Transfusion - Mississippi Valley Regional Blood Center

Report
MVRBC Technical
Advisory Committee:
Jan. 17, 2012
Louis M. Katz MD
Louis M. Katz MD
Mississippi Valley Regional Blood Center
Davenport, IA
EVP, Medical
Affairs
Mississippi Valley Regional Blood Center
Adj. clinical professor, IM/ID, UIHC Carver College of Medicine
Premise(s) of blood management

“Blood still kills”

Blood still costs money, and
transfusion costs much more

Growing evidence supports much
more restrictive transfusion
strategies than used in most venues
Why are restrictive triggers appropriate?
primum non nocere

SHOTs woefully under-reported

Description of putative “new” serious hazards


Pro-inflammatory
Immunosuppressive

Large prospective trials (TRICC, TRIPICU, PINT,
FOCUS, TRACS) demonstrate outcomes at least
as good using restrictive triggers

Positive impact of liberal triggers on functional
outcomes not demonstrated in (FOCUS)

Activity costs of transfusion
Residual risk from RBC transfusion
Carson et al. Submitted. 2012
Global Red Cell Utilization Rates: 2008-09
RBCs per 1,000 Population
60
50
40
30
20
10
0
Source: D Devine et al.: International Forum/Inventory Management
Vox Sanguinis 2009
TRICC: Primum non nocere?
Restrictive (7 gm) Liberal (10 gm)
Mortality
Length
of stay
n=418
%
n=420
%
p
30 day
78
18.7
98
23.3 .11
60 day
95
22.7
111
26.5 .23
Hospital
93
22.2
118
28.1 .05
ICU
11.010.7
11.5 11.3
.53
Hospital
34.8 19.5
35.5 19.4
.58
“A restrictive strategy of red-cell transfusion is at least as
effective as and possibly superior to a liberal strategy in
critically ill patients.” NEJM. 1999.
FOCUS results
Units transfused
Liberal trigger
(n=1007)
Restrictive trigger
(n=1009)
1866
652
(97% transfused) (41.5% transfused)
Median units
2 (IQ 1-2)
0 (IQ 0-1)
1 outcome
35.2%
34.7%
60 day mortality
7.6%
6.6%
In-hosp ACS, death
4.3%
5.2%
Readmit, fall, fatigue, function
Carson et a.l. NEJM. 2011
No differences
Costs of surgical RBC transfusion
$522
Austria
Activity-based cost
$154
RBC acquisition cost
$611
Switzerland
$194
$726
Rhode Island
$203
$1,183
New Jersey
$248
$0
$200
Shander et al. Transfusion. 2010.
$400
$600
$800 $1,000 $1,200
Getting the ground ready

Admin and doc buy-in (oh, and trust)
 Center
 Hospital



Clinical people who know their way around
medical documentation at the facilities
Access and IT resources
Simple (reproducible) data requirements
What we have done








Initial pitch(es) to admin and medical in support of
conservative transfusion
Confidentiality in writing
IT preparation to find the records we need
Record review
Data analysis and reporting
Multiple presentations of the data
Process development to the level they allow
Reaudit (just starting)
MVRBC RBC trigger audits
•
Descriptive manual chart audit of RBC units
given. Generally during a single quarter
•
Record ordering physician and specialty
•
Hemoglobin on admission, at time of 1st order
(i.e. “transfusion trigger”) and after transfusion
•
Documentation of bleeding in medical record
•
DRG, ICD-9
•
Hypothesis generating
Number of units ordered for 1st transfusion
3000
2500
16 audits at 14 hospitals
(or systems)
2000
1500
1000
500
0
0
1
2
3
4
6
Units
8
10
Hgb in nonbleeding patients during episode of care
16
14
12
10.6
10.5
Grams 10
9.1
8
8.3
TRICC
6
4
Admission 1st transfusion
Post-1st
DC
Hemoglobins associated with perioperative transfusions
20
18
16
14
12
12.1
Grams
10
8
10.6
10.4
8.4
FOCUS
6
4
Admission 1st transfusion
After 1st
DC
Hbg trigger with no bleeding: by hospital
14
12
10
Grams 8
8.2 8.3 8.1
8.1
7.7 7.7
8.5 8.4
8.1
8.6 8.8
8.5
7.9
8.4
7.6
7.9
TRICC
6
4
2
1
2
3
4
5
6
7
8
9
10 10A 11 12 13 14 9A
Hbg trigger for perioperative transfusion: by hospital
14
12
10
9.1
8.8
8.4
Grams 8
8.7
8.3
8.6
8.1
8.4
8.1
8.5
8.7
8.0
8.0 FOCUS
7.6
6
4
2
1
2
3
5
6
7
8
9
10 10A 11 12 14 9A
Caveat emptor

Reliable as our ability to find info in the record

Confounders (e.g. cardio-respiratory
compromise, severity of illness) not sought
(TRICC says don’t matter)

Acuity of intra-operative bleeding hard to assess

DRG/ICD-9 numbers too small for real analysis

Denominators can be hard to get, especially for
inter-hospital comparisons

Retrospective, manual audits
AIM-II software: “concurrent”, automated audits
AIM-II software
Conclusions

Transfusion in acute hemorrhage best left to judgment
at the bedside consensus guidelines

91% of non-bleeders transfused above TRICC

76% with operative bleeding transfused above FOCUS

Attention to non-bleeding & periop patients with an
emphasis on EBM will reduce RBC use

Discharge hemoglobin levels suggest that an emphasis
on single unit transfusions will be useful

Reduction = direct $$ and clinical savings
Barriers


Lack of basic training in transfusion medicine at all
levels
 “This is how Dr. Osler taught me to do it…”
 “My patients are sicker…”
 “I’ve never seen TRALI…”
Resources for real-time decision support and
intervention
 Process
 IT support (including AIM-II?)
 Clinical (“real docs”) champions
Barriers





Hospital-acquired infections
Falls
Med errors
Readmissions etc., etc.… ad nauseum.
This is about getting on the priority
menu for resources (people and time)
 (TJC was supposed to fix this)

similar documents