Full text of DOI:10.1007/BF03020384

1155
Neuroanesthesia and Intensive Care
Point of care and central laboratory determina-
tions of the aPTT are not interchangeable in sur-
gical intensive care patients
[Les examens visant à déterminer le TCA, réalisés au chevet du malade ou dans un laboratoire
central, ne sont pas interchangeables pour les patients des soins intensifs chirurgicaux]
Martine Ferring MD, Guido Reber PhD,* Philippe de Moerloose MD,* Paolo Merlani MD, Marc Diby RN,
Bara Ricou MD
Purpose: The objective of the study was to compare a bedside
whole blood activated partial thromboplastin time (aPTT) per-
formed by a point of care (POC) apparatus (CoaguCheck® Pro) in
surgical intensive care (SIC) patients with a conventional aPTT
obtained from the central laboratory.
Objectif : Comparer le temps de céphaline activé (TCA) du sang total
réalisé au chevet du malade (CDM), au moyen d’un CoaguCheck®
Pro, avec le TCA traditionnel obtenu d’un laboratoire central, pour des
patients aux soins intensifs chirurgicaux (SIC).
Méthode : Les mesures prospectives concomitantes des deux TCA ont
été faites pour 233 échantillons prélevés auprès de 46 patients suc-
cessifs admis aux SIC après une opération cardio-vasculaire ou
abdominale majeure.
Methods: The prospective concomitant measurements of the two
aPTT were performed in 233 samples from 46 consecutive
patients admitted after cardiovascular or major abdominal surgery.
Results: Inter-operator, inter-instrument and inter-cartridge variabil-
ity of the new device measured in three healthy volunteers and in
nine patients in stable condition (controls) was low (F test: P=0.86).
The agreement by Bland and Altman between POC and central lab-
oratory aPTT (-20.2 ± 18.8 sec) was not satisfactory. The agree-
ment between POC and central laboratory aPTT in patients after
surgery was worst (-17 ± 33.1 sec). Heparin treatment or timing of
blood sampling after intensive care admission (<48 hr vs >48 hr)
did not influence the agreement. The correlation between POC or
central laboratory aPTT and anti-factor Xa activity was poor (r2
0.077 and 0.181 respectively). The test which correlated the best to
heparin doses was anti-factor Xa activity (r2 0.714).
Résultats : La variabilité inter-opérateurs, inter-instruments et inter-
éprouvettes du nouvel appareil étudié auprès de trois volontaires en
bonne santé et neuf patients de condition stable (témoins) a été faible
(Test F : P = 0,86). Le test de concordance de Bland et Altman entre
le TCA (-20,2 ± 18,8 s) réalisé au CDM ou en laboratoire n’était pas
satisfaisant. Ce même test réalisé chez des patients après l’opération
a été pire (-17 ± 33,1 sec). Le traitement à l’héparine ou la chronolo-
gie de l’échantillonnage prélevé après l’arrivée aux soins intensifs (<
48 h vs > 48 h) n’a pas eu d’effet sur la concordance. Il n’existait
qu’une très faible corrélation entre le TCA fait au CDM ou en labora-
toire et l’activité de l’anti-facteur Xa (r2 0,077 et 0,181 respective-
ment). Le test le mieux corrélé aux doses d’héparine a été l’activité de
l’anti-facteur Xa (r2 0,714).
Conclusion: POC aPTT and central laboratory aPTT showed a
poor agreement in SIC patients admitted after surgery, although in
healthy volunteers or in control patients, this agreement was bet-
ter. The best test to monitor heparin treatment in this setting was
anti-factor Xa activity.
Conclusion : Le TCA fait au CDM et le TCA du laboratoire central
n’ont affiché qu’une faible concordance chez des patients des SIC
admis après l’intervention, même si cette concordance a été meilleure
pour des patients témoins et des volontaires en bonne santé. Dans ces
circonstances, le meilleur test pour suivre le traitement avec héparine
a été l’activité de l’anti-facteur Xa.
From the Divisions of Surgical Intensive Care, Department of Anesthesiology, Pharmacology and Surgical Intensive Care and the
Hemostasis unit,* Division of Angiology and Hemostasis, Department of Medicine, University Hospital of Geneva, Geneva, Switzerland.
Address correspondence to: Professor Philippe de Moerloose, Hemostasis Unit, University Hospital, Geneva, 1211 Geneva 14,
Switzerland. Phone: 0041-22-37 27 752; Fax: 0041-22-37 29 777; E-mail: philippe.deMoerloose@hcuge.ch
Accepted for publication May 29, 2001.
Revision accepted August 20, 2001.
CAN J ANESTH 2001 / 48: 11 / pp 1155–1160

1156
CANADIANJOURNAL OF ANESTHESIA
OAGULATION disorders are observed
commonly in surgical intensive care (SIC)
patients. Coagulation tests are, therefore,
performed frequently, particularly the acti-
POC and central laboratory aPTT and anti-factor Xa
activity
The aPTT was determined both in the central laborato-
ry (conventional method) and in the SIC (POC). Blood
was drawn into a syringe by direct venipuncture when-
ever possible, or through an arterial catheter from which
10 mL (ten times the dead space) of blood4 were with-
drawn prior to collection of the sample in order to min-
imize its contamination with heparin from the
continuous irrigation systems attached to the invasive
catheter. A drop of whole blood was deposited on the
prewarmed aPTT cartridge of the POC, and the remain-
der dispensed into the tube for central laboratory assess-
ment. The maximum delay between blood sampling and
central laboratory measurement was 60 min.
For the aPTT determined in the central laboratory,
plasma (citrated blood, 0.129 mol·L– 1sodium citrate)
was obtained by centrifugation at 3500 x g for ten
minutes at 15–20°C. Actin (Dade Behring) was used
as the source of phospholipid and activator (ellagic
acid). The tests were performed with automated coag-
ulometers, either a Behring Coagulation System or a
Behring Coagulation Timer (all from Dade Behring,
Marburg, Germany). Maximal measuring time was set
at 200 sec, the reference range being 25 to 31 sec.
Heparin concentrations were evaluated with an
anti-factor Xa (anti-FXa) chromogenic assay
(Berichrom Heparin, Dade Behring) and the Behring
Coagulation System. Since bovine antithrombin is also
present in the reagent, the test is independent of its
actual concentration in the sample.9 The target values
depends on the clinical situation (therapeutic ranges
from 0.3 to 0.6 U·mL– 1).
The POC provided by CoaguCheck® Pro is a battery-
powered, portable laser photometer using a phospholipid
(soybean phosphatide) as platelet substitute and a bovine
brain sulfatide as activator of intrinsic coagulation. 45 µL
samples of capillary, arterial or venous whole blood can
be used. Instructions for use were explained to the nurs-
ing team during 20 min teaching sessions, three times per
week, for four weeks. The monitors were checked every
two days (20 measurements) with two level control car-
tridges. Each two weeks, a lyophilized whole blood inter-
nal quality control was conducted. Reference interval was
established previously by the manufacturer (range 20.2 to
40.8 sec). Three portable devices were utilized. Inter-
operator, inter-instrument and inter-cartridge variability
of POC aPTT was determined by simultaneous analysis
of the same blood sample drawn from 12 control patients
by three different nurses.
C
vated partial thromboplastin time (aPTT) which is
used widely because it allows the detection of various
coagulation abnormalities as well as the monitoring of
heparin therapy.
The time from venipuncture to the availability of
aPTT results from the central laboratory can be very
long.1–4 Such delays hinder rapid diagnosis of coagu-
lation disorders and adequate anticoagulation to be
achieved rapidly. Rapid knowledge of aPTT results
should allow the implementation of measures to
decrease bleeding as well as the prevention of throm-
botic complications.
A portable bedside point of care instrument (POC:
CoaguCheck®
Pro,
Boehringer
Mannheim
Diagnostics, USA, Roche Diagnostics, Switzerland),
which allows aPTT measurements in three minutes, has
been shown to be accurate in both controls, anticoagu-
lated and/or coronary unit patients1,2,5–7 and therefore
may be of interest also in SIC patients. However, there
are only few data available on the usefulness and relia-
bility of such a device for these patients.4,8 The objective
of this study was to evaluate prospectively the accuracy
of POC in comparison with central laboratory aPTT
after cardiovascular or abdominal surgery.
Patients and methods
Patients and data collection
Consecutive patients admitted to the SIC at Geneva
University Hospital after cardiovascular or major
abdominal surgery were enrolled from July 12 to
August 20, 1999. Patients with known previous coag-
ulation disorders were excluded. Patients were tested
at six-hour intervals from admission until achievement
of normal aPTT values given by the central laborato-
ry. I V heparin was then started either for prophylactic
or therapeutic use. The aPTT was thereafter deter-
mined routinely once daily and four hours after any
change of heparin dosing.
To assess the inter-operator, inter-instrument and
inter-cartridge variability of aPTT measured by POC,
three healthy volunteers and nine SIC stable patients
(before SIC discharge, without known bleeding disor-
der and without coagulation abnormalities) were test-
ed (controls).
The study was approved by the Ethical Committee
for Human Research of our institution.
Data were analyzed according to the administration
of heparin, aprotinin, aspirin and clopidogrel before
SIC admission.

Ferring et al.: aPPT MONITORING I N THE ICU
1157
Statistical analysis
TABLE I Patients characteristics at SIC admission (n=46)
Statistical analysis was performed on a Macintosh
computer utilizing Statview II software (Abacus
Concepts, Berkeley, California, USA). Results were
analyzed either by Bland and Altman1 0for comparison
of similar assays and by linear regression analysis for
different assays. To be interchangeable, we deter-
mined a priori an upper limit of ten seconds as an
acceptable difference between the two methods, as it
would not modify clinical management. Inter-opera-
tor, inter-instrument and inter-cartridge variability of
POC aPTT were assesed by F test. We planned to ana-
lyze >200 samples to demonstrate in different sub-
groups of 25 samples a r >0.4 with P <0.05.
Age
Sex
Body weight
Diagnosis
years (mean SD)
males/females (n/n)
kg (mean SD)
61 14
33/13
76 18
n
30
26
4
16
2
10
1
3
Cardiovascular surgery
Cardiac surgery
Aortic surgery
Abdominal surgery
Hepatic surgery
Intestinal surgery
Pancreatic surgery
Biliar surgery
Therapy before SIC admission
Heparin
n
38
26
14
6
Protamine
Aprotinin
Aspirin
Clopidogrel
For calculation purposes, when a sample was not clot-
table, an arbitrary 200 sec clotting time was assigned
(three samples with POC aPTT and two samples with
central laboratory aPTT and two with both tests).
1
TABLE II Correlation between CoaguCheck® Pro aPTT, cen-
tral laboratory aPTT, anti-FXa assay and heparin dosage
Results
SIC patients
Comparison
Number of
samples
Correlation
coefficient r2
Forty-six patients were included in the study. Table I
shows the patients’ characteristics on admission. Two
hundred thirty-three aPTT measurements were real-
ized, 181 from cardiovascular and 52 from abdominal
surgery patients. For each point, the difference
between the clotting times obtained with the two
assays was plotted as a function of their mean. Only
clottable samples are represented. As shown in Figure
1, agreement between central laboratory aPTT and
POC aPTT was poor (-17 33.1 sec). There was an
important dispersion of values and a trend towards
longer clotting times with POC. When the patients
were divided according to the type of surgery, the
agreement was slightly better in abdominal (-23.3
25.5 sec) than in cardiovascular surgery patients
CoaguCheck® aPTT/
heparin dosage
CoaguCheck® aPTT/
anti-FXa assay
Central laboratory aPTT/
heparin dosage
Central laboratory aPTT/
anti-FXa assay
136
71
0.035
0.077
0.061
0.181
0.714
136
71
Anti-FXa assay/
heparin dosage
52
aPTT=activated partial thromboplastin time; anti-FXa=anti-factor
Xa.
(-15.2
35 sec). The mean difference between the
two methods was lower in samples collected in the ini-
tial 48 hr (-11.5 33 sec) as compared to samples col-
lected after 48 hr (-27.5 31 sec). Heparin did not
influence the agreement (-20.8 27 sec).
hepatic alterations, anemia, thrombocytopenia and
temperature alterations before or during SIC did not
alter the agreement between central laboratory and
POC aPTT (data not shown).
POC as well as central laboratory aPTT values cor-
related poorly with heparin dosage (r2 0.035 and r2
0.061 respectively) and anti-FXa (r2 0.077 and 0.181
respectively). The best correlation with heparin
dosage was obtained with anti-FXa activity measure-
ments (r2 0.714; Table II).
Aprotinin administration (n=14) during extracor-
poreal circulation or protamine administration (n=26)
before SIC admission did not influence significantly
the degree of agreement between central laboratory
and POC aPTT (data not shown). Antiplatelet agent
(aspirin and clopidogrel), renal insufficiency/failure,
In all the subgroups, the 95% of differences
between the two methods fell outside the a priori
assigned
10 sec difference required to accept the
interchangeability of the methods.
Controls: Healthy volunteers and stable SIC patients
POC aPTT measurements performed by three nurses
(three devices) on the same samples showed good cor-
relation between each other (F test: P=0.86). In these
samples, the agreement between mean POC aPTT
and central laboratory aPTT values was better (-20.2
18.8 sec) than in control SIC patients (Figure 2).

1158
CANADIANJOURNAL OF ANESTHESIA
FIGURE 1 Agreement between activated partial thromboplastin time measured by CoaguCheck® Pro point of care and by central labo-
ratory in different situations (A-F). The bias represents the systematic error between the two assays (bold line), the mean difference
2
standard deviations represents the limit of agreement or the 95% confidence interval (dotted line). The confidence interval of the differ-
ence between the two methods falls in all subgroups outside the a priori required upper limit of 10 sec (grey area) to accept the inter-
changeability of methods. Samples that were not clottable with one or both assays were omitted.
A: Samples drawn from all patients; B: Samples drawn from cardiovascular patients; C: Samples drawn from abdominal surgery patients; D:
Samples obtained during initial 48 hr surgical intensive care unit (SICU) stay; E: Samples obtained 48 hr after SICU admission; F:
Samples drawn from patients on heparin therapy.
The correlation between central laboratory aPTT and
heparin doses or anti-FXa activity was good (r2 0.932
and r2 0.841 respectively) and the correlation between
mean POC aPTT values and heparin doses or anti-FXa
activity was satisfactory (r2 0.700 and r2 0.484).
of subgroups of patients and of samples drawn in dif-
ferent clinical conditions, or under different therapeu-
tic regimens, could not identify a major factor
influencing the agreement or the correlation.
The discrepancy between central laboratory and
POC aPTT values may be due to the variable reduc-
tions in multiple coagulation factors in patients at
their arrival to SIC. It can hardly be due to mishan-
dling of the new device since variability of measure-
ments made by three well-trained nurses was low. The
agreement between POC aPTT and central laborato-
ry aPTT in controls were found to be better, confirm-
ing previous data.1–7,11–14
Our results are partly in line with other studies
which have evaluated POC in cardiac surgery patients.
Reich et al.8 studied 141 patients and considered an
accuracy of 10% and a precision of 25% as clinical-
ly acceptable. The mean error for the aPTT values was
plus seven seconds and the 95% confidence interval for
the bias was -21 to + 35 sec, which were considered
Discussion
Our study in patients just admitted in SIC after
surgery shows that the agreement between aPTT val-
ues obtained with POC and those from the central
laboratory is poor, in contrast to the results obtained
in controls. No acceptable agreement was found when
considering either all patients, subgroups of patients
or controls. Indeed, the data presented in Figure 1
indicate that the two assays give discordant results.
POC aPTT tend to be longer than central laboratory
aPTT and this difference tends to increase with pro-
longed aPTT. Heparin and all other treatments tested
or the time after SICU admission did not appear to be
associated with the poor agreement observed. Analysis

Ferring et al.: aPPT MONITORING I N THE ICU
1159
sec. Thus, all studies have shown an agreement with a
confidence interval of the difference between the two
methods superior to the a priori determined accept-
able upper limit of ten seconds.
It is well known that using the same plasma, various
commercial aPTT reagents give different results.18
This is particularly true for aPTT testing in
heparinized patients where deviations up to 200%
have been reported.1 9 Werner et al.2 0 studied the
effect of analytical uncertainty of conventional and
bedside assays of aPTT on clinical decisions in heparin
therapy. Despite an overall agreement of about 70%,
the bedside coagulation instrument correctly classified
only seven (39%) of 18 specimens with subtherapeutic
concentrations of heparin. In summary, our results are
not unexpected, particularly given the non standard-
ized nature of the aPTT assay, the numerous variables
that can influence aPTT immediately after surgery and
the different responses of both methods to factor defi-
ciencies and heparin.17
Monitoring heparin anticoagulation in patients
who have undergone surgery associated with extracor-
poreal circulation represents a challenge since global
tests are affected by heparin administration, possible
excess protamine, acquired factor(s) deficiency, dis-
seminated intravascular coagulation, isolated primary
hyperfibrinolysis as well as use of volume expanders.
The relationship between the degree of anticoagula-
tion and the extent of clotting time prolongation of
global tests like the aPTT is not easy to establish,
because of a wide variability in confounding factors, in
individual response and also in differences in aPTT
reagents and devices. Chromogenic anti-factor assays
for determination of heparin concentration allow to
estimate the actual contribution of heparin to the
degree of anticoagulation induced by both the drug
and the hemostatic abnormalities. Therefore, dose
adjustments of heparin anticoagulation are difficult to
achieve with the aPTT only, especially in this type of
patients. This may explain the better correlation
observed between heparin dosage and anti-FXa activ-
ity than between heparin dosage and either POC
aPTT or central laboratory aPTT, which is due prob-
ably to the impact of low AT III and coagulation fac-
tor levels not affecting anti-FXa activity.
FIGURE 2 Agreement between point of care and central labora-
tory activated partial thromboplastin time in healthy volunteers
and control patients. The bias represents the systematic error
between the two assays (bold line), the mean difference 2 stan-
dard deviations represents the limit of agreement or the 95% con-
fidence interval (dotted line). The confidence interval of the
difference between the two methods falls outside the a priori
required upper limit of 10 sec (grey area) to accept the inter-
changeability of methods.
unsatisfactory for patient management. Nuttall et
al.,1 5 using also a Ciba Corning 512 Coagulation
Monitor (the previous model of CoaguCheck® Pro),
enrolled 100 patients following cardiopulmonary
bypass surgery and found a correlation coefficient (r)
of 0.95 between the portable device and central labo-
ratory measurements in arterial samples drawn ten
minutes after heparin neutralization. This result is far
better than the one found in our study for samples
without heparin (r 0.354). However, in a subsequent
analysis of 148 patients, the same authors1 6 reported a
positive and negative predictive value for bleeding ten-
dency after cardiopulmonary bypass with bedside
aPTT devices of only 33% and 89% respectively. Boldt
et al.4 studied 80 patients and found an agreement of
-2.8 16.5 sec with POC. The time of sampling and
the influence of heparin were not mentioned. Careful
examination of their data indicates that the scatter of
values is about three fold less than ours. Nevertheless,
the 95% confidence interval reported implies a total
possible difference up to 32 sec for any single mea-
surement, which seems considerable.
In conclusion, the agreement between POC and
central laboratory aPTT was poor in SIC patients after
surgery but better in controls. These two tests give
potentially different information and only a prospec-
tive trial comparing their usefulness for predicting
clinical events could allow to establish their respective
utility in postoperative SIC patients.
In the study by Despotis et al.,1 7 the bias analysis of
the aPTT performed by the central laboratory or the
bedside device revealed an agreement of +10 28.8

1160
CANADIANJOURNAL OF ANESTHESIA
12 Zabel KM, Granger CB, Becker RC, et al. Use of bed-
side activated partial thromboplastin time monitor to
adjust heparin dosing after thrombolysis for acute
myocardial infarction: results of GUSTO-I. Am Heart
J 1998; 136: 868–76.
13 Becker RC, Ball SP, Eisenberg P, et al. A randomized,
multicenter trial of weight-adjusted intravenous heparin
dose titration and point-of-care coagulation monitor-
ing in hospitalized patients with active thromboembol-
ic disease. Am Heart J 1999; 137: 59–71.
Acknowledgements
The authors thank the nursing team of the Surgical
Intensive Care, with special attention to Sylvain
Senechal and Sandrine Bourquin for their active partic-
ipation in this study. The authors are grateful to Miss
F.M. Hannes from Roche Diagnostics, Switzerland, for
providing graciously the CoaguCheck® Pro devices.
References
1 Becker RC, Cyr J, Corrao JM, Ball SP. Bedside coagula-
tion monitoring in heparin-treated patients with active
thromboembolic disease: a coronary unit experience.
Am Heart J 1994; 128: 719–23.
14 Eiswirth G, Walch S, Bommer J. New bedside test for
monitoring anticoagulation during hemodialysis. Artif
Organs 1998; 22: 346–8.
15 Nuttall GA, Oliver WC, Beynen FM, Dull JJ, Murray
MJ, Nichols WL. Intraoperative measurement of activat-
ed partial thromboplastin time and prothrombin time
by a portable laser photometer in patients following
cardiopulmonary bypass. J Cardiothorac Vasc Anesth
1993; 7: 402–9.
16 Nuttall GA, Oliver WC, Beynen FM, Santrach PJ,
Strickland RA, Murray MJ. Determination of normal
versus abnormal activated partial thromboplastin time
and prothrombin time after cardiopulmonary bypass. J
Cardiothorac Vasc Anesth 1995; 9: 355–61.
17 Despotis GJ, Santoro SA, Spitznagel E, et al. On-site
prothrombin time, activated partial thromboplastin
time, and platelet count. A comparison between whole
blood and laboratory assays with coagulation factor
analysis in patients presenting for cardiac surgery.
Anesthesiology 1994; 80: 338–51.
2 Despotis GJ, Santoro SA, Spitznagel E, et al. Prospective
evaluation and clinical utility of on-site monitoring of
coagulation in patients undergoing cardiac operation. J
Thorac Cardiovasc Surg 1994; 107: 271–9.
3 Granger CB, Hirsh J, Califf RM, et al. Activated partial
thromboplastin time and outcome after thrombolytic
therapy for acute myocardial infarction. Results from
the GUSTO-I trial. Circulation 1996; 93: 870–8.
4 Boldt J, Walz G, Triem J, Suttner S, Kumle B. Point-of-
care (POC) measurement of coagulation after cardiac
surgery. Intensive Care Med 1998; 24: 1187–93.
5 Reiner JS, Coyne KS, Lundergan CF, Ross AM. Bedside
monitoring of heparin therapy: comparison of activated
clotting time to activated partial thromboplastin time.
Cathet Cardiovasc Diagn 1994; 32: 49–52.
6 Ruzicka K, Kapiotis S, Quehenberger P, et al.
Evaluation of bedside prothrombin time and activated
partial thromboplastin time measurement by coagula-
tion analyzer CoaguCheck Plus® in various clinical set-
tings. Thromb Res 1997; 87: 431–40.
18 Pedersen JT, Pincus MR, Rapiejko JA. Comparison of
activated partial thromboplastin time reagents. Lab
Med 1988; 19: 421–4.
19 Shojania AM, Tetreault J, Turnbull G. The variations
between heparin sensitivity of different lots of activated
partial thromboplastin time reagent produced by the
same manufacturer. Am J Clin Pathol 1988; 89:
19–23.
7 Solomon HM, Mullins RE, Lyden P, Thompson P,
Hudoff S. The diagnostic accuracy of bedside and labo-
ratory coagulation. Procedures used to monitor the
anticoagulation status of patients treated with heparin.
Am J Clin Pathol 1998; 109: 371–8.
20 Werner M, Gallagher JV, Ballo MS, Karcher DS. Effect
of analytic uncertainty of conventional and point-of-
care assays of activated partial thromboplastin time on
clinical decisions in heparin therapy. Am J Clin Pathol
1994; 102: 237–41.
8 Reich DL, Yanakakis MJ, Vela-Cantos FP, DePerio M,
Jacobs E. Comparison of bedside coagulation monitor-
ing tests with standard laboratory tests in patients after
cardiac surgery. Anesth Analg 1993; 77: 673–9.
9 Teien AN, Lie M. Evaluation of an amidolytic heparin
assay method: increased sensitivity by adding purified
antithrombin III. Thromb Res 1977; 10: 399–410.
10 Bland JM, Altman DG. Statistical methods for assess-
ing agreement between two methods of clinical mea-
surement. Lancet 1986; I: 307–10.
11 Ansell J, Tiarks C, Hirsh J, McGehee W, Adler D,
Weibert R. Measurement of the activated partial
thromboplastin time from a capillary (fingerstick) sam-
ple of whole blood. A new method for monitoring
heparin therapy. Am J Clin Pathol 1991; 95: 222–7.