Challenges to Quality and Quality Measurement

1
Challenges to Quality and Quality Measurement

• Randall D. Cebul, M.D.
• Center for Health Care Research Policy
• Professor of Medicine
• Case School of Medicine at MetroHealth Medical
  Center
  
• March 31, 2004

2
Institute of Medicines Definition of Quality

• The degree to which health services for
  individuals and populations increase the
  likelihood of desired health outcomes and are
  consistent with current professional knowledge.
• Institute of Medicine. 2001. Crossing the
  Quality Chasm.

3
Why Should We Care? Selected Indicators(http//ww
w.iom.edu)

• 44-98K Americans die from medical errors
  annually
  
• Medical errors kill more people annually than
  breast cancer, AIDS, or MVAs
  
• Hospital medication errors cost 2B annually
• Lag between demonstration of effectiveness and
  incorporation in care 17 years
  
• Only 55 of adults receive recommended care
  little difference between preventive or chronic
  care
  
• 18K Americans die from heart attacks annually
  because they do not receive appropriate
  preventive Rx
  
• 43M uninsured Americans have consistently worse
  outcomes, and are more likely to die prematurely,
  than insured Americans.

4
Quality Language and Acronyms and Related
Approaches to Measurement

• Structure Process Outcome (Donabedian)
• Patient Safety/Medical Errors/Adverse Events
  (IOM)
  
• EMR Electronic medical records
• TRIP Translating Research into Practice
  (AHRQ), EBM Evidence-based Medicine (Sackett)
  
• CQI Continuous Quality Improvement
  (IHI/Berwick)
  
• Report Cards (HEDIS, CHQC)
• Risk-adjusted Outcomes
• Volume-outcome/Practice Makes Perfect (Leapfrog)

5
Case A

• 45 y.o. M seen in follow-up of ED visit for
  recent onset breathlessness, found to have CHF
  and poorly controlled DM and HBP.
  
• He was last seen in clinic 3 years ago, when his
  diabetes was complicated only by mild
  proteinuria and he was begun on a beta blocker
  for HBP
• Laboratory A1C 7.0 Urinary microalbumin 50
  Creatinine 1.2
  
• In the interval, he lost his job at LTV and his
  health insurance, and could not pay for his
  medications.

6
Case A

• Findings from the ED Visit
• CHF (CXR, echocardiogram, BNP)
• DM poorly controlled (A1C 12.0)
• Blood pressure poorly controlled (180/110)
• Chronic renal insufficiency (creatinine 4.0, K
  5.3)
  
• Medications were restarted and he was referred to
  urgent care clinic.
  
• Quality of care issues and consequences were
  discussed with the resident.
  

7
Case A Quality MomentsTRIPLag, Uninsurance

• TRIP Lag We dont Practice what we Publish
• 2001 not begun on ACE Inhibitor (TRIP)
• 2004 too late to start harmsbenefits
• Lack of health insurance is a barrier to access,
  and therefore is a quality problem
  
• Discontinuity in care d/t loss of health
  insurance
  
• Reduced functional status and shortened life
  expectancy
  
• Steep cost trajectory

8
Case B

• 48 y.o. F presents for an exacerbation of asthma
  precipitated by bronchitis. She is a 2 ppd
  cigarette smoker.
  
• She has health insurance with a highly regarded
  HMO that does not cover smoking cessation-related
  medications.
  
• Conversations were held with the HMO.

9
Case B Quality MomentsBusiness Matters and
Quality

• Optimal Quality is Inhibited by Business
  Matters
  
• MD Adverse Selection
• Cigarette smokers cost more other HMOs do not
  cover smoking cessation we would get all the
  smokers premiums would have to increase
  
• Admin Time horizon Free Rider
• Enrollee turnover
• ROI whose investment, whose return?
• The free rider
• HMO 1 invests in prevention
• Patients insurance transfers to HMO 2
• Benefit of prevention accrues to HMO 2

10
Case C

• 36 y.o F admitted moribund with hemorrhagic
  stroke due to ruptured aneurism.
  
• She had previously discussed with her husband and
  parents her desire not to undergo heroic measures
  should something calamitous like this happen to
  her.
• Her husband represented her views to the hospital
  at the time of admission and Do Not Resuscitate
  orders were written.
  
• She died peacefully on the 3rd day.

11
Case C Quality MomentsChallenges in Using
Mortality

• All Deaths are Not Alike
• Death with dignity/patient choice should not be
  counted against provider hospital or physician
  in report cards
  
• Risk-adjustment Typically Does Not Reflect This
  
  
• 20 of all stroke patients in Cleveland are DNR
  by day 2.
  
• 60 of All stroke deaths by 30 days are among
  Early DNR patients

12
Case D

• 75 y.o. patient with CHF received her flu vaccine
  at her local elder center.
  
• Is there a quality/measurement problem?

13
Case D Quality MomentMissing Data

• Absence of Documentation is Not the Same as
  Documentation of Absence
  
• Out-of-system utilization is a problem for
  quality measurement.
  
• Typical problems
• Flu vaccine among the elderly (Missed good
  behavior)
  
• Patients readmitted to Hosp B after complication
  associated with care in Hosp A (Missed bad
  outcome)

14
Case E

• The Leapfrog Group (www.leapfroggroup.org) and
  other Business Coalitions are Motivating
  (Evidence-based) Hospital Referrals on Volume
  Thresholds

15
Volume-Mortality Relationship (p- Cleveland Significant, but Noisy, Relationship
16
Case E Quality MomentVolume Thresholds are Not
Supportable

• Volume-Outcome Links are Associations
• Often noisy Thresholds difficult to identify
• Not clearly causally linked
• Fail to consider implications at local levels
• Low volume but good hospitals may be put out of
  business
  

17
Take Home Messages

• The medium is the message what we measure (or
  dont measure) may matter
  
• Lack of Access is a quality problem
• Measuring risk-adjusted outcomes seems easy,
  but its not
  
• Measuring process seems hard, and it is
• Volume doesnt tell much about quality of care
• Systems changes are costly, but may be worth it