The Effect of In Vitro Hemodilution with Gelatin,
Dextran, Hydroxyethyl Starch, or Ringer’s Solution
on Thrombelastograph姞
Georg A. Petroianu, Priv. Doz. Dr. med.*, Jie Liu, Dr. med.*, Wolfgang H. Maleck,
Cathrine Mattinger, Dr. med.†, and Wolfgang F. Bergler, Priv. Doz. Dr. med.‡
Departments of *Pharmacology and Toxicology and †Otorhinolaryngology, University of Heidelberg at Mannheim,
Mannheim; ‡Department of Anesthesiology, Ludwigshafen City Hospital Germany
To determine the effects of progressive in vitro hemodi-
lution with various plasma substitutes on whole blood
coagulation, blood was obtained from six healthy vol-
unteers. The Thrombelastograph® (TEG; Haemoscope
Corp., Morton Grove, IL) variables of reaction time, co-
agulation time, maximum amplitude, and growth an-
gle were determined. The following plasma substitutes
were tested: two gelatin solutions (4% gelatin polysuc-
cinate and 5.5% oxypolygelatin); two dextrans (10%
dextran 40 and 6% dextran 60); and five hydroxyethyl
starch (HES) preparations (6% HES 70/0.5– 0.55, 3%
HES 200/0.5, 6% HES 200/0.5, 10% HES 200/0.5, and
6% HES 450/0.7). Ringer’s solution was also tested to
assist analyzing the intrinsic effect of colloid molecules
on blood coagulation. The dilution ratios of citrated
blood volume to plasma substitute volume were 10:2,
10:4, and 10:10. Blood coagulation was affected by
plasma substitutes when the dilution ratios of citrated
blood volume to colloid solution volume were 10:4 and
C
olloid solutions are used to restore intravascular
fluid volume. To avoid the risks associated with
transfusion of allogenic blood products and to
limit the high costs of albumin solutions, synthetic col-
loid solutions are used as an alternative for blood loss
replacement (1). Presently available are gelatin solutions,
dextran, and hydroxyethyl starch (HES) preparations (2,3).
When choosing a colloid for a specific purpose, its
therapeutic value must be carefully weighed against
its risk for adverse effects, such as impairment of renal
function, allergy, and impairment of coagulation by
Supported, in part, by a grant from Forschungsfonds der Fakultät
für Klinische Medizin, Mannheim der Universität Heidelberg,
Mannheim, Germany.
Accepted for publication December 6, 1999.
Address correspondence and reprint requests to Priv. Doz. Dr. med.
Georg Petroianu, University of Heidelberg at Mannheim, Department
of Pharmacology and Toxicology, Maybach St. 14-16, 68169 Mann-
heim, Germany. Address e-mail to petroia@rumms.uni-mannheim.de.
©2000 by the International Anesthesia Research Society
0003-2999/00
Arzt‡,
10:10. TEG variables did not change significantly after
in vitro hemodilution with lactated Ringer’s solution.
The tested gelatin solutions showed less intrinsic effect
on blood coagulation than other plasma substitutes. All
HES preparations showed similar intrinsic effects as 6%
dextran 60. The plasma substitute of 10% dextran
40 had the strongest effect on coagulation. Coagulation
time was the most markedly affected TEG variable.
Blood coagulation may be compromised when the dilu-
tion ratio of blood volume to colloid solution volume is
⬎10:4. Whereas gelatin solutions have less intrinsic ef-
fect on blood coagulation, 10% dextran 40 has the stron-
gest effect on coagulation. Implications: Blood coagu-
lation may be compromised when the dilution ratio of
blood volume to colloid solution volume is ⬎10:4.
Whereas gelatin solutions have less intrinsic effect on
blood coagulation than hydroxyethyl starch or dextran,
10% dextran 40 has the strongest effect on coagulation.
(Anesth Analg 2000;90:795–800)
dilution of plasma coagulation factors, as well as spe-
cific effects on certain plasma components and cellular
elements (4). Prolonged bleeding time has been re-
ported after using dextran or HES preparations (4 – 6).
The purpose of this study was to determine the
effects of progressive in vitro hemodilution with vari-
ous plasma substitutes on whole blood coagulation.
We tested two gelatin solutions, two dextrans, and
five HES preparations. Blood coagulation was as-
sessed by using the Thrombelastograph® (TEG; Hae-
moscope Corp., Morton Grove, IL), a global and func-
tional coagulation test that measures the interaction of
the clotting cascade and platelets in whole blood (7,8).
Methods
Citrated blood was obtained from six healthy volun-
teers (one men/five women). All volunteers had nor-
mal coagulation, normal renal and hepatic function by
Anesth Analg 2000;90:795–800 795
796 CARDIOVASCULAR ANESTHESIA PETROIANU ET AL. ANESTH ANALG
COLLOIDS AND COAGULATION 2000;90:795–800

1948 and has been extensively described in the litera-

Figure 1. Thrombelastograph® (TEG; Haemoscope Corp., Morton
Grove, IL) enables a global assessment of hemostatic function. It
documents the interaction of platelets with the protein coagulation
cascade from the time of the initial platelet-fibrin interaction,
through platelet aggregation, clot strengthening and fibrin cross-
linkage to eventual clot lysis. Split point (Sp) is the distance from
recording start to the first sign of divergence. Reaction time (R) is
the interval between recording start and the time at which the
amplitude of the TEG reaches 2 mm. It represents the rate of
thromboplastin generation and reflects the function of the intrinsic
coagulation system. Coagulation time (K) is the time from R to a
level of clot firmness of 20 mm. The coagulation time reflects the
function of the intrinsic system, platelets, and fibrinogen. Maximum
amplitude (MA) represents the largest amplitude reached and is a
function of the elasticity of the blood clot. The platelet function,
fibrinogen, and factor XIII have an influence on the MA. Growth
angle (␣) is the clot formation rate or the speed with which a solid
clot forms and is known to be a function of fibrinogen and platelets.
history and none was receiving anticoagulant or anti-
platelet medication. Their partial thromboplastin time,
prothrombin time, coagulation factor V activity, and
their baseline TEG values were in the normal range.
Within 4 h of sample taking, the respective dilutions
with the various solutions were prepared and TEG
variables, reaction time (R), coagulation time (K), max-
imum amplitude (MA), and growth angle (␣) were
determined by using recalcified, celite-activated
blood, and the computerized coagulation analyser
TEG. The method was developed first by Hartert (9) in
ture (7,8,10). A schematic TEG trace is shown in Figure
1. Reaction time is the time from when the blood is
placed in the cuvette until an amplitude of 2 mm is
reached. The R value is the time necessary for initial
fibrin formation, representing intrinsic clotting. The K
value, defined as the time interval from the end of the
R value until the amplitude of the TEG tracing, is
20 mm. It represents a measure of the speed at which
a thrombus of a certain solidity develops; it is a meas-
ure of the rapidity of fibrin formation and cross-
linking. The MA is the greatest amplitude achieved on
the TEG and reflects the strength of the fibrin clot; it
depends on platelet number and function as well as
fibrinogen levels. The ␣ value is measured as the slope
of the outside divergence of the tracing from the point
of the R value. It is expressed in degrees and denotes
the speed at which a clot is formed and cross-linked
(7,8).
The tested plasma substitutes are summarized in
Table 1. Ringer’s solution was also tested to separate
dilutional effects from the intrinsic effect of plasma
substitute molecules on blood coagulation. The pH of
all solutions was adjusted with 8.4% NaHCO3 to 7.4 to
achieve a physiological pH for effective hemostatic
mechanism. The dilution ratios of citrated blood vol-
ume to colloid solution volume were 10:2, 10:4, and
10:10. In an adult with approximately 6 L of blood
volume, this would be roughly equivalent to blood
losses of 1, 2, and 3 L, respectively, replaced with
infusion. TEG variables R, K, MA, and ␣ were mea-
sured on each dilution of colloid solution.

Table 1. Summary of Tested Solutions

Substitution
Generic name Manufacturer Substance MW g/100 mL degree
Gelafusal-N Serum-Werk Bernburg, Gelatin 30,000 4 No
Bernburg, Germany polysuccinate
Gelifundol Biotest Pharma, Oxypolygelatine 30,000 5.5 No
Dreieich, Germany
Onkovertin N B. Braun Melsungen, Dextran 40,000 10 No
Melsungen, Germany
Onkovertin B. Braun Melsungen, Dextran 60,000 6 No
Melsungen, Germany
Expafusin Pharmacia & Upjohn, HES 70,000 6 0.5–0.55
Erlangen, Germany
Haes-steril Fresenius-Klinik, Bad HES 200,000 3 0.5
Homburg, Germany
Haes-steril Fresenius-Klinik, Bad HES 200,000 6 0.5
Homburg, Germany
Haes-steril Fresenius-Klinik, Bad HES 200,000 10 0.5
Homburg, Germany
Plasmasteril Fresenius-Klinik, Bad HES 450,000 6 0.7
Homburg, Germany
Ringer’s solution DAB 7 B. Braun Melsungen, Electrolyte
Braun Melsungen, Germany
HES ⫽ hydroxyethyl starch, DAB ⫽ Deutsches Arzneibuch, MW ⫽ molecular weight.
ANESTH ANALG CARDIOVASCULAR ANESTHESIA PETROIANU ET AL. 797
2000;90:795–800 COLLOIDS AND COAGULATION

Table 2. Effect of Progressive in vitro Hemodilution (10:2, 10:4, and 10:10)a with Two Gelatin Solutions, Two Dextrans,
and Five HES Preparations on TEG Parameters in Healthy Volunteers
R (min) K (min) MA (mm) ␣ (°)
Absolute Baseline Absolute Baseline Absolute Baseline Absolute Baseline
value % value % value % value %
Baseline 10.8 ⫾ 3.0 3.0 ⫾ 0.7 58.7 ⫾ 4.4 70.7 ⫾ 4.8
Ringer
10:2 9.2 ⫾ 4.7 84 ⫾ 26 3.1 ⫾ 0.9 102 ⫾ 5 54.7 ⫾ 5.0 93 ⫾ 7 71.8 ⫾ 5.4 102 ⫾ 6
10:4 9.8 ⫾ 5.8 87 ⫾ 28 3.7 ⫾ 1.2 123 ⫾ 23 51.8 ⫾ 6.0 88 ⫾ 5* 67.3 ⫾ 6.2 95 ⫾ 7
10:10 9.0 ⫾ 2.6 85 ⫾ 18 5.1 ⫾ 2.0 169 ⫾ 36* 40.5 ⫾ 4.8 69 ⫾ 10* 63.0 ⫾ 8.2 89 ⫾ 9
Gelatin
4% Gelafusal-N
10:2 7.3 ⫾ 1.2 71 ⫾ 11* 3.7 ⫾ 0.7 124 ⫾ 11* 51.7 ⫾ 5.8 88 ⫾ 6* 68.8 ⫾ 2.8 99 ⫾ 6
10:4 6.9 ⫾ 2.5 63 ⫾ 10* 4.3 ⫾ 0.6 179 ⫾ 28* 45.8 ⫾ 4.6 78 ⫾ 6* 66.3 ⫾ 4.8 95 ⫾ 5
10:10 8.5 ⫾ 1.6 81 ⫾ 13 7.9 ⫾ 1.5 269 ⫾ 38* 32.4 ⫾ 4.4 55 ⫾ 4* 56.5 ⫾ 8.8 76 ⫾ 8*
5.5% Gelifundol
10:2 7.5 ⫾ 1.1 73 ⫾ 12* 3.3 ⫾ 0.9 107 ⫾ 8 55.3 ⫾ 3.4 95 ⫾ 6 70.1 ⫾ 4.8 99 ⫾ 2
10:4 7.1 ⫾ 1.7 68 ⫾ 12* 4.0 ⫾ 0.9 135 ⫾ 11* 48.9 ⫾ 4.8 83 ⫾ 6* 67.9 ⫾ 4.8 96 ⫾ 2
10:10 9.2 ⫾ 1.8 88 ⫾ 16 6.6 ⫾ 0.9 225 ⫾ 28* 36.3 ⫾ 5.9 63 ⫾ 8* 56.4 ⫾ 3.9 80 ⫾ 3*
Dextrans
10% Onkovertin N
10:2 8.8 ⫾ 2.8 84 ⫾ 28 6.0 ⫾ 0.6 208 ⫾ 49* 40.8 ⫾ 6.2 69 ⫾ 7* 57.6 ⫾ 2.8 82 ⫾ 6*
10:4 9.8 ⫾ 3.6 92 ⫾ 29 8.8 ⫾ 0.8 290 ⫾ 76* 37.1 ⫾ 5.4 63 ⫾ 8* 46.1 ⫾ 2.3 66 ⫾ 4*
10:10 14.3 ⫾ 2.6 138 ⫾ 27 22.3 ⫾ 4.5 758 ⫾ 155* 23.5 ⫾ 1.9 40 ⫾ 4* 24.5 ⫾ 5.0 36 ⫾ 5*
6% Onkovertin
10:2 8.5 ⫾ 2.9 81 ⫾ 22 5.5 ⫾ 0.6 189 ⫾ 33* 46.0 ⫾ 3.7 79 ⫾ 3* 58.6 ⫾ 3.9 83 ⫾ 4*
10:4 7.7 ⫾ 1.9 73 ⫾ 14* 7.8 ⫾ 1.1 265 ⫾ 35* 40.0 ⫾ 5.0 68 ⫾ 4* 52.3 ⫾ 4.1 74 ⫾ 2*
10:10 10.9 ⫾ 2.5 104 ⫾ 21 13.3 ⫾ 2.5 453 ⫾ 78* 27.0 ⫾ 2.9 46 ⫾ 4* 39.8 ⫾ 4.1 57 ⫾ 7*
HES Preparations
6% Expafusin
10:2 5.9 ⫾ 1.6 59 ⫾ 21* 4.1 ⫾ 1.1 135 ⫾ 6* 52.8 ⫾ 6.6 90 ⫾ 11 66 ⫾ 5.9 93 ⫾ 4
10:4 6.9 ⫾ 1.2 68 ⫾ 23 5.6 ⫾ 0.4 193 ⫾ 36* 42.2 ⫾ 5.6 72 ⫾ 8* 60.8 ⫾ 2.9 86 ⫾ 5*
10:10 8.6 ⫾ 2.3 83 ⫾ 25 10.6 ⫾ 2.4 357 ⫾ 42* 31.9 ⫾ 4.8 54 ⫾ 6* 43.0 ⫾ 4.6 61 ⫾ 3*
3% Haes-steril
10:2 6.8 ⫾ 2.7 64 ⫾ 17* 3.5 ⫾ 0.7 118 ⫾ 4* 53.1 ⫾ 5.5 91 ⫾ 6 69.0 ⫾ 4.8 98 ⫾ 6
10:4 7.0 ⫾ 2.2 67 ⫾ 18* 4.6 ⫾ 1.1 155 ⫾ 31* 46.7 ⫾ 5.1 79 ⫾ 4* 64.3 ⫾ 5.9 91 ⫾ 5
10:10 8.7 ⫾ 2.0 80 ⫾ 21 8.9 ⫾ 1.4 303 ⫾ 33* 31.0 ⫾ 4.2 53 ⫾ 5* 52.2 ⫾ 4.4 74 ⫾ 3*
6% Haes-steril
10:2 6.8 ⫾ 1.2 66 ⫾ 19* 3.9 ⫾ 0.9 135 ⫾ 36 50.8 ⫾ 7.1 86 ⫾ 10 66.8 ⫾ 6.4 95 ⫾ 8
10:4 8.2 ⫾ 1.5 79 ⫾ 19 5.8 ⫾ 1.3 198 ⫾ 33* 44.2 ⫾ 6.8 75 ⫾ 9* 56.4 ⫾ 5.8 81 ⫾ 5*
10:10 9.5 ⫾ 0.7 94 ⫾ 25 11 ⫾ 1.7 374 ⫾ 38* 31.2 ⫾ 3.8 53 ⫾ 6* 41.8 ⫾ 4.4 59 ⫾ 5*
10% Haes-steril
10:2 7.4 ⫾ 2.0 71 ⫾ 19 4.9 ⫾ 0.7 168 ⫾ 22* 47.0 ⫾ 5.1 80 ⫾ 5* 61.7 ⫾ 3.6 88 ⫾ 5*
10:4 7.3 ⫾ 1.0 72 ⫾ 20 6.8 ⫾ 0.8 232 ⫾ 41* 38.2 ⫾ 6.7 65 ⫾ 8* 55.2 ⫾ 3.8 78 ⫾ 2*
10:10 11.6 ⫾ 1.1 114 ⫾ 31 13.4 ⫾ 2.0 461 ⫾ 86* 28.9 ⫾ 5.0 50 ⫾ 9* 35.8 ⫾ 2.3 51 ⫾ 5*
6% Plasmasteril
10:2 7.1 ⫾ 1.6 68 ⫾ 17* 4.5 ⫾ 0.6 154 ⫾ 26* 50.3 ⫾ 4.5 86 ⫾ 8 64.4 ⫾ 4.5 91 ⫾ 6
10:4 7.6 ⫾ 2.0 73 ⫾ 24 5.7 ⫾ 1.1 193 ⫾ 30* 43.3 ⫾ 6.0 74 ⫾ 8* 59.8 ⫾ 5.6 86 ⫾ 7*
10:10 9.3 ⫾ 1.6 91 ⫾ 25 11.0 ⫾ 1.4 377 ⫾ 55* 32.7 ⫾ 4.3 56 ⫾ 6* 40.3 ⫾ 3.8 57 ⫾ 5*
Values are mean ⫾ sd (n ⫽ 6).
Paired sign test was performed to compare the significance of the difference between solution-treated samples and baseline values.
a
Blood volume:solution volume.
R ⫽ reaction time, K ⫽ coagulation time, MA ⫽ maximum amplitude, ␣ ⫽ growth angle, HES ⫽ hydroxyethyl starch, TEG ⫽ Thrombelastograph威
(Haemoscope Corp., Morton Grove, IL).
* P ⬍ 0.04, significantly different from baseline.

All data were presented as mean ⫾ sd in absolute Results

values, percentages of baseline (Table 2), or percent-
ages of values of lactated Ringer’s solution (Table 3).
The paired sign test was performed to compare the
significance of the difference between plasma
substitute-treated samples and baseline, as well as
lactated Ringer’s solution.
To specifically assess the intrinsic effect of plasma
substitute molecules on blood coagulation as opposed
to a dilutional effect, hemodilution with lactated Ring-
er’s solution was first analyzed (Table 2). At 10:10 he-
modilution ratio, compared with baseline values, K
798 CARDIOVASCULAR ANESTHESIA PETROIANU ET AL. ANESTH ANALG
COLLOIDS AND COAGULATION 2000;90:795–800

Table 3. Effect of Progressive in vitro Hemodilution (10:2, 10:4, and 10:10)a with Two Gelatin Solutions, Two Dextrans,
and Five HES Preparations on TEG Parameters in Comparison with Hemodilution with Ringer’s Solution in
Healthy Volunteers
R (min) K (min) MA (mm) ␣ (°)
Absolute Ringer Absolute Ringer Absolute Ringer Absolute Ringer
value % value % value % value %
Ringer
10:2 9.2 ⫾ 4.7 3.1 ⫾ 0.9 54.7 ⫾ 5.0 71.8 ⫾ 5.4
10:4 9.8 ⫾ 5.8 3.7 ⫾ 1.2 51.8 ⫾ 6.0 67.3 ⫾ 6.2
10:10 9.0 ⫾ 2.6 5.1 ⫾ 2.0 40.5 ⫾ 4.8 63.0 ⫾ 8.2
Gelatin solutions
4% Gelafusal-N
10:2 7.3 ⫾ 1.2 92 ⫾ 35 3.7 ⫾ 0.7 122 ⫾ 14 51.7 ⫾ 6.0 95 ⫾ 9 68.8 ⫾ 2.8 96 ⫾ 7
10:4 6.9 ⫾ 2.5 78 ⫾ 23 4.3 ⫾ 0.6 122 ⫾ 17 45.8 ⫾ 4.6 89 ⫾ 6* 66.3 ⫾ 4.8 99 ⫾ 5
10:10 8.5 ⫾ 1.6 99 ⫾ 30 7.9 ⫾ 1.5 166 ⫾ 40* 32.4 ⫾ 4.4 81 ⫾ 14* 53.1 ⫾ 3.4 86 ⫾ 13
5.5% Gelifundol
10:2 7.5 ⫾ 1.1 95 ⫾ 41 3.3 ⫾ 0.9 105 ⫾ 8 55.3 ⫾ 3.4 99 ⫾ 12 70.1 ⫾ 4.8 98 ⫾ 4
10:4 7.1 ⫾ 1.7 83 ⫾ 28 4.0 ⫾ 0.9 112 ⫾ 13 48.9 ⫾ 4.8 94 ⫾ 6 67.9 ⫾ 4.8 102 ⫾ 7
10:10 9.2 ⫾ 1.8 108 ⫾ 33 6.6 ⫾ 0.9 138 ⫾ 31 36.7 ⫾ 5.5 90 ⫾ 15 56.4 ⫾ 3.9 96 ⫾ 14
Dextrans
10% Onkovertin N
10:2 8.8 ⫾ 2.8 112 ⫾ 59 6.0 ⫾ 0.6 206 ⫾ 53* 40.8 ⫾ 6.2 74 ⫾ 7* 57.6 ⫾ 2.8 81 ⫾ 7*
10:4 9.8 ⫾ 3.6 115 ⫾ 48 8.8 ⫾ 0.8 252 ⫾ 42* 37.1 ⫾ 5.4 72 ⫾ 11* 46.1 ⫾ 2.3 69 ⫾ 7*
10:10 14.3 ⫾ 2.6 166 ⫾ 41 22.3 ⫾ 4.5 465 ⫾ 125* 23.5 ⫾ 1.9 59 ⫾ 6* 24.5 ⫾ 5.0 39 ⫾ 6*
6% Onkovertin
10:2 8.5 ⫾ 3.0 110 ⫾ 58 5.5 ⫾ 0.6 186 ⫾ 36* 46.1 ⫾ 3.7 85 ⫾ 5* 58.6 ⫾ 3.9 82 ⫾ 6*
10:4 7.7 ⫾ 1.9 92 ⫾ 40 7.8 ⫾ 1.1 223 ⫾ 52* 39.9 ⫾ 5.0 77 ⫾ 5* 52.3 ⫾ 4.1 78 ⫾ 6*
10:10 10.9 ⫾ 2.5 124 ⫾ 24 13.5 ⫾ 2.5 280 ⫾ 72* 27.0 ⫾ 2.8 68 ⫾ 10* 39.8 ⫾ 4.1 65 ⫾ 12*
HES Preparations
6% Expafusin
10:2 5.9 ⫾ 1.6 80 ⫾ 44 4.1 ⫾ 1.1 133 ⫾ 8* 52.8 ⫾ 6.6 97 ⫾ 9 66 ⫾ 6.0 92 ⫾ 4
10:4 6.9 ⫾ 1.2 85 ⫾ 35 5.6 ⫾ 0.4 160 ⫾ 31* 42.2 ⫾ 5.6 82 ⫾ 6* 60.8 ⫾ 2.9 91 ⫾ 7
10:10 8.6 ⫾ 2.3 98 ⫾ 25 10.6 ⫾ 2.4 217 ⫾ 35* 31.9 ⫾ 4.8 80 ⫾ 15* 43 ⫾ 4.6 69 ⫾ 6*
3% Haes-steril
10:2 6.8 ⫾ 2.7 81 ⫾ 27 3.5 ⫾ 0.7 116 ⫾ 8 53.1 ⫾ 5.5 97 ⫾ 10 69 ⫾ 4.8 96 ⫾ 6
10:4 7.0 ⫾ 2.2 82 ⫾ 34 4.6 ⫾ 1.1 130 ⫾ 36 46.7 ⫾ 5.1 90 ⫾ 6 64.3 ⫾ 5.9 96 ⫾ 11
10:10 8.7 ⫾ 2.0 99 ⫾ 22 8.9 ⫾ 1.4 188 ⫾ 47* 31.0 ⫾ 4.2 78 ⫾ 16* 52.2 ⫾ 4.4 84 ⫾ 11
6% Haes-steril
10:2 6.8 ⫾ 1.2 88 ⫾ 43 3.9 ⫾ 0.9 129 ⫾ 39 50.8 ⫾ 7.1 94 ⫾ 9 66.8 ⫾ 6.4 94 ⫾ 11
10:4 8.2 ⫾ 1.5 97 ⫾ 44 5.8 ⫾ 1.3 163 ⫾ 23* 44.2 ⫾ 6.8 86 ⫾ 7 56.4 ⫾ 5.8 85 ⫾ 5*
10:10 9.5 ⫾ 0.7 112 ⫾ 27 11 ⫾ 1.7 223 ⫾ 53* 31.2 ⫾ 3.8 74 ⫾ 9* 41.8 ⫾ 4.4 71 ⫾ 10*
10% Haes-steril
10:2 7.4 ⫾ 2.0 91 ⫾ 28 4.9 ⫾ 0.7 165 ⫾ 27* 47 ⫾ 5.1 86 ⫾ 5* 61.7 ⫾ 3.6 86 ⫾ 5*
10:4 11.6 ⫾ 1.1 90 ⫾ 35 6.8 ⫾ 0.8 192 ⫾ 30* 38.2 ⫾ 6.7 73 ⫾ 6* 55.2 ⫾ 3.8 82 ⫾ 7*
10:10 11.6 ⫾ 1.1 136 ⫾ 36 13.4 ⫾ 2.0 285 ⫾ 78* 28.9 ⫾ 5.0 71 ⫾ 5* 35.8 ⫾ 2.3 58 ⫾ 11*
6% Plasmasteril
10:2 7.1 ⫾ 1.6 87 ⫾ 32 4.5 ⫾ 0.6 152 ⫾ 30* 50.3 ⫾ 4.5 92 ⫾ 9 62.8 ⫾ 4.9 90 ⫾ 6
10:4 7.6 ⫾ 2.0 84 ⫾ 25 5.7 ⫾ 1.1 158 ⫾ 35* 43.3 ⫾ 6.0 85 ⫾ 7* 59.8 ⫾ 5.6 90 ⫾ 8
10:10 9.3 ⫾ 1.6 105 ⫾ 36 11 ⫾ 1.4 225 ⫾ 51* 32.7 ⫾ 4.3 82 ⫾ 18 40.3 ⫾ 3.8 66 ⫾ 11*
Values are mean ⫾ sd (n ⫽ 6).
Paired sign test was performed to compare the significance of the difference between solution-treated samples and Ringer’s solution-treated samples.
a
Blood volume:solution volume.
R ⫽ reaction time, K ⫽ coagulation time, MA ⫽ maximum amplitude, ␣ ⫽ growth angle, HES ⫽ hydroxyethyl starch, TEG ⫽ Thrombelastograph威
(Haemoscope Corp., Morton Grove, IL).
* P ⬍ 0.04, significantly different from baseline.

increased and MA decreased in statistical significance.
However, all values were within the normal range
(R ⫽ 7–14 min, K ⫽ 3–7 min, MA ⫽ 40 – 60 mm, ␣ ⫽
40°-70°) as reported in the literature (8).
Compared with baseline values, progressive in vitro
hemodilution with plasma substitutes compromised
TEG variables, especially K in values greater than
7 min, and MA in values smaller than 40 mm. At
10:10 hemodilution the values were outside the nor-
mal range (Table 2). Although some values showed
statistically significant changes at 10:2 and 10:4 he-
modilution groups, they were within the normal
ANESTH ANALG
2000;90:795–800
range. Onkovertin N (dextran) 10% showed the stron-
gest effect on TEG variables (mostly on K).
The intrinsic effect of plasma substitute molecules
on TEG variables was assessed by comparing the dif-
ference between lactated Ringer’s solution diluted
blood and plasma substitute diluted blood (at same
degree of hemodilution, Table 3). Two gelatin solu-
tions, Gelafusal-N (gelatin polysuccinate) in acetated
Ringer’s solution, and Gelifundol (oxypolygelatin)
5.5%, showed less intrinsic effect on blood coagulation
than other plasma substitutes. Other tested solutions
resulted in a longer K, decreased MA, and clot ␣,
mostly at 10:4 and 10:10 hemodilution. Onkovertin
(dextran) 6%, Onkovertin N 10%, and Haes-steril
(HES) 10% (200/0.5) showed a more pronounced ef-
fect than other tested solutions.
Discussion
Our results demonstrate that blood coagulation is af-
fected by plasma substitutes in vitro when the dilution
ratios of citrated blood volume to colloid solution
volume are 10:4 and 10:10. The tested gelatin solu-
tions, Gelafusal-N 4% and Gelifundol 5.5%, showed
less intrinsic effect on blood coagulation than other
plasma substitutes. Onkovertin N 10% had the stron-
gest effect on coagulation despite its smaller molecular
weight than Onkovertin 6%. Among all HES prepara-
tions, Haes-steril 10% showed a slightly stronger an-
ticoagulant effect than other HES solutions. Therefore,
it appears that the number of plasma substitute mol-
ecules (concentration) is more important than the mo-
lecular weight or the degree of substitution of mole-
cules in HES preparations. Although K was the most
markedly affected TEG variable, MA and rapidity of
angle ␣ were also compromised. This pattern is com-
patible with platelet dysfunction and with previous
reports of a decrease in all factor VIII moieties pro-
duced by HES and dextran infusion (5,11,12).
Our results concerning the effect of Ringer’s solution
on TEG variables are comparable with the results of
Niemi et al. (13) who reported that acetated Ringer’s
solution improved coagulability at 20 vol % concentra-
tions; however, it impaired coagulation at 50 vol % con-
centration. The decreased coagulability at 50 vol % con-
centration was considered a dilutional effect. We found
only a minor effect of Ringer’s solution except for a
prolonged K and reduced MA in the 10:10 dilution.
The effect of colloids on hemostasis may potentially
cause an alteration of the following hemostatic mech-
anisms: 1) a dilution effect and accompanying changes
in blood viscosity and platelet count; 2) an effect on
activity of blood coagulation factors and fibrinolysis;
and 3) an influence on platelet function.
Gelatin produces no extensive coating effect and
therefore characteristically has only a minimal effect
CARDIOVASCULAR ANESTHESIA PETROIANU ET AL. 799
COLLOIDS AND COAGULATION
on platelet function, probably a dilution effect (2). HES
will reduce levels of the coagulation factors, fibrino-
gen, factor VIII, and von Willebrand’s factor, and re-
duce platelet function. It is hypothesized that complex
polysaccharide precipitates certain coagulation fac-
tors, making the factors unavailable to the coagulation
cascade. The effects of dextrans on the coagulation
pathway are similar to those of HES (2,4,14).
Our results are comparable with that of Mortier et
al. (15). They reported that although modified fluid
gelatin 4% did not change TEG variables (at 50% he-
modilution), profound hemodilution (50%) with 10%
dextran 40 and 6% HES compromised TEG variables,
namely, a marked decrease in the MA and ␣, and an
increase in K. Karoutsos et al. (16) investigated the in
vivo effects of 3.5% modified gelatin and 6% HES
200/0.62 on TEG. TEG variables did not change in the
HES group; however, they showed a state of hyper-
coagulability with a significant decrease in R and R ⫹
K and an increase in ␣. Ruttmann et al. (17) showed a
similar coagulation profile in an in vitro study. HES in
vivo hemodilution (equal to a 10:2 dilution) decreased
MA; however, it did not affect other TEG variables
(18).
We conclude that blood coagulation may be com-
promised when the dilution ratio of blood volume to
colloid solution volume is ⬎10:4. Gelatin solutions
have less intrinsic effect on blood coagulation than
either HES or dextran. Onkovertin N 10% has the
strongest effect on coagulation. When bleeding is not a
concern and thromboprophylaxis is to be achieved,
use of HES or dextran may offer some advantages.
However, we suggest that dextrans (especially 10%
dextran 40) and HES preparations should be used
with caution when bleeding would potentially be of
serious consequence to the patient.
References
1. Stoelting RK. Blood substitutes. In: Stoelting RK, ed. Pharma-
cology and physiology in anesthetic practice. 2nd ed.
Philadelphia: JB Lippincott, 1991:577–9.
2. Dieterich HJ, Groh J, Peter K, Fickweiler E. Volumenersatzlö-
sungen. In: Hartig W, ed. Moderne infusionstherapie, 7th ed.
Munich: W. Zuckschwerdt Verlag, 1993:561–72.
3. Jamnicki M, Zollinger A, Seifert B, et al. Compromised blood
coagulation: an in vitro comparison of hydroxyethyl starch 130/
0.4 and hydroxyethyl starch 200/0.5 using thromboelastogra-
phy. Anesth Analg 1998;87:989 –93.
4. Ljungström KG. Colloid safety: fact and fiction. Bailliére’s Clin-
ical Anaesthesiology 1997;11:163–77.
5. Dietrich W. Einflu␤ von kolloidalen volumenersatzmitteln auf
die blutgerinnung. Anästhesiol Intensivmed Notfallmed
Schmerzther 1998;33:266 – 8.
6. Harke H, Pieper C, Meredig J, et al. Rheologische und gerin-
nungsphysiologische untersuchungen nach infusion von HÄS
200/0.5 und dextran 40. Anaesthesist 1980;29:71–7.
7. Mallett SV, Cox DJA: Thromboelastography. Br J Anaesth 1992;
69:307–13.
8. Tuman KJ, Spiess BD, McCarthy RJ, et al. Effects of progressive
blood loss on coagulation as measured by thromboelastogra-
phy. Anesth Analg 1987;66:856 – 63.
800 CARDIOVASCULAR ANESTHESIA PETROIANU ET AL.
COLLOIDS AND COAGULATION
9. Hartert H: Blutgerinnungstudien mit der thrombelastographie,
einen neuen untersuchungsverfahren. Klinische Wochenschrift
1948;16:577– 83.
10. Petroianu GA, Werth WA, Maleck WH, et al. Thromboelastog-
raphy in male, pregnant and fetal minipigs. Scand J Lab Anim
Sci Suppl 1998;25:221–3.
11. Conroy JM, Fishman RL, Reeves ST, et al. The effects of desmo-
pressin and 6% hydroxyethyl starch on factor VIII:C. Anesth
Analg 1996;83:804 –7.
12. Blaicher AM, Reiter WJ, Blaicher W, et al. The effect of hydroxy-
ethyl starch on platelet aggregation in vitro. Anesth Analg 1998;
86:1318 –21.
13. Niemi TT, Kuitunen AH. Hydroxyethyl starch impairs in vitro
coagulation. Acta Anaesthesiol Scand 1998;42:1104 –9.
14. Warren BB, Durieux ME. Hydroxyethyl starch: safe or not?
Anesth Analg 1997;84:206 –12.
ANESTH ANALG
2000;90:795–800
15. Mortier E, Ongenae M, de Baerdemaeker L, et al. In vitro eval-
uation of the effect of profound haemodilution with hydroxy-
ethyl starch 6%, modified fluid gelatin 4% and dextran 40 10%
on coagulation profile measured by thromboelastography. An-
aesthesia 1997;52:1061– 4.
16. Karoutsos K, Nathan N, Lahrimi A, et al. Thromboelastogram
reveals hypercoagulability after administration of gelatin solu-
tion. Br J Anaesth 1999;82:175–7.
17. Ruttmann TG, James MFM, Viljoen JF. Haemodilution induces a
hypercoagulable state.Br J Anaesth 1996;76:412– 4.
18. Ruttmann TG, James MFM, Aronson I. In vivo investigation into
the effects of haemodilution with hydroxyethyl starch (200/0.5)
and normal saline on coagulation. Br J Anaesth 1998;80:612– 6.