Health Care Policy: Comparing and Contrasting the Options

1
Health Care PolicyComparing and Contrasting the
Options

• Presented by David W. Allen, Jr.
• [phone]
• david_at_allenpie.com

2
The Economics of Health Care

• Responds in accordance with well-established
  market dynamics like supply and demand
• Providers and consumers economic behavior is
  rational
• Follow the money to understand health care

3
The Path to 2004

• 1880s Chancellor Bismark implements government
  funded health care in Germany
• 1910s Teddy Roosevelt runs on platform that
  includes national health care
• 1930s FDRs national health care plans defeated
  by AMA Kaiser-Permanente HMO established

4
The Path to 2004

• 1940s IRS rules employer sponsored health
  insurance is deductible
• 1950s Growth of group health insurance
• 1960s Medicaid and Medicare
• 1970s Staff Model HMOs
• 1980s Group and Network Models HMOs
• 1990s Point-of-Service HMOs
• 2000s All managed care all the time

5
Health Care in 2004

• The Good
• Best medical care in the world
• The Bad
• The uninsured (43.6 million uninsured)
• Unbalanced allocation of health care
• Mediocre public health
• Costs

6
Health Care Costs

• In 2003, 1.7 trillion spent on health care
• 15.3 of GDP dedicated to health care
• In Minnesota, healthcare consumes 11,000 per
  household
• Costs increasing faster than overall inflation
• Estimated to reach 22,000 per household by 2010
• Waste
• Estimates are that 25 to 50 of provider costs
  are associated with administration
• Estimates are that a quarter or more of all
  health care costs are associated with ineffective
  or unnecessary care

7
Health Policy Options

• Tweak the Current System
• Single Payer System
• Consumer-Driven System

8
Criteria for Grading the Options

• What impact on health care quality?
• What impact on costs?
• What impact on universal access?

9
Tweaking the Current System

• Major Challenges
• Controlling costs
• Correcting imbalances
• Covering the uninsured

10
Tools of the Managed Care Trade

• Staff Model
• Universal Budget
• Group, Network Models
• Comprehensive Capitation
• Point-of-Service
• Fee Withholds
• Utilization Management
• All providers
• Disease Management
• Pay-for-Performance

11
Controlling Costs under the Current System

• We the undersigned are united in our belief that
  a unique opportunity now exists to address the
  crisis of quality facing the nations health
  system The strategic concept of paying for
  performance a bedrock principle in most
  industries has begun to emerge in health care
  in a variety of experiments The inertia of the
  health care system could easily overwhelm nascent
  efforts to raise average performance levels out
  of mediocrity Decisive change will occur only
  when Medicare creates financial incentives that
  promote pursuit of improved quality.
• Excerpt from Paying for Performance Medicare
  Should Lead an open letter published in the
  November/December 2003 Health Affairs and signed
  by 15 prominent health care experts including Don
  Berwick, Paul Ellwood, Alain Enthoven, George
  Halvorson, Ken Kizer and Uwe Reinhardt.

12
Controlling Costs under the Current System

• BCBSM recently launched two new outcomes-based
  provider incentive programs and their purpose
  is rewarding provider performance relative to
  proven clinical outcomes... including a goal of
  increasing appropriate use of generic drugs.
• --Douglas Hiza, M.D., medical director of BCBSM,
  as quoted in Minnesota Physician October 2003.

13
Controlling Costs under the Current System

• The number of health plans using
  pay-for-performance initiatives is estimated to
  represent 30 percent of all HMO membership
  nationwide. Whatever performance measures are
  used, its important that they be justifiable in
  both clinical and financial outcomes Plans have
  the data. Plans are the communications
  interchange among consumers, providers, and
  employers. Plans pay out the money. The era of
  managed authorization is fading, and few will
  miss it. Pay for performance and incentives for
  quality outcomes are the steppingstones to a new
  type of reimbursement for network providers.
• Quality outcomes and pay for performance
  Evolution in provider reimbursement, by Charles
  Fazio, M.D., Minnesota Physician, February 2004.

14
Three Ways to Measure Quality

• Measures of Outcomes, e.g.
• Post-operative mortality or morbidity rates
• Clinical metrics (e.g. cholesterol levels)
• Measures of Processes, e.g.
• Hemoglobin checks on diabetics
• Prescription of beta blockers post heart attack
• Patient Satisfaction

15
Measuring Outcomes is Very Difficult

• Requires medical record, not just claims
• Expensive
• Intrusive, may violate patient privacy
• May not be cost-effective
• Requires risk adjustment
• Failure to adequately adjust for risk penalizes
  providers who care for the sickest patients

16
NY DoH Effort to Measure Heart Surgery Outcomes

• Measures mortality rate following coronary artery
  bypass surgery
• Audits medical records
• Considers Patient Risk Factors (e.g. left
  ventricular function, prior heart attack)
• Considers co-morbidities (e.g. obesity, diabetes)
• Result some heart surgeons are avoiding sickest
  patients

17
Maine Effort to Measure Substance Abuse Treatment
Outcomes

• Performance-based Contracting
• Increased funding for best substance abuse
  providers
• Reduced funding or contract cancellation for
  poorly performing providers
• Result Sickest patients could not get treatment

18
Measuring Processes is also Problematic

• Relatively few agreed upon standards of care
• According to a 1998 NEJM report, only 15 20 of
  medical care can be justified on the basis of
  rigorous scientific data
• Only measures patients who access the process
  difficult to measure excluded patients
• There is little evidence correlating good process
  to good outcomes

19
Patient Satisfaction is Not a Surrogate for
Quality

• It is well established that sicker patients are
  more critical of their care givers
• Questions about outcomes suffer same deficiencies
  as outcome studies
• Questions about processes suffer same
  deficiencies as process studies
• Responses often have little relationship to
  quality

20
Case Study Pay-for-Performance in Minnesota

• The Minnesota Community Measurement Project
• Medicas Pay-for-Performance Plans

21
Minnesota Community Measurement Project

• Initiated in 2001 by the Minnesota Council of
  Health Plans
• Steering Committee of eleven, representing seven
  health plans, MCHP, NCQA, and two medical groups
• Advisory group of 16 physicians representing 15
  medical groups

22
Minnesota Community Measurement Project

• Initially focused on diabetes care
• Other studies may include childhood
  immunizations, well child visit rates, breast
  cancer, cervical cancer, Chlamydia screening
  rates, hypertension, asthma, and depression

23
MCMP Diabetes Study

• Forty-nine medical groups participating, all with
  60 or more diabetic patients enrolled in the
  seven participating health plans
• Blinded results for 2001 and 2002 have been
  shared with each group
• Measures apparently include percent of patients
  classified by HbA1c, LDL-cholesterol, blood
  pressure, over 40 taking aspirin, tobacco, and
  screened for kidney and eye.

24
MCMP Diabetes Study

• Study takes into account insurance coverage
• Study apparently does not take into account age,
  sex, health status, co-morbidities, or any other
  factors than insurance coverage
• Expressed intention is to publicly release data
  in late 2004 as a report card.
• Some health plans may alter reimbursement based
  on results

25
Problems with MCMP Diabetes Study

• Diabetes care standards were developed to assist
  physicians in improving their practices, not as
  the basis of comparing physicians
• There are many variables that effect these
  scores, not just quality of physician care
• The Diabetes Quality Improvement Project has
  stated that measures such as those being used by
  MCMP are not appropriate for comparing plans or
  providers

26
Problems with MCMP Diabetes Study

• Hemoglobin A1c levels are correlated with
  demographic variables (patient age and sex),
  physician and site, socioeconomic status
  (including income, education, and employment
  status), duration of diabetes, and health status
  measures.
• Timothy P. Hofer, The unreliability of
  individual physician report cards for assessing
  the costs and quality of chronic disease, JAMA
  1999 2812098-2105, 2099

27
Other Hofer Conclusions

• Physicians with poor scores can dramatically
  improve their scores by getting rid of the
  patients with the top 5 percent of HbA1c levels
• Physician treatment accounts for only 3 of HbA1c
  variation
• To achieve 80 reliability requires a sample size
  of at least 100 patients

28
LDL-Cholesterol also does not correlate well with
Quality

• One study
• 27 of diabetic patients had LDL-Cholesterol
  above 129 mg/dL
• Of these patients, 51 were being treated
  appropriately with statins or statins were
  contraindicated
• Of the remaining 49, 24 had factors beyond the
  physicians control limiting treatment options
• From Eve A. Kerr et all., Building a better
  quality measure Are some patients with poor
  quality actually getting good care? Medical
  Care 2003 41 1173-1182

29
MCMP Conclusions

• Ironically, methodology is not evidence-based
• Results will most likely reflect factors beyond
  physician control, rather than quality. For
  example
• Doctors with lower income patients will probably
  have lower scores
• Patients with poor drug benefits (e.g., statin
  coverage) will probably have lower scores
• Process is probably diverting resources away from
  quality services
• May create incentive to game by turning away
  sickest patients

30
Medicas Pay for Performance

• The quality and outcomes payments are based on
  the volume of commercial, fully insured CMS-1500
  claims for eligible network clinics Medicas
  first pay-for-performance program, for generic
  drugs, was implemented in 2003 and continues this
  year. The measurement method is based on the
  number of claims filed and the rate of generic
  drugs prescribed
• Quality outcomes and pay for performance
  Evolution in provider reimbursement, by Charles
  Fazio, M.D., Minnesota Physician, February 2004.

31
Medicas Pay for Performance

• In 2004, Medica is adding five additional
  performance based incentives (four
  physician-based and one hospital-based)
  Pediatric Asthma The measurement is the
  percentage of eligible patient medical records
  containing an asthma action plan. Diabetes The
  measurement is Community Measurement Project
  (CMP) criteria around optimal diabetes
  management
• Quality outcomes and pay for performance
  Evolution in provider reimbursement, by Charles
  Fazio, M.D., Minnesota Physician, February 2004.

32
Medicas Pay for Performance

• Low back pain The measurement is the percentage
  of eligible cases that follow Institute for
  Clinical Systems Improvement (ICSI) guidelines
  for imaging in the first six weeks. Chlamydia
  The measurement is the percentage of sexually
  active women ages 13 to 25 whose medical records
  indicate they were offered and/or received a
  screening. Patient safety This hospital-based
  pay-for-performance plan will use the Leapfrog
  Group measurements around computerized order
  entry (CPOE).
• Quality outcomes and pay for performance
  Evolution in provider reimbursement, by Charles
  Fazio, M.D., Minnesota Physician, February 2004.

33
Medicas Pay for Performance

• Leans more towards cost-effectiveness than
  quality
• Favors larger provider organizations
• Measures processes rather than outcomes
• May create disincentives for caring for the sick
• Myopic broad measurement is an impossibility

34
Summary Problems with Pay for Performance

• Quality cant be measured effectively by health
  plans
• Erode patient confidentiality
• Tend to focus on cost-effectiveness rather than
  quality
• Not evidence-based (e.g., DOER)
• Efforts to reward quality frequently result in
  penalizing those who care for the sickest patients

35
Summary Tweaking the Current System

• Principal strategy (PFP) wont work
• Disease Management by Insurers wont work
• Costs wont be controlled
• Imbalances will continue
• No good solution to problem of uninsured
• Public health problems likely to worsen

36
Davids Grades for Tweaking the Current System

• Quality of Care B (inconsistent)
• Affordability F (tremendous waste)
• Universality D (43 million uninsured)
• Overall Grade C-

37
Government Financed Health Care

• Much of the developed world made this choice a
  century ago
• How is it working?

38
Government Financed Health Care is better at
universal coverage
Source Karen Donelan et al., All payer, single
payer, Managed care, no payer Patients
perspectives in three Nations, Health Affairs
1996 15(2) 254-265
39
Canadians Wait for Health Care

• At any given time, more than 5 of all Canadians
  are on a waiting list from some kind of medical
  service
• Recent waiting times 5.5 months for heart
  bypasses, 5.7 months for hernia repairs, 7.3
  months for cholecystectomies, 6.4 months for
  hemorrhoidectomies, 8.3 months for varicose vein
  treatments, 3.7 months for hysterectomies, 7.1
  months for prostatectomies
• The waiting time for surgical referrals increased
  by 7.3 from 2002 to 2003

40
Specialty Care Suffers

• Outpatient surgery is discouraged
• CT scanners and advanced imaging is usually
  available only in hospitals
• Huge volumes of general practitioner visits are
  managed by rationing time spent, limiting access
  to diagnostic lab and x-ray
• In Canada, 50 of diabetics are undiagnosed. 5
  of diabetic diagnoses are made by optometrists
  observing retinal damage.

41
Government Financing Susceptible to Politics

• In British Columbia, residents of Vancouver and
  Victoria receive 37 more physician services per
  capita than other residents of the province,
  including 5.5 times more services from thoracic
  surgeons, 3.5 more services from psychiatrists,
  and 2.5 more services from dermatologists,
  anesthetists, and plastic surgeons.

42
Burgeoning International Market for Government
Health Care Refugees

• According to the Frasier Institute, 1.4 of the
  Canadian population leaves Canada to seek care
  elsewhere
• Border area U.S. providers attract many Canadian
  patients
• One half of the University of Washington Medical
  Centers In Vitro Fertilization patients are
  Canadian

43
Burgeoning International Market for Government
Health Care Refugees

• Other countries are also attracting refugees from
  government run health care
• Thailand treated 308,000 patients from abroad in
  2002
• Singapore treated 200,000 patients from abroad in
  2002 and aims to serve 1 million per year by 2010
• India treated 10,000 patients from abroad in 2002
  and estimates this will be a 1 billion business
  by 2012

44
Other Implications of Government Financed Health
Care

• Diminished research and development most medical
  advances originate in the United States
• Lower pay rates doctors in the U.S. earn an
  average of two to three times more than doctors
  in Canada.
• Canada has a net loss of about 500 doctors per
  year to the U.S.
• Canada has chronic labor problems as
  short-staffed and underpaid employees are
  dissatisfied

45
Other Implications of Government Financed Health
Care

• One Size Fits All is not good health care
• The Fosamax (alendronate) example
• Consumer choice is better

46
Davids Grades for Government Financed Health
Care

• Quality D (access to primary care avoids F)
• Affordability C (inefficient, but mechanisms for
  control)
• Universality A (covers everyone)
• Overall Grade C

47
Consumer-Driven Health Care

• The IRS has encouraged first dollar coverage and
  group health insurance
• First dollar coverage isnt insurance
• First dollar coverage creates a moral hazard
• First dollar coverage creates an impression that
  health care is free
• Group health insurance subsidizes unhealthy
  behavior
• True insurance is important
• In any given year, 20 dont use any health care
  while 1 consume 27 of health care resources

48
Consumer-Driven Health Care

• New IRS regulations are finally making it
  possible for many to buy true insurance (for
  catastrophes) and pay routine, predictable costs
  out-of-pocket
• MSAs
• HRAs
• HSAs

49
Changing Consumer Behavior

• Key Elements
• Individual Freedom
• Information
• Financial control
• Encourages rational consumption
• Works in markets everywhere
• Cash and Counseling Experience

50
Changing Provider Behavior

• Lasik eye surgery example
• Outcomes include
• Improved responsiveness to patients
• Less overbuilding of high tech
• Correcting the imbalance of health services
• Less emphasis on Big
• Patient options
• Optimizing cost, convenience, service, and
  specialization

51
The Reduction of Waste

• Reduction of administrative waste
• The dismantling of managed care
• 12 administration
• Disease management and other programs
• Provider overhead
• Efficiency is rewarded
• Reduction of clinical inefficiencies
• Doctor and patient relationship restored

52
Accomplishing Universality

• For the 180 million Americans who currently have
  private health insurance
• Allow pre-tax dollars to be put in HSAs
• Require that they maintain health coverage
• Impose tax penalties equal to premium cost if
  evidence of coverage isnt provided then
  automatically enroll them in the safety net
• Cost 0
• Savings Immense

53
Accomplishing Universality

• For the 43 million Americans who are currently
  uninsured
• Impose income sensitive tax penalties if they
  dont secure individual coverage
• Provide safety net option if they cant get
  their own coverage
• Make certain that safety net option
• Uses Cash and Counseling programs
• Builds cash value so theyll migrate to their own
  policies
• Cost Substantial (100 billion?)
• Savings Substantial

54
Accomplishing Universality

• For the 74 million Americans covered by Medicaid
  and Medicare
• Also enroll them in the safety net option (with
  same features as described above)
• Provide incentives similar to those for everyone
  else to secure individual coverage
• Cost 0 (already in federal budget)
• Savings Substantial (Many will migrate to their
  own coverage)

55
Davids Grades for Consumer-Driven Health Care

• Quality A (Extra credit for responsiveness)
• Affordability B (Consumers decide, but low
  income households lower grade)
• Universality B (Complex but doable)
• Overall Grade B

56
Summary

• There isnt any panacea
• Consumer-driven health care is optimal if Quality
  and Affordability are more important than
  Universality
• Challenge is how to get there from here