Blood Management
The Impact of Blood Transfusions on
Surgical Outcomes
Peter Mazolewski, MD, FACS
LtCol USAR
Cynosure Summit
May 21, 2012
Disclosures
• NO pharmacy
• Paid consultant for Strategic Healthcare
Group
– Comprehensive Provider for Blood
Management Consultation
2
What is Blood Management?
• Blood Management is evidence-based transfusion
practices which result in safe and efficient use of
blood products
• Why Blood Management?
–
–
–
–
–
Inappropriate and excessive transfusions
No formal blood management training for clinicians
Measurable risks associated with EACH transfusion
Judicious use of blood reduces patient complications
Blood is a limited resource
4
What We Have Learned is
Not What We Now Know
risk
BENEFIT
• What we learned in residency and medical
school is far different than what the present
literature states
5
Current Understanding of
Transfusion Practices
RISK
benefit
In most clinical situations in which we are
administering blood, the risk of transfusion far
exceeds the benefit of transfusion
6
Patient Populations
• Hemodynamically stable patients with blood
component deficiencies, i.e. anemia, low platelets,
abnormally elevated INR
• Acutely injured trauma patients with immediate life
threatening hemorrhage
– Consider GI hemorrhage which is life threatening
– Consider Post operative life threatening bleeding
– Consider person with unresponsive lactic acidosis
7
Changing Dogma Creates Confusion
Two developing dogmas which are different
than what we had learned previously:
– Hemodynamic stable anemic pts even with cardiac
disease:
OBSERVE
– Acutely bleeding trauma: EARLY RESUSCITATION with
BLOOD COMPONENTS
• Then turn off after resuscitation
8
How Can We Make a Change?
• Medicine is always changing—not only at
periphery, but also at its core
• Treatment with blood is only one example
• Other examples
–
–
–
–
Hyperventilation in Head Injury
B-blockade in CHF
Hormonal Replacement Therapy
Radiation Therapy in Breast Cancer??
• But this particular change is going to be
difficult
9
Why Difficult to
Change?
• Crosses all specialties and all doctors who
were taught the same thing:
risk
BENEFIT
10
Risks (Historical)
• Primary Infectious (vs. secondary)
– Viral (HBV, HCV, and HIV)
– Bacterial (plts, 1:1000 – 1:2000)
• Transfusion Reactions
– Fever, Flushing
– Tachycardia
– Aggressive Immune response that is life threatening
(RARE)
• Perception of “Episodic” Risk
– Few and far between
– Clinicians have not taken seriously--in the sense that
the benefit always outweighs these episodic rare
complications
11
Risks (Current)
• TRIM (Transfusion Related ImmunoModulation)
– Immunosuppression
• Secondary Infection
– Pneumonia, wound site infections, central line sepsis, etc
• Reduced Survival in Cancer?*
• Increased ICU and Hospital length of stay
– Immunostimulation?
• TRALI (Transfusion Related Acute Lung Injury)
• ARDS, MODS
• Late Mortality, most notably from cardiac literature
– Vs. Early Mortality which may be also mechanical in nature (plugging, acute
pathophysiology)
– Dose Response Pattern (as opposed to episodic)
• Each transfusion carries a measurable insult
• Insults increase with number of transfusions
• Also dose-dependent with respect to storage time
12
*Vamvakas EC. Transfusion 1996.
Benefit (Historical)
• Enhanced Oxygen Delivery
• Improved overall Morbidity and Mortality
• Prevent Cardiac morbidity (angina and MI)
13
Paradoxical Reality
• While O2 delivery (DO2) is increased, tissue oxygen
consumption (VO2)either stays the same or worsens*
– VO2 independent of D02
• “RBC STORAGE LESION” (This is NOT fresh BLOOD!!!)
– Loss of 2,3 DPG which shifts Oxy-hgb curve making it difficult to unload
oxygen in the tissues
– Discoid RBC now spherical, broken (rheologic properties)^
– Increased RBC clumping and aggregation
– Decoupling of NO-Hgb** (free Hgb scavenges also NO)
– Pro-Inflammatory Nature of Stored Blood
– Which is pro thrombotic in nature
» Increase 3-6fold PAI after 35 days, increased ADP which stimulates
platelets, Nitric Oxide inhibits platelet activation
*Tsai et al, Transfusion 2004
^Almac et al, Best Pract Res Clin Anesth 2007
**Crawford et al. Blood 2006. Reynolds et, Proc
Natl Acad Sci 2007
Quality Issues: Storage Defects and
Microvascular Perfusion
 Decreased 2,3- DPG, ADP
 Build-up of cytokines, free
Hgb, K+, debris (BRMs) 1,2
 Poor deformability3
Concentration (pg/ml)
200
150
IL-1
IL-8
IL-6
*
100
***
*
*
50
0
Day 1
Day 10
Day 20
Day 30
Day 40
3
Kristiansson- ActaAnesthScand 1996;40
Fransen- Chest 1999;116
3 Hovav- Transfusion 1999; 39
1
2
15
Stored
PRBCs
behave in
this manner!
Anemic
patients with
NL RBCs can
at least
behave in
these ways
Schechter, Gladwin, NEJM April 10, 2003
Tissue oxygenation after
Transfusion In Trauma
Kiraly. J Trauma. July 2009
18
Blood Transfusion Video
Benefit (Historical)
• Enhanced Oxygen Consumption (NO!!!)
• Improved overall Morbidity and Mortality
– Trauma is exception with MTP (Massive
Transfusion Protocol)
• Less Cardiac Events
20
Current Risk/Benefit Ratio
RISK
BENEFIT
21
Risks (Current)
• TRIM (Transfusion Related ImmunoModulation)
– Immunosuppression
• Secondary Infection
– Pneumonia, wound site infections, central line sepsis, etc
• Reduced Survival in Cancer?*
• Increased ICU and Hospital length of stay
– Immunostimulation?
• TRALI (Transfusion Related Acute Lung Injury)
• ARDS, MODS
• Late Mortality, most notably from cardiac literature
– Vs. Early Mortality which may be also mechanical in nature (plugging, acute
pathophysiology)
– Dose Response Pattern (as opposed to episodic)
• Each transfusion carries a measurable insult
• Insults increase with number of transfusions
• Also dose-dependent with respect to storage time
22
*Vamvakas EC. Transfusion 1996.
Immunosuppression of
PRBCs
Effect of blood transfusions
on subsequent kidney transplants.
Opelz G, Sengar DP, Mickey MR, Terasaki PI. Transplant
Proc 1973; 5:253-9
Opelz reevaluated again in 1997 (prospective randomized) with modern
immunosuppressive therapy and still found a significant impact of PRBC
transfusion on renal allograft acceptance
Did this Information Initiate Quality
Controls in Blood Banking?
• I argue “NO.” (With respects to immunomodulation)
• The only current measure of “quality” with regards to RBC
transfusion is if 70% of the RBCs survive in the circulation
at 24hrs post transfusion. (Developed in 1940’s)
• Except for leukoreduction (which is not universal),there has
been no attempt to diminish or to measure cellular debris
or mediators
• Therefore it is not surprising that we see a multitude of
complications as a result of PRBC transfusions.
24
Infection and Transfusion in
Surgical Procedures
* Significant in
N=589
*
N=868
*
*
N=109
*
N=267
N=343
N=197
*
Colorectal
Adapted from: Despotis G.
Murphy
Vignal
Tartter
Spinal
Jensen
Trauma
Jensen
N=238
Triulzi
40
35
30
25
20
15
10
5
0
Transfusion
Multivariate Analysis
Edna
Incidence of Infection (%)
No Transfusion
Cardiac
25
•
•
•
•
4892 Patients at initial ICU admission
1518 mechanically ventilated >48hrs
VAP in 20.5%
Multivariate Analysis transfusion
independently increased risk for VAP
• OR 1.89
26
Adjusted hazard ratio
Transfusion and Nosocomial Infections
1.8
1.6
serious bacterial
infection
1.4
pneumonia
1.2
n=9598
1
1
2
3
4
5+
Units Transfused
Carson et al, Transfusion 1999;39
27
Transfusion and
Length of Stay
25
22.5
21
20
18
17
17
16
15
Vent
Time
ICU
Stay
Hospital
Stay
Days
14
12.5
12
10
10
10
8
7
6
n = 576
5
2.5
0
0
1-2
3-4
5-6
>6
Units Transfused
Shapiro et al, J Trauma 2003;55
28
1
2
Basran et al- Anesth Analg 2006;103
Koch et al- NEJM 2008;358
LOS
In Hospital Mortality
ARF
Quality Issues: Duration of Blood Storage
and Outcomes
29
Transfusion Requirements in Critical Care
(TRICC) Study
Restrictive
<7 g/dl
Liberal
<10 g/dl
p-Value
Overall 30-day
Mortality
18.7%
23.3%
p=0.11
Pts < 55 years
old
5.7%
13.0%
p=0.02
APACHE II score
≤20
8.7%
16.1%
p=0.03
Cardiac Disease
20.5%
22.9%
p=0.69
In-hospital
mortality
22.2%
28.1%
p=0.05
Outcomes
30
Is Anemia beneficial??
• Yes it is.
• Anemia lessens viscosity and therefore lessens
resistance.
• Diminished venous resistance results in improved
venous return.
• Diminished arterial resistance results in lower SVR.
• Together both these mechanisms result in a
significantly higher cardiac output.
Tsai, et al. Transf Apher Sci. 2010 August.
31
CRIT Study
• 2004, 284 ICUs, 4892 Pts in prospective, observational
cohert
• Transfusion was an independent predictor of ICU length of
stay, increased complications, and increased mortality
(1.65 adjusted mortality ratio) with propensity scoring (1059
pts transfused to 1059 matched no transfusion)
• 65% pts received transfusions with hgb>8.0 gm/dl
• <10% pts received transfusions with hgb<7.0 gm/dl
Corwin et al. Crit Care Med. 2004. Vol 3, No 1
Transfusion and Mortality in Trauma
(Pre MTP era)
33
Malone et al. J Trauma. May 2003
NSQIP Data (adjusted odds ratio)
0 PRBC
1PRBCS
2PRBCS
# pts
120,389
1,343
1,903
SSI
Reference
1.02
1.25
1.12
1.04
1.24
1.25
1.29
1.5
1.32
1.38
Reference
UTI
Pneumonia
Reference
Sepsis or
Septic Shock
Mortality <30
days
Reference
Reference
Bernard AC, et al. jamcollsurg 2008
Surgical Outcomes: one unit
transfusion
Ferrarris, et al. Arch Surg. 2012;147:49-55
35
Benefit (Historical)
• Enhanced Oxygen Delivery (NO!!!)
• Improved overall Morbidity and Mortality
(NO!!!)
– Because we are transfusing too many
patients where there is simply no benefit. If
there is a benefit, we cannot see it.
– Overall, trend for outcomes is significantly
worse
– Trauma is exception with MTP (Massive
Transfusion Protocol)
Less Cardiac Events
36
Current Risk/Benefit Ratio
RISK
benefit
37
Blood Transfusion &
Percutaneous Coronary Intervention
Doyle,et al. JACC. 2009. Vol 53
39
Cardiac Surgery and Blood
Transfusion
• Northern New England Cardiovascular Disease
Study Group
• 8 Centers evaluated 9079 consecutive patients
between 2001-2004
• Cox Proportional Hazard and Propensity Analysis
• 3254 patients exposed to one or two units PRBCs
and compared to those who had none
• 16% increased mortality out to 6mos
• Surgenor et al. Cardiovascular Anesthesiology. 2009
42
TRACS
• Prospective Randomized Trial in Cardiac Elective Surgery
• JAMA Oct 13, 2010
• Compared Liberal (>10gm/dl) and Restrictive (>8gm/dl)
strategies in Blood Management
• 250pt each arm
• 78% (liberal) vs 47% (restrictive) pts transfused
• No difference in outcomes
• When looking at those transfused, independent risk factor
for clinical complications or death @ 30days (HR for each
add’l unit 1.2)
• Avoid transfusion solely to correct low hgb
FOCUS Trial
• 2016 high risk cardiovascular pts undergoing hip
fracture surgery prospectively randomized into a
restrictive vs. liberal transfusion policy
• Restrictive 8gm/dl and symptomatic
• Liberal 10gm/dl
• Median Age 82
• Non-inferiority study
Carson, et al. NEJM. Dec 2011.
44
FOCUS Trial—RESULTS
• Primary Outcome: inability to walk 10ft
unassisted at 60d and mortality
– 34.7% restrictive and 35.2% liberal
• All other secondary measures of morbidity did not
reach statistical significance (trend towards
restrictive group)
– Very similar percentages seen in other high powered
observational studies using propensity scoring.
• 65% less blood used in restrictive group
• 97% of patients transfused in liberal group
45
RBC Transfusion in CVT Surgery
and Quality of Life
• Patient’s perspective
• RBC and Component transfusion therapy and
functional health-related Quality of Life
• 7321/12536 pts over 4yr period (1995-99)
completed self administered DASI preop and
postop (6-12mos)
• Regression analysis showed that postoperative
functional status after CVT surgery was
incrementally worse with more PRCs and platelets
transfused
Kock, et al. Soc Th Surg. 2006
Benefit (Historical)
• Enhanced Oxygen Delivery (NO!!!)
• Improved overall Morbidity and Mortality
(NO!!!)
– Because we are transfusing too many
patients where there is simply no benefit. If
there is a benefit, we cannot see it.
– Trauma is exception with MTP (Massive
Transfusion Protocol)
• Less Cardiac Events (NO!!!)
– Actually more!!!
48
Current Risk/Benefit Ratio
RISK
benefit
49
Difficult Dilemma
• Anemia and acute coronary events carry a high
mortality
• However, transfusion has not been shown to
confer a benefit or improvement in survival
• IN FACT, many studies show that patients do
worse with transfusion
• What to do?
• This data should provide enough comfort for
clinicians NOT to feel need to transfuse in anemic
hemodynamically stable patients with cardiac
disease
Is Hgb 7.0 the trigger?
It is “the trigger to CONSIDER” transfusion…..
TASK FORCE-EAST AND
ACCM
• NO “TRANFUSION TRIGGER” based on hgb level
• Critically ill patients, HD stable anemia, except possibly those w/acute
myocardial ischemia, a “restrictive” strategy (hgb < 7gm/dl) is equally as
effective as a “liberal” strategy (hgb < 10gm/dl) LEVEL I evidence
• In critically ill pts on the ventilator, the above also holds true. LEVEL II
• In resuscitated, critically ill trauma pts, the above holds true. LEVEL II
• In critically ill patients with stable cardiac disease, the above holds true.
LEVEL II. (very important: prevention of ischemia not supported by
literature)
• RBC transfusions should be given as single unit. LEVEL II
• RBC transfusions should not be considered as an absolute method to
improve tissue oxygen consumption in critically ill. LEVEL II
• Napolitano Crit Care Med. Dec 2009.
Indicators for Considering RBC
Transfusion (in absence of bleeding)
Normovolemic anemia (Hgb≤7) WITH
signs or symptoms of inadequate
oxygen delivery
Napolitano et al. Critical Care Medicine 2009;37:3124-3157
Pape, A., et al: Blood Transfus 7:250-8, 2009
What are signs or symptoms of
inadequate oxygen delivery?
Signs
1. ScvO2 < 70% [nl=80%] (central
2.
3.
4.
5.
6.
7.
line)
SvO2 < 65% [nl=75%] (PA
catheter)
Low cerebral or tissue oximetry
Base deficit – ABG
Lactic acidosis - lab
ST changes - EKG
↓ LV contractility by trans esoph
echo
1.
2.
3.
4.
5.
SYMPTOMS
Mental status
alteration
Dyspnea
Chest pain
New arrhythmias
Tachycardia (VERY
RELATIVE)
-Main Provider Excuse
Pape, A., et al: Blood Transfus 7:250-8, 2009
Napolitano et al. Critical Care Medicine 2009;37:3124-3157
How Can we immediately
improve the situation?
• If we follow transfusion guidelines published by large
societies (and remove less blood), we will transfuse far
less blood (>50%) than currently given.
• This will diminish the need to store blood for 42 days and
perhaps we can survive with a smaller pool of blood which
is younger, fresher, and significantly safer (goal < 21days).
• In addition, we need to improve our science regarding
blood banking which ameliorates the effects of the
“storage lesion.”
57
Benefits (Current)
• Survival?
– Trauma. Yes, emerging in the setting of a Massive
Transfusion Protocol
• 1:1:1 ratio (FFP:RBC:Plts)
• But data emerging that 1:1:1 is resulting in excessive MODS
without survival benefit
• 1:2-3 Ratio considered optimal by some at this time…but
literature quickly evolving
• Evidence indicates that in the setting of MTP, less blood is used
on each pt during an entire hospitalization.
Dente, et al. J of Trauma, 2009
58
Massive Transfusion (MTP)
Improves Survival
• Instituted a MTP in 1:1:1 ratio PRBC:FFP:PLTS
• Prospective Non Randomized (“Unethical”)
• No faculty turnover and no other significant changes
in treatment
• Compared Pre-MTP with MTP
• 73 MTP vs 84 similar pre-MTP patients
• 24hr mort: 17% vs 36%, MTP vs Pre-MTP
• 30day mort: 34% vs 55%
• Early deaths from coagulopathic bleeding: 4/13
MTP, 21/31 Pre-MTP
Dente, et al. J of Trauma. 2009
59
MTP Questions Remain
• Optimal ratios of blood component resuscitation?
• Trying to mimic military experience of fresh whole blood
• Vanderbilt experience 3:2 PRBC:FFP, 5:1 PRBC:Plts
showed survival experience
• No dose dependent survival beyond these ratios
• Grady 1:1and 2:1for both FFP and Plts; showed similar
survival advantage for pts who received 2:1 and 3:1
• Late effects of excessive blood leading to later mortality?
MODS, ARDS, Infection, etc
• Call it early and diminish amount of uncrossmatched blood
• Turn it off immediately and institute restrictive approach
(<7gm hgb/dl) to blood component therapy
60
Current Risk/Benefit Ratio
RISK
benefit
61
THANK YOU
Correcting Coagulopathy–
Plasma
Why do we transfuse?
• Replace volume and coagulation factors during massive
transfusion
• Prevent further or future bleeding in patients
undergoing invasive procedures
• Reverse Coumadin therapy in patients with or without
bleeding
• Address isolated coagulation factor abnormalities
63
INR and FFP
• INR Established to monitor Coumadin effectiveness
• Estimated INR of FFP is 1.3 – 1.7 (mean about 1.4)
• Deitcher showed INR 1.3 – 1.9 mean factor:
–
–
–
–
FII 31-65%
FV 40-70%
FVII 22-60%
All levels consistent with adequate concentrations for
hemostasis
• Makes sense since “therapeutic” INR is at least 2.5 or
3.0 INR
Understanding Derivation
of INR
• Most thromboplastins in use in the US prior
to 2000, had an ISI of greater than 2.0
• Based on this understanding, equivalent
INRs pre-2000 and post-2000 was 1.5 =
2.25
• This certainly fits with the literature and
clinically
Tavares, et al. Transfusion April 2011;51:754-761
FFP Effect on INR
• Holland and Brooks
– INR change = .37 (pretransfusion INR) - 0.47
– INR > 1.7 before transfusion
• 50% of patients showed improvement in INR after FFP
transfusion
– INR 1.3 – 1.6
• Supportive treatment of underlying medical condition lowers
PT the same as FFP
FFP Effect on INR
• Abdel-Wahab, prospective transfusions in
patients with INR 1.1 – 1.85
–
–
–
–
–
324 plasma units given
121 patients
Less than 1% normalization of INR
15% corrected only half way to normal
No retrospective correlational to bleeding
INR and bleeding CVT Surgery
• Pawan, et al
– Cardiac Surgical Literature
– 4 groups based on INR: <1.5, 1.5-2.0, 2.02.5,>2.5
• No evidence of increased bleeding in group 2 pts,
i.e. those with INR 1.5-2.0
Roger Williams Hospital
Providence RI
• With the concept that 2.0 INR is the “new” 1.5,
they built an evidence based program around
this transfusion “threshold.”
• Dropped their FFP usage 80%!!!!
• Not one reported adverse event attributed to
bleeding or lack of use of FFP to prevent
bleeding
• Inpatient Mortality decreased over this time
period---cause and effect?????