Transcript Transfusion _10-18-13_revised PMG

BLUE CROSS BLUE SHIELD OF MICHIGAN
CARDIOVASCULAR CONSORTIUM
VASCULAR INTERVENTIONS COLLABORATIVE- (BMC2 VIC)
P. Michael Grossman, MD
Director, BMC2 PVI
Associate Professor
University of Michigan Hospitals and Health Center
October 2013
Focus for Today
 Update on our recent expansion and addition
of new sites
 Brief overview of data from the PVI part of our
Collaborative
 History of transfusion in BMC2
 Future direction
What is the “Blue Cross Blue Shield of Michigan
Cardiovascular Consortium Vascular Intervention
Collaborative?”
 I only have a few
minutes, so from
now on…..
 BMC2 VIC
What is BMC2 VIC?
 Regional, physician-run, multidisciplinary




collaboration
Partnership with Blue Cross Blue Shield of Michigan
Assess and improve quality of care and outcomes of
PVI patients
Overcome barriers of traditional market and
academic competition
Collect, organize, audit, and analyze data
 Report procedural variables and outcomes
 Individual physician operators
 Institutions
 Support quality improvement initiatives
4
Which Physicians Participate?
 Consecutive cases from all vascular
interventions in participating hospitals
 Truly multidisciplinary
 Vascular surgery
 Interventional radiology
 Interventional cardiology
2014 PVI – 47 Participating Hospitals/Outpatient Centers
New PVI Outpatient Center Participants
 Michigan Vascular Center
 Michigan Outpatient Vascular Institute
BMC2 PVI/VIC
DATA OVERVIEW
PVI Patient Characteristics
PVI Patients Have Multiple Co-Morbidities
2012 (%)
n=8927
Female Gender
Average Age (years)
Current Smoker
Hyperlipidemia
Hypertension
Diabetes Mellitus
CHF
COPD
CVD/TIA
CAD
Renal Failure on Dialysis
Body Mass Index
Anemia
43.8
68.5
40.5
86.8
91.3
47.7
20.8
28.7
29.2
59.7
5.7
30.1
43.5
Percent of PVI Procedures by Arterial Bed
BMC2 PVI Outcome Trends:
2007 – YTD Q2 2013
23%
Percent
38%
14%
44%
6-Month Medication Compliance*
2009 – 2012
Percent
RX compliance at 6 months with discharge medications
*Medication Compliance – Patient continues to take the medications
as prescribed at discharge after the PVI procedure
12
6-Month Medication Compliance*
Not Discharged with RX: 2009 – 2012
RX use at 6 months for patients not discharged
on RX (and no documented contra-indication)
*Medication Compliance – Patient continues to take the medications
as prescribed at discharge after the PVI procedure
37
Transfusion
HISTORY IN BMC2
BMC2 “Historical” Quality
Improvement Approach






Collect data
Audit data
Report data to participants
Benchmarked to collaborative
Good performers congratulated
Underperformers scrambled to meet the mark or
ignore the reports
 Communication with physician champions and site
data coordinators
 “Trickle Down” Quality Improvement
Early PVI Transfusion Data
Early Transfusion Data
Quality Improvement Initiative: 2008
 Initiative focused on decreasing peri-
procedural bleeding complications and
post-procedural transfusion
 Weight-based heparin anticoagulation dosing
 Initial procedural heparin dose should be ≤ 60 U/kg
 Check intra-procedural ACT
 Heparin to achieve an ACT of 200 to 250 seconds
 Follow guidelines for blood transfusion
18
From “Trickle Down” to “Aim for a Target”
“…Setting a target resulted in better performance and significant cost
savings (in janitorial services).”
Quality Improvement Goals: 2009-2010
Physician Advisory Committee approved creation of
numeric PVI goals based on Collaborative data and
goal to smooth practice variation across sites and
reduce outliers:
 ASA at Discharge > 90%
 Statin at Discharge >75%
 Vascular Complications < 4%
 Post PVI Transfusion < 7%
 Contrast Induced Nephropathy < 7%
A Quality Improvement Team was Created
 QI team performed 24 site visits over a 2 year
time period
 Visited sites where outcomes were low
 Discussed a detailed questionnaire regarding site
processes for peripheral vascular procedures
 Obtained site protocols and order sets
 Identified and starred the very best practices from
these visits
 Worked with the QI Team to merge the best
practices together into a BMC2 compendium of
“Best Practices”
QI Site Visit Questionnaire - Transfusion
April 2010 BMC2 Meeting
 All BMC2 meeting to obtain a comprehensive
review of the outcome literature, ways to prevent
complications and patient management
 Transfusion was one of the areas of focus
 “Best Practice” protocols distributed to all sites –
transfusion protocols included
 Added focus on patients transfused with Hgb >8.0
 Additional quarterly hospital QI Reports were also
created with detailed outcome specific
information
Transfusion as an Adverse Outcome
 Independent risk factor for mortality in
several cardiovascular studies




Transfusion associated with bleeding
Increased hospital stay
Increased cost
Other risks of transfusion:
 RBCs in stored blood act as NO “sinks”
 Increase vasoconstriction, platelet aggregation
 Stored RBCs are low in 2,3-diphosphoglyceric acid
and have high oxygen affinity
 Increased inflammatory mediators
 Exacerbation of myocardial ischemia
24
Transfusion Literature
 ACS Patients who received a blood transfusion had
markedly worse 30-day mortality compared with those
who did not receive a transfusion.
Rao SV, et. al., JAMA 2004; 292:1555–1562
 Patients undergoing lower extremity revascularization;
found a higher risk of postoperative mortality, pulmonary,
and infectious complications after receiving intraoperative
blood transfusion.
O’Keeffe SD, et. al. J Vasc Surg 2010; 51:616-21
25
Transfusion Literature (cont.)
 Independent association between RBC transfusion and
increased risk of 1-year mortality.
 PCI Patients receiving blood > 35 days old had significantly
worse 1-year survival rates compared with patients
receiving blood < 35 days old and patients not transfused.
Kim, P., et. al. Clinical Cardiology 2007; 30:II-35-43
 In patients undergoing cardiac surgery, transfusion of red
cells that had been stored for more than 2 weeks was
associated with a significantly increased risk of
postoperative complications as well as reduced short-term
and long-term survival.
Koch CG et al. N Engl J Med 2008; 358:1229-1239
26
2010 BMC2 Best Practice Protocols
2010 BMC2 Best Practice Protocols
2010 BMC2 Best Practice Protocols
2008 - 2010 Transfusion Rate – 6% to 7%
RESULTS FROM OUR POST-PROCEDURE
TRANSFUSION QI INITIATIVE?
2011 Survey: Value of BMC2 Best Practices
49 Responses:
 75% had obtained the protocols
 66% thought the protocols added value to
quality improvement in the BMC2 Collaborative
 46% stated the protocols were used at their site
to support quality improvement
 Several others mentioned that quality
improvements were “in process”
BMC2 PVI Outcome Trends for Transfusion:
2007 – YTD Q2 2013
Transfusion Rate (%)
52%
Transfusion
RECENT PVI SURVEY RESULTS
SEPTEMBER/OCTOBER 2013
15 RESPONDENTS
Recent PVI Survey Results
Do you transfuse at a threshold of:
Recent PVI Survey Results
Do you use clinical parameters to decide when
to transfuse in addition to Hgb? If so:
Recent PVI Survey Results
Does your hospital employ techs, nurses, PAs or
other professionals who obtain femoral artery
access in the cath lab?:
Recent PVI Survey Results
In a patient with a post operative NSTEMI and
Hgb = 9.5 would you transfuse?
Recent PVI Survey Results
In a post procedure patient with CAD and
without symptoms of ischemia and a Hgb = 7.5
would you transfuse this patient?
Transfusion
FUTURE DIRECTION
Predictors of PVI Post-Procedure Transfusion
All Variables - Excludes hybrid procedures
Significant Variables (P-value < 0.05)
Variable
Odds Ratio
Lower Limit CI
Upper Limit CI
Female
1.981
1.759
2.230
BMI
0.984
0.975
0.994
Hx CHF
1.322
1.156
1.512
Hx COPD
1.370
1.213
1.548
Hx CVD/TIA
1.284
1.140
1.446
Hx CAD
1.223
1.049
1.425
Hx Prior PCI
0.865
0.752
0.994
Hx Prior CABG
0.790
0.683
0.913
Current/Recent GI Bleed
2.128
1.618
2.799
Hx Afib
1.246
1.081
1.437
Hx Renal Failure/Current Dialysis
1.417
1.156
1.738
Hx Renal Transplant
0.549
0.315
0.955
Predictors of PVI Post-Procedure Transfusion (cont.)
All Variables - Excludes hybrid procedures
Significant Variables (P-value < 0.05)
Variable
Odds Ratio
Lower Limit CI
Upper Limit CI
IV Heparin Post-Procedure
5.763
5.010
6.630
Indication = Claudication
0.803
0.709
0.909
Indication = Limb Ischemia
2.118
1.853
2.421
Pre-Procedure Anemia (WHO def.)
4.645
3.996
5.399
Pre-Procedure Creatinine >1.5
1.249
1.072
1.454
Mesenteric Procedure
2.078
1.483
2.913
Heparin Dose (>=60 units/kg)
1.214
1.069
1.378
C stat 294.7, P-value 0.6381, AUC 84.05%
Variables: Preliminary Risk Model

















Family History of premature CAD
Hyperlipidemia
Hypertension
Diabetes
CHF
Significant Valve Disease
COPD
CVD/TIA
CAD
Prior PCI
Prior MI
Prior CABG
Current/Recent GI Bleed
Afib
Other Atherosclerotic Disease
Renal Failure
Renal Transplant

















Gender
Age
BMI
Pre Procedure Aspirin
Pre Procedure Clopidogrel
Pre Procedure Prasugrel
Pre Procedure Statin
Pre Procedure Cilostazol
Pre Procedure Coumadin
Procedure Type: Upper Extremity
Procedure Type: Lower Extremity
Procedure Type: Mesenteric
Procedure Type: Renal
Indication = Claudication
Indication = Limb Ischemia
Pre Procedure Anemia (WHO definition)
Pre Procedure Creatinine >1.5
43
Preliminary Risk Model
Includes patient history and pre-procedure variables only – Excludes hybrid procedures
Significant Variables (P-value < 0.05), AUC 81.12%
Variable
Hx CHF
Hx COPD
Hx CVD/TIA
Hx CAD
Hx Prior PCI
Hx Prior CABG
Current/Recent GI Bleed
Hx Afib
Hx Renal Failure/Current Dialysis
Hx Renal Transplant
Female
BMI
Pre-Procedure Warfarin
Mesenteric Procedure
Upper Extremity Procedure
Indication = Claudication
Indication = Limb Ischemia
Pre-Procedure Anemia (WHO definition)
Pre-procedure Creatinine >1.5
Odds Ratio
1.328
1.316
1.281
1.179
0.861
0.798
2.186
1.181
1.370
0.559
2.002
0.980
1.365
2.694
2.064
0.738
2.589
4.927
1.196
Lower Limit CI
1.165
1.168
1.141
1.010
0.748
0.692
1.678
1.017
1.121
0.326
1.782
0.971
1.151
1.565
1.188
0.648
2.245
4.247
1.031
Upper Limit CI
1.514
1.483
1.438
1.377
0.990
0.919
2.847
1.372
1.674
0.958
2.248
0.989
1.617
4.638
3.585
0.840
2.985
5.715
1.388
Post-Procedure Transfusion: OE Plot (Blinded)
Collaborative Post-Procedure Transfusion OE Plot
Future Direction
 Implementation of risk model for transfusion
into quarterly reports
 Issuance of “final” 2013 BMC2 Best Practice
Protocols this winter – will contain Post op MI,
SSI, and transfusion in a surgical setting.
 Draft 2013 Best Practices are in your folders
today.
47
BMC2 Welcomes Dr. Carson
We are looking forward to hearing his presentation later
this morning and are grateful to him for joining us today.
Jeffrey L Carson, MD
Vice Chair, Research
Richard C Reynolds Professor of Medicine
Chief, Division of General Internal Medicine
Rutgers Robert Wood Johnson Medical School
New Brunswick, New Jersey 08901