Transcript The Vascular Quality Initiative: Using PSO Registry Data to Improve

The Vascular Quality Initiative:
Using PSO Registry Data to Improve Care
Adam W. Beck, M.D., FACS
Chair, Arterial Quality Council
Society for Vascular Surgery Patient Safety Organization
Disclosures
• No relevant disclosures
Launched by Society for Vascular Surgery
in 2011
• Mission: To improve the quality, safety, effectiveness
and cost of vascular health care by collecting and
exchanging information.
• 3 Components:
– National Registries in a Patient Safety Organization
– Regional Quality Improvement Groups
• Based on Vascular Study Group of New England, 2002
– Web-based data collection - reporting system
Patient Safety Organization (Patient Safety Act)
• Allows patient identified information to be collected
for quality improvement without informed consent
• Protects work product (any comparative data) from
discovery to encourage honest reporting
• Precludes comparative data to be used for physician
disciplinary purposes or marketing
• Allows non-identifiable data to be published
– Statistical de-identification of patient, provider, hospital
• Ideal vehicle for quality improvement registry
National Registries in a Patient Safety Organization
• Carotid disease
– Endarterectomy and stenting
• Aortic disease
– Open and endovascular abdominal aneurysm repair
– Endovascular repair thoracic aorta
• Lower extremity arterial disease
– Bypass, interventional procedures, amputation
•
•
•
•
Medical Management (in development)
Dialysis access
Vena cava filters
Varicose veins
Advantages of SVS PSO Registry Data
• Allows data from all patients to be included
– Not biased by those who only give consent
• Much more detailed information than claims data
– Pre-, intra-, and post-op variables (> 150 per procedure)
• One year follow-up for key outcomes
– Completed in physician’s office
• All consecutive procedures – allows rate calculation
– Audited against hospital and physician claims data
• Longer follow-up with matched Medicare Claims
– Survival also from Social Security Death Index
Value of Patient Safety Organization
Number of Participating Centers
375
350
325
300
275
250
225
200
175
150
125
100
75
50
25
0
PSO
• Reassure increasingly
diverse and larger
centers that their data
were confidential,
protected and would
only be used for quality
improvement.
VQI Participating Centers
Big Data: 270,000
Procedures, 7,500
per Month
Hospital Types
Community
32%
37%
Academic
31%
Teaching Affiliate
376 Centers, 46 States + Ontario
Total Procedures
Captured
(as of 2/1/2016)
271,819
Carotid Endarterectomy
62,986
Carotid Artery Stent
10,141
Endovascular AAA Repair
25,135
Open AAA Repair
7,861
Peripheral Vascular
Intervention
85,682
Infra-Inguinal Bypass
28,585
Supra-Inguinal Bypass
9,676
Thoracic and Complex
EVAR
5,614
Hemodialysis Access
23,376
Lower Extremity
Amputations
4,740
IVC Filter
4,778
Varicose Vein
3,245
Physician-Driven, Multi-Specialty Patient Safety Organization
2500 Specialists
All Procedures
1600 Specialists
Interventional Procedures
4%
5%
10%
11%
17%
Cardiology
47%
Vascular
Surgery
17%
Radiology
Vascular Surgery
Cardiology
Cardiac Surgery
26%
Cardiology
38%
Vascular
Surgery
26%
Radiology
Radiology
General Surgery
Other
Vascular Surgery
Cardiology
Radiology
General Surgery
How to Grow and Sustain a Self-Funded PSO
• Leverage “Big Data” from the national registry to power
analyses of processes that lead to best outcomes
• Promote physician practice change, ownership and
connection to the PSO/registry by developing smaller,
regional quality improvement groups
New England QI Group Model - 2002
• Semi-annual meetings of physicians, nurses,
researchers and administrators
• Analyze variation in process and outcomes among
regional centers
• Discuss potential causes for variation
• Develop quality improvement projects in areas
where substantial variation exists
• Promote ownership, collaboration, and greater
opportunity to translate data into practice change
Network of
17 Regional Quality Groups
AK
HI
Semi-annual meetings, Review variation
Regional quality improvement projects
Regional Groups: Lessons Learned
• Comparative feedback stimulates practice change
– Physicians are naturally competitive
– We all want to improve our results
– We all want to have the best results
• Most vascular patients should be on a statin pre-op
– Record statin use
– Feedback results to surgeons
VSGNE Pre-op Statin Use 2004
Initial 25 Surgeons
VSGNE Pre-op Statin Use 2007
Initial 25 Surgeons
Dec-98
Mar-99
Jun-99
Sep-99
Dec-99
Mar-00
Jun-00
Sep-00
Dec-00
Mar-01
Jun-01
Sep-01
Dec-01
Mar-02
Jun-02
Sep-02
Dec-02
Mar-03
Jun-03
Sep-03
Dec-03
Mar-04
Jun-04
Sep-04
Dec-04
Mar-05
Jun-05
Sep-05
Dec-05
Mar-06
Jun-06
Sep-06
Dec-06
Mar-07
Jun-07
Sep-07
Dec-07
Mar-08
Percent
Pre-op Statin Use VSGNE 2003-2008
100%
90%
80%
Started QI
Initiative
70%
30%
2003
2004
78%
70%
62%
60%
54%
50%
40%
Developed
Letters to PCPs
2005
Set QI Goal
= 80%
20%
2006
2007
2008
Regional Groups: Lessons Learned
• Comparative feedback stimulates practice change
• Large dataset can answer important clinical questions
– Should I use protamine to reverse heparin during
carotid endarterectomy?
• Protamine reverses anticoagulant used during surgery, to
promote clotting
– Re-operation for bleeding: 1.7%
• Concern about causing too much clotting: stroke, MI
– Low frequency events cannot be studies in small
series and randomized trials are unrealistic
VSGNE Surgeon Practice
4587 Total CEAs
Protamine
No Protamine
2087 (46%)
2500 (54%)
-Stone et al, J Vasc Surg, 2010
Reoperation for Bleeding
% Patients
1.7%
*P=0.001
0.6%
-Stone et al, J Vasc Surg, 2010
Thrombotic Complications
% Patients
*P=NS
-Stone et al, J Vasc Surg, 2010
Regional Groups: Lessons Learned
• Comparative feedback stimulates practice change
• Large dataset can answer important clinical questions
• Trusted analyses, reports can rapidly change practice
– Physicians have ownership of regional group data
– Protamine data were presented to regional group
and published
VSGNE Protamine Use During CEA
Protamine use
increased from 46%
before 2009 to 61%
after 2009 (P<.001).
Presented - Published
Results in 2009
Regional Groups: Lessons Learned
•
•
•
•
Comparative feedback stimulates practice change
Large dataset can answer important clinical questions
Trusted analyses, reports can rapidly change practice
Changed practice can improve outcomes
Protamine Use and Bleeding
P=.003
-Patel et al, J Vasc Surg 2013
Current QI Projects in VQI Regional Groups
•
•
•
•
•
•
•
•
•
Increasing use of antiplatelet and statin use pre-op and at DC
Decreasing myocardial infarction after arterial procedures
Optimizing graft type choice for leg bypass
Enhancing recovery after lower extremity amputation
Reducing length of stay after VQI procedures
Improving long term follow up of patients in VQI
Reducing preventable causes of readmissions
Preventing contrast-induced nephropathy after arteriography
Increasing smoking cessation after major arterial procedures
Improving Quality Across VQI Regions
• COPI Reports
– Center Opportunity Profile for Improvement
• Analyze and report variation in outcome
• Multivariable model to define causes of outcome
• Individual report to each center:
– How they compare with others for the outcome and
each factor associated with the outcome
– Provides a customized, actionable improvement plan
for each center
In-Hospital Surgical Site Infection after Infrainguinal Bypass
• Significant variation
found across VQI
participating centers
and regions
• Risk factors associated
with SSI:
– Operation > than
220 minutes
– Transfusion > 2 units
PRBC
– Skin prep not
chlorhexidine
30%
COPI Report for SSI after Lower Extremity Bypass
29
COPI Report for SSI after Lower Extremity Bypass
30
31
12/2/13
11/2/13
85%
10/2/13
90%
9/2/13
8/2/13
7/2/13
6/2/13
60%
5/2/13
65%
4/2/13
79%
3/2/13
2/2/13
1/2/13
12/2/12
70%
11/2/12
75%
10/2/12
9/2/12
8/2/12
7/2/12
6/2/12
5/2/12
4/2/12
3/2/12
2/2/12
1/2/12
Percentage
Chlorhexidine Skin Prep Use
100%
95%
93%
80%
COPI
Report
55%
50%
Percentage
Percentage
Centers with Most Improvement in Chlorhexidine Use
32
% Patients with Length of Stay > 1 Day after
Elective Carotid Endarterectomy
8,000 Procedures, 2011 to 2012
(Excludes inhospital deaths, previous ipsilateral CEA, concombinant CABG)
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Observed
Expected
Center Process Variation
*******
********
**
*****
*** **
*** ** *
**
*
VQI Centers
Adjusted for: age, gender, race, hypertension, diabetes,
pre-op beta blocker, CAD, CHF, COPD, stress test, discharge ASA,
discharge statin, prior radiation therapy, pre-op MRA
*
***
** ***
*** ***
**********
****
******
Observed significantly
different than
expected:
* p<0.05
**p<0.01
Factors Associated with LOS> 1 Day after CEA
Unexplained
11%
Patient Factors
18%
Process Factors
7%
Major Adverse
Events
23%
IV Meds Required
for Hyper- or
Hypotension
31%
Low Surgeon
Volume
10%
Value of VQI Participation
• Does participation in VQI (receiving benchmark
reports, attending regional meetings, etc.) improve
patient outcomes?
Late Survival after Major Arterial Procedures
• 50,000 Patients in VQI who underwent
– Leg bypass / intervention, oAAA / EVAR, CEA / CAS
• Evaluated pre-operative and discharge medications:
– Antiplatelet agent (ASA, PY212 inhibitors)
– Statins (HMG-CoA reductase inhibitors)
• Outcomes analyzed:
– Effect on patient survival
– Variation across centers
– Impact of participation in VQI
-De Martino et al, J Vasc Surg, 2015
Effect of Discharge Medications on Survival
81% Both
75% AP
68% Statin
55% None
26% Absolute improvement in 5-year
survival when patients are
discharged on AP & Statin
P<0.001 SE < 0.1
Years
-De Martino et al, J Vasc Surg, 2015
Variation in % Patients Discharged on
Anti-platelet and Statin
100%
VQI Mean = 76%
30%
VQI Centers
Patients on Antiplatelet and Statin Pre-op and Discharge
Based on Center Years Participation in VQI
Number of Years Participating in VQI
New VQI Initiatives
• Evaluating appropriateness of treatment
Appropriate Treatment
• Appropriate treatment requires not only good early and
late outcomes, but also:
– Correct patient selection
– Correct procedure selection
• VQI provides an opportunity to analyze variation in
patient and procedure selection
– Feedback data to centers
– Goal: regression toward the mean and reduced variation
• Current data show large variation!
Carotid Artery Treatment by Stent (vs endarterectomy)
50%
All VQI Centers Mean = 13% Stent Rx
45%
40%
35%
Procedure Selection Variation
30%
25%
20%
15%
10%
5%
0%
1%
VQI Centers
47%
Carotid Artery Treatment in Asymptomatic Patients
100%
90%
All VQI Centers Mean = 77% Asymptomatic
80%
70%
60%
50%
Patient Selection Variation
40%
30%
29% Asymptomatic
20%
10%
0%
VQI Centers
100%
Variation in Ultrasound Criteria for Severe Disease
600
550
500
450
Mean PSV (cm/sec)
400
Disease Severity Judged by Blood
Flow Velocity in Narrowed Artery
PSV = 550 cm/sec
More Severe Disease:
Fewer Procedures
Median Velocity= 360 cm/sec
350
300
Patient Selection Variation:
Substantial Opportunity
250
200
150
100
50
0
Less Severe Disease:
More Procedures
PSV = 50 cm/sec
VQI Centers
New VQI Initiatives
• Evaluating appropriateness of treatment
• Evaluating new medical devices in real world practice
FDA, Device Manufacturer Collaboration
• Many devices used to treat vascular disease (eg, stents)
• Devices approved after limited testing in selected
patients in centers of excellence
• In practice, devices may be used outside indications for
use and in less experienced centers
• Data about device performance in real world practice
can inform physicians, FDA and manufacturers, to
optimize treatment benefits for patients
• VQI is sharing non-identifiable data regarding new
device outcomes with FDA and manufacturers
SVS PSO
Organizational Structure
SVS PSO
Organizational
Chart
Governing
Council
Executive Committee
Communication
Committee
Arterial Quality
Council
Venous Quality Council
Vascular Technology
Council
Registry Committees
Registry Committees
Industry Project Steering
Committees
Research Advisory
Committee
Ad Hoc Committees
Cost Integration
Long Term
Follow-up
EHR - Epic
Claims Data
Integration
Roles of SVS PSO and M2SVQI
SVS PSO
• Registry data specifications and
definitions
• Oversight of data use, including
research project approval
• Regional group administration
• Advanced data analytics
• COPI reports
• Regulatory interaction
• Industry project approval and
oversight
M2S
• Technology platform
• Data entry mechanism
• Reporting/analytics
• EMR integration
• Sales, marketing, and contracting
• Registry operations
• Site training and support
• Claims validation
• PQRS administration
• Industry project management
SVS PSO
Organizational
Chart
Governing
Council
Executive Committee
Communication
Committee
Arterial Quality
Council
Venous Quality Council
Vascular Technology
Council
Registry Committees
Registry Committees
Industry Project Steering
Committees
Research Advisory
Committee
Ad Hoc Committees
Cost Integration
Long Term
Follow-up
EHR - Epic
Claims Data
Integration
Over 80 PeerReviewed
Publications
187 Approved Projects
26 Approved in Q4 2015
SVS PSO
Organizational
Chart
Governing
Council
Executive Committee
Communication
Committee
Arterial Quality
Council
Venous Quality Council
Vascular Technology
Council
Registry Committees
Registry Committees
Industry Project Steering
Committees
Research Advisory
Committee
Ad Hoc Committees
Cost Integration
Long Term
Follow-up
EHR - Epic
Claims Data
Integration
Matching Medicare Claims Data with VQI Registry
• Goal: determine late outcomes after initial vascular treatment
using Medicare claims related to subsequent treatment
• Dartmouth Institute for Health Policy and Clinical Practice
• Match claims data identifiers with VQI patient identifiers:
– CEA, CAS, EVAR, oAAA, TEVAR, PVI, leg bypass datasets
– Match rates > 90% for procedures done 2003-2013, age > 65
• Social Security Death Index also matched in patients of all ages
• Data available for VQI based research and industry projects
Matching Medicare Claims Data with VQI Registry
• Major outcomes and time to event after index procedure:
• Carotid (CEA and CAS)
– Re-admission, death, stroke, re-intervention (CEA or CAS), imaging
• Aorta (oAAA, EVAR, TEVAR)
– Re-admission, death, re-intervention (open or EVAR), abdominal
surgery for complication after oAAA
• Leg (Infra, Supra, PVI)
– Re-admission, death, re-intervention (open or EVAR), imaging,
amputation (major, minor)
SVS PSO
Organizational
Chart
Governing
Council
Executive Committee
Communication
Committee
Arterial Quality
Council
Venous Quality Council
Vascular Technology
Council
Registry Committees
Registry Committees
Industry Project Steering
Committees
Research Advisory
Committee
Ad Hoc Committees
Cost Integration
Long Term
Follow-up
EHR - Epic
Claims Data
Integration
Estimating Cost for Procedures Performed in VQI
• Goal: Provide centers with anonymous benchmark comparison of
treatment costs when adjusted for disease severity and procedure
details already collected in VQI
• Method: Partnered with MedAssets, a healthcare performance
improvement company that focuses on finance and operations
• Pilot Project:
– 17 VQI hospitals submitted UB-04 claims for 2014 EVARs
– Using VQI clinical data, cases grouped as standard or complex, and
with and without post-treatment complications
– MedAssets calculated costs from charge data using usual methods
• Currently preparing reports of total and component cost
• Plan to offer for all procedures for all centers
Sample Cost Report
SVS PSO
Organizational
Chart
Governing
Council
Executive Committee
Communication
Committee
Arterial Quality
Council
Venous Quality Council
Vascular Technology
Council
Registry Committees
Registry Committees
Industry Project Steering
Committees
Research Advisory
Committee
Ad Hoc Committees
Cost Integration
Long Term
Follow-up
EHR - Epic
Claims Data
Integration
Use of VQI for Post Approval Surveillance
• Aligns with FDA recommendations to use registries to
collect real world data for post approval surveillance
• “FDA believes that
device registries should
serve as the foundation
of our National Medical
Device Postmarket
Surveillance System.”
Procedures and Devices Captured
Procedure
Devices Included
Carotid Endarterectomy
Patch Type
Carotid Artery Stent
Stent, Protection Device Type
Endovascular AAA Repair
Stent Graft Type
Open AAA Repair
Prosthetic Graft Type
Peripheral Vascular Intervention
Balloon/Stent / Atherectomy Devices
Infra-Inguinal Bypass
Prosthetic Graft Type
Supra-Inguinal Bypass
Prosthetic Graft Type
Thoracic and Complex EVAR
Stent Graft Type
Hemodialysis Access
Graft Type
Lower Extremity Amputation
Bandage, prosthesis type
IVC Filter
Device Type
Varicose Veins
Ablation Device, Sclerosants
Medical Management PAD (in development)
Pharmacologic Agents
Use of VQI for Post Approval Surveillance
• Aligns with FDA recommendations to use registries to
collect real world data for post approval surveillance
• Leverages an existing network of 2500 specialists in
350 potential centers experienced with data collection
Specialist Types
11%
Hospital Types
5% 4%
32%
37%
17%
47%
17%
31%
Vascular Surgery
Radiology
Cardiology
Academic
Teaching
Community
Use of VQI for Post Approval Surveillance
• Aligns with FDA recommendations to use registries to
collect real world data for post approval surveillance.
• Leverages an existing network of 350 potential centers
and 2600 specialists experienced with data collection
• Uses existing VQI data forms supplemented as needed
with additional variables, follow-up time points
– Customized content for participating sites
– New sites can join VQI for project if not already member
Use of VQI for Post Approval Surveillance
• Aligns with FDA recommendations to use registries to
collect real world data for post approval surveillance
• Leverages an existing network of 350 potential centers
and 2600 specialists experienced with data collection
• Uses existing VQI data forms supplemented as needed
with additional variables, follow-up time points
• Informed consent and IRB approval NOT required
– Standard of care practice and PSO Quality Project
– Allows all patients treated in center to be captured
Use of VQI for Post Approval Surveillance
• Aligns with FDA recommendations to use registries to
collect real world data for post approval surveillance
• Leverages an existing network of 350 potential centers
and 2600 specialists experienced with data collection
• Uses existing VQI data forms supplemented as needed
with additional variables, follow-up time points
• Informed consent and IRB approval NOT required
• SVS PSO contracts with sites, facilitates recruitment
– Uniform contract, no separate site negotiations
Use of VQI for Post Approval Surveillance
• Aligns with FDA recommendations to use registries to
collect real world data for post approval surveillance
• Leverages an existing network of 350 potential centers
and 2600 specialists experienced with data collection
• Uses existing VQI data forms supplemented as needed
with additional variables, follow-up time points
• Informed consent and IRB approval NOT required
• SVS PSO contracts with sites, facilitates recruitment
• Result = Rapid, Efficient, Flexible, Cost-Effective
Aug-14
Sep-14
Oct-14
Nov-14
Dec-14
Jan-15
Feb-15
Mar-15
Apr-15
May-15
Jun-15
Jul-15
Aug-15
Sep-15
Oct-15
Nov-15
Dec-15
Jan-16
Feb-16
Mar-16
Patient Enrollment
TEVAR Dissection Project Enrollment Rate
Existing Network of VQI Centers Allows Rapid Enrollment
400
Actual
350
300
250
200
Expected
150
100
50
0
Project Steering Committee
• Scientific Steering Committee is appointed by SVS PSO
with input from sponsor to insure experienced users
• Steering Committee reviews data and interacts with
sites to adjudicate any data questions during analysis
• Steering Committee takes responsibility for scientific
presentation and publication of final data analysis in
consultation with sponsor
• Coordinated by specific Project Manager at M2S
Protocol Development, Site Selection
• Collaboration between SVS PSO Steering Committee, FDA and the
sponsor helps design a pragmatic surveillance protocol
– Data collection requirements that fit real world registries
– Follow-up time points appropriate for standard of care
• Steering Committee selects optimal sites for project
– Physicians who have been trained to use the new device
– Sites which meet resource requirements for data entry
– Sites with known accurate and timely VQI data collection
• M2S Project Team monitors site performance, provides feedback
Data Ownership
• SVS PSO provides sponsor with copy of “line by line” data
– De-identified for patient, provider, hospital
– Dates converted to intervals since procedure, age, etc.
• Sponsor may use data for analysis, regulatory requirements
and marketing
– Marketing content must be approved by SVS PSO when
source of data is cited to insure accurate interpretation
• Steering Committee produces scientific presentations,
publications, in consultation with sponsor
Additional Industry Uses for VQI Data:
Supplementing Historical Data to Reduce Project Length
• Substantial data exists in VQI but might be missing some custom
variables needed for a specific device project
• It is possible for sites to retrieve and add custom data elements
to previously entered data forms, including additional follow-up
• VQI identifies relevant procedures, invites sites to supplement
data for required number of procedures
• This allows projects to rapidly accrue procedures and shorten
time required for follow-up and study completion
• One recent project completed, developed efficient on-line
mechanism for sites to identify and enter needed data
Additional Industry Uses for VQI Data:
Using Registry Data to Expand Device Indications
• Many devices have been applied to arteries or patients not
tested in initial trials, and thus used “off-label”
– Biliary stents in peripheral arteries, veins
– EVAR devices for short, angulated aortic neck
• Individual physicians make decisions based on limited data
• FDA recognizes need for a more efficient method to expand
indications potentially based on comprehensive registry data
• Initial project underway to establish additional indication in PVI
treatment of SFA disease, without IDE requirement
Additional Industry Uses for VQI Data:
Use of VQI for Premarket Clinical Trials
• Must be conducted according to 21CFR 812 (IRB approval,
informed consent, investigational site monitoring, investigator
and sponsor responsibilities, adverse event reporting and
submission of periodic reports to FDA and IRBs)
• Benefit to leverage existing VQI network and avoid double data
entry by sites into both VQI registry and industry CRF
• Uses existing VQI data forms supplemented as needed with
additional variables, follow-up time points
• Opportunity to reduce study cost and time simply by using VQI
for electronic data capture
Additional Industry Uses for VQI Data:
FDA Medical Device Reporting
• Current MDR system is costly, does not capture all device events,
and provides no denominator to calculate event rates
• Potential exists to use registry to identify potential adverse
device-related events:
– Capture “potential” serious device-related events by:
• Death, re-intervention, excess imaging studies after index procedure
• Use both registry and claims data
– Fund centers to provide additional information about events
• Drill down to establish cause of event and any device contribution
• Current discussions with FDA about potential pilot project
Conclusions
•
•
•
•
•
The PSO structure facilitates growth of QI initiative
Regional groups promote physician ownership, trust
Comparative data stimulates practice change
Big Data provides new information
COPI reports provide actionable, site specific
opportunities for change that improves outcomes
• Variation in appropriateness can be measured and will
hopefully lead to better patient, procedure selection
• PSO data can serve multiple stakeholders for optimal
efficiency
One Platform. One Data Set. Many Stakeholders.
SVS PSO
M2S
Hospitals
Research
Projects
Physicians
Device
Companies
FDA
VQI Registry
Stakeholders
CMS &
Other
Payers
EHR
Companies
InterNational
Registries
NIH
NINDS
Patients