Practice Assessment: the Michigan Experience

1
Practice Assessmentthe Michigan Experience

• M. Ashraf Mansour, M.D.
• Division of Cardiovascular Surgery
• Spectrum Health, Grand Rapids, MI

2
Background

• ACS-NSQIP
• MSQC
• Data collection
• Hospital-specific outcomes data
• Practitioner data
• Opportunities for improvement

3
Program Overview

• The ACS NSQIP is an outcomes-based, data-driven,
  risk-adjusted surgical quality improvement
  program, which empowers surgeons and medical
  centers to report reliably their outcomes and
  potentially improve care and lower costs
• Roots in the Veterans Health Administration 14
  years of operating experience
• 128 VA hospitals national participation in the
  ACS NSQIP growing daily
• 110,000 cases entered per year. To date, over a
  million total in the VA database 100,000 in the
  ACS database
• ACS expansion of the private-sector
  programOctober, 2004

4
Evolution of the Program

• The ACS NSQIP has grown steadily over the years,
  meticulously building and proving its models and
  methodology across a spectrum of medical center
  environments

1985
86
87
88
89
90
91
92
93
94
95
96
97
98
99
2000
01
02
03
04
05
Pilot study at three private-sector hospitals
NVASRS
VA hospitals under scrutiny
Congress passes law 99-166
AHRQ grant 14 academic medical centers
Over 40 hospitals enrolled Additional 100 in the
application process
HCFA reports unadjusted comparative mortality
rates
Inception of VA NSQIP
4 community hospitals join
ACS expansion of private-sector InitiativeACS
NSQIP
5
VA Hospitals Under Scrutiny

• In the late 1980s, the VA faced a barrage of
  criticism over the quality of surgical care in
  their 128 hospitals
• At issue
• operative mortality rates
• perceived comparisons to national (private-sector
  norms)
• In response, Congress passed U.S. Public Law
  99-166 mandating the VHA to report its surgical
  outcomes annually
• on a risk-adjusted basis to factor in a patients
  severity of illness
• compared to national averages

6
Unique Position

• While the VA knew there were no national averages
  or risk-adjusted surgical models, they recognized
  their unique position from which to create them

DIFFICULT
IDEAL
ENVIRONMENT
Community Hospitals
VA Hospitals
Academic Hospitals
Mixture of new legacy systems little
integration
Mixture of new legacy systems little
integration
Advanced homogeneous IT systems
Information Systems
Chief of Surgery has moral authority over docs,
but
Defined structure
Centralized authority clear chain-of-command
Organization Structure
Doctors in private practice not staffed on-site
On-staff on-site
On-staff on-site
Doctor Location
Communication w/ Other Hospitals
Little communication w/ other hospitals
Little communication w/ other hospitals
Established channels among VA hospitals
7
The VA Response

• The VA rigorously collected, modeled and analyzed
  their datain fact, quality of the data is a
  hallmark of the NSQIP
• Standardized data definitions
• Dedicated a nurse reviewer in each hospital to
  capture preoperative, intraoperative and 30-day
  outcome variables
• Annual audits of each sites data
• Distribution and sharing of blinded data with and
  between sites
• Created feedback mechanism for best practices and
  implementation of focused quality improvement
  initiatives
• Risk-adjustment models for outcomes of surgery
• Used stepwise forward logistical regression to
  identify the preoperative risk factors predictive
  of outcomes
• Developed observed v. expected models for
  surgical morbidity and mortality

8
(No Transcript)
9
(No Transcript)
10
MSQC Goals

• SSI Best Practice study-national
• Best PracticesPeri operative BT
• Best PracticesGlycemic Control
• Best Practices SCIP-7

11
Disclaimerthis is a work in progress! ?
12
MSQC structure

• Hospital participation
• Surgeon Champion
• Full time Nurse to abstract data
• Audit
• P4P
• Hospital Surgeon specific data
• Impact on Quality

13
ACS-NSQIP Site Visits
14
MSQC Workstation Preview
Links to MSQC Website and ACS NSQIP Website
15
All cases will go to MSQC database, user will
determine which cases go to ACS NSQIP
16
Special MSQC Project Fields
17
Putting the QI in ACS-NSQIP

• Results from the ACS-NSQIP Best Practice
  Initiative

18
Best Practices Study Initiative SSI
19
Putting the QI in ACS-NSQIP

• Establishment of a reliable measurement system
• Define variation in performance
• Identify best performers
• Identify best practices
• Distribute the information

20
20 Low Outliers
21
13 High Outliers
22
Compare and contrast Low and High outlier
hospitals

• 20 Low Outlier 22,031 cases
• 13 High Outlier 15,428 cases

23
Putting the QI in ACS-NSQIP

• Establishment of a reliable measurement system
• Define variation in performance
• Identify best performers
• Identify best practices-but how?
• Distribute the information

24
What was different?

• Structural Characteristics Low (n20) High
  (n13) P
• Trainee/bed ratio 0.25 0.61 0.0001
• Emergency Cases 11.2 13.9 lt.0001

25
Operative Duration
Low High
min min p

• Operative duration 102.783.9 128.3104.3 25 lt.00
  01
• x work RVU 14.42 15.60 8 lt.0001

26
(No Transcript)
27
Site visits Summary-low outlier

• No trainees
• Little turnover surgeons or nurses
• No travelers
• Remarkably efficient
• Leadership support
• Very positive safety culture
• Ease of communication
• Few breaks during surgery

28
Best Practices -MSQC

• What are we trying to do?
• Decrease mortality
• Decrease morbidity

29
But how?

• Pick interventions which would influence many
  various outcomes
• ? Avoidance of Blood Transfusions
• ? Avoidance of Hyperglycemia
• ? Avoidance of Hypothermia

30
(No Transcript)
31
(No Transcript)
32
Table 1 MSQC PostOp Outcomes All Hospitals
Anemic vs. Non-Anemic (General and Vascular
Cases, excl. Emergent, All Years)
33
Most risk factors for SSI are not alterable

• Hyperglycemia is

34
Frank et al 1997 JAMA 2771127

• Perioperative Maintenance of Normothermia Reduces
  the Incidence of Morbid Cardiac Events A
  Randomized Clinical Trial

35
MSQC Measurement Parameters A tentative list

• Appropriate initial dose of antibiotics
• Redosing gt 3 h
• cases transfused
• units transfused/case
• cases transfused Hctgt30
• anemic cases operated
• Duration surgery
• Teamwork intervals
• High FiO2
• Strict glycemic control

36
BMC2

• University of Michigan quality initiative for
  coronary interventions
• Hospital specific data

37
Variation among Centers
38
Peripheral Vascular QualityImprovement
Initiative (PVQI2)

• Endovascular procedures
• Vascular surgeons, Cardiologists and
  Interventional Radiologists
• Data collection form
• Quarterly reports
• Hospital and physician data

39
Outcome measurements

• Deaths
• Complications
• MI
• Limb ischemia
• Emergent OR
• Blood transfusion
• Renal failure
• Amputation

40
Use of closure devices

• Pseudoaneurysm rate
• Device failure
• Management of access site

41
Summary

• Participation in QI initiatives is becoming
  mandatory
• Implications for hospital reimbursement
• Physician data used for privileges
• Report cards will be used for MOC
• Data use transparency and full disclosure to the
  public?

42
Acknowledgements

• Darrell (Skip) Campbell, M.D.
• MSQC and NSQIP
• PVQI2