COAGULATION AND TRANSFUSION MEDICINE
Original Article
In Vitro Analysis of Shed Blood From
Patients Undergoing Total Knee
Replacement Surgery
ROBERT C. BLAYLOCK, MD, 1 KATHLEEN S. CARLSON, MT(ASCP), 2
JAMES M. MORGAN, MD, 3 GARY 0 . TOBIN, MT(ASCP), 2 GARY D. REEDER, BA, 4
AND HAROLD B. ANSTALL, MD 1
One method used to obtain autologous blood includes collection of
wound drainage postoperatively. The decision to wash wound drainage
before infusion is left to individual institutions. The composition of
blood collected from joint spaces has not been adequately evaluated.
Wound drainage from total knee replacement was collected into a cardiotomy reservoir, without anticoagulation, for 4 hours after surgery.
Coagulation parameters were evaluated on the washed supernatant and
unwashed supernatant. The most significant findings were the following: 1. Modified prothrombin time: washed supernatant and unwashed
supernatant were substituted for tissue thromboplastin reagent. The
unwashed supernatant initiated fibrin formation (mean, 108 seconds),
whereas the washed supernatant did not (mean, > ISO seconds, P =
.01). 2. Euglobulin lysis times: Mixtures containing 50% normal
plasma and 50% washed supernatant or unwashed supernatant were
used to determine plasmin activation (unwashed supernatant + normal
plasma = 24 minutes; washed supernatant + normal plasma = 106
minutes; P = .03). Lower euglobulin lysis times indicates increased
plasmin activity. 3. Fibrin degradation products: Concentrations were
significantly elevated in unwashed supernatant (mean 10,240 /ig/mL)
versus washed supernatant (mean 5 ftg/mL, P = .02). Fibrin degradation products are inhibitors of fibrin formation and platelets.
The authors conclude the unwashed supernatant from wound drainage collected after total knee replacement contains activated components of the soluble coagulation and fibrinolytic systems, and these
substances can be significantly reduced with washing. (Key words: Autologous transfusion; Coagulopathy; Shed blood; Wound drainage) Am
J Clin Pathol 1994;101:365-369.
The use of homologous blood carries many potential risks for
the recipient, including transfusion-transmitted diseases, transfusion reactions, isoimmunization, bacterial contamination
and graft-versus-host disease. Autologous blood is recognized
as the safest blood product available to patients of all age
groups1"5 and can be obtained by predeposit, preoperative hemodilution, intraoperative salvage, or the collection of postoperative "shed blood."' Although the use of shed blood has become popular in patients undergoing orthopedic surgical
procedures, it has also been used effectively in reducing the
need for homologous blood in various clinical situations.6"7 A
recent review article, which discussed the overall safety of
transfusing washed blood collected intraoperatively, included
the following statement: ". . . with the advent of modern cell
salvage systems and the washing of salvaged blood, reports of
coagulopathy associated with intraoperative salvage have
largely disappeared." 8 Blood collected postoperatively from
surgical drains may be transfused with, or without, additional
processing, such as washing. The only current requirement
from the American Association of Blood Banks Standards
states that blood collected postoperatively must be transfused
within 6 hours of initiating the collection process.9
The contents of shed blood from joint spaces have not been
evaluated as thoroughly as mediastinal drainage. Collection devices are available that facilitate the easy collection and reinfusion of wound drainage from joint spaces. Washing before infusion is optional. Concerned about potential thromboplastic
substances in the supernatant of unwashed shed blood, we performed in vitro analysis of supernatant from washed-versusunwashed blood collected from joint space after total knee replacement.
From the ' University of Utah Medical Center. Sail Lake City, Utah:
Associated Regional and University Pathologists, Inc.. Salt Lake City.
Utah; 3Utah Orthopedics Specialists. Salt Lake City. Utah: and4Electromedics. Inc.. Englewood. Colorado.
2
MATERIALS A N D METHODS
Study
Manuscript received January 29, 1993; accepted February 24, 1993.
Address reprint requests to Dr. Blaylock: Associate Medical Director. Transfusion Services, Department of Pathology, M/S: 5C130
SOM. University of Utah Medical Center, 50 North Medical Drive,
Salt Lake City, UT 84132.
Design and Preparation
of
Samples
Patients undergoing total knee replacement were selected for
this study. Because shed blood from this patient population is
not currently salvaged at our institution, patients were not denied an autologous blood product. Blood was collected for the
365
366
COAGULATION AND TRANSFUSION MEDICINE
Original Article
first 4 hours after surgery into an Electromedics, Inc. (Englewood, CO) cardiotomy reservoir with less than 80 mmHg of
vacuum. No anticoagulant was used in the collection device.
The use of an anticoagulant in collection devices is optional,
because the plasma becomes serum before reaching the reservoir. Samples for analysis were obtained from the patient at the
end of the surgical procedure and from the collection device
before and after washing.
Blood was collected postoperatively from each patient using
standard venipuncture technique. Venous blood was collected
using Becton Dickinson VACUTAINER Systems (Rutherford,
NJ) and anticoagulated (1:10) with .105 M sodium citrate.
Plasma was collected after centrifugation (3000 rpm for 10 minutes), using a Beckman TJ-6 (Beckman Instruments, Inc.,
Brea, CA). Aliquots of plasma were dispensed into plastic
tubes, capped and frozen at -60 °C, and stored at —75 °C.
Before use, tubes were defrosted at 37 °C.
Drainage blood from the collection device wasfiltered,using
the 40-^m microaggregate filter supplied with the cardiotomy
reservoir. The filtered drainage blood was then washed with 1 L
.9% saline using a Cobe 2991 automated cell processor (Cobe
Laboratories, Inc., Lakewood, CO), following established procedures for washed red blood cells.10 Samples for analysis were
collected both before and after washing.
Coagulation Testing
Functional assays of coagulation included a prothrombin
time (PT), activated partial thromboplastin time (APTT), fibrinogen, factor V, and factor VIII. These factors were selected
because they are actually consumed in the clotting process.
Clotting times were determined using the BBL fibrometer
(Becton-Dickinson, Cockeysville, MD) and the COAG-AMATE X2 (Organon Teknika, Durham, NC). Functional levels of plasminogen and antithrombin III were determined using the DuPont ACA (DuPont Co., Wilmington, DE). Assays
used to evaluate the fibrinolytic system included fibrin(ogen)
degradation products (FDP; (Wellcome Diagnostics, Burroughs
Wellcome Co., Research Triangle Park, NC) and D-dimer
(American Bioproducts, Parsippany, NJ). Quantitative determinations of plasma Protein C (American Diagnostica, Inc.,
Greenwich, CT) and Protein S (American Bioproducts, Parsippany, NJ) antigens were made, using the rocket Immunoelectrophoresis method described by Laurell."
To estimate the differences in content of activated products
of the fibrinolytic system contained in shed blood both before
and after washing, a euglobulin clot lysis time (ELT) was performed.12"14 To perform the ELT, supernatant from the unwashed shed blood was mixed with an equal volume of normal
control plasma. The procedure was repeated with samples
from shed blood, following washing. The mixtures were acidified (pH 5.3), to extract the euglobulin portion of the plasma,
recalcified using .025 M CaCl2, and the time for clot lysis was
determined. Reduced ELTs are consistent with increased fibrinogen) lysis.
To ascertain if contents of shed blood had the ability to initiate the coagulation cascade, a test that we called "modified
prothrombin time" was performed. This term was used because the thromboplastin normally used to initiate a prothrombin time was replaced by supernatant from unwashed shed
blood, washed shed blood, and plasma from the patient, collected postoperatively. Pooled normal plasma from 20 persons
(.1 mL) was mixed with the supernatant of unwashed shed
blood (. 1 mL) and recalcified, using .025 M CaCl2 after incubation for 5 minutes at 37 °C. Supernatant from the washed
product and plasma from the patient were also-used in attempts to initiate fibrin formation with the normal 20-pool
plasma.
Additional
Testing
The volume of blood contained in the reservoir was measured after 4 hours of collection into the device. Complete
blood counts were performed on the patient following surgery,
and on both the pre- and postwash shed blood products.
RESULTS
Seven cases were evaluated, with the data summarized in
Table 1. Coagulation parameters were measured on patient
plasma collected postoperatively. These parameters showed
elevated FDP and D-dimer levels, providing evidence of systemic activation of the fibrinolytic system.
Student's Mest was used to compare differences between
means of the prewash and postwash supernatant. The prewash
supernatant initiated clot formation in normal plasma, using a
modified PT; the postwash supernatant did not initiate clot
formation (P = .01). Remarkable in the prewash assays was the
absence of clotting factors, except for a small amount of factor
VIII. The prewash PT and APTT were both prolonged beyond
150 seconds, which was consistent with no fibrinogen being
detected. A correction of the PT and APTT was performed,
using a mixture of 50% normal plasma and 50% pre- and postwash samples. The corrected PT and APTT (PT = 12.7 seconds; APPT = 27 seconds), using a prewash sample, were below the normal range for the assays (PT = 13.3-15.2; APTT =
32-48 seconds). In contrast, the corrected PT and APTT values
(PT = 15.4 seconds; APTT = 44 seconds), using the postwash
supernatant, were within the normal range. This shows evidence of hypercoagulability with the unwashed supernatant.
The euglobulin portion of normal plasma lysed in an average
time of 24 minutes when mixed with prewash material, compared with 106 minutes with postwash fluid (P - .03). Fibrin
degradation products were significantly more concentrated in
the prewash (10,240 jig/mL) compared with postwash (5.7 fig/
mL) fluid (P = .02).
The postwash supernatant had extremely low levels of all
measured analytes (Table I), as has been previously reported.7"
DISCUSSION
A strategy for the use of autologous blood should be evaluated for all patients undergoing elective surgery, and should
include maximizing the safety of the autologous product provided. One potential risk with using unwashed shed blood is the
induction of a coagulopathy. Blood collected from mediastinal
drainage has been transfused unprocessed for years, without a
high frequency of patient complication.715 However, recent
studies in patients receiving unwashed mediastinal drainage
have shown significant changes in coagulation parameters.16,17
The authors of these studies concluded that limits should be
placed on the amount of unwashed mediastinal drainage transfused to a patient. Very little published information is available
concerning the safety of unwashed blood collected from joint
spaces after orthopedic procedures. The studies that have evalu-
A.J.C.P. March 1994
>150
>150
1.2
13.9
0.0
> 150
20.1
108.8
0.0
>150
32-48
> 150
> 150
2.9
33.4
MOD PT APTT
(sec)
(sec)
0.0
1.0
0.0
0.0
1.0
8.9
16.9
49.8
183.1
Factor
VIII %
150-350 71-140 50-200
0.0
0.0
0.0
1.0
21.2
70.7
0.0
102.0
Factor
V/o
287.4
%
FIB mg
0-5
5.6
5.7
6.144.0
10.240.0
23.6
17.3
FDP
Uglml
0-5
3.1
5.3
3.2
19.8
60-150
0.0
85-122
0.0
65-122
0.0
0.0
2.0
39.7-48.3
77.6
101.3
3.7-10.1
1.2
140-440
1.8
2.7
0.7
13.5
12.0
7.4
2.9
2.5
42.1
49.8
8.9
265.4
30.1
3.6
37.5
Platelets
WBC
Hematocrit °lo lOOOImcl lOOOImcl
181.7
340.1
29.1
20.4
6.7
617.5
84.6
26.0
98.6
Vol.
(ml)
66.7
16.6
87.3
Protein Protein
S°/o
C°/o
76.7
8.6
45.3
10.7
8.7
48.6
94.9
86.9
PLG % AT III %
90-240 75-124
46.8
108.0
8.9
1.341.9
2.8
24.0
57.0
189.2
ELT
(min)
1.846.4
3.0
2.1
D-Dimer
Ug/ml
PT = prothrombin lime: MOD PT = modified prothrombin time: APTT = activated partial thromboplastin time: FIB = fibrinogen; FDP = fibrin(ogcn) degradation products; ELT =
cuglohulin clot lysis time: PLG = plasminogen: AT 111 = antithrombin 111: TKR = total knee replacement.
Ret", range 13.3-15.2
SD
Mean
Post-wash
SD
Mean
Prc-wash
SD
Mean
Palicnt
PT
(sec)
T A B L E 1 . I N V I T R O A N A L Y S I S OF SHED B L O O D F R O M T K R
s
¥
s
*-* p
a
a. *
to
3
H
3 >
n
CO
>
a.
f
& o
5"
Analysis of
368
COAGULATION AND TRANSFUSION MEDICINE
Original Article
aled the use of shed blood from orthopedic procedures have
measured the homologous blood requirements of patients receiving shed blood. An evaluation of the coagulation and fibrinolytic system was either not performed 1819 or was only partially executed.20 One cannot assume that the contents of shed
blood from the mediastinum and a joint space are equal, and
patient complications following the transfusion of unwashed
shed blood from joint spaces are emerging.21 These complications have included severe airway edema, extreme bouts of
hypotension, febrile reactions, and one death, which occurred
in a man who had been evaluated for a deep-venous thrombosis 5 hours after reinfusion of an unwashed product. 22 Adverse
reactions occurred in 25% of patients receiving unwashed
drainage, and no adverse effects were seen in patients receiving
a washed product. The author of this study concluded that the
contents of shed blood need to be evaluated if unwashed drainage continues to be considered as a clinical option. 22
Experimental evidence in animals exists that demonstrates
changes in the hemostatic and thrombotic balance after the
infusion of unwashed shed blood.23 Factors that enable the
transfusion of an unwashed product include the rapid dilution
of potentially dangerous activated components of the coagulation and fibrinolytic systems and reliance on naturally occurring compensatory mechanisms (ie, ATIII, Protein C and S). In
theory, enough deleterious product could be infused into the
patient that would overwhelm the systemic hemostatic and
thrombotic balance, resulting in a coagulopathy.
In this study, patients in the recovery room were found to
have systemic elevation of FDPs, which are potent inhibitors of
clot formation and platelet function. Infusing 617 mL (the
average volume drainage collected in this study) of unwashed
shed blood with an FDP concentration of 10,240 Mg/mL could
significantly elevate the systemic FDP concentration. In addition, the unwashed product contains components of the fibrinolytic system that caused the ELTs to be 82 minutes shorter,
compared with the same assay performed with the supernatant
of the washed product. Both the high concentration of FDPs
and elevated levels of fibrinolytic components found in unwashed blood could lead to increased bleeding in a patient.
Another characteristic of the unwashed product that concerned us was its ability to initiate clot formation in normal
plasma. Systemic activation of the soluble coagulation system
in a patient would certainly add morbidity, thus removing any
benefits associated with autologous blood. Other evidence
showing the hypercoaguability of this blood is the overcorrection of the PT and APTT, from greatly prolonged values (> 150
seconds) to less than the normal range for the assays.
Unwashed blood from joint spaces does not contain elements of value to the patient other than red cells. Clotting
factors are missing, and others have shown that platelets contained in shed blood are not functional.24 This study has shown
that washing can improve the quality of shed blood from joint
spaces. Others have demonstrated the efficacy of washing for
removal of fibrinolytic by-products. 25 Washing removed most
of the contents of the prewash serum that could potentiate a
coagulopathy and reduced FDP levels to near zero. Postwash
blood could not initiate fibrin formation in normal plasma,
and the activity of fibrinolytic substances were greatly reduced,
as evidenced by the longer period of time needed for clot lysis.
Although unwashed shed blood from joint spaces is
currently being used without widespread reports of clinical
complications, patients are not receiving the safest blood product available. A patient with a hemostatic and thrombotic sys-
tem that is out of balance could be pushed into a severe coagulopathy with the supernatant of unwashed shed blood. The
logistics and expense of washing shed blood seem justified until
further evidence exists proving the safety of unwashed blood
from joint spaces. The majority of activated hemostatic and
thrombotic components are removed during the washing procedure.
Although this study did not use an anticoagulant in the collection device, other studies have shown that coagulation and
fibrinolytic factors are activated before reaching the collection
device 715 ; therefore, the efficacy of the anticoagulant is reduced. Currently, the collection of shed blood is being performed with and without the use of an anticoagulant.
The in vitro analysis of shed blood from a joint space has
demonstrated that the coagulation system has been activated,
and fibrinogen and other clotting factors have been depleted.
Normal plasma can be stimulated to clot when supernatant
from unwashed shed blood is used in place of thromboplastin
in a modified PT, but supernatant from the unwashed product
does not initiate fibrin formation. Unwashed shed blood contains activated products of the fibrinolytic system, as evidenced
by decreased ELTs and increased FDP and D-dimer levels.
Although washing shed blood removes most of the components and by-products of the fibrinolytic and coagulation systems, the in vivo effects on the systemic hemostatic and thrombotic systems produced by transfusing unwashed shed blood
from joint spaces has not been ascertained.
This study, involving blood from joint spaces after total knee
replacement, has shown the average blood volume available to
the patient after washing is 181 mL with a 78% hematocrit.
This represents 60% to 75% of a unit of packed red blood cells,
which may or may not reduce the need for homologous blood.
This volume of blood may not be sufficient to justify transfusion, regardless of whether the blood has been washed, but this
paper is not designed to argue that question. Despite the small
sample size of this study, there are some disturbing trends identified in the data analysis of the unwashed shed blood. Use of
this product would negate its benefit if it were to induce a
coagulopathy.
Acknowledgment. Funding for this study was provided by Electromedics, Inc., Englewood, Colorado.
REFERENCES
1. The National Blood Resource Education Program Expert Panel.
The use of autologous blood. JAMA 1990;263:414-417.
2. Toy PTCY, Strauss RG, Stehling LC, et al. Predeposit autologous
blood for elective surgery. A national multicenter study. A' Engl
J Med 1987;316:517-520.
3. Silvergleid AJ. Safety and effectiveness of predeposit autologous
transfusions in preteen and adolescent children. JAMA
1987;257:3403-3404.
4. Surgenor DM. The patient's blood is the safest blood. N Engl J
Med 1987;316:542-544.
5. Pindyck J, Avorn J, Kuriyan M, et al. Blood donation by the elderly: Clinical and policy considerations. JAMA 1987;257:
1186-1188.
6. SchafF HV, Hauer JH, Bell WR, et al. Autotransfusion of shed
mediastinal blood after cardiac surgery: A prospective study. J
Thome Cardiavasc Surg 1978;75:632-641.
7. Hartz RS, Smith JA, Green D. Autotransfusion after cardiac operation; assessment of hemostatic factors. J Thome Cardiovasc
Surg 1988;96:178-182.
A.J.C.P. • March 1994
BLAYLOCK ET AL.
In Vitro Analysis of Shed Blood From Total Knee Replacement
8. Williamson KR, Taswell HF. Intraoperative blood salvage: A review. Transfusion 1991;31:662-675.
9. Widmann FK, ed. Standards for Blood Banks and Transfusion
Services. 14th ed. Arlington, Va: American Association of
Blood Banks, 1991, p. 47.
10. O'Connor Wooten MJ. Use and analysis of saline washed red
blood cells. Transfusion 1976; 16:5:464-468.
11. Laurell CD. Electroimmunoassay. Scan J Clin Lab Invest
1972;29; 124:21-27.
12. Kowalski E, Kopec M, Niewiarowski S. An evaluation of the
euglobulin method for the determination of fibrinolysis. J Clin
Pathol 1959; 12:215.
13. Buckell M. The effect of citrate on euglobulin methods of estimating fibrinolytic activity. J Clin Pathol 1958; 11:403.
14. Human Blood Coagulation Haemostasis and Thrombosis. 2nd ed.
Biggs R, ed. Oxford, UK: Blackwell Scientific, 1976, Appendix
2, p. 724.
15. Hauer JM, Thurer RL. Controversies in autotransfusion. Vox
Sang 1984;46:8-12.
16. Kongsgaard UE, Tollofsrud S, Brosstad R, et al. Autotransfusion
after open heart surgery: Characteristics of shed mediastinal
blood and its influence on the plasma proteases in circulating
blood. Acta Anaesthesiol Scan 1991 ;35:71-76.
17. Fuller J A, Buston BF, Picken J, et al. Haematological effects of
reinfused mediastinal blood after cardiac surgery. Med J Aust
1991:154:737-740.
369
18. Gannon DM, Lombardi AV Jr, Mallory TH, et al. An evaluation
of the efficacy of postoperative blood salvage after total joint
arthroplasty. A prospective randomized trial. J Arthroplasty
1991;6:109-114.
19. Trammel TR, Fisher D, Brueckmann FR, et al. Closed-wound
drainage systems. The Solcotrans Plus versus the Stryker-CBC
ConstaVAC. Orthop Rev 1991;20:536-542.
20. Groh GI, Buchert PK, Allen WC. A comparison of transfusion
requirements after total knee arthroplasty using the Solcotrans
autotransfusion system. J Arthroplasty 1990;5:281-285.
21. Woda R, TezlafF JE. Upper airway edema following autologous
blood transfusion from a wound drainage system. Can J
Anaesth 1992;39:290-292.
22. Clements DH, Sculco TP, Burke SW, et al. Salvage and reinfusion
of postoperative sanguineous would drainage. J Bone Joint Surg
1992;74:646-651.
23. Kinglsey JR, Valeri CR, Peters H, et al. Citrate anticoagulation
and cell washing for intraoperative autotransfusion in the baboon. Am J Surg 1976; 113:717-721.
24. Hauer JM, Thurer RI, Weintraub RM. Platelet function in shed
mediastinal blood following cardiac surgery (Abstract). Joint
Congress of the International Society of Haematology and International Society of Blood Transfusion, Budapest, 1982.
25. Lawrence-Brown MM, Couch C, Halliday M, et al. D-dimer levels
in blood salvage for autotransfusion. Aust N Z J Surg
1989:5967-5970.
Vol. 101 - N o . 3