Who Needs to be Transfused… and Who Does Not?
Jerry E. Squires MD, PhD
Department of Pathology and Laboratory
Medicine
Medical University of South Carolina

A RISK / BENEFIT DECISION
What are these risks??

Risks of
Transfusion
Transfusion-
Transmitted
Infections
Transfusion
Reactions
Mis-
Transfusion

• The Potential List is LONG
– HIV
– HBV
– HCV
– HTLV
– Syphilis
– WNV
– The “others:”
• Chagas
• Parvo B19
• Malaria
• Babesiosis
• Etc., Etc., Etc.
• The Actual Risk is SMALL
– Hepatitis C 1:1,935,000
– Hepatitis B 1:400,000
– HIV 1:2,135,000

Risk Comparison

Reaction Type Predominant
Symptoms
Allergic Urticaria
Anaphylactic
Febrile
TRALI
TACO
Acute
Hemolytic
Urticaria
Hypotension
(shock)
Fever (>1 o C)
Resp. Distress
Pul. Edema
Resp. Distress
Pul. Edema
Fever
Pain
Hemolysis
Hypotension
Cause Treatment/
Prevention
Type 1
Hypersensitivity
Antihistamine
IgA Deficiency Supportive / IgA deficient products or washed
Cytokines in store products
WBC/HLA aby in donor product
Volume
Antipyretic
LR Products
Supportive
Red Cell Aby
Diuresis
Manage I/O
Patient/Sample
Identity
Prognostic
No sequellae
Must receive
IgA deficient products
No sequellae
No sequellae
No sequellae
Frequency
1:100
1:20,000
1:300
1:5000
?????
1:33,000

Serious Hazards of Transfusion (SHOT)
* Voluntary reporting system for adverse reactions in UK
DTR (28)
4.6%
PTP (2)
0.3%
ATR (68)
11.1%
TRALI (23)
3.8% TTI (3)
0.5%
• IBCT????
– Wrong blood (18%)
• ABO, Rh, “luck”
– Pre-transfusion testing error (4.5%)
• Aby screen, DAT, wrong sample
– Transplant blood type error (0.5%)
• ABO error
– Error in specification (29%)
• Irradiation, CMV, Antigennegative
– Inappropriate transfusion (14%)
• Wrong component transfused
– Unsafe transfusion (16%)
• Improper storage, outdated
– RhIg administration error (18%)
• Late, wrong patient, outdate
IBCT (485)
79.6%

Transfusion-Transmitted
Disease
• Overall risk of TTI now less than 11.5/1,000,000
• Risk of HIV and HCV now approximate 1:2,000,000
• Risk reduction is due to donor screening and testing improvements
Transfusion Reactions
• Allergic and febrile reactions remain the most common
• The risk of TRALI is now approximately 1:5000 and is the most common cause of transfusion associated death reported to the FDA
• Male predominant plasma products reduce the risk of
TRALI
Mis-Transfusion
• Mis-transfusion due to error is the most common type of adverse transfusion problem reported to national hemovigilance programs (~1 : 6000 transfusions)
• Most acute hemolytic reactions resulting in death result from ABO incompatibility which is usually due to mistransfusion

• Typically, discussions of transfusion risk centers on the 3 areas that have already been mentioned:
– Transfusion Reactions
– Transfusion Transmitted Disease
– Mis-Transfusion
• But, is there another transfusion risk that should be added into the risk / benefit equation?

• TRICC: Transfusion
Triggers in Critical Care
(Hebert PC et al, 1999)
– RCT (1994-1997)
– Liberal Txf arm:
• 420 patients
• Txf trigger 10.0 g/dL
– Restrictive Txf arm:
• 418 patients
• Txf trigger 7.0 g/dL
– Primary end-point: 30 day mortality from all causes
Outcomes:
* 30-day mortality similar in both arms
* Mortality advantage for restrictive txf. for patient <55 y or APACHE score 2 or less

Cancer Recurrence

of Transfusion
• Every 2 seconds someone in the US gets a transfusion
• 30,000,000 blood components are transfused every year in the US
• 4,500,000 people are transfused in the US every year
• 1 out of 7 hospital admissions gets a transfusion
So, in spite of the risks— small though they may be—we must think that transfusion is providing some BENEFIT to our patients…..

• It is estimated that as many as 25% of the red cell transfusions in the US are unnecessary.
• The question is not whether transfusion is required in the care of many patients…
• The question is which patient should be transfused; or in which patient will a transfusion be potentially life-saving and in which patient will a transfusion be lifeshortening?

Who Needs It?
• Patient evaluation
– Organ ischemia (CV disease)
– Patient coagulopathy
• Laboratory evaluation
Hgb. < 6 • RC usually indicated
Hgb. 6-10 • RC used based on clinical setting
Hgb.>10 • RC rarely indicated
• Estimated blood loss
– Visual inspection of surgical field
– Sponge counts
– Suction ASA Guidelines
Anesthesiology 2006

• Methods to reduce RC use
– Anemia
• Tolerance of lower Hgb
• Pharmacologic approach
– One unit at a time
• Reduce “2-unit” transfusion orders without Hgb/Hct
• Joint Commission guidance
– Reduce blood draws
• Iatrogenic anemia (ICU patients ~45 mL/day)

Final Considerations
• Evidence of benefit from RBC transfusion is hard to find
• Most benefit is assumed and not clinically proved
• Some patients benefit from blood transfusion, but we need to do a better job of determining who they are
• Giving MORE blood is NOT better
• Many red cell transfusions are probably unnecessary
• Patients transfused when it is unnecessary get all the
RISK and NO BENEFIT

Who Needs It?
• What do you hope to accomplish?
– To prevent or stop bleeding due to thrombocytopenia
• Practice Guidelines:
– Anesthesiology 2006; 105: 198-208
– “In surgical or obstetric patients with normal platelet function, platelet transfusion is rarely indicated if the platelet count is known to be greater than 100,000 X
10 9 ”
– “…and is usually indicated when the platelet count is below 50 X 10 9 ”

Who Needs It?
• Prophylactic Platelet Transfusion:
– Transfusion of platelets to non-bleeding patients with “low” platelet counts—to prevent thrombocytopenic hemorrhage
• The Questions:
– Is prophylactic platelet transfusion necessary?
– If so, what is a safe and effective platelet transfusion trigger?

Relationship Between Platelet Count and Bleeding
• Gaydos LA et al, 1962
– 92 nontransfused thrombocytopenic patients
– % days with bleeding
• Slichter SJ and Harker LA, 1978
– 20 aplastic thrombocytopenic patients
– Fecal blood loss

Are Prophylactic Platelet Transfusions Necessary?
• Friedmann, AM et al (2002)
– Multiple logistic regression analysis of the frequency of bleeding as a function of platelet count in 2942 thrombocytopenic patients
– Conclusion: first morning platelet count or lowest daily platelet count did NOT correlate with bleeding frequency
• Wandt, H et al (2006)
– A comparison of therapeutic versus prophylactic platelet transfusion in BMT patients
– Conclusion: therapeutic transfusion resulted in NO increase in bleeding episodes (and reduced platelet use by
50%)

Who Needs It?
• If prophylactic platelet transfusions are used in non-
bleeding patients…
– What is a safe and effective platelet transfusion trigger?
1.
Rebulla P et al. 1997
– Adult patients with acute leukemia in first remission induction; randomized into 2 groups:
– Lower threshold: 10 X 10 9 / L
» 3.1% of days with significant bleed
» 21.5% decrease in plt use
– Higher threshold: 20 X 10 9 /L
» 2% of days with significant bleed
2.
Wandt H et al. (2006)
– 105 pateints with acute leukemia; randomized into 2 groups:
– Lower threshold: 10 X 10 9 /L
» 17% of patients with bleeding complications
– Higher threshold: 20 X 10 9 /L
» 18% of patients with bleeding complications

• Slichter S 2007:
– “…the consensus of medical opinion is that a plt count of at least 50 X 10 9 /L should be maintained.”
– “Unfortunately there are no definitive studies to substantiate this plt transfusion trigger.”
– “…patients with intracranial bleeding and during and following neurosurgical procedures should have plt counts maintained at >100 X
10 9 /L”
• Platelet Transfusion in
Patients Undergoing Invasive
Procedures
Bishop et al (1987)
 95 patients with acute leukemia undergoing 167 surgical procedures
 70% of procedures were classified as “major” (e.g. laparotomy, thoracotomy, hip replacement, AK amputation)
 Results: no procedure-related deaths or excess bleeding when the platelet count ≥ 50 X 10 9 /L

• Current Platelet Transfusion Recommendations:
– Invasive procedures:
• 50,000/µL
• Neurologic procedures? (100,000/µL????)
– Prophylactic (nonbleeding patient):
• 10,000/µL
• Fever, Sepsis may benefit from a “higher” trigger

Another Aspect of Bleeding Risk
• Hematocrit and Bleeding
Risk:
– Valeri et al, 2001
• The hematocrit may play a role in bleeding risk particularly in thrombocytopenic patients
• In normal volunteers, plateletpheresis which reduced platelet count significantly did
not affect bleeding time (right bars)
• But the removal of red cells reducing the Hct from 41% to
35%, almost doubled the bleeding time
• Conclusion: maintain the Hct. In thrombocytopenic patients

Who Needs It?
• The US“love affair” with plasma:
RBC and FFP Use in US (1982-2001)
Year
1982
1989
RBC
(X 10 6 )
11.5
12.1
FFP
(X 10 6 )
1.9
2.2
FFP :
RBC
1 : 6.6
1 : 5.5
1994
1999
2001
11.1
12.4
13.9
2.6
3.3
3.9
1 : 4.3
1 :3.7
1 : 3.6

Who Needs It?
US
NZ
Country
France
Annual RC and FFP Use
RBC Unit
(X 10 3 )
RBC Units
(per 1000 population)
2,100 34.4
FFP Units
(X 10 3 )
FFP Units
(per 1000 population)
242 4.0
FFP:RBC
1 : 8.5
UK 2,700 45.3
385 6.5
1 : 7.0
13,900
125
49.5
32.1
3900
21.3
13.9
5.5
1 : 3.6
1 : 5.9

Who Needs It?
American Society of Anesthesiologists
Practice Guidelines (1996):
1.
Urgent reversal of Warfarin Therapy
2.
Correction of known coagulation factor deficiencies when specific concentrates are unavailable
3.
Correction of microvascular bleeding in presence of elevated (>1.5 x normal) PT or PTT
4.
Correction of microvascular bleeding secondary to coagulation factor deficiency in patients transfused with more than one blood volume when PT and PTT cannot be obtained promptly
5.
FP should be given in doses calculated to achieve a minimum of 30% of plasma factor levels
6.
FP is contraindicated for augmentation of plasma volume

Who Needs It?
• Assumptions in the Use of Plasma:
– Abnormalities of PT / INR correlate with the risk of bleeding
– Plasma transfusion can correct the abnormal PT /
INR thereby reducing (or eliminating) the risk of bleeding
Are these assumptions correct?????

• Does a prolonged PT (INR) correlate with a risk of bleeding?
IT DEPENDS !!!!!

Does the PT / INR Predict Bleeding Risk?
• Wahab OI et al, 2006
– Compared estimated blood loss in 121 patients with mildly elevated PT / INR
– Result: no correlation between PT / INR and blood loss
• Ewe K, 1981
– Compared liver bleeding time to PT in patients undergoing liver biopsy
– Result: no correlation between PT and liver bleeding time

Does the PT / INR Predict Bleeding Risk?

Does the PT / INR Predict Bleeding?

Coagulation Factor Hemostatic Levels for Surgery (%)
VIII
IX
X
II
V
VII
20-30
20
10
40
30
20
Sensitivity of PT
Reagent (%)
28
52
44
N/A
N/A
49
Bottomline: the PT (INR) will be prolonged even when there are adequate levels of coagulation factors to mediate normal hemostasis

• Can the PT (INR) be used as an indicator of bleeding risk?
• Agarwal et al (2012)
– Method:
• 20 consecutive acute liver failure patients
• Measured PT, TEG, individual pro- and anti-coagulatant factors, thrombin generation

• Agarwal et al 2012
– Results:
• PT significantly prolonged (50.7 s ±7.2) and did not correlate with TEG results
• TEG: 20% hypocoagulable; 45% normal; 35% hypercoagulable
• Reduction in plasma levels of BOTH procoagulants and natural anticoagulants but a significant increase in plasma Factor VIII and vWF
• NO bleeding and NO blood transfusions

• Agarwal et al (2012)
– Comments:
• “…the perception of a bleeding diathesis with progressive
ALF as indicated by standard clotting tests (PT) is not substantiated by a more comprehensive assessment…using
TEG”
• TEG indicates a “more balanced” coagulation state in these patients
• PT affected by VII, X, V, II, fibrinogen and does not assess anticoagulant, platelet and endothelial contributions to coagulation
Bottomline: Perhaps the PT (INR) while assessing coagulation factor deficiencies in bleeding patients, but it may also give an overly simplified assessment of bleeding risk

• Given that there is a question as to whether
Plasma Transfusion will actually PREDICT a patient’s bleeding risk (especially at INR levels
<2)….
• What is the capacity of Plasma Transfusion to
CORRECT a prolonged PT / INR?

Does Plasma Correct a Prolonged PT/INR?
• Youseef et al, 2003
– 80 patients with cirrhosis and elevated PT
– Indications for plasma:
• 41% prophylaxis
• 59% active bleeding
– Dose:
• 75% received 2-4 units
• 25% received >4 units
– Result: with plasma, 89% of patients failed to correct PT

Does Plasma Correct a Prolonged PT / INR?
• Holland LL and Brooks
JP, 2006
– In adult and pediatric patients the lowering of an INR less that 1.7 with
FFP infusion is minimal

Does Plasma Correct a Prolonged PT/INR?
• Wahab OI et al, 2006
– 121 patients with a mildly elevated INR (1.1-
1.85)
– The transfusion of plasma to patients with mildly elevated PT / INR results in partial normalization of PT in a minority of patients and fails to correct the PT in
99% of patients

Transfuse Transfuse

• All transfusions carry some risk, but the most significant risks may be:
– Mis-transfusion
– Patient outcomes
• Your best guide to transfusion is the patient’s clinical condition; laboratory values (hemoglobin, PT, INR, platelet count) are of marginal use at best
• It is estimated that up to 25% of RBC transfusions are unnecessary, the question therefore is not about eliminating transfusion, but rather about choosing who and when—when will transfusion save a life and when might it shorten it!