General Articles

Prolonged International Normalized Ratio Correlates

with a Large Intravascular Fluid Balance After Major
Abdominal Surgery
Michal Barak, MD*
Oded Jurim, MD
Ronit Tal, MD
Yeshayahu Katz, MD, DSc

We performed a prospective randomized study of 32 patients who had undergone

pancreaticoduodenectomy and did not receive blood during and after surgery. The
patients were prospectively assigned to two groups related to fluid balance in the
immediate postoperative period. Group 1 (14 patients) were maintained at a
positive intravascular fluid balance of 0 1000 mL; Group 2 (18 patients) were
maintained at a positive balance of 1000 2000 mL. Complete blood counts and
coagulation tests (International Normalized Ratio) and activated partial thromboplastin time (aPTT) were performed at three time points: the day before surgery, on
arrival at the postanesthesia care unit, and on leaving the postanesthesia care unit
(16 h later). There were significant differences in International Normalized Ratio
values between the groups with deterioration during the time they were in the
postanesthesia care unit but not in aPTT values. Positive correlation was found
between the amount of positive fluid balance and International Normalized Ratio
prolongation, but not with aPTT, suggesting that restricted intravascular fluid
balance is beneficial for preservation of coagulation after major abdominal surgery.
(Anesth Analg 2006;103:448 52)

assively transfused trauma patients may suffer

from disordered hemostasis as a result of thrombocytopenia or coagulopathy (13). Earlier studies used
transfusion with whole blood (1); however, a similar
tendency was found in more recent studies in which
packed red blood cells were used. In patients who lost
more than 50% of their blood volume during elective
surgery, dilution of coagulation factors, rather than
dilution thrombocytopenia, was the initial cause of
increased surgical bleeding (4). We hypothesized that
the dilution effect of IV fluid is dependent on the
volume infused. Previously, we (5) found significant
correlation between deteriorating coagulation and the
administration of more than 3 L of crystalloids during
abdominal surgery. There was also correlation between administration of more than 500 mL of colloid
and prolonged prothrombin time. The purpose of the

From the Departments of *Anesthesiology and Surgery B,

Rambam Medical Center; Department of Surgery A, Carmel Medical Center; Laboratory for Anesthesia, Pain and Neural Research,
Rappaport Faculty of Medicine, TechnionIsrael Institute of Technology, Haifa; Department of Anesthesiology, Haemek Medical
Center, Afula, Israel.
Accepted for publication April 11, 2006.
Address correspondence and reprint requests to Yeshayahu
Katz, MD, DSc, Chairman, Department of Anesthesiology, Haemek
Medical Center, Afula 18101, Israel. Address e-mail to
ykatz18@hotmail.com.
Copyright 2006 International Anesthesia Research Society
DOI: 10.1213/01.ane.0000223677.34513.88

448

present study was to investigate the effect of postoperative intravascular fluid balance on the coagulation
profile after major abdominal surgery.

METHODS
We conducted a prospective study over 18 mo on
adult patients scheduled for elective pancreaticoduodenectomy with combined epidural and general anesthesia. The study was approved by our local Ethics
Committee and all the patients provided signed informed consent. Exclusion criteria were history of
coagulopathy, treatment with anticoagulants, aspirin,
or herbal medications, preoperative anemia, abnormal
preoperative coagulation tests or platelet count, renal
insufficiency (preoperative creatinine 2.0 mg/dL),
morbid obesity, and administration of blood or blood
products (packed red blood cells, fresh-frozen plasma)
before, during, or after surgery.
During the studied time period, 37 patients underwent pancreaticoduodenectomy. Five patients received blood during the operation and were excluded
from the study. We collected demographic and medical information, surgical and anesthetic information,
events during and after surgery, and blood tests for
each patient.
All patients had standard identical care before and
during the operation. One surgeon (OJ) performed the
operation with the same team of assistants. The patient was allowed to drink clear water until 2 h before
surgery. Oral premedication included diazepam 10
mg and metoclopramide 10 mg 1 h before anesthesia.
Vol. 103, No. 2, August 2006

On arrival in the operating room, an IV line was

placed for fluid and drug administration. The patient
was continuously monitored with electrocardiogram,
pulse oximeter, end-tidal CO2 capnograph, invasive
arterial blood pressure, central venous pressure, urinary catheter, and esophageal thermometer. Body
temperature was kept within 35.8 36.8C using an
active air warming blanket. Warm lactated Ringers
solution was administered for maintenance based on
the formula 4:2:1 mL kg1 h1, supplemented with
an additional 5 mL kg1 h1 for external losses
when the abdomen was open (6). In addition, each
patient received 7 mL/kg of hydroxyethyl starch 6%
in normal saline (HAES-steril 6%, Mw 200 kDa, MS
0.5, C2/C6 ratio 5:1; Fresenius Kabi Deutschland
GmbH, Bad Homburg, Germany) during the operation. A thoracic epidural catheter (T57) was placed
and analgesia was maintained continuously, using 6
mL/h of bupivacaine 0.125% with fentanyl 5 g/mL
during the surgery and for 3 days afterwards. General
anesthesia was induced with midazolam 0.03 mg/kg,
fentanyl 0.001 mg/kg, and thiopental 35 mg/kg
followed by vecuronium 0.1 mg/kg for muscle relaxation. Isoflurane 0.7%1.1% in a mixture of air and
oxygen was used for maintenance of anesthesia. Operative blood loss was estimated by counting sponges
and measuring the volume collected in the suction
bottles. Blood loss of 500 mL was not replaced by
additional fluids.
All patients were tracheally extubated at the end of
the surgery while in the operating room and then
transferred to the postanesthesia care unit (PACU)
where they stayed for 16 h. Patients who experienced
pain (visual analog score 3) received epidural
supplements or IV meperidine 20 mg.
In the PACU, the basic standard fluid administration was lactated Ringers solution 100 mL/h. In our
previous experience, we observed a tendency towards
diminished urinary output a few hours after surgery.
Therefore, 6 h after the end of this surgery, each
patient received HAES 6% 500 mL and furosemide 5
mg.
The intravascular fluid balance during the 16-h
period was randomized to one of the following two
options: a positive fluid balance of 0 (1000) mL
(Group 1) or a positive balance of (1000)(2000)
mL (Group 2). Patients were randomized according to
an arbitrary discretion of one of four attending anesthesiologists in the PACU, who were not study investigators. Their instructions of intravascular fluid balance were primarily a set standard, regardless of the
patient. The PACU nurses and the investigators were
not blinded as to these variables. Every 4 h the PACU
nurse calculated the intravascular fluid balance from
arrival at the PACU as follows: input (standard fluid
administration plus solutions given for drug administration) minus output (urine and drain output). If the
balance was less than targeted, the nurse gave lactated
Ringers solution 250 mL in addition to the basic fluid
Vol. 103, No. 2, August 2006

protocol. If the balance was above the upper limit of

the target, the patient received furosemide 5 mg. The
final intravascular fluid balance on leaving PACU to
the ward was recorded.
During surgery and PACU stay, mean arterial
blood pressure was kept above 60 mm Hg. If lower,
ephedrine 5 mg was administered. Urinary output
was targeted to be more than 0.3 mL kg1 h1. The
decision to give blood (packed red blood cells unit)
was based on estimated blood loss, the patients vital
signs, and repeated blood tests (the lowest hemoglobin level allowed was 9.0 g%).
Complete blood count and coagulation tests were
performed at the following time points: on the day
before surgery, on arrival at PACU, and on leaving
PACU to the ward (16 h apart). Blood samples were
collected in EDTA tubes for complete blood count and
citrate for coagulation assays. Analysis included hemoglobin and platelet counts (Technicor H3; Bayer,
Germany), International Normalized Ratio (INR) (normal, 0.751.3), and activated partial thromboplastin
time (aPTT) (normal, 27 40 s) (STA-compact; Diagnostica, Stago, France). Serum creatinine was measured every other day; for the study we used data
before surgery and at discharge from hospital.
Statistical analysis was performed using Statistica 6
(StatSoft, Tulsa, OK). Intergroup frequency differences
were compared using Pearsons 2 test. Normally distributed parametric variables were compared with the
Students t-test or one-way analysis of variance. Data of
the hemoglobin, platelet count, and coagulation tests
were compared with repeated-measures analysis of variance univariate tests. Wherever normality was violated
(by the Mauchley Sphericity Test), Greenhouse & Geisser and Huynh-Feldt adjustments were applied. Analysis of covariance of the remaining demographic and
clinical variables, with each time point of the coagulation
test data as dependant variable, was performed. Correlations were analyzed by Pearsons method. Significance
level was set at P 0.05.

RESULTS
The clinical characteristics of the patients, anesthesia, and surgical details are shown in Table 1. No
significant difference was found between groups in
these variables. No patient had bleeding complications or required reoperation. There was no inhospital mortality. Postoperative complications are
shown in Table 1. On discharge from hospital, no
patient had a serum creatinine above 2.0 mg/dL.
Blood test results (hemoglobin and platelet count) are
summarized in Table 2. On arriving at PACU, hemoglobin was significantly less than preoperative hemoglobin
for each study group (P 0.01 for both groups). Also, on
leaving PACU hemoglobin was significantly less than
preoperative hemoglobin (P 0.01 for both groups). In
each time point there was no significant difference in
hemoglobin values between Group 1 and Group 2.
2006 International Anesthesia Research Society

449

Table 1. Baseline Clinical Characteristics, Anesthetic and Surgical Data

Age (yr)
Male/female
Weight (kg)
ASA II /III
Co-morbidity
Diabetes (n)
Cardiovascular diseases (n)
Pulmonary diseases (n)
Preoperative creatinine (mg/dL)
Duration of surgery (min)
Estimated blood loss during surgery (mL)
Intraoperative lactated Ringers solution
input (mL)
Fluid balance in PACU (mL)
Length of stay in hospital (days)
Postoperative complications
Respiratory complications (atelectasis,
pneumonia)
Sepsis
Abdominal abscess
Wound infection
Minor stroke
Values are median (range), n, or mean

Group 1 Balance
0 (1000) mL
(n 14)

Group 2 Balance
(1000)(2000) mL
(n 18)

59 (4883)
8/6
76 (61100)
5/9

65 (2786)
10/8
70 (4590)
7/11

4
7
5
0.77 0.23
420 (310480)
280 (220500)

6
9
4
0.75 0.17
445 (340540)
310 (250410)

2450 (21504250)
555 (165(980))
16 (1130)

2350 (18504500)
1830 (12002000)
17 (828)

2
2
2
3
0

3
2
0
1
1

SD.

There was a similar trend in the platelet count, with no

significant difference between groups
INR and aPTT changed significantly in each group
over time. The preoperative mean INR was 1.04 for
both groups. On arriving at the PACU, it was 1.10 and
1.11 for Groups 1 and 2, respectively. On leaving the
PACU, the mean INR was 1.15 for Group 1 and 1.27
for Group 2 (Fig. 1). Comparing INR on arriving at the
PACU to INR on leaving the PACU, the change was
significant within each group (P 0.01).
The same trend was found in aPTT values. The
preoperative mean aPTT was 33.0 s for Group 1 and
32.9 s for Group 2. On arrival at the PACU, it was 33.9
and 34.2 for Groups 1 and 2, respectively. On leaving
the PACU, aPTT was 37.8 for Group 1 and 40.0 for
Group 2 (Fig. 2). There were significant differences

between preoperative aPTT and aPTT on leaving the

PACU and between aPTT on arrival at the PACU and
on leaving the PACU within each group (P 0.01).
At two time points, preoperatively and on arrival at
the PACU, INR values were similar in both groups (P
0.90 and 0.52, respectively). However, on leaving the
PACU, there was a significant difference between
groups (P 0.001) (Fig. 1). No significant difference was
found in the aPTT values of the groups at the 3 time
points (P 0.95, 0.83, and 0.51, respectively) (Fig. 2).
We defined the change in INR during PACU stay
(INR on leaving PACU/INR at PACU arrival) as INR

Table 2. Hemoglobin and Platelet Count on Day Before

Surgery, on Arrival at PACU and on Leaving PACU for the Ward

Hemoglobin (g%)
Preoperative
Arrival at PACU
Leaving PACU
Platelet count
(103/mm3)
Preoperative
Arrival at PACU
Leaving PACU

Group 1
Balance
0 (1000) mL
(n 14)

Group 2
Balance
(1000)
(2000) mL
(n 18)

13.5 1.4
11.7 1.0*
11.2 1.3*

13.0 1.6
11.1 1.2*
10.5 1.5*

239 117
213 101
220 82

308 131
249 110
228 93

Values are mean sd

* P 0.01 compared with preoperative value.

450

Prolonged INR Correlates with Fluid Balance

Figure 1. International Normalized Ratio (INR) at three time

points: on the day before surgery, on arrival in the postanesthesia care unit (PACU), and on leaving the PACU for two
study groups. Data are mean 95% confidence interval.
*P 0.01 Group 1 versus Group 2.
ANESTHESIA & ANALGESIA

Figure 2. Activated partial thromboplastin time (aPTT) at

three time points: on the day before surgery, on arrival at the
postanesthesia care unit (PACU), and on leaving the PACU
for two study groups. Data are mean 95% confidence
interval. No statistical significance was found.

quotient. We found significant correlation between the

total intravascular fluid balance on leaving the PACU
and the INR quotient (r 0.54, P 0.004) (Fig. 3). No
correlation was found between total fluid balance on
leaving PACU and aPTT quotient (aPTT on leaving
PACU/aPTT on PACU arrival) (r 0.04; P 0.8).

DISCUSSION
For many years IV fluid administration has been one
of the fundamentals of patient management, intended to
maintain hemodynamic stability and prevent hypotension and hypoperfusion (7). Solutions used as intravascular volume replacement leak to the extravascular
space (nonfunctional fluids), resulting in increased total
body water (8). This impaired homeostasis is further
exacerbated by perioperative events, such as decreased
blood oncotic pressure, aldosterone and antidiuretic
hormone secretion triggered by stress, third space in the

Figure 3. Correlation between International Normalized

Ratio (INR) quotient (INR on leaving the postanesthesia care
unit [PACU]/INR on arrival to PACU) and total fluid
balance on leaving the PACU. The correlation is statistically
significant (r 0.54; P 0.004).
Vol. 103, No. 2, August 2006

traumatized/surgical compartment, and patient immobilization (9,10). Excess extravascular fluid may decrease
tissue oxygenation, deteriorate pulmonary function, and
disrupt wound and anastomosis healing (11). These
pathophysiologic events have clinical implications, as
has been demonstrated (1214). Brandstrup et al. (14)
showed that generous fluid administration resulted in
increased postoperative cardiopulmonary complications, tissue-healing complications, and mortality. In
contrast, Holte et al. (15) showed that generous crystalloid administration improved postoperative organ function and recovery and shortened hospital stay after
laparoscopic cholecystectomy. Although relating to
many aspects of perioperative outcome, coagulation has
not been dealt with (1215).
Our study focused on the effect of intravascular
fluid balance on coagulation during the immediate
postsurgical period. Several perioperative factors,
such as mild or moderate hypothermia, blood administration, and trauma per se may affect coagulation
(3,16 18). Hemodynamic instability with low hepatic
perfusion during surgery reduces production of coagulation factors and degradation of activated factors
(19). Earlier studies examined the influence of various
types of solutions on hemostasis. It was found that
colloids reduce clot strength and platelet function
(20 23). In patients who lost 50% of their blood
volume during surgery, the use of crystalloid solutions and packed red blood cells caused coagulation
factor dilution (4). Indeed, there is no scientific evidence for the occurrence of dilution coagulopathy in
minor bleeding, where no blood products are used
(24,25). In our study, we excluded patients who received blood products and used a uniform composition of solutions to determine the net effect of the
amount of infused solutions on coagulation. The body
temperature of all patients was kept strictly within
normal limits to avoid the effect of hypothermia. We
found that the patients who had restricted intravascular fluid balance exhibited preserved INR compared
with those who had a larger fluid balance. Moreover,
we showed correlation between the fluid balance and
deterioration of INR during the PACU stay. On leaving the PACU, 4 patients would have been considered
to have an increased INR in Group 2 compared with
none in Group 1. It should be noted that the limit of
normal INR is somewhat arbitrary, as INR is a continuous scale. Knowing the dose (fluid) versus response (INR) relationship is the key point, and the
number that crosses a particular threshold value is less
important. We believe that understanding the sensitivity of INR to extra intravascular fluid is important;
without it the treatment might be fresh-frozen plasma
instead of waiting and avoiding excessive fluid.
The aPTT values were similar between our two
patient groups. In earlier studies, the infusion of large
volumes of fluids resulted in significant increase of
both PT and aPTT (2,4,5). However, our study was
designed to test the effect of relatively small volumes.
2006 International Anesthesia Research Society

451

We propose that the extrinsic pathway (INR) is more

vulnerable to the negative effect of fluid administration than the intrinsic pathway (aPTT). Factor VII may
be responsible for the susceptibility to fluids because
of its short half-life being the rate-limiting component
of the clotting cascade and the main noncellular
humoral factor of the extrinsic pathway (26). Conversely, a hypercoagulable state occasionally occurs
after surgery, apparent by postoperative deep vein
thrombosis and pulmonary embolism and by abnormal thromboelastogram (27,28). The contradiction between hypercoagulability and prolonged INR may be
explained by the fact that both coagulation cascade
and the fibrinolytic system are damaged simultaneously.
A major concern in managing the patient with a
restricted fluid regimen is maintaining renal function
(29). Data regarding renal function in controlled hypovolemic conditions is found in studies of liver
resections and thoracotomies, where a restricted fluid
regimen is the standard of care. From these studies it
is apparent that the risk of renal damage in dry
regimen is very small (30 33). In our study, we had no
case of renal failure or permanently increased serum
creatinine. Thus, we conclude that the risk of causing
renal failure by fluid restriction is not high.
In summary, we found that there is correlation
between a positive intravascular fluid balance and
deterioration of INR in the immediate postoperative
period. It is reasonable to believe that policy should
aim at a restricted fluid regimen if preserved coagulation is desired. Further investigation is necessary to
establish definitive end-points for fluid management.
REFERENCES
1. Miller RD, Robbins TO, Tong MJ, Barton SL. Coagulation
defects associated with massive blood transfusion. Ann Surg
1971;174:794801.
2. Reed 2ndRL Ciavarella D, Heimbach DM, et al. Prophylactic
platelet administration during massive transfusion. Ann Surg
1986;203:408.
3. Murray DJ, Olson J, Strauss R, Tinker JH. Coagulation changes
during packed red cell replacement of major blood loss. Anesthesiology 1988;69:83945.
4. Murray DJ, Pennell BJ, Weinstein SL, Olson JD. Packed red cells
in acute blood loss: dilutional coagulopathy as a cause of
surgical bleeding. Anesth Analg 1995;80:33642.
5. Barak M, Rudin M, Vofsi O, et al. Fluid administration during
abdominal surgery influences on coagulation in the postoperative period. Curr Surg 2004;61:45962.
6. Prough D, Mathru M. Acid-base, fluids, and electrolytes. In:
Barash PG, Cullen BF, Stoelting RK, eds. Clinical anesthesia, 3rd
ed. Philadelphia: Lippincott-Raven, 1997:1656.
7. Cosnett JE. The origins of intravenous fluid therapy. Lancet
1989;1:76871.
8. Shires GT III, Barber A, Shires GT. Fluids and electrolytes
management of the surgical patient. In: Schwartz, ed. Principles
of surgery. New York: McGraw-Hill, 1999;5377.
9. Irvin TT, Modgill VK, Hayter CJ, et al. Plasma-volume deficits
and salt and water excretion after surgery. Lancet
1972;2:115962.
10. Holte K, Sharrock NE, Kehlet H. Pathophysiology and clinical
implications of perioperative fluid excess. Br J Anaesth
2002;89:62232.

452

Prolonged INR Correlates with Fluid Balance

11. Rosenthal MH. Intraoperative fluid management: what and

how much? Chest 1999;115:106s12s.
12. Lobo DN, Bostock KA, Neal KR, et al. Effect of salt and water
balance on recovery of gastrointestinal function after elective
colonic resection: A randomized controlled trial. Lancet
2002;359:18128.
13. Lowell JA, Schifferdecker C, Driscoll DF, et al. Postoperative
fluid overload: not a benign problem. Crit Care Med
1990;18:72833.
14. Brandstrup B, Tonnesen H, Beier-Holgersen R, et al. Danish
Study Group on Perioperative Fluid Therapy. Effects of intravenous fluid restriction on postoperative complications: comparison of two perioperative fluid regimens. Ann Surg
2003;238:6418.
15. Holte K, Klarskov B, Christensen DS, et al. Liberal versus
restrictive fluid administration to improve recovery after laparoscopic cholecystectomy: a randomized, double-blind study.
Ann Surg 2004;240:8929.
16. Farkash U, Lynn M, Scope A, et al. Does prehospital fluid
administration impact core body temperature and coagulation
functions in combat casualties? Injury 2002;33:10310.
17. Kirkpatrick AW, Chun R, Brown R, Simons RK. Hypothermia
and the trauma patient. Can J Surg 1999;42:33343.
18. Brohi K, Singh J, Heron M, Coats T. Acute traumatic coagulopathy. J Trauma 2003;54:112730.
19. Amitrano L, Guardascione MA, Brancaccio V, Balzano A. Coagulation disorder in liver disease. Semin Liver Dis
2002;22:8396.
20. Ng KF, Lam CC, Chan LC. In vivo effect of hemodilution with
saline on coagulation: a randomized controlled trial. Br J Anaesth 2002;88:4702.
21. Turkan H, Ural AU, Beyan C, Yalcin A. Effects of hydroxyethyl
starch on blood coagulation profile. Eur J Anaesthesiol
1999;16:1569.
22. Martin G, Bennett-Guerrero E, Wakeling H, et al. A prospective
randomized comparison of thromboelastographic coagulation
profile in patients receiving lactated Ringers solusion, 6%
hetastarch in a balanced-saline vehicle, or 6% hetastarch in
saline during major surgery. J Cardiothorac Vasc Anesth
2002;16:4416.
23. Innerhofer P, Fries D, Margreiter J, et al. The effect of perioperatively administered colloids and crystalloids on primary
platelet-mediated hemostasis and clot formation. Anesth Analg
2002;95:85865.
24. Boldt J. New light on intravascular volume replacement regimens: what did we learn from the past three years? Anesth
Analg 2003;97:1595604.
25. Kwan I, Bunn F, Roberts I. WHO Pre-Hospital Trauma Care
Steering Committee. Timing and volume of fluid administration
for patients with bleeding. Cochrane Database Syst Rev
2003;3:CD0002245.
26. Roberts HR, Monroe DM, Escobar MA. Current concepts of
hemostasis: implications for therapy. Anesthesiology
2004;100:72230.
27. Bergqvist D, Lowe GD, Berstad A, et al. Prevention of venous
thromboembolism after surgery: a review of enoxaparin.
Br J Surg 1992;79:4958.
28. Ng KF, Lam CCK, Chan LC. In vivo effect of haemodilution
with saline on coagulation: a randomized controlled trial. Br J
Anaesth 2002;88:47580.
29. Campbell IT, Baxter JN, Tweedie IE, et al. IV fluids during
surgery. Br J Anaesth 1990;65:7269.
30. Melendez JA, Arslan V, Fischer ME, et al. Perioperative outcomes of major hepatic resections under low central venous
pressure anesthesia: blood loss, blood transfusion, and the risk
of postoperative renal dysfunction. J Am Coll Surg
1998;187:6205.
31. Neal JM, Wilcox RT, Allen HW, Low DE. Near total esophagectomy: The influence of standardized multimodal management
and intraoperative fluid restriction. Reg Anesth Pain Med
2003;28:32834.
32. Moller AM, Pedersen T, Svendsen PE, Engquist A. Perioperative
risk factors in elective pneumonectomy: the impact of excess
fluid balance. Eur J Anaesth 2002;19:5762.
33. Jordan S, Mitchell JA, Quinlan GJ, et al. The pathogenesis of
lung injury following pulmonary resection. Eur Respir J
2000;15:7909.

ANESTHESIA & ANALGESIA