utilization measures are lower
• Fewer physicians, nurses, and hospital beds per
  capita than OECD median
• Fewer office visits, acute care bed days, shorter
  inpatient bed stays than OECD median
• MRI/CT scans equal to OECD median
• Prices are higher
• Oregon insurance CEOs focus on unit price
  increases
• Payments to providers
• Quality of services
• Some of this is good, some of it is questionable

6
Why do health care costs go up?
7
Why do health care costs go up?

• Costs are high, but will get higher
• In the US and in the OECD
• The rate of cost increases is similar across
  countries
• Just hurts us more because our baseline levels
  are so high to begin with
• What drives health care costs up?
• Lots of little reasons
• One big one.

8
Technological change

• New procedures, drugs, equipment
• Many of which lead to longer, healthier lives
• All of which increase total health care costs
• Example
• 1956 heart disease death
• 2006 heart disease 40,000 life
• Spending related to new technology
  (procedures/drugs/devices) accounts for 50 to
  75 of increases in spending

9
The Cycle of Unaccountability
Drug / Device
My job is innovation that helps people . . . its
up to the doctors to control use.
We want to pay for the right things, but theres
little data and saying no jeopardizes our
relationships.
My job is doing everything I can to help my
patient.
Safety, not cost-effectiveness, is my job.
Consumer-Patient
I want the best of everything. Dont ask me to
pay more.
Source Galvin
10
Uncompensated care and cost shifting in Oregon
11
Uncompensated care in hospitals in Oregon
Last year of OHP open enrollment
OHP implementation
12
Uncompensated care in Oregon(preliminary
estimates)

• 2004 hospital uncompensated care 299M
• Total uncompensated care for 2004 estimated to be
  425M
• What is the burden on those with commercial
  insurance?
• Approximately 6 - 9 of 2004 Oregon family
  premium of 9,906

13
Health reform the cost-shift

• Cost shifting not a viable long-term strategy
• An inefficient hidden tax
• Implicit agreement to support catastrophic care
  over preventive care
• Adds to the increasing cost of commercial
  premiums and erosion of employer-sponsored health
  insurance
• The magnitude of uncompensated care in Oregon is
  large
• Substantial savings for employers/employees from
  policies that cover the uninsured
• But need to consider polices to insure savings
  actually get to employers/employees

14
Variations in care
15
Variations

• The Wennberg variations
• Pick your procedure (Back surgery, MRIs, CABG,
  Vioxx) and your region (states, counties with
  states)
• E.g., Medicare's costs per enrollee by region
  varied from 4,500 to nearly 12,000 in 2003
• Better outcomes not associated with higher
  spending
• Estimates of 20 - 30 of spending could be
  eliminated
• Big savings how to capture it?
• More rigorous use of evidence-based medicine
• Investment in Information Technology
• Better coordination of care

16
Chronic Ilnesses
17
Spending on chronic disease

• 5 of the population accounts for 56 of health
  care expenditures
• Fastest area of health care cost growth
• Bodenheimer Can we decrease costs for our
  sickest patients by 50?
• Large theoretical savings from disease
  management/EMR/HIT
• Care Management Plus model at OHSU
  nurse-based care management IT for patients
  with multiple chronic illnesses

18
Markets and competition
19
A lot of interest in what markets and competition
can do for health care

• This is a natural response
• Markets are the American way
• Concern about moral hazard
• Consumers arent consumers
• More shopping would lead to better utilization
  and/or lower prices
• Focus on consumer-driven health plans (CDHP),
  high deductibles, health savings accounts (HSAs)

20
Markets supply side and demand side

• Supply side
• Focus on the provider/health plan
• Ex ante price setting
• Demand side
• Focus on the patient/consumer
• Ex post price setting

21
Supply side the evidence

• Focus on provider
• Real (inflation-adjusted) health care spending
  was flat for much of the 1990s
• Complaints from providers patients
• But no observed quality/outcome problems
• How did managed care do it?
• Most savings came from rate reductions provider
  discounts
• Not from gatekeeping, better utilization review
  or other ways of managing care
• Were there process improvements from providers?
• Some but a lot of focus on achieving
  counterbalancing market power
• Some lessons from prepaid group model
• Freedom from FFS chances to innovate (group
  visits)
• Some evidence of process improvements, costs
  savings

22
Demand side the evidence

• Yes, in fact, moral hazard exists
• BUT savings smaller than you would think
• Co-payments/deductibles have the biggest impact
  on access, not on price
• Whether or not you go
• Not how much you pay once you are there.
• Estimated savings if everyone moved into Health
  Savings Account
• Range of 2.5-7.5
• One-time only savings does not do much for the
  technology problem
• Evidence on HSA take-up
• Co-payments for poor/Medicaid populations?

23
Can markets tackle long-term growth?

• In 2007, TramGenix releases a cure for
  Alzheimers. Cost 20,000/year
• This is great! (and cost-effective by
  conventional standards)
• 50K Oregonians with Alzheimers, another 26K with
  related disease
• Implies an additional 3000 in health premiums or
  taxes for an Oregon family of four
• Best estimate adds another 100K to 200K to
  uninsured through increased premiums
• This is bad!
• It is very difficult to manage a drug that costs
  20,000
• (or 100,000) with no substitute
• Is there a market solution for this problem?

24
Summarizing markets

• If markets have been successful at cost control,
  it has been primarily by extracting discounts
  from providers (supply side)
• i.e., impact on price not quantity
• Public programs can do this, too
• Evidence on savings from consumerism is real
  but so far relatively small
• Markets dont have a great answer for the
  technology-cost relationship
• Markets dont do subsidies

25
Covering the Uninsured the Cost to Oregon
26
OHPC modeling based on 3 building blocks to
expand coverage

• Individual health insurance requirement/mandate
• Extending publicly financed coverage and
  insurance premium subsidies to more Oregonians
• Health Insurance Exchange

27
Assumptions

• Reform occurs in 2008.
• 100 coverage (0 uninsurance)
• OHP eligible to any individual with income lt200
  FPL
• Uncompensated care is estimated to be 540
  million per year in Oregon
• Crowd-out is estimated to be 25
• No subsidies for those currently covered with ESI
  (firewall)
• Subsidies for commercial premiums are such that
  the individuals spending on premiums is capped
  according to the following schedule
• Individual with incomes between 100 and 200 FPL
  have spending for premiums capped at 720 for
  adults and 360 for children
• Individual with incomes between 200 and 300 FPL
  have spending for premiums capped at 1,440 for
  adults, 720 for children
• Individual with incomes above 300 FPL do not
  have spending caps on their premium spending

28
Basic structure

• Model has three components
• Enrollment
• Medicare
• Medicaid (by PERC)
• Commercial (ESI/Individual)
• Uninsured (by FPL)
• Spending on health services
• Medicare (ok data)
• Medicaid by PERC (good data)
• Commercial (weak data)
• Uninsured (decent estimate based on hospital
  uncompensated care)
• Cost of coverage
• Medicaid
• Based on spending administrative overhead
  federal match
• Commercial
• Based on spending admin. overhead allows for
  savings from reduced uncompensated care

29
Outputs of interest defined as

• Total state spending (OHP/Medicaid plus premium
  subsidies)
• Federal match (non-Medicare spending)
• Employer spending
• Individual spending on premiums
• Results are annual figures (not biennium)
• Results are for adults children

30
Current snapshot of the uninsured
31
The next step implementing policy

• UNINSURED assume universal coverage (i.e., no
  uninsured)
• Individuals under 100 FPL
• Assume 100 moved into OHP/Medicaid
• Individuals at 100-200 FPL
• Assume 80 moved into Medicaid
• Assume 20 purchase ESI with 80 premium subsidy
• Individuals at 200-300 FPL
• Assume 50 purchase individual insurance with 50
  premium subsidy
• Assume 50 purchase ESI with 50 premium subsidy
• Individuals at gt300 FPL
• Assume 34 purchase individual insurance (no
  subsidy).
• Assume 66 purchase ESI (no subsidy).

32
Integration with Employer Sponsored Insurance
(ESI)J.Gruber

• Low income pool how to treat those with ESI?
  Three alternatives
• 1) Firewall MA approach but 30,000 are
  excluded from affordable coverage
• 2) Premium assistance
• sounds attractive, since many uninsured are
  offered ESI leverage employer dollars
• But it is actually incredibly expensive

33
Premium Assistance Facts J.Gruber

• Fact 1 Among those who are offered ESI below
  300 of poverty, vast majority take it
• Below 100 of poverty of all offered, only 25
  uninsured
• 100-200 of poverty 13 uninsured
• 200-300 of poverty 6 uninsured
• Implication if you offer premium assistance to
  low income populations, most of those eligible
  already have coverage!
• Great for horizontal equity not for coverage

34
Premium Assistance Facts J.Gruber

• Fact 2 Among those offered ESI who are
  uninsured, price sensitivity is very low
• After all, these individuals were already offered
  a very large subsidy and declined!
• These are folks who dont want insurance
• Fact 3 If you subsidize employee contributions
  for a sizeable share of employees, employers will
  raise those contributions!

35
Premium Assistance Implications J.Gruber

• Simple example 1000 persons below 300 of
  poverty offered insurance at 2000/year 100 of
  them are uninsured
• Offer premium assistance of 1000/person
• 750 of 900 already taking ESI take assistance
• 25 of 100 not offered ESI take assistance
• Cost 775,000
• Newly covered 25 persons
• Costs/Newly covered 31,000!
• Not unreasonable Grubers study of impact of
  Section 125 for Federal employees found cost per
  newly insured of 31,000 to 84,000

36
Alternative 3 Vouchers J.Gruber

• Allow employees to come to the pool with employer
  dollars
• In theory, same as premium assistance
• In practice, perhaps less expensive because
  employees who are covered are reticent to drop
  that coverage and move to the pool
• But still expensive per newly insured
• Bottom line Hard choices on low income ESI
  eligible
• Our estimates assume a firewall

37
Individual market coverage and income
38
ESI market and income
39
When does crowd-out from ESI happen?

• Some happens with job turn-over
• But biggest threat is likely to be firms with
  large of low-wage employees
• Approximately 150K to 200K receiving ESI from
  firms where the majority of employees are low
  wage (lt10 hr)
• About average for the country

40
Results
41
Outputs of interest defined as

• Total state spending (OHP/Medicaid plus premium
  subsidies)
• Federal match (non-Medicare spending)
• Employer spending
• Individual spending on premiums
• Results are annual figures (not biennium)
• Results are for adults children

42
Where do the uninsured go?
43
(No Transcript)
44
Results spending
45
Rationale

• State spending up 548
• 496 from OHP/Medicaid enrollment
• 52M for subsidies going to previously uninsured
• Employer spending relatively flat
• Greater number of employees covered, but some
  savings from reduced uncompensated care
• Individual spending slight decline
• More people covered, but large number 100-300
  FPL with ESI who become eligible for new subsidies

46
Thank you

• and questions?
• 503.494.1989
• mcconnjo_at_ohsu.edu