Prospective monitoring of plasma and platelet transfusions in a large teaching hospital results in significant cost reduction
Plasma and platelets (PLTs) are often transfused to correct mild to moderately abnormal laboratory values. Our objective was to reduce unnecessary plasma and PLT transfusions to nonbleeding patients by prospective triage and education of end users in evidence-based hemostasis and transfusion medicine practices. Using the Parkland Memorial Hospital's transfusion service and admission database as the data source, this study comprises the comparison of transfusion data on plasma and PLT use between pre- (2000-2002) and posttriage (2003-2006) periods. Yearly transfusion and wastage data on red blood cells (RBCs), plasma, and PLTs and yearly hospital admissions, trauma visits, and surgical procedures were extracted retrospectively for the study. The study revealed that implementation of triage resulted in a significant reduction of plasma (60%) and PLT (25%) transfusions, saving more than $3,000,000 over 4 years. Prospective triage and evidence-based transfusion practice education reduced unnecessary plasma and PLT transfusions and health care costs.
Prospective monitoring of plasma and platelet transfusions in a
large teaching hospital results in signiﬁcant cost reduction_2413487..492
Ravindra Sarode, Majed A. Refaai, Karen Matevosyan, James D. Burner, Scott Hampton, and
BACKGROUND: Plasma and platelets (PLTs) are often
transfused to correct mild to moderately abnormal labo-
ratory values. Our objective was to reduce unnecessary
plasma and PLT transfusions to nonbleeding patients
by prospective triage and education of end users in
evidence-based hemostasis and transfusion medicine
STUDY DESIGN AND METHODS: Using the Parkland
Memorial Hospital’s transfusion service and admission
database as the data source, this study comprises the
comparison of transfusion data on plasma and PLT use
between pre- (2000-2002) and posttriage (2003-2006)
periods. Yearly transfusion and wastage data on red
blood cells (RBCs), plasma, and PLTs and yearly hospi-
tal admissions, trauma visits, and surgical procedures
were extracted retrospectively for the study.
RESULTS: The study revealed that implementation
of triage resulted in a signiﬁcant reduction of plasma
(60%) and PLT (25%) transfusions, saving more
than $3,000,000 over 4 years.
CONCLUSIONS: Prospective triage and evidence-
based transfusion practice education reduced unneces-
sary plasma and PLT transfusions and health care
Appropriate use of precious blood components
results in cost-effective transfusion therapy and
reduces transfusion-related complications.1
Physicians unfamiliar with evidence-based
transfusion guidelines transfuse blood components
inappropriately to correct mild abnormalities of coagu-
lation tests, thus contributing to adverse outcomes and
increased health care costs.2,3 This stresses the national
blood supply, which has decreased overall as a result of
increased use in complex medical/surgical procedures
and a decrease in blood donation caused by increasingly
stringent FDA blood safety regulations.4,5
Although transfusion-transmitted infectious diseases
have been dramatically reduced, other complications
persist, that is, transfusion-related acute lung injury
(TRALI), the most common cause of transfusion-related
deaths in the United States in the past 2 years,6and
transfusion-associated circulatory overload (TACO),
which is potentially fatal in the elderly and young chil-
dren.7Inadequate exposure to transfusion medicine
practices results in underrecognized and underreported
The Joint Commission requires hospitals to moni-
tor transfusion practices.9Unfortunately, this is usually
ABBREVIATIONS: BURC =Blood Utilization Review
Committee; PT =prothrombin time; PTT =partial thromboplas-
tin time; TACO =transfusion-associated circulatory overload;
TP =thawed plasma.
From the Department of Pathology and Division of
Hematology-Oncology, The University of Texas Southwestern
Medical Center, and the Department of Pathology, Parkland
Health and Hospital System, Dallas, Texas.
Address reprint requests to: Ravindra Sarode, MD, The Uni-
versity of Texas Southwestern Medical Center, 5323 Harry Hines
Boulevard, CS3.114, Dallas, TX 75390-9073; e-mail: ravi.sarode@
Received for publication April 23, 2009; revision received
July 20, 2009, and accepted July 23, 2009.
Volume 50, February 2010 TRANSFUSION 487
retrospective and occurs too late to effect change. Park-
land Memorial Hospital is a 960-bed, Level 1 designated
trauma center and teaching hospital for the University
of Texas Southwestern Medical Center (Dallas, TX). After a
comprehensive literature review, in 2003 the Blood Utili-
zation Review Committee (BURC) established transfusion
guidelines for blood components and prospective triage
mechanisms for plasma and platelets (PLTs) for non–
operating room orders.10-13 Our goal was to educate
clinicians who place orders, especially house staff, on
evidence-based transfusion guidelines, thus reducing
unnecessary plasma and PLT transfusions for correction
of mild coagulation test abnormalities. This 7-year study
analyzed pre- and posttriage plasma and PLT usage.
MATERIALS AND METHODS
BURC guidelines for plasma and PLTs
are given in Table 1; triage protocols
for plasma and PLTs are given in Figs. 1
and 2. To facilitate emergency plasma
transfusion, Parkland Memorial Hospi-
tal uses thawed plasma (TP), which is
fresh-frozen plasma (FFP) kept thawed
at 1 to 6°C. Once thawed, FFP must be
discarded after 24 hours, whereas TP
can be held for up to 5 days at 1 to 6°C.
Except for Factor (F)VIII, FFP and TP
are equivalent in terms of factor con-
centrations.14 Because FVIII, an acute-
phase reactant, is generally elevated in
patients with coagulopathy, including
liver disease, we use TP to avoid delay
in transfusion and to reduce wastage.
Blood bank personnel reviewed the
most recent, relevant laboratory values,
that is, prothrombin time (PT), interna-
tional normalized ratio, partial throm-
boplastin time (PTT), and PLT count.
Orders not meeting guidelines were
triaged to the on-call transfusion medi-
cine resident (pathology residents on
transfusion medicine service rotation
during weekdays and on-call residents
after hours and on weekends) who con-
tacted the ordering physician to discuss
the clinical scenario and transfusion
indications. They mutually decided to
1) transfuse the product; 2) cancel the
order; 3) reduce the quantity of blood
component; 4) transfuse a different
product, for example, cryoprecipitate
rather than plasma or PLTs; or 5) use
alternative therapy, for example, pro-
thrombin complex concentrate rather
than plasma for warfarin reversal15 and desmopressin
or cryoprecipitate for uremic bleeding.16 Appropriate
additional laboratory tests were recommended to assess
the hemostatic defect and evaluate replacement therapy.
In complex situations the transfusion medicine fellow and
attending were also consulted. All calls were discussed
at transfusion medicine morning rounds, where further
plans for transfusion therapy were made and signed off
by an attending. The data were presented at bimonthly
This study includes annual data on blood compo-
nents used from 2000 to 2006 on red blood cells (RBCs),
PLTs, and plasma as well as data on annual hospital
admissions, trauma, and surgical procedures.
TABLE 1. Transfusion guidelines
1. Laboratory screening criteria:
a. INR >1.6 in presence of bleeding (with or without prolonged PTT) or
b. Isolated, prolonged PTT due to FXI deﬁciency
2. Thrombotic thrombocytopenic purpura
3. Neurosurgical patients in the perioperative setting with an INR >1.4
4. Reversal of warfarin effect (prothrombin complex concentrate preferred)
5. Bleeding patients with coagulopathy
B. PLT transfusion
1. Prophylaxis against bleeding—PLT count <10 ¥109/L
2. Neonate—PLT count <50 ¥109/L
3. Bedside procedure—PLT count <50 ¥109/L
4. Kidney or liver biopsy—PLT count <50 ¥109/L
5. Bronchoscopy without biopsy—PLT count <50 ¥109/L
6. Bronchoscopy with biopsy—PLT count <75 ¥109/L
7. Intra-/postoperative bleeding—PLT count <50 ¥109/L
8. Clinical bleeding with dysfunctional PLTs:
PLT count <50 ¥109/L (medical)
PLT count <100 ¥109/L (surgical)
9. Neurosurgery—PLT count <100 ¥109/L
10 TP received
INR result unavailable/
or >12 hours old?
Operating room? 1. Notify resident
2. Fill order and wait for
INR < 1.6?
*INR > 5.0?
> 6 TP filled within
previous 12 hours?
Fig. 1. TP triage protocol. *Suprathermic INR generally seen with warfarin therapy
and hospital guidelines recommend prothrombin complex concentrate.18
SARODE ET AL.
488 TRANSFUSION Volume 50, February 2010
The chi-square analysis was performed to compare
the baseline period of 2000 through 2002 to 2006 for the
percentage of decreased use and wastage for RBCs, TP,
and PLTs. A software program was used (WINKS, TexaSoft,
Cedar Hill, TX).
Other measures undertaken: R. Sarode provided
hospitalwide education, including grand rounds on
hemostasis, laboratory evaluation of hemostasis, pitfalls
of unnecessary testing, appropriate blood component
therapy, and adverse effects of transfusions emphasizing
TRALI. We developed several protocols for appropriate
and timely blood component therapy (massive trans-
fusion protocol, warfarin reversal protocol, etc.) as well as
for utilization of other hemostatic agents for better hemo-
stasis, for example, prothrombin complex concentrate
for warfarin reversal and recombinant FVIIa in the
massive transfusion protocol. The hospital’s transfusion
medicine Web site provided additional educational
information and all protocols.
Table 2 shows transfusion data of RBCs, plasma, and PLTs
from January 2000 through December 2006. There were a
total 10,692 plasma units transfused in 2000. With triage
implementation in 2003, plasma transfusions decreased
signiﬁcantly by 29% (7597 units) and continued to
decrease over the next 3 years reaching a 60% decrease
(4288 units) in 2006. Concurrently, plasma wastage was
reduced by 93% from 675 units in 2000 ($50,625) to only 45
units in 2006 ($3375; Table 2); this was attributed to the
change from FFP to TP in March 2001. From 2003 through
2006 the mean savings in actual blood bank charges was
$231,000 per year ($55/unit). Compared to 2000, $352,000
was saved in 2006 alone. The total savings from reduced
wastage was $143,000 (approx. $35,750/year), with further
savings in blood administration costs (technician time,
transfusion sets, nursing time, premedication, etc.) of
$100 per unit, or $1,680,900 (approx. $420,000 annually)
from 2003 through 2006.
There were 1518 PLT “doses” (pooled random-donor
PLTs or apheresis PLTs) transfused in 2000; this gradually
decreased to 1133 doses (25%) in 2006 after triage
(Table 2). Total savings in blood bank acquisition costs
for posttriage PLTs (last 4 years) was $804,650 (approx.
$201,163 annually) when compared
with pretriage years and savings from
avoiding PLT wastage of $148,500
($37,125/year). PLT administration cost
savings ($100/dose) reached $146,300
for the same period of time ($36,576
Although the approval rate was
63% for TP and 60% of PLTs—mostly
for presurgical coagulopathy or for
low PLT counts (>50 ¥109/L for surgery
and >100 ¥109/L for neurosurgery)—the
actual number of triages were reduced
from 40 to 50 per week in 2004 to three to
four calls per week in late 2008, most
likely again due to a successful educa-
tional effort. We also denied requests
for inappropriately ordered product,
for example, PLTs instead of plasma and
Two doses of
1. Issue third dose
2. Notify resident
>20 ¥ 109/L
Two doses of
within 24hr period?
Fig. 2. PLTs triage protocol.
TABLE 2. Yearly blood components use and wastage
Year RBCs Wastage (%) TP Wastage (%) PLT doses Wastage (%)
2000 22,274 174 (0.78) 10,692 675 (6.3) 1518 130 (8.6)
2001 21,516 167 (0.78) 9,827 316 (3.2) 1518 116 (7.6)
2002 20,052 169 (0.84) 10,325 145 (1.4) 1302 56 (4.3)
2003 17,826 116 (0.65) 7,597 87 (1.1) 1141 88 (7.7)
2004 18,657 70 (0.38) 7,370 116 (1.6) 1264 98 (7.8)
2005 18,601 126 (0.68) 6,704 35 (0.5) 1071 33 (3.1)
2006 19,475 96 (0.49) 4,288 45 (1.0) 1133 31 (2.7)
Percentage decrease (over 7 years)* 12.6 45.0 59.9 93.3 25.4 76.2
* Comparing 2000 through 2002 baseline data to 2006 for use and wastage for all components, p <0.001 (chi-square).
TRIAGE OF PLASMA AND PLTS
Volume 50, February 2010 TRANSFUSION 489
vice versa, TP for prolonged PTT due to heparin, or lupus
anticoagulant. The reasons for nonapproval for plasma
and PLTs are given in Figs. 3 and 4, respectively.
RBC transfusion and admission data
We used RBC transfusion data from 2000 to 2006 as a
measure of blood bank services. Because of a very high
volume of RBC transfusions compared to other compo-
nents and well-established transfusion triggers, we do not
triage RBC orders. From 2000 to 2006, RBC transfusions
also decreased by 12.6% (secondary to “appropriate
transfusion therapy” educational effort). There was a 45%
wastage reduction during this same period (Table 2). This
reduction occurred despite a 2.4% increase in hospital
admissions in 2006 compared to 2000, with increased
trauma admissions (14.7%) and surgical procedures
(17.3%; Table 3). The gradual reduction in both PLT
and RBC wastage is attributed to better inventory
management; however, increased wastage of RBCs in 2005
and plasma and PLTs in 2004 is due to an overlap in
switching blood suppliers.
This study shows signiﬁcant reduction in transfusion
of plasma (60%) and PLTs (25%) over a 4-year period
after implementation of a prospective triage process.
This occurred despite concurrent rise in hospital admis-
sions, trauma, and surgical procedures. Wastage of pre-
cious blood components was commensurately decreased.
These combined measures resulted in dramatic savings
exceeding $3,000,000, owing to end-user education via
grand rounds, prospective triage resulting in one-to-one
interaction with house staff, and implementation of
structured protocols, for example, massive transfusion
protocol17 and warfarin reversal.18
Because TRALI and TACO are underrecognized
and underreported, we could not document decreased
adverse outcomes owing to reduced transfusion of
plasma-containing products. However, we can speculate
that 60% decrease in plasma transfusion would be asso-
ciated with a 60% reduction in TRALI and TACO.
5% 4% 2%
Fig. 3. Reasons for nonapproval of TP. Clockwise: 39%, coagu-
lation variables adequate for hemostasis; 19%, coagulation
tests old/pending/procedure not done; 16%, warfarin reversal;
8%, on hold; 7%, heparin contamination; 5%, ordered in error;
4%, wrong patient; 2%, heparin reversal.
Fig. 4. Reasons for nonapproval of PLTs. Clockwise: 56%, PLT
count adequate for hemostasis; 12%, no recent PLT count;
11%, on hold; 8%, ordered in error; 5%, idiopathic thrombocy-
topenic purpura/heparin-induced thrombocytopenia/
thrombotic thrombocytopenia purpura; 4%, wrong patient;
4%, refractory patient with no bleeding.
TABLE 3. Yearly hospital data
Year Hospital admissions Trauma admissions Surgical procedures
2000 41,679 4,105 13,511
2001 42,426 3,953 13,078
2002 41,260 4,242 13,801
2003 41,084 4,686 14,582
2004 41,425 4,769 15,341
2005 40,005 4,670 14,858
2006 42,682 4,708 15,847
Percentage increase (over 7 years)* 2.4 14.7 17.3
* Comparing 2006 data to 2000 data.
SARODE ET AL.
490 TRANSFUSION Volume 50, February 2010
In most hospitals blood components are released
with limited or no oversight by transfusion medicine phy-
sicians unless there is a shortage of blood components.
This contrasts sharply with pharmacy practices, where
an inappropriate drug or dose request is investigated. In
smaller hospitals, with minimal usage of blood products,
this may be an insigniﬁcant issue. Without oversight in
medium to large medical centers, however, uncontrolled
release of blood products presents a variety of challenges,
among which are increased adverse transfusion reactions,
increased health care costs, depletion of inventory, and
overburdening of the community’s blood supply.
Most transfusions in nonbleeding patients are
performed to correct an abnormal laboratory value, espe-
cially in surgical patients, where even mildly abnormal
coagulation tests result in plasma transfusion.2Screening
tests for coagulation were developed to identify the cause
of bleeding in patients with a bleeding tendency. Impor-
tantly, they do not predict hemostasis in a nonbleeding
patient. Mild to moderate abnormalities of PT and PTT
show no correlation with surgical bleeding.19 Studies show
that surgical bleeding correlates with the experience of
This decrease in component use underscores the
important role transfusion medicine residents and fellows
play in house staff education at our institution. Plasma
and PLT triage calls to on-call residents have decreased
signiﬁcantly over the past 2 years. Pathology residents
have gained skills in negotiating the risks and beneﬁts of
blood products and promoting use of alternate safe and
In conclusion, implementation of prospective triage
resulted in substantial plasma and PLT usage reductions,
despite increased admissions, trauma cases, and surgical
procedures over the same period. To improve overall
health care efﬁcacy and costs, continued education
in transfusion medicine and hemostasis practices for
medical students, house staff, and practicing clinicians
We are grateful to all the pathology residents and transfusion
service technicians who participate in the triage process.
CONFLICT OF INTEREST
None of the authors has a conﬂict of interest with regard to this
1. Alving B, Alcorn K. How to improve transfusion medicine.
A treating physician’s perspective. Arch Path Med 1999;
2. Segal JB, Dzik WH. Paucity of studies to support that
abnormal coagulation test results predict bleeding in the
setting of invasive procedures: an evidence-based review.
3. Holland L, Sarode R. Should plasma be transfused prophy-
lactically before invasive procedures? Curr Opin Hematol
4. Whitaker BI, Sullivan M. The 2007 nationwide blood
collection and utilization survey report. Bethesda (MD):
American Association of Blood Banks; 2005.
5. Whitaker BI, Green J, King M, Leibeg L, Mathew SM,
Schlumpf KS, Schreiber GB. The 2007 national blood
collection and utilization survey report. Bethesda (MD):
American Association of Blood Banks; 2007.
6. Food and Drug Administration. Fatalities reported to FDA
following blood collection and transfusion. Silver Spring
(MD): FDA; 2007.
7. Popovsky MA. Transfusion-associated circulatory overload:
the plot thickens. Transfusion 2009;49:2-4.
8. Popovsky MA. Transfusion-related acute lung injury and
transfusion-associated circulatory overload. ISBT Sci Ser
9. Comprehensive Accreditation Manual for Hospitals
(CAMH): the ofﬁcial handbook. 2009 Edition. Oakbrook
Terrace, IL: JCAHO/Joint Commission Resources; 2009.
10. de Moerloose P. Laboratory evaluation of hemostasis
before cardiac operations. Ann Thorac Surg 1996;62:
11. Wandt H, Ehninger G, Gallmeier WM. New strategies
for prophylactic platelet transfusion in patients with
hematologic diseases. Oncologist 2001;6:446-50.
12. Carson JL, Hill S, Carless P, Hebert P, Henry D. Transfusion
triggers: a systematic review of the literature. Transfus Med
13. Lozano M, Cid J. Consensus and controversies in platelet
transfusion: trigger for indication, and platelet dose.
Transfus Clin Biol 2007;14:504-8.
14. Downes KA, Wilson E, Yomtovian R, Sarode R. Serial mea-
surement of clotting factors in thawed plasma stored for 5
days. Transfusion 2001;41:570.
15. Makris M. Optimisation of the prothrombin complex con-
centrate dose for warfarin reversal. Thromb Res 2005;115:
16. Hedges SJ, Dehoney SB, Hooper JS, Amanzadeh J, Busti AJ.
Evidence-based treatment recommendations for uremic
bleeding. Nat Clin Pract Nephrol 2007;3:138.
17. O’Keeffe T, Refaai M, Tchorz K, Forestner JE, Sarode R. A
massive transfusion protocol to decrease blood compo-
nent use and costs. Arch Surg 2008;143:686-90; discussion
18. Holland L, Warkentin TE, Refaai M, Crowther MA,
Johnston MA, Sarode R. Suboptimal effect of a three factor
prothrombin complex concentrate (proﬁlnine-SD) in
correcting supratherapeutic INR due to warfarin overdose.
TRIAGE OF PLASMA AND PLTS
Volume 50, February 2010 TRANSFUSION 491
19. Davis JW, Davis IC, Bennink LD, Hysell SE, Curtis BV,
Kaups KL, Bilello JF. Placement of intracranial pressure
monitors: are “normal” coagulation parameters necessary?
J Trauma 2004;57:1173-7.
20. Bernard RW, Stahl WM. Subclavian vein catheterizations:
a prospective study. I. Non-infectious complications.
Ann Surg 1971;173:184-90.
21. Smith TP, McDermott VG, Ayoub DM, Suhocki PV,
Stackhouse DJ. Percutaneous transhepatic liver
biopsy with tract embolization. Radiology 1996;198:
22. DeLoughery TG, Liebler JM, Simonds V, Goodnight SH.
Invasive line placement in critically ill patients: do hemo-
static defects matter? Transfusion 1996;36:827-31.
SARODE ET AL.
492 TRANSFUSION Volume 50, February 2010