The Need to Broaden Our CV Quality Metrics

1
The Need to Broaden Our CV Quality Metrics

• Eric D. Peterson, MD, MPH
• Director of CV Outcomes Research and Quality
• Duke Clinical Research Institute (DCRI)
• Associate Vice Chair of Quality for DOM
• Duke University Medical Center

2
Last Years AHQA Talk Ending Business as Usual
in ACS Care

• Documented the wide gulf between ACS Guideline
  recommendations and community practice
• Paradoxical care
• Significant care disparities
• Providers variability (MD specialty, hospital)
• Correlated process adherence with patient outcome
  
• Provided examples of collaborative QI projects
• Demonstrated that QI workshospitals that
  improved care overtime had progressively better
  outcomes

3
This Year Key Points

• Begin by setting the stage
• Propose expanding scope of quality assessment
• Improve existing performance metrics
• Need for balancing and goal-based metrics
• Importance of patient safety
• Moving beyond the hospital
• Transitional and long-term care
• Patient compliance and involvement in care
• Appropriateness and cost issues
• Impending procedure and imaging crisis
• End with few thoughts on potential roles of QIOs

4
Where Are We in 2005?

• CMS
• Faces the elderly explosion
• Need to constrain costs while assuring quality
• MDs
• Know that budget caps coming
• P4P seems like a acceptable option if done fairly
• QIOs
• Ever expanding expected duties
• Increased pressure to prove worth

5
The Future of Medicare Spending

• Medicare enrollment
• 2002 40 million
• 2030 77 million N Engl J Med 2001344928-31
• Workerbeneficiary ratio
• 2002 4.0 1
• 2030 2.3 1 WWW.whitehouse.gov
• Healthcare Costs
• 2005 1.8 Trillion
• 2014 3.6 Trillion (19 of US economy!)
• USA Today Feb 24, 2005

6
Annual Report to Congress, the Medicare Board of
Trustees

• The financial status of the fund has
  deteriorated significantly, with asset exhaustion
  projected to occur in 2019.
• Projected Medicare costs would exceed those for
  Social Security in 2024

-- March 23, 2004
7
These Issues Create a Imperative for
Experimentation

• Public reporting
• Pay for Performance
• Promotion of IT Infrastructure
• Coverage under Protocol
• LVAD, AICD

8
Healthcare Consumerism Driving Public Reporting
9
Rise in Public Consumerism in Medicine
10
(No Transcript)
11
Lessons from Initial Efforts in Public Reporting
PAs CV Consumers Guides

• History
• 1992 PA CABG consumer guide
• 1996 1st AMI consumer guide
• Survey of physicians
• 10 reported discussing Guide with pts
• 87 Guide had minimal effect on referral
• 60 felt since guide, increased difficulty
  getting tough cases operated on

Schneider, Epstein NEJM 1996335251-6
12
Will Patient Consumerism Drive Quality?

• Care choice often made under urgent or emergent
  conditions
• Many markets lack local competitors
• Data often not accurate or understandable
• Patients often not empowered
• Selection dictated by referring MD or payer
• And both have potential conflicts of interest

13
Pay for Performance Movement

• 35 health plans covering more than 30 million US
  patients have programs tying performance with
  bonus payments
• Mark McClellan CMS Director, suggests that Pay
  for Performance based compensation will account
  for 20-30 of physician compensation in the next
  5 years

Wall Street Journal Sept 17th, 2004
14
Do Pay Incentives Work?Experience from Hawaii
Medical Services Association
15
CMS Oct 2002-2003
www.cms.gov
D/C Instructions 37 LVF Assessment 99.4
16
Unintended Consequences of Incentive Systems

• As with any program designed to bring about a
  certain change, unintended consequences present a
  worry Paul Shekelle
• (Financial) incentives for a team backfires,
  creating competition and depressing morale,
  output and quality Edward Deming

17
Potential Unintended Consequences of P4P

• Gaming ones results
• Doctoring charts not patients
• Avoiding difficult care cases
• Care disparities may actually increase
• Major costs associated with profiling
• Provider IT, data collection, QA personnel
• Profiler Data harvest/analysis, reporting,
  audits
• This is NOT going to improve patient care
• Test taking mentality
• Underemphasize non-targeted care processes

18
Potential Unintended Consequences (2)

• Safety concerns
• Can lead to rushed-initiation and over-treatment
• Doesnt emphasis achieving treatment targets
• Are drugs titrated? Goals achieved?
• Loss of innovation
• May actually inhibit adoption of new therapies
• Leaving those most in need behind!
• Might actually accentuate gulf between have and
  have not hospitals

19
Promotion of IT as the Answer to Quality

• A growing revolution is transforming the
  everyday practice of medicine. Owing more to
  laptops than lab coats, this is an information
  revolution that with forever change the way
  doctors make decisions.

Millenson, ML. Demanding Medical Excellence
Doctors and Accountability in the Information
Age 1997
20
Major Issues for IT Adoption

• 7-10 years to achieve wide-scale adoption
• Slow process due to
• Small providers reap little to no benefit
• General reluctance to make change
• Lack of data integration
• Other barriers to implementation pt. privacy
• Total cost 21.6 43.2 billion
• Dr. Brailers recent budget 0

SOURCE Connecting for Health. Financial, Legal
and Organizational Approaches to Achieving
Electronic Connectivity in Healthcare. Markle
Foundation. October 2004
21
Even When Here, IT Only Part of Solution

• Significant efforts will be needed to convert
  data into useable media for measurement/change
• Integration of data across care setting
• Meaningful, timely provider feedback
• Assisted by decision support tools
• Integrated with other QI efforts
• And avoid information overload

22
Moving to Improve CV Quality in This New Era

• Improve existing performance metrics
• Broadening view of quality
• Patient safety
• Longitudinal care issues
• Promoting a Culture of Quality
• Encouraging patient involvement in process
• Appropriateness, CE care

23
Characteristics of Ideal Performance Indicator

• Evidence based
• Measurable
• Large net impact on public health
• Effect of treatment on outcome
• Patients eligible for measure
• How large is gap in care?
• Consider whether measure should be assessed as
  process or targeted outcome
• Consider need for balancing metric

24
Are We Emphasizing the Right Issues?CMS CORE
Quality Indicators MI

• Aspirin at Arrival
• Aspirin at Discharge
• Thombolytic received lt30
• PCI received lt120
• Beta-Blockers at Arrival
• Beta-Blockers at Discharge
• ACE Inhibitor for LV Dysfunction
• Smoking Cessation

25
Use of Reperfusion Therapies in Eligible Patients
RTreperfusion therapy
Barron HV, et al. Circulation 1998
26
Door to Balloon Times By Transfer Status
228
NRMI Transfer-In Patients
Minutes (Median)
171
NRMI Non-Transfer-In Patients
111
100
27
Need for Balancing Metrics

• Inappropriate Treatment
• e.g., CHF pts receiving antibiotics lt4 hrs
• Adverse events
• e.g., CABG pts w/ early extubation requiring
  re-intubation
• Over-testing
• of CHF pts w/ documented low EF who receive
  second echo study in 6 months
• Diagnostic Yield
• of clean cardiac catheterizations

28
Emphasize Process or Target?

• Smoking Cessation
• e.g., receiving advice or quitting
• Primary PCI
• e.g., time to balloon or procedural success
• Lipid Lowering or Blood Pressure control
• e.g., prescribed drug or at goal
• CHF Instructions
• e.g., given weight and dietary instructions vs
  3 month readmission rate

29
Building a Culture of Quality Flexibility in
Selection of QI Metrics?

• A central tenant of QI is that involved parties
  engage in selection of QI goals, yet current
  system imposes external metrics
• If a state QIO already succeeding in area, should
  they have flexibility to look at different
  issues?
• Similarly, if hospital is above threshold for
  basics, shouldnt they self define novel QI
  metrics?
• and be rewarded based on their success in
  changing their prospectively defined areas?

30
Broadening Our Quality Metrics Safety To err
is human

• 17 year old girl, end-stage restrictive
  cardiomyopathy
• Blood type O
• Received heart-lung transplant from type A donor
• Redo heart-lung transplant February
• Died February 22

31
Medication Errors on Cardiology Service
Allen LaPointe N, Archives Int Medicine
20031631461-1466
32
Need New Paradigm for SafetyActive
Preparation

• Terrorist treats stimulated a new strategy
• Proactively considering potential bad events,
• Develop systems to prevent,
• Then actively test system
• Example Airline security
• Cant we proactively prepare for safety concerns
  in medicine?
• Identify Top 20 safety issues
• Hospitals could develop different systems to
  prevent
• Then develop scenarios to test robustness of
  system

33
Quality and Safety Merge Bleeding Risks in ACS
Data Through Q2 2004 CRUSADE data (n74,271)
Excluded CABG, transfer outs, missing data
34
Non-CABG Related Transfusions by Number of Drugs
and Age
Data Through Q2 2004 CRUSADE data (n74,271)
Excluded CABG, transfer outs, missing data
35
Appropriate Dosing of Acute Medications

• UF Heparin
• Use weight-based dosing
• Bolus 60-70 U/kg ? Infusion 12-15 U/kg/hr
• LMW Heparin Enoxaparin
• Use weight-based dosing
• ? to 1 mg/kg SC q24 hr, if CrCl lt 30 cc/min
• GP IIb-IIIa Eptifibatide
• ? Infusion to 1.0 ug/kg, if CrCl lt 50 cc/min
• GP IIb-IIIa Tirofiban
• ? Bolus to 6 ug/kg, if CrCl lt 30 cc/min
• ? Infusion to 0.05 ug/kg/min, if CrCl lt 30 cc/min

Dosing information collected in CRUSADE
beginning Q1 2004
36
Excessive Dosing of Antithromboticsin ACS
Patients
Q1-Q2 2004 CRUSADE data
37
Association of Dosing Combinations Heparin GP
IIb/IIIa and Transfusion
Among patients receiving both Heparin and GP
IIbIIIa
38
The Train Speeds Up.Shrinking In-hospital ACS
Care
39
Transitional ACS Care Not missing the steps

• In
• Lab
• Revasc?
• Other Rx?

• Hospital
• ED
• Admit?
• CCU?
• Transfer?

• Pre-
• Discharge
• Right meds
• Right pt
• Education

• 3-Mo Eval
• Re-assess EF
• Lipids at goal?
• On right meds?
• On right dose?
• Depression?
• Other risks
• addressed?

• CCU
• Acute Cath?
• Tx to Floor?

40
Long-term Medical Adherence The Duke Experience
Patients taking Medication ()
Baseline
95
86
65
55
1 Year
78
67
47
40
Patients taking med at baseline (Green Bars), and
at one year (Orange Bars).
41
(No Transcript)
42
Friday, May 14, 2004
Weight-loss strategy pays off

• Local MD, frustrated after years of failing to
  help patients lose weight, offers to pay them out
  of his own pocket.
• Dollar for Pound For every pound lost, they
  got 1
• Results 200 patients
• 70 retention in weight loss program
• 80 of those in program lost weight
• Average wt loss 8.8 pounds
• Cost him 1,000 but worth every dollar!

43
PCI Appropriateness?
CP1027346-1
44
PCI Appropriateness by Hospital
Process Measures

• NY State 1990
• Random sample of 1306 pts undergoing PCI
• Appropriateness based on modified Delphi Method
• Independent angio readings suggests another 10
  may also be inappropriate as well

Hilborne LH et al. JAMA 1993269761-5
45
Potential Impact of Inappropriate PCI
Process Measures

• Close High Mortality Hospital
• 600,000 PCI/yr in US
• 15 of PCI at Hi Mort Hosp
• 1.3 x mortality risk at Hi Mort Center vs not
• 0.8 in-hospital mortality for elective PCI

• Avoid Inappropriate PCI
• 600,000 PCI/yr in US
• 6 inappropriate and 38 uncertain
• 25 of uncertain PCI are actually inappropriate
• 0.8 in-hospital mortality for elective PCI

744 deaths avoidable by eliminating inappropriate
PCI
216 avoidable deaths (by closing High Mort
Centers)
46
Appropriate Nuclear Stress Tests?
CP1027346-1
47
Diagnostic studies Always Cause a Combination of

• Good Effects

• Bad Effects

CostRisk of test Delay in diagnosis Pt
concerns, more testing
Correct early diagnosis
48
The Current Cycle of Clinical Diagnostic Imaging
New Imaging Modality
Technological advances imaging studies
performed at select sites
Observational studies on diagnostic accuracy in
specific populations
Standard Imaging Modality
Off-Label use
Guideline Recommendations
Recent direct consumer (patient and physician)
marketing
Widespread use Re-imbursement
49
ACC/AHA Recommendations on Imaging
50
Evidence behind Imaging Recommendations
51
The Cycle of Diagnostic Imaging New Model
New Imaging Modality
Studies on diagnostic accuracy in specific
populations
Incorporate into randomized studies of different
imaging strategies for common presentations
Education and Feedback
Outcome Data (Time to Diagnosis cost /diagnosis
Guideline Recommendations
Electronic Image data capture / tied to billed
indication
Appropriate Testing
52
Thoughts on Future of QIOs Role in Quality
Improvement

• Traditionally, QIOs
• Motivate hospital engagement in QI
• Major role with those ready for change
• Data collection QI Training
• Yet limited rigorous evaluation of impact
• New World
• All centers motivated to improve (by P4P) and
  many good QI programs already available
• Major role with those w/o internal skills for QI
• Data auditing partner with others for QI
• Quantitative evaluation of what works!

53
Thoughts on Future Role for QIOs in
Information Technology

• Proposed
• Requirement to drive IT adoption
• Promote integration in centers
• Perhaps a better role
• Others will take lead on development and
  dissemination of IT
• But QIO quantitative skills key to use data to
  monitor quality of care
• And should use new innovation in QI efforts

54
Thoughts on Future Role for QIOs On Broader
Quality Roles

• Role in Patient Safety
• Need to broaden metrics.. did you do it right!
• Active assessment of readiness?
• Role in Transitional Care
• Prime role to track pts across care systems
• Role with Patients
• Should QIOs work directly with Medicare
  beneficiaries to promote healthy lifestyles?
• Role in Appropriateness
• Need to assure our resources are well spent
• (but dont tell my colleagues I suggested this!)

55
Skate to where the puck is going, not where it
is. Wayne Gretsky