Report Cards, P4P, EMRs, and Disease Management

1
Report Cards, P4P, EMRs, and Disease Management

• An Analysis of Managed Care 2.0

2
The debate about quality has been corrupted in
two ways

• Quality problems have been exaggerated this is
  usually accomplished by confusing inferior
  quality with access barriers.
• Discussion of QI has been limited to those
  activities which plans can conduct (e.g.,
  financial incentives, report cards). QI which
  leaves out plans (e.g., public health, ending the
  nurse shortage) gets less attention.

3
Example of exaggeration of the quality problem

• Extensive research has documented that all
  three forms of clinical quality problems
  underuse, overuse, and misuse are ubiquitous in
  American medicine. (p. 166).
• Elise C. Becher and Mark R. Chassin, Improving
  the quality of health care Who will lead?
  Health Affairs 200120(5)164-179, 166.

4
Becher and Chassin offered this proof of
ubiquitous inferior quality

• A 1998 Rand literature review finding 30-40
  underuse and 20-30 overuse, and malpractice
  studies finding, 1 misuse.
• But a far more extensive Rand study (2003) found
  46 underuse and 11 overuse.
• Overuse and misuse obviously involve provider
  error. But underuse may not.

5
Rand reported 46 underuse and 11 overuse

• But Rand made no attempt to determine what
  caused underuse and overuse.
• Examples of Rand findings for diabetics
• 24 had A1c measured every six months
• 14 had annual eye exam
• 23 had urine protein checked annually
• 56 received dietary and exercise counseling
• 45 had follow-up visit every six months.

6
Researchers ignored underuse until late 1990s

• Most health services research to date has been
  directed at identifying and reducing excessive
  utilization. Little attention has been given to
  underuse of care.
• Two scholars at the RAND Corporation (R. L.
  Kravitz and M. Laouri, Measuring and averting
  underuse of necessary cardiac procedures A
  summary of results and future directions, Joint
  Commission Journal on Quality Improvement
  199723268-76).

7
Example of misuse of the 2003 Rand study
(conflating quality and access)

• Despite the extensive investment in
  developing clinical guidelines, most clinicians
  do not routinely integrate them into their
  practices. In a recent study of US adults,
  Elizabeth McGlynn and colleagues found that more
  than half did not receive the recommended
  care.
• Dan Mendelson and Tanisha V. Carino,
  Evidence-based medicine in the United States De
  rigueur or dream deferred? Health Affairs
  200524133-136, 134.

8
Another example of the misuse of the Rand study

• Research has shown that physicians incorporate
  the latest medical evidence into their treatment
  decisions 50 percent of the time (McGlynn et al,
  2003).
• US Department of Health and Human Services,
  Office of National Coordinator for Health
  Information Technology, The Decade of Health
  Information Technology Delivering
  Consumer-Centric and Information-Rich Health
  Care, July 21, 2004, 3.

9
Another example of misuse of the Rand study

• Physicians deliver recommended care only about
  half of the time. (citing McGlynn et al.)
• Richard Hillestad et al., Can electronic
  medical record systems transform health care?
  Potential health benefits, savings, and costs,
  Health Affairs 2005241103, 1110.
• This article, also by Rand scholars, was funded
  by the computer industry hailing the benefits of
  EMRs.

10
Rand facilitated misunderstanding our results
need no risk adjustment

• We primarily chose measures of processes as
  indicators, because they represent the activities
  that clinicians control most directly, and
  because they do not generally require risk
  adjustment.
• Elizabeth McGlynn et al., The quality of
  health care delivered to adults in the United
  States, New England Journal of Medicine
  20033482635-45, 37.

11
Outcome and process measures

• Outcome measures reflect changes in patient
  health. Examples mortality rates after surgery,
  cholesterol level, and ability to carry out
  activities of daily living.
• Process measures reflect how well providers
  comply with standards of care. Examples percent
  of children vaccinated, and percent of diabetics
  given eye exams.

12
Underuse is affected by factors outside physician
control

• No health insurance or insurance with pre-ex
  exclusions or out-of-pocket payments
• Other barriers (patient values, low income,
  illiteracy, immobility, transportation, daycare,
  change in residence or insurance).

13
Evidence that health insurance affects underuse
by diabetics

• Athough an estimated 35 percent of those with
  health coverage had received a blood glucose
  test, a cholesterol test, eye exam, foot exam,
  and influenza vaccination, just 14 percent of
  those without health coverage received the same
  set of services.
• US GAO, Managing Diabetes Health Plan Coverage
  of Services and Supplies, February 2005, 19.

14
Evidence that patient behavior affects process
measures

• Three-fifths of elderly Medicare beneficiaries
  who receive an appropriate recommendation for
  cholecystectomy fail to have it done
• half of insured patients who should, according to
  a stress test, have an angiogram do not get it
  and
• a fourth of insured patients who, according to
  their angiogram, should have angioplasty or
  bypass surgery receive neither.
• Sources SM Asch et al., Measuring underuse of
  necessary care among elderly Medicare
  beneficiaries using inpatient and outpatient
  claims, JAMA 20002842325-2333 (cholecystectomy
  bullet) PP Garg et al., Understanding
  individual and small area variation in the
  underuse of coronary angiography following acute
  myocardial infarction, Med Care 200240614-626,
  and M Laouri et al., Underuse of coronary
  angiography Application of a clinical method,
  Int J Qual Health Care 1997915-22 (angiogram
  bullet) LL Leape et al., Underuse of cardiac
  procedures Do women, ethnic minorities, and the
  uninsured fail to receive needed
  revascularization? Ann Internal Med
  1999130231-233, and M Laouri et al., Underuse
  of coronary revascularization procedures
  Application of a clinical method, J Am Coll
  Cardiol 199729891-897.

15
Patient refusal has been documented in studies of

• warfarin for atrial fibrillation,
• aspirin for heart attack,
• hypertension medication,
• vaccines for influenza and pneumonia,
• blood glucose tests,
• colorectal cancer screens, and
• radiation therapy for cancer.
• Sources SD Weisbord et al., Is warfarin really
  underused in patients with atrial fibrillation?
  J Gen Intern Med 200116743-749 J ONeil, A
  small step for womens hearts, New York Times,
  February 22, 2005, D6 BS Bloom, Continuation of
  initial antihypertensive medication after one
  year of therapy, Clin Ther 199820671-681 PR
  Dexter et al., Inpatient computer-based standing
  orders vs physician reminders to increase
  influenza and pneumococcal vaccination rates A
  randomized trial, JAMA 20042366-2371 VS
  Elliott VS, Researchers call for more diabetes
  testing, American Medical News, September 22/29,
  2003, 19 LC Walter et al., Pitfalls of
  converting practice guidelines into quality
  measures Lessons learned from a VA performance
  measure, JAMA 20042912466-2470 N Bickel et
  al., The quality of early-stage breast cancer
  care, Ann Surg 2000220-224..

16
(Patient refusal cont.)

• Patient refusal accounted for 59 percent of the
  underuse of colorectal cancer screens among
  Veterans Affairs patients.
• At a 2005 meeting of the American Heart
  Association, investigators reported on a study
  which found that doctors recommended aspirin on a
  daily basis to about 95 percent of women who had
  suffered heart attacks and stroke, but that only
  54 percent of the heart-attack patients and 43
  percent of the stroke patients complied with the
  recommendation.
• Sources Walter et al., op cit.(colorectal
  bullet) ONeil, op cit. (aspirin bullet)

17
Thus, current research permits us to say

• Overuse occurs 11 of the time and
• Misuse (malpractice) occurs lt1 of the time.
• Underuse due to provider failure occurs some
  unknown percent of the time.
• These figures reveal serious problems, but they
  do not add up to ubiquitous.

18
Exaggerating the problem of inferior providers
serves insurance industry

• Insurance industry has used the picture of inept
  providers to promote managed care.
• QI that does not assume inept providers and/or
  which insurance companies cannot do that is,
  which does not fall under the rubric of managed
  care gets much less attention.

19
Managed care is not the only way to improve
quality

• Other methods with more substantial evidence to
  support them include
• Ending the nurse shortage
• ending waiting times for emergency services
• insuring the uninsured and under-insured
• conducting public education campaigns re
  appropriate medical care and the effects of
  unhealthy behavior
• rolling back the excesses of managed care
• measuring and sharing performance results
  privately with providers
• conducting controlled trials and other forms of
  traditional research to find new treatments and
  to evaluate the efficacy of existing treatments.

20
Managed care has gone through two stages

• Managed Care 1.0 relied on
• financial incentives (capitation and bonuses),
  and
• utilization review and drug formularies.
• Managed Care 2.0 relies on
• report cards, which facilitate P4P, and
• disease management.

21
Definition of terms

• Report cards Any document purporting to measure
  the quality of care given by particular providers
  which is used to reward or punish providers.
• Pay for performance Any method of paying
  providers based on grades on report cards.

22
(Definitions cont.)

• Report card advocates propose that providers be
  rewarded and punished by
• market forces (plans, employers, and patients
  avoid low-scoring providers and patronize
  high-scoring providers), and/or
• pay for performance (insurers pay low
  scorers less, high scorers more).

23
DMAAs definition of DM

• Activities conducted by third parties that
• Identify people with certain diseases by
  examining their medical records or claims
• Rely on evidence-based practice guidelines
• Educate patients (may include surveillance)
• Measure processes and outcomes and report the
  results to patients and providers.
• Source Disease Management Association of
  America http//www.dmaa.org/definition.html,
  accessed February 9, 2006.

24
Another definition of DM

• Disease management is the latest catchphrase
  in the ever-evolving American health care
  spectacle. Disease management is a
  systematic, population-based approach to identify
  persons at risk, intervene with specific programs
  of care, and measure clinical and other
  outcomes.
• Thomas Bodenheimer, Disease management
  Promises and pitfalls, New Eng J Med
  19993401202-1205, 1202.

25
Report cards are now advocated simultaneously
with

• Interoperable electronic medical records (EMRs)
  (aka, regional and national health information
  networks) and
• Pay-for-performance methods of reimbursement in
  order to reward high scorers and punish low
  scorers.

26
Interoperable EMRs are advocated in order

• To facilitate collection of medical records on
  all Minnesotans/Americans all the time, and
• To risk adjust scores on report cards.
• Risk adjustment refers to the process of
  adjusting scores on report cards to reflect
  differences in patient health and other factors
  outside of provider control.

27
In sum, Managed Care 2.0 means

• (1) Report cards, which require
• interoperable EMRs and
• pay-for-performance methods of reimbursement and
• (2) Disease management.

28
Managed Care 2.0 appeared in the wake of the
failure of MC 1.0

• Events of the past year demonstrate beyond a
  doubt that managed care has failed and failed
  dismally. The greatest single ethical crisis
  facing American health care as we move into new
  year is what to do about it.
• Art Caplan, director of the Center for
  Bioethics at the University of Pennsylvania ("In
  2001, managed care our No. 1 health crisis,"
  MSNBC, December 21, 2001 http//www.msnbc.com/news
  /671464.asp, accessed December 23, 2001).

29
(Failure of MC 1.0 cont.)

• Managed care is basically over. People hate it,
  and it's no longer controlling costs. Health-care
  inflation is now back in the double digits. So if
  it's not saving money, then why should we have
  it? But like an unembalmed corpse decomposing,
  dismantling managed care is going to be very
  messy and very smelly, and take awhile.
• George Lundberg, former editor of JAMA who as
  recently as 1996 had co-authored an article
  defending managed care (Linda Marsa, Former JAMA
  editor laments the state of medical care, Los
  Angeles Times, March 26, 2001, http//www.latimes.
  com/print/health/200103 26/t000026016.html,
  accessed March 28, 2001).

30
MUHCCs position on report cards and
pay-for-performance

• Quality Report cards and P4P have not been shown
  to improve quality, and some research indicates
  they harm patients.
• Cost Report cards and P4P have not been shown to
  save money, and may raise costs.
• Small-scale report card and P4P experiments
  should be conducted report cards P4P should not
  implemented on a wide scale.

31
MUHCCs position on EMRs

• Quality EMRs may enhance quality in some clinics
  and hospitals. Evidence does not support the
  claim that making EMRs interoperable will improve
  quality.
• Cost Evidence does not support the claim that
  EMRs, with or without interoperability, will
  reduce cost.
• Providers should not be required by government,
  or given financial incentives financed by taxes,
  to buy EMR hardware and software.

32
MUHCCs position on disease management

• Quality DM has been shown to improve quality.
• Cost The evidence does not warrant the claim
  that DM will save money.
• Because DM can improve quality, research on
  effective means of DM should continue, and
  effective DM programs should be covered by
  insurance or delivered through public health
  agencies.

33
Report cards

• The following slides examine the claims made for
  report cards, pay-for-performance, and electronic
  medical records.

34
Governor claims report cards will improve
quality, reduce costs

• Rewarding providers for improved health
  outcomes and encouraging patients to use the best
  providers will not only help contain costs, it
  will improve the quality of care, Pawlenty
  said. (Governor Pawlenty unveils Smart Buy
  Alliance to slow health care costs and improve
  quality, press release, November 29, 2004,
  http//www.governor.state.mn.us, accessed
  November 30, 2004).

35
The Legislature claims report cards improve
quality, cut costs

• Minnesota Statutes Sec. 62J.43, signed by
  Governor Pawlenty on May 29, 2004, says
• To improve quality and reduce health care
  costs, state agencies shall encourage the
  adoption of best practice guidelines. The
  commissioner of health shall facilitate access to
  quality of care measurement information to
  providers, purchasers, and consumers by
  disseminating information on adherence to best
  practices care by physicians and other health
  care providers.

36
Governor-Legislature claims rely on three
assumptions

• (1) Report cards improve quality more often than
  they damage quality
• (2) Quality improvements inevitably lead to cost
  reductions
• (3) The cost reductions achieved by report cards
  will outweigh the cost of producing report cards.
  

37
There is little evidence that report cards
improve quality

• Despite extensive adoption of quality
  measurement and reporting, little research
  examines the effect of public reporting on the
  delivery of health care, and even less examines
  how report cards may improve care. The
  potential negative consequences of public
  reporting are largely unexplored.
• Rachel M. Werner and David A. Asch, The
  unintended consequences of publicly reporting
  quality information, JAMA 20052931239-44, 39.

38
Report cards could damage quality three ways

• (1) By being inaccurate (steering patients to
  inferior doctors)
• (2) By inducing doctors to reject sicker
  patients
• (3) By inducing doctors to shift resources from
  unmeasured to measured patients.

39
Report cards can be accurate for some things,
e.g., vacuum cleaners

• Consumer Reports report card on vacuum
  cleaners
• Offers grades on 38 vacuum cleaners on a
  five-point scale (from excellent to poor).
• 3 quality measures
• - cleaning (carpet, bare floors, w/ tools)
• - other results (ease of use, noise, emissions)
• - features (bag, brush, manual pile adj,
  weight)
• Kenmore (Sears) got 79 points, Sanyo Performax
  and Panasonic Fold NGo got 53

40
But patients are not floors, and doctors are not
vacuum cleaners

• Comparisons of quality are not useful if the
  playing field is not level, that is, if the
  conditions under which quality is measured are
  not the same.
• Keeping the playing field level is much easier to
  do while measuring the quality of vacuum cleaners
  than it is while measuring doctors and hospitals.

41
Many factors outside provider control influence
health outcomes

• Factors that influence health outcomes that are
  outside of provider control include
• Patient health status prior to treatment
• Patient insurance status (presence of
  deductibles and co-pays no coverage for service
  being measured no coverage at all)
• Patient income, education and values.

42
Failure to measure health status affects scores

• The next slide illustrates how scores on
  hospitals can be distorted when differences in
  patient health are measured only crudely. It
  shows that when stage of illness at admission
  was ignored, 18 of 65 hospital units scored above
  or below average, but when it was factored in,
  only 6 scored above or below average.

43
Hospital mortality rates vary depending on stage
of illness

• Hospital mortality rates for 13 hospitals and
  five conditions under HCFA and Green-Wintfeld
  Models
• Actual Mortality Rate HCFA Model Green-Wintfeld
  Model
• Above expected range 8 2
• Within expected range 47 59
• Below expected range 10 4
• Total 65 65
• Low-risk heart disease, severe acute heart
  disease, cancer, stroke, and pulmonary disease
• HCFA adjusted mortality rates for only a few
  of the factors that could have affected patient
  mortality that were outside hospital control
  (risk adjustment included age, sex, diagnoses
  other than the principal diagnosis, number of
  hospitalizations in the past 12 months, referral
  source (physician or nursing home), and urgency
  of admission (emergent, urgent, or elective)).
  Green-Wintfeld added to the HCFA adjusters an
  adjustment for stage of principal diagnosis at
  admission.

44
Income affects preventive services for insured
patients

• Lower SES socioeconomic status patients had
  lower compliance with Pap smears, mammograms, and
  diabetic eye exams, and were less likely to have
  a referral or make any office visit. These
  income effects are not confined to the poorest
  patients but span the entire socioeconomic
  spectrum.
• Peter Franks et al., Effects of patients and
  physician practice socioeconomic status on the
  health care of privately insured managed care
  patients, Medical Care 200341842-852, 842
• Patients were all insured by the same plan,
  described as the largest local managed care
  organization in the ten-county area surrounding
  Rochester, New York.

45
Quality-of-care scores for diabetics vary
depending on measure of quality

• (1) LDL cholesterol under 130 73
• (2) Measure (1) doctor has responded to high
• reading, patient has contraindications to
  statins 87
• (3) Measures (1) (2) other factors 90
• Other factors included patient refuses to
  take lipid-lowering medications lipid management
  low priority or difficult to address no primary
  care visit after high reading has active care
  elsewhere other interventions tried within six
  months of high reading (diet, exercise, or other
  lipid-lowering drug).
• Source Eve Kerr et al., Building a better
  quality measure Are some patients with poor
  quality actually getting good care? Medical
  Care 2003411173-1182.

46
Experts say risk adjustment of report card grades
is essential

• The interpretation of medical outcomes is
  further complicated by the need to make
  adjustments for comorbidity and the intensity and
  state of the patients illness a far from
  trivial undertaking. Paul Ellwood (Outcomes
  management A technology of patient experience,
  New England Journal of Medicine19883181549-1556)
  .
• The importance of co-morbidity must be
  stressed.... If co-morbidity is not considered,
  there will always be the potential for individual
  providers to be unjustly accused of poor
  quality because of patient selection. Richard
  W. Asinger, MD (Constructive use of clinical
  databases, The Medical Journal of Allina,
  1996(1)31-34, 32).

47
(Experts say risk adjustment is essential, cont.)

• Case-mix adjustments are made in almost all
  profile analyses to account for the differences
  in provider performance attributable solely to
  differences in the populations served (p. 764).
  Risk adjustments contribute vitally to reducing
  unfair profile evaluations (p. 765). Cindy L.
  Christiansen and Carl N. Morris, Improving the
  statistical approach to health care provider
  profiling, Ann Intern Med 1997127764-768.
• Accurate risk adjustment is necessary for
  observational and health services research,
  including comparison of outcomes of different
  treatments and quality assessment. Jay F.
  Piccirillo et al., Prognostic importance of
  comorbidity in a hospital-based cancer registry,
  JAMA 200429124241-47.

48
(Experts say risk adjustment is essential, cont.)

• We found that patient characteristics were
  315 times more important than hospital
  characteristics in predicting mortality after
  simple surgery, so small errors in risk
  adjustment may loom large compared to hospital
  differences.
• Jeffrey H. Silver and Paul R. Rosenbaum, A
  spurious correlation between hospital mortality
  and complication rates The importance of
  severity adjustment, Medical Care
  199735OS77-OS92, Supplement, OS87.

49
Unadjusted report cards damage access for sicker
patients

• Performance-based contracting gave providers of
  substance abuse treatment financial incentives to
  treat less severe OSA Office of Substance Abuse
  clients in order to improve their performance
  outcomes. Fewer OSA clients with the greatest
  severity were treated in outpatient programs with
  the implementation of PBC performance-based
  contracting.
• Yujing Shen, Selection incentives in a
  performance-based contracting system, Health
  Services Research 200338535-552, 535.

50
Even risk-adjusted report cards can damage access
for sicker diabetics

• We found that if those physicians with the
  worst profiles . . . for 1991 managed to
  discourage the patients with the top 5 of HbA1c
  levels (representing only 1-3 patients per
  physician) from returning to their panel, they
  would in most cases achieve a panel HbA1c profile
  in 1992 that would be substantially improved than
  average. . . . . Thus, the patients HbA1c levels
  from the previous year proved a far better
  predictor of what a patients HbA1c level would
  be in the current year, better than . . . our
  case-mix adjusters. Manipulating their patient
  pool, based on a patients prior year HbA1c
  level, is the easiest way for physicians to have
  a substantial improvement in their profile
• Timothy P. Hofer et al., The unreliability of
  individual physician report cards for assessing
  the costs and quality of chronic disease, JAMA,
  19992812098-2105, 2103 emphasis added.

51
New Yorks heart surgery report card

• First physician-specific report card
• Grades performance of hospitals and surgeons on
  heart surgery using 30-day mortality as quality
  measure
• Considered most accurate report card in America
• Has been more carefully examined that any other
  report card

52
New York heart surgery report card is the gold
standard

• New York States measurement and publication of
  coronary artery bypass graft (CABG) surgery
  mortality rates has emerged as a model in the
  campaign for useful performance data. The
  reality is that these measures of performance are
  the best available, and that substantial
  improvements are not likely for some years.
• Stephen F. Jencks, Clinical performance
  measurement -- a hard sell, JAMA
  20002832015-2016, 2015, 2016.

53
NY heart surgery report card is rigorously
adjusted

• 72 risk factors are adjusted
• They include
• number of coronary arteries occluded and degree
  of occlusion
• previous heart attack
• hemodynamic state just prior to surgery (ability
  to maintain blood pressure)
• chronic obstructive pulmonary disease
• kidney failure
• smoking history (last two weeks, last year)

54
NY report card is expensive

• The New York Department of Health pays for
• five full-time equivalent staff maintaining the
  database... and
• a utilization review agent to audit a sample
  of 50 cases from half the hospitals each year.
• The three dozen heart surgery hospitals in NY pay
  for
• data coordinators to collect and maintain their
  databases most hospitals have a full-time
  coordinator dedicated to this task.
• Source Edward L. Hannan et al., Public release
  of cardiac surgery outcomes data in New York
  What do New York state cardiologists think of
  it? Am Heart J 199713455-61, 62)

55
Results of 1998-2000 NY report card on 34 CABG
hospitals

• Statewide 30-day mortality average 2.32
• Three hospitals had higher-than-expected rates
• Two hospitals had lower-than-expected rates
• 29 hospitals had expected rates
• Source New York Department of Health, Adult
  Cardiac Surgery in New York State, 1998-2000,
  http//www. health.state.ny.us/nysdoh/heart/pdf/19
  98_2000) cabg.pdf, accessed January 16, 2005.

56
Results of 2000-2002 NY report card on 36 CABG
hospitals

• Statewide 30-day mortality average 2.27
• Three hospitals had higher-than-expected rates
• Three hospitals had lower-than-expected rates
• 30 hospitals had expected rates
• Source New York Department of Health, Adult
  Cardiac Surgery in New York State,2000-2002.

57
Outliers on 1998-2000 and 2000-2002 NY hospital
CABG reports

• 1998-2000 2000-2002
• High mortality rates
• Albany Med Ctr (4.08) Buffalo General (4.67)
• Ellis Hosp (6.13) Mount Sinai (4.86)
• Mount Sinai (6.01) NY Hospitals Ctr (4.31)
• Low mortality rates
• Lenox Hill (1.15) St. Josephs (0.90)
• Winthrop U Hosp (1.10) Staten Island (0.82)
• Vassar Brothers (0.00)

58
Rates for 1998-2000 NY hospital outliers two
years later

• 1998-2000 2000-2002
• Albany Med Ctr 4.08 2.83
• Ellis Hosp 6.13 3.29
• Mount Sinai 6.01 4.86
• Lenox Hill 1.15 2.02
• Winthrop U Hosp 1.10 2.78

59
Change in outlier status among 156
surgeons,1998-2000 to 2000-2002 report

• 156 surgeons met criteria for grading in
  1998-2000 report 21 (13) were outliers
• 14 had higher-than-expected mortality rates
• 7 had lower-than-expected mortality rates
• All 21 outliers were graded in 2000-2002 report,
  but in that period only 6 of these 21 were
  outliers
• Criteria were either 200 operations during
  this period, or at least one operation in each of
  1998, 1999, and 2000.
• Source Calculations by Kip Sullivan based on
  data in New York Department of Health, Adult
  Cardiac Surgery in New York State, 1998-2000,
  http//www. health.state.ny.us/nysdoh/heart/pdf/19
  98_2000) cabg.pdf, accessed January 16, 2005.

60
Study suggested New York report card improves
quality

• Odds of death from CABG surgery in
• NY relative to rest of US, 1994-1999 0.67
• Source Edward L. Hannan et al., Provider
  profiling and quality improvement efforts in
  coronary artery bypass graft surgery, Medical
  Care 2003411164-1172, Table 4, 1170 (subjects
  were Medicare beneficiaries risk adjustment was
  done with 12 adjusters from administrative data)

61
But the study in the preceding slide was poorly
done

• The study in the preceding slide is not
  credible because it examined mortality rates only
  among New Yorkers who underwent CABG surgery. The
  study did not attempt to determine if NY surgeons
  were refusing to perform surgery on sicker heart
  patients. The next several slides indicate that
  is what happened.

62
Recent studies find NY report card damages health
overall

• Our results show that report cards on heart
  surgeons led to increased expenditures for both
  healthy and sick patients, marginal health
  benefits for healthy patients, and major adverse
  health consequences for sicker patients. Thus, we
  conclude that report cards reduced our measure of
  welfare over the time period of our study (p.
  577). More severely ill patients experienced
  dramatically worsened health outcomes (p. 583).
  David Dranove et al., Is more information
  better? The effects of report cards on health
  care providers, Journal of Political Economy
  2003111555-588.

63
Reason NY report card induces surgeons to reject
sicker patients

• Mandatory reporting mechanisms inevitably
  give providers the incentive to decline to treat
  more difficult and complicated patients (p.
  581). Report cards led to a decline in the
  illness severity of patients receiving CABG in
  New York relative to patients in states without
  report cards (p. 583).
• David Dranove et al., Is more information
  better? The effects of report cards on health
  care providers, Journal of Political Economy
  2003111555-588.

64
(NY report card induces surgeons to reject sicker
patients, cont.)

• The December 19, 1991 Newsday article stated
  that several NY surgeons warned that some
  surgeons were turning down difficult cases to
  protect their statistics (p. 410). An article
  appeared in the New York Times entitled Faint
  hearts. As fate would have it, a woman was
  turned down for surgery because she had a fresh,
  large myocardial infarction. Her daughter was a
  reporter for the New York Times. After great
  difficulty, the daughter eventually found a
  surgeon who would operate on her mother (p.
  411).
• Bradley J. Harlan, Statewide reporting of
  coronary artery surgery results A view from
  California, J Thorac Cardiovasc Surg
  2001121(3)409-17.

65
(NY report card induces surgeons to reject sicker
patients, cont.)

• The incentive to refuse treatment for high-risk
  patients has created a kind of spiritual crisis
  in the field of cardiac surgery. Heart surgeons
  are shrinking from taking on the toughest cases
  because of statistics.
• Sandeep Jauhar (When doctors slam the door
  Under the current system, a doctors reputation
  may depend on his or her willingness to turn away
  a dying man, New York Times Magazine, March 16,
  2003, 30, 34).

66
Even the best surgeons dont trust the NY report
card

• There is nothing that separates me from the
  rest of the people on the list, Dr. Jeffrey
  Gold said. And even though Dr.Gold is ranked at
  the top of the 1994 report, he has qualms about
  it. Im concerned about the predictability of
  it, he said. I certainly would not use it as
  the sole way of selecting an institution or a
  surgeon.
• Elisabeth Bumiller (Death rankings shake New
  York cardiac surgeons, New York Times, September
  6, 1995, A1, B11)

67
New Yorks angioplasty report card is having a
similar effect

• An overwhelming majority of cardiologists 79
  in New York say that, in certain instances, they
  do not operate on patients who might benefit from
  heart surgery, because they are worried about
  hurting their rankings on physician scorecards
  issued by the state, according to a survey
  released today.
• Marc Santora, Cardiologists say rankings sway
  choices on surgery, New York Times, January 11,
  2005, A18.

68
Report cards cause resource shifts to services
being graded

• Although paying for high quality is an
  innovation with obvious potential benefits, it
  may also lead to the misallocation of
  resources. The medical director at one of
  Californias largest managed-care organizations
  described the problem succinctly 'Everybody's
  doing what they are required to do in responding
  to the quality measurements that are being used.
  Every ounce of energy is being diverted to
  responding to these not one ounce of energy is
  going to any other aspect of quality.
• Lawrence Casalino, The unintended consequences
  of measuring quality on the quality of medical
  care," New England Journal of Medicine
  19993411147-1150, 1147.

69
(NYs angioplasty report card, cont.)

• The patient population in the Michigan
  angioplasty registry had a significantly higher
  frequency of comorbidities. A case selection
  bias driven by the fear of public reporting of
  higher mortality rates in New York was one
  possible explanation .
• Mauro Moscucci et al., Public reporting and
  case selection for percutaneous coronary
  interventions, J Am Coll Cardiology
  2005451759-65.

70
(Report cards cause resource shift, cont.)

• If providers face a number of tasks and
  resources are limited, then effort will be
  allocated toward those tasks that are explicitly
  rewarded, taking resources away from other
  activities. Inevitably, ... the dimensions of
  care that will receive the most attention will be
  those that are most easily measured and not
  necessarily those that are most valued.
• Meredith B. Rosenthal et al., Paying for
  quality Providers incentives for quality
  improvement, Health Affairs 200423(2)127-141,13
  9.

71
(Report cards cause resource shift, cont.)

• From the present study which found HMOs were
  less likely to detect colorectal cancer early
  and the earlier breast cancer study which
  found HMOs were more likely to detect breast
  cancer early one can infer that the incentives
  of health plans are to allocate resources to
  those activities upon which they are measured.
  This suggests that preventive screening for
  conditions such as colorectal cancer that are not
  required to be in a report card (such as HEDIS)
  are more likely to be neglected.
• Anna Lee-Feldstein et al., Health care factors
  related to stage at diagnosis and survival among
  Medicare patients with colorectal cancer, Med
  Care 200240362-374, 374.

72
Example of a shift in resources triggered by
report cards

• It may seem that an optimal performance
  standard would be to maximize the percentage of
  patients who have an HbA1c lt7.0. Such a standard
  may divert a health systems attention from
  treating poorly controlled patients to
  disproportionately focusing on the larger numbers
  of patients who are slightly above this cutoff.
• Rodney A. Hayward et al., Quality improvement
  initiatives, Diabetes Care 2004 27 (Suppl.
  2)B54-B60, B56.

73
Reports on number of procedures do not pose risks
report cards do

• For a few procedures, evidence exists that
  quality is higher at hospitals that do high
  volumes of those procedures. Reports on the
  number of procedures do not create the three
  report card risks
• (1) Inaccuracy
• (2) Doctors avoiding sicker patients
• (3) Doctors shifting resources away from
  unmeasured to measured services

74
Practice makes perfect rule has been found for

• Treatment for AIDS (strong correlation)
• Pancreatic cancer surgery (strong)
• Esophageal cancer surgery (strong)
• Abdominal aortic aneurysm surgery (strong)
• Congenital heart disease surgery (strong)
• Coronary-artery bypass surgery (weak)
• Coronary angioplasty (weak correlation)
• Carotid endarterectomy (weak)
• Other types of surgery for cancer (weak)
• Some orthopedic procedures (weak)
• Treatment of low-birth-weight and premature
  babies (weak)
• Source Kenneth W. Kizer, The volume-outcome
  conundrum, New England Journal of Medicine
  20033492159-2161.

75
Review

• We have reviewed the first of three assumptions
  that have to be true in order for report cards to
  work that report cards improve quality of care.
  Report cards can damage quality three ways
• (1) By being inaccurate
• (2) by inducing providers to refuse to treat
  sicker patients (regardless of how accurate the
  report card is) and
• (3) by inducing providers and plans to shift
  resources away from unmeasured services.

76
We turn now to the last two assumptions about
report cards

• (2) Quality improvements inevitably lead to cost
  reductions
• (3) The cost reductions achieved by report cards
  outweigh the cost of producing report cards.

77
Quality improvement does not inevitably lead to
lower costs

• Although it's a widely held belief that
  quality health care leads to lower costs,
  insurers have no data that directly measures
  return on investment of their P4P
  pay-for-performance programs. Healthleaders
  (Paula DeWitt , The new incentive plan, March
  2004, http//www.healthleaders.com/magazine/cover.
  php? contentid53006, accessed April 10, 2004)

78
(Quality improvement does not lead inevitably to
lower costs, cont.)

• Results of this study show that it is possible
  to increase SFDs symptom free days in children
  with asthma. However, the improvements were
  realized with an increase in the costs associated
  with asthma care.
• Archives of Pediatrics and Adolescent Medicine
  (S.D. Sullivan et al., A multisite randomized
  trial of the effects of physician education and
  organizational change in chronic asthma care
  Cost-effectiveness analysis of the Pediatric
  Asthma Care Patient Outcomes Research Team II
  (PAC-PORT II), 2005159428-434, 428).

79
(Quality improvement and costs, cont.)

• Right from the start, it has been one of the
  great illusions that quality and cost go in
  opposite directions. There remains very little
  evidence of that.
• Donald Berwick, President and CEO, Institute
  for Healthcare Improvement (A deficiency of
  will and ambition A conversation with Donald
  Berwick, Health Affairs, Web Exclusive,
  January-June 2005, W5-1-W5-9, 7)

80
Report card infrastructure will be expensive

• To achieve an NHIN (National Health Information
  Network) would cost 156 billion in capital
  investment over 5 years and 48 billion annual
  operating costs or a total of about 400 billion
  over 5 years, or 2 of total spending.
• Note This is infrastructure only. The cost of
  grading thousands of services provided by
  hundreds of thousands of providers is extra.
• Rainu Kaushal et al., The costs of a National
  Health Information Network, Ann Int Med
  2005143165-73, 165

81
Report cards on providers suffer defects similar
to those on schools

• No Child Left Behind report cards on schools
  have been criticized for the same reasons
  provider report cards have
• They dont adjust for factors outside school
  control and are therefore inaccurate
• they shift resources away from unmeasured
  services and
• they are costly.

82
Bipartisan group concluded NCLB impedes quality
improvement

• The underlying problem is that all schools
  are measured equally, regardless of differences
  in socioeconomic factors or unique challenges
  the schools face (p 15). Schools are
  reluctant to accept transfers because they fear
  it would increase their chance of failing (p.
  22)
• National Conference of State Legislatures, Task
  Force on NCLB, Final Report, February 2005.

83
Governor assumes Alliance can measure quality
accurately

• The Smart Buy Alliance will adopt uniform
  methods of measuring quality of care and will
  purchase health care based upon those
  measurements. Consumers and purchasers cannot
  make good decisions in the marketplace without
  access to easy-to-understand information about
  health care ... quality. The Alliance will
  require health plans and providers to participate
  in efforts to make such information available.
  The Community Measurement Project is an
  example of the type of information to be made
  available. (Governor Pawlenty unveils Smart
  Buy Alliance to slow health care costs and
  improve quality, press release, November 29,
  2004, http//www.governor.state.mn.us, accessed
  November 30, 2004).

84
Diabetes quality measures, Community Measurement
Project

• None of these measures is risk-adjusted
• (1) patients with HbA1c less than or equal to
  8.0 (and 7.0) OUTCOME
• (2) patients with LDL-cholesterol less than 130
  (and 100) OUTCOME
• (3) patients with blood pressure less than
  130/85 (and 130/80) OUTCOME
• (4) patients over age 40 taking aspirin
  PROCESS
• (5) patients known to be nonsmokers OUTCOME
• (6) patients with annual screening for kidney
  and eye complications PROCESS
• (7) A composite of the first five measures
• An outcome measure is one that measures the
  effect of treatment on patient health. Survival
  after surgery is an example of an outcome
  measure. So too is reported pain level in
  arthritis patients following drug treatment. The
  outcome measures shown above are sometimes called
  intermediate outcome measures because they are
  not equivalent to absence of disease but are
  rather physiological indicators that serve as
  rough predictors of health in the future. A
  process measure is one that measures how
  frequently doctors complied with a recommended
  process, such as taking blood pressure or
  administering beta blockers after a heart attack.
• Source Gail M. Amundson and John Frederick,
  Medical group quality data a reality,
  MetroDoctors The Journal of the Hennepin and
  Ramsey Medical Societies, January/February 2004,
  17-19.

85
HMO advocates have called for report cards for 35
years

• A performance reporting system of proven
  reliability would be developed and installed to
  provide both individual consumers and quantity
  buyers (e.g., HEW) with accurate information on
  the comparative performance of alternative
  sources of health care. (HMOs would be required
  to make such information available.)
• Paul M. Ellwood et al. (Health maintenance
  strategy, Medical Care 19719291-298, 297).

86
(HMO advocates have called for report cards,
cont.)

• The development of an effective system of
  collecting and disseminating data on quality and
  outcomes is an essential component of a health
  care reform strategy. Such a strategy will allow
  the monitoring of the impact of cost containment
  initiatives on health care quality. . . . The
  Commission and the Commissioner of Health will
  work collaboratively to collect and disseminate
  comparative data on the quality of services
  provided by providers, health plans, and ISNs in
  order to facilitate competition and continuously
  improve systemwide health care quality.
• Minnesota Health Care Commission (Containing
  Costs in Minnesotas Health Care System A Report
  to Governor Arne H. Carlson and the Minnesota
  Legislature, January 25, 1993, 28).

87
High-deductible advocates also call for report
cards

• Consumer-directed health care supposes a new
  formulation one driven by consumers with
  cash-in-hand, demanding to know for themselves
  who is the best urologist in town, how do I get
  the most value for the money Im spending?
  Information systems to support this movement will
  grow exponentially. But the information ... is
  not an end to itself. The real revolution will
  come when health-care consumers use that
  information to reward higher quality and punish
  the mediocre.
• Greg Scandlen, Galen Institute (How
  consumer-driven health care evolves in a dynamic
  market, Health Services Research
  2004391113-1118, 1117)

88
But accurate report cards are almost nonexistent

• "We have no assurances that the competition of
  health plans . . . will reward those who
  deliver higher quality care. . . . Purchasers
  and consumers have not, so far, rewarded or
  punished plans based on quality. . . . If
  purchasers and consumers had tools that allowed
  them to buy on quality, ... the thinking that lay
  behind the original HMO movement may still play
  out"
• Paul M. Ellwood, Jr. and George D. Lundberg,
  ("Managed Care A Work in Progress," Journal of
  the American Medical Association
  19962761083-1086, 1085).

89
(Accurate report cards are almost nonexistent,
cont.)

• Physician profiles are not and may never be
  ready for public consumption.
• Andrew Bindman,Can physician profiles be
  trusted? JAMA 1999281 2142-2143, 2143)

90
(Accurate report cards are almost nonexistent,
cont.)

• Hospital profiling remains an unproven strategy
  for improving outcomes of care.
• David W. Baker et al., Mortality trends during
  a program that publicly reported hospital
  performance, Medical Care 200240879-90, 879.

91
Quality can be improved without report cards

• The Cooperative Cardiovascular Project induced
  large improvements in quality of care of heart
  attack patients in four pilot states by giving
  doctors feedback (at the hospital, in seminars,
  by phone, and by mail).
• Improvements included increased use of aspirin
  (84 to 90) and beta blockers (47 to 68), and
  reduced one-year mortality (32.3 to 29.6).
• Source Thomas A. Marciniak et al., Improving
  the quality of care for Medicare patients with
  acute myocardial infarction Results from the
  Cooperative Cardiovascular Project, JAMA
  199821791351-1357

92
(Quality improvement without report cards, cont.)

• Other methods of improving quality without
  report cards include
• (1) Traditional research
• (2) Establishing universal health insurance
• (3) Reducing drug prices
• (4) Ending the nurse shortage
• (5) Public health programs.

93
Electronic medical records (EMRs)

• The following slides demonstrate that the
  evidence does not support the claim that
  interoperable EMRs will improve quality or reduce
  costs.

94
Advocates claim EMRs can do it all

• By computerizing health records, we can avoid
  dangerous medical mistakes, reduce costs, and
  improve care.
• George W. Bush, State of the Union Address,
  January 20, 2004 (quoted in Rainu Kaushal et al.,
  The costs of a National Health Information
  Network, Ann Int Med 2005143165-173, 165).

95
(Advocates claims re EMRs cont.)

• It is widely believed that broad adoption of
  electronic medical records (EMR) systems will
  lead to major health care savings, reduce medical
  errors, and improve health.
• Richard Hillestad et al., Can electronic
  medical record systems transform health care?
  Potential health benefits, savings, and costs,
  Health Affairs 20051103-1117, 1103.

96
Proponents make three claims

• (1) EMRs save time
• (2) EMRs improve doctors decisions
• (3) EMRs facilitate the production of report
  cards which in turn improve quality.
• None of these claims have been proven.

97
EMRs have not been shown to save time for
providers

• With the exception of pharmacy settings, there
  is little consistent evidence that IT
  information technology systems save time for
  providers. In some instances, the literature
  suggests the reverse.
• Medicare Payment Advisory Commission (Report to
  Congress New Approaches in Medicare, June 2004,
  163)

98
(EMRs dont save time, cont.)

• Only 13 of 100 trials evaluated the impact
  of the CDSS clinical decision support systems
  on clinician workflow, with more than half of
  these CDSSs requiring more time and effort from
  the user compared with paper-based methods.
• Amit X. Garg et al., Effects of computerized
  clinical decision support systems on practitioner
  performance and patient outcomes A systematic
  review, JAMA 20052931223-1238, 1226.

99
EMRs have not been shown to improve health

• Fifty-two trials of clinical decision support
  systems assessed patient outcomes . Only 7
  trials reported improved patient outcomes.
• Amit X. Garg et al., Effects of computerized
  clinical decision support systems on practitioner
  performance and patient outcomes A systematic
  review, JAMA 20052931223-1238, 1231.

100
(EMRs and health, cont.)

• In 2001, the Agency for Healthcare Research and
  Quality determined that 14 safety practices had
  greater strength of evidence regarding their
  impact and effectiveness than any practice which
  relied on IT. These include such low-cost items
  as appropriate provision of nutrition and use
  of maximum sterile barriers while placing central
  intravenous catheters to prevent infections.
• Medpac (Report to Congress New Approaches in
  Medicare, June 2004, 162)

101
Some studies report harm done by computers

• We found that a widely used CPOE computerized
  physician order entry system facilitated 22
  types of medication error risks. Examples include
  fragmented CPOE displays that prevent a coherent
  view of patients medications, pharmacy inventory
  displays mistaken for dosage guidelines, and
  inflexible ordering formats generating wrong
  orders.
• Ross Koppel et al., Role of computerized
  physician order entry systems in facilitating
  medication errors, JAMA 20052931197-1203.

102
NHIN advocates favorite studies are opinions,
not evidence

• Two papers cited frequently by EMR advocates
• Richard Hillestad et al., Can electronic
  medical record systems transform health care?
  Potential health benefits, savings, and costs,
  Health Affairs 20051103-1117, 1103.
• Jan Walker et al., The value of health care
  information exchange and interoperability,
  Health Affairs Web Exclusives, January-June 2005
  24, Suppl. 1)W5-10-18.

103
Hillestad et al.

• Conclusion Fully standardized HIEI health
  care information exchange and interoperability
  could yield a net value of 77.8 billion per
  year.
• According to an accompanying paper, savings
  would amount to 1.6 percent of health spending in
  2019 (Clifford Goodman, Do it for the quality,
  1125)

104
(Hillestad et al. cont.)

• Authors are part of the Rand HIT Project.
• Funded by Cerner, GE, Hewlett-Packard, Johnson
  and Johnson, and Xerox.
• Their methods were extraordinarily biased
• The currently useful evidence is not robust
  enough to make strong predictions, and we
  describe our results only as potential.
• We chose to interpret reported evidence of
  negative or no effect of HIT as likely being
  attributable to ineffective or not-yet effective
  implementation.

105
Walker et al.

• Conclusion Net savings from national
  implementation of fully standardized
  interoperability between providers and five other
  types of organizations could yield 77.8 billion
  annually, or approximately 5 percent of the
  projected 1.661 trillion spent on US health care
  in 2003 (W5-10)

106
(Walker et al. cont.)

• Funded by the Foundation for the eHealth
  Initiative, which is funded by the computer and
  insurance industries among others.
• We convened a panel of nationally known
  experts. With relatively little research and
  literature on the value of HIEI health care
  information exchange and interoperability, the
  panelists played an important role.

107
Disease management

• The following slides demonstrate that disease
  management (DM) is promoted by insurance
  companies and DM vendors, and that the evidence
  does not support the claim that disease
  management will reduce health care costs.

108
Disease Management Association of Americas
board, 2006

• Lifemasters Jefferson Medical College
• Wellpoint Dept of Mental Health, TN
• Geisinger Health Plan American Healthways
• McKesson Health Solutions Air Logix
• Matria Healthcare Magellan Health Services
• Caremark Rx Sanofi-Aventis
• Fibrogen Kaiser Permanente
• Pitney-Bowes American College of
• Astra-Zeneca Pharmaceuticals Cardiology

109
DM was begun by the drug industry

• The boom in DM was initiated by the
  pharmaceutical industry. By 1995, most
  pharmaceutical manufacturers had unveiled a
  variety of DM programs. Merck-Medco Managed
  Care sells its diabetes DM program to
  employers and plans , identifying patients
  with diabetes through its 51-million-person
  pharmacy data base.
• Thomas Bodenheimer, Disease management
  Promises and pitfalls, New Eng J Med
  19993401202-1205, 1202.

110
No evidence that disease management saves money

• On the basis of its examination of
  peer-reviewed studies of disease management
  programs, CBO finds that to date there is
  insufficient evidence to conclude that disease
  management programs can generally reduce the
  overall cost of health care services.
• Congressional Budget Office (An Analysis of the
  Literature on Disease Management Programs,
  October 13, 2004, http//www.cbo.gov/showdoc.cfm?i
  ndex5909sequence0, accessed September 25,
  2005)

111
(DM doesnt cut costs, cont.)

• Although interest in disease management
  programs is growing, evidence of their clinical
  and cost effectiveness remains limited. Without
  many attractive alternative mechanisms to control
  costs, many employers are adopting disease
  management despite the lack of evidence.
• Center for Studying Health System Change
  (Ashley Short et al., Disease management A leap
  of faith to lower-cost, higher-quality health
  care, October 2003, Issue Brief No. 69, 3)

112
(DM doesnt cut costs, cont.)

• Despite high expectations, evidence of both
  disease management and case management programs
  success in controlling costs and improving
  quality remains limited.
• Center for Studying Health System Change
  (Ashley Short et al., Disease management A leap
  of faith to lower-cost, higher-quality health
  care, Issue Brief No. 69, October 2003).

113
(DM doesnt cut costs, cont.)

• A growing number of DM programs offer to
  monitor patients with chronic conditions and help
  avoid dangerous complications. But the long-term
  cost effectiveness of such programs has been hard
  to measure. ... There is a chance DM programs
  could actually raise costs.
• Wall Street Journal (Laura Landro, Does
  disease management pay off, October 20, 2004,
  D4).

114
(DM doesnt cut costs, cont.)

• Weve made real progress in keeping people
  healthier who have chronic illnesses, says
  Edward Wagner with Group Health Cooperatives
  Center for Health Studies in Seattle. But we
  still dont know definitively what the economic
  impacts of disease management are. Dr. Wagner
  expresses skepticism about outsourced
  disease-management programs.
• Wall Street Journal (