Coagulation and Transfusion Medicine / AUTOMATED VON WILLEBRAND ACTIVITY ASSAY

A New Automated Screening Assay for the Diagnosis

of von Willebrand Disease
Raneem O. Salem, PhD,1 and Elizabeth M. Van Cott, MD1,2
Key Words: von Willebrand factor activity; Automated; Ristocetin cofactor
DOI: 10.1309/CEPND3LFHQ87XU4D

Abstract
A new, automated assay for von Willebrand factor
(vWF) activity has recently become commercially
available (HemosIL vWF activity assay,
Instrumentation Laboratories, Lexington, MA). We
prospectively studied 61 specimens from 58 patients
undergoing laboratory testing for suspicion of von
Willebrand disease with this new method, in
comparison with the established ristocetin cofactor
method. Assays for factor VIII and vWF antigen were
also performed using an established method on an
MDA-180 coagulation analyzer (bioMrieux, Durham,
NC) and a new method on an ACL TOP coagulation
analyzer (Instrumentation Laboratories). Blood types
were determined. The results showed no significant
difference between the assays for factor VIII (mean,
97% for MDA-180 and ACL TOP; P = .494) or vWF
antigen (mean, MDA-180, 109%; ACL TOP, 111%; P =
.766). The mean result for the ristocetin cofactor assay
was 106% vs 93% with the automated vWF activity (P
= .007). The automated activity assay was 100%
sensitive and 86% specific for detecting vWF
abnormalities and seems to be a suitable screening test.
Abnormal results should be followed up with a
ristocetin cofactor activity assay for confirmation.
Further study is recommended to confirm these
conclusions.

730
730

Am J Clin Pathol 2007;127:730-735

DOI: 10.1309/CEPND3LFHQ87XU4D

von Willebrand disease (vWD) is the most common

inherited bleeding disorder in humans, occurring in approximately 1% of the population.1 von Willebrand factor (vWF) is
essential for normal blood coagulation because it mediates
platelet adhesion to injured endothelium, and it binds and protects factor VIII in the circulation. vWD is a caused by a quantitative or qualitative defect of vWF, leading to easy bruising,
epistaxis, excessive postoperative or posttraumatic bleeding,
gum bleeding, and menorrhagia. Diagnosis can be difficult,
requiring correlation of clinical findings with multiple laboratory test results and family studies. Laboratory testing is complicated by the fact that vWF is an acute phase reactant, such
that vWF increases with stress, injury, surgery, pregnancy, and
oral contraceptive use and in the newborn period. Type 1 vWD
is a quantitative deficiency, and type 2 vWD variants are qualitative deficiencies. Type 3 vWD is a rare, severe quantitative
deficiency, with a prevalence of 1 in 1 million. Laboratory
tests for vWD are necessary not only for diagnosis but also for
determining the subtype of disease. Initial tests include factor
VIII activity, vWF antigen, and ristocetin cofactor (a vWF
activity assay).
In addition to these complexities with vWD testing, the
ristocetin cofactor assay is labor-intensive and requires technologist expertise. An easier, automated vWF activity assay
could simplify laboratory testing for vWD. The HemosIL
vWF activity assay (Instrumentation Laboratories, Lexington,
MA) is a latex turbidimetric immunoassay designed for use on
Instrumentation Laboratory analyzers as an aid in the diagnosis and classification of vWD. The reagent contains latex particles coated with a monoclonal anti-vWF antibody directed
against the platelet binding site on vWF (the platelet glycoprotein Ib receptor). If vWF in the patient plasma is normal, it
American Society for Clinical Pathology

Coagulation and Transfusion Medicine / ORIGINAL ARTICLE

binds to and agglutinates the latex particles. The degree of

agglutination is directly proportional to vWF activity in the
patient plasma and is measured by light transmittance, which
decreases when aggregates form.
The main objective of this study was to evaluate the
HemosIL vWF activity assay as a screening assay for vWD
and to compare it with the ristocetin cofactor assay as the reference method, in conjunction with vWF antigen, factor VIII,
and blood type results. In addition, 2 vWF antigen methods
and 2 factor VIII methods were compared. We studied the
hypothesis that the automated vWF activity assay (along with
vWF antigen and blood type) would function as a screening
assay for vWD such that normal results with the automated
activity assay would require no further testing for vWD (eliminating the need for a ristocetin cofactor assay), whereas
abnormal results with the automated activity assay would
result in reflex testing with a ristocetin cofactor assay for
definitive vWD assessment.

Materials and Methods

We prospectively studied 61 specimens from 58 consecutive patients (3 patients had predesmopressin and postdesmopressin specimens submitted) undergoing laboratory testing
for suspicion of vWD at the Massachusetts General Hospital
(Boston) coagulation laboratory with the HemosIL vWF
activity assay and compared results with those of a reference
(ristocetin cofactor) method. In addition, factor VIII and vWF
antigen assays were performed using an established (reference) method on the MDA-180 coagulation analyzer
(bioMrieux, Durham, NC) and a new method on the ACL
TOP coagulation analyzer (Instrumentation Laboratories).
Blood types were determined because vWF levels are lower
with blood type O compared with other blood types. Nine
parts of freshly drawn venous blood were collected into one
part of 3.2% trisodium citrate. The paired assays were run at
the same time using specimens that had undergone the same
number of freeze-thaw cycles (0 to 1). Institutional review
board approval was obtained for this study. Assays were performed in accordance with National Committee for Clinical
Laboratory Standards (now Clinical and Laboratory Standards
Institute) documents H21-A42 and H51-A3 and as recommended by the manufacturer. Document H51-A does not discuss automated latex immunoassays for vWF in detail.
Factor VIII and vWF Antigen
On the MDA-180 analyzer, factor VIII was determined
using factor VIIIdeficient plasma from Precision Biologic
(Dartmouth, Canada) and Platelin L activated partial thromboplastin time (PTT) reagent from bioMrieux; vWF antigen was
determined by using Liatest vWF (Diagnostica Stago, Asnires,

France). On the ACL TOP analyzer, factor VIII was determined

by using factor VIIIdeficient plasma and SynthaSIL PTT
reagent from Instrumentation Laboratories; vWF antigen was
determined by using the HemosIL vWF latex immunoassay.
For vWF antigen, patient plasma was prediluted 1 part plasma
plus 2 parts diluent by the ACL TOP. Latex reagent was then
added by the ACL TOP, and agglutination was measured at 405
nm as a decrease in light transmittance.
vWF Activity
The reference method was the ristocetin cofactor assay,
which measures the ability of patient plasma to agglutinate
formalin-fixed platelets in the presence of ristocetin. The test
was performed on a PACKS4 aggregometer (Helena
Laboratories, Beaumont, TX) using lyophilized platelets
(Helena Laboratories), and ristocetin (ABP [American
Biochemical and Pharmaceutical], Marlton, NJ). A 4-point
reference curve was prepared with each run. Patient plasma
was diluted 1:2 and 1:4, and both dilutions were assayed as
described in National Committee for Clinical Laboratory
Standards document H51-A.3
The automated vWF activity assay was performed on the
ACL TOP, using the HemosIL vWF activity assay. Patient
plasma was prediluted 1 part plasma plus 2 parts diluent by the
ACL TOP. The ACL TOP then added 150 L of latex reagent
to 30 L of plasma, and agglutination was measured at 405 nm
as a decrease in light transmittance.
Quality Control
Quality control was performed according to the laboratorys operating procedure and as recommended by the manufacturers for all assays used in this study. Two levels of controls for
each assay were tested with each assay run immediately before
assaying the patient specimens. Control results had to be within accepted limits before assaying the patient specimens.
Definitions of Abnormalities in vWF Laboratory Test
Results
Results were considered normal if the activity was more
than 60% and the vWF activity/vWF antigen ratio was more
than 0.7. If results did not meet these criteria, they were categorized as vWD (see Criteria for Diagnosing vWD below)
or as a nondiagnostic abnormality. Nondiagnostic abnormalities were defined as follows: (1) a vWF activity/vWF antigen
ratio of less than 0.7 and/or (2) activity more than the lower
limit of normal for the blood type but 60% or less (ie, vWD
cannot be excluded).
The lower limits of normal for vWF activity were defined
as 36% for type O, 48% for type A, 57% for type B, and 64%
for type AB.4 Results were also analyzed using ABO-specific
ranges determined on an ACL TOP (lower limit of normal,
40.3% for type O and 48.8% for nontype O).
Am J Clin Pathol 2007;127:730-735

American Society for Clinical Pathology

731

DOI: 10.1309/CEPND3LFHQ87XU4D

731
731

Salem and Van Cott / AUTOMATED VON WILLEBRAND ACTIVITY ASSAY

Criteria for Diagnosing vWD

For the purposes of the analysis, vWD type 1 was defined
as a ristocetin cofactor result less than the normal range for the
blood type with a normal ratio. vWD type 2 was defined as
ristocetin cofactor result less than the normal range for the
blood type with an abnormal ratio (<0.7). Multimer analysis
was performed for patients with results that were suggestive of
type 2 vWD.
Statistics
Paired, 2-tailed t tests were used to compare results
between methods. Bland-Altman plots were constructed to
assess for bias between methods.

Results
The mean age of patients was 37.0 years (range, 2-89
years), and 61 specimens were analyzed. Two specimens were
excluded owing to interference by hemolysis (n = 1) or hyperlipidemia (n = 1) in the ACL TOP automated vWF activity
assay. Five specimens were excluded because the vWF level
was elevated, and there was insufficient specimen to repeat the
ACL TOP vWF activity assay at a higher dilution. In the
Bland-Altman plots and t tests, data from these 7 specimens
were included if appropriate (ie, factor VIII and vWF antigen
results were valid for comparison in these specimens). For the
remaining 54 specimens, the data set was complete.
Assay Comparisons
The mean values for factor VIII, vWF antigen, and vWF
activity are given in Table 1. Results for factor VIII and vWF
antigen obtained from the MDA-180 were similar to the results
obtained from the ACL TOP, with no significant differences.

Factor VIII (%)

vWF antigen (%)
vWF activity (%)
Ratio (vWF activity/vWF antigen)*

50

100 150 200 250

Factor VIII Activity (%)

ACL
TOP

97
109
106
0.96

97
111
93
0.88

P
.494
.766
.007
.003

vWF activity values on the ACL TOP were on average lower

than the ristocetin cofactor results (P = .007). The ratio of
vWF activity to the MDA-180 vWF antigen, which can be
used to differentiate vWD subtypes, was lower on average
when using ristocetin cofactor activity (on the PACKS4 analyzer) than when using the vWF activity on the ACL TOP (P
= .007). Substituting ACL TOP vWF antigen results for
MDA-180 vWF antigen results did not alter the ratio findings
(P = .0003; data not shown). There was no significant difference between the ristocetin cofactor/vWF antigen (MDA-180)
ratio and the ristocetin cofactor/vWF antigen (ACL TOP)
ratio.
Bland-Altman plots did not reveal evidence for bias
between the factor VIII assay performed on the MDA-180
and on the ACL TOP Figure 1A. Similarly, Figure 1B
shows no bias between the vWF antigen assay performed on
the MDA-180 and that on the ACL TOP. Figure 1C shows that
ristocetin cofactor activity values were often higher than the
value obtained for vWF activity on the ACL TOP, suggesting

30
20
10
0
10 0
20
30
40
50

100

200

vWF Antigen (%)

300

Difference (Risto ACL)

100
50
0
0
50
100
150

Reference
Method

vWF, von Willebrand factor.

* For the reference method (MDA-180 coagulation analyzer, bioMrieux, Durham ,
NC), ristocetin cofactor/vWF antigen; for the ACL TOP coagulation analyzer
(Instrumentation Laboratories, Lexington, MA), automated vWF activity/vWF antigen.

B
Difference (MDA ACL)

Difference (MDA ACL)

Table 1
Assay Comparisons: Mean Values for Factor VIII Activity,
vWF Activity, and vWF Antigen on the ACL TOP vs the
Reference Method in 54 Cases

100
50
0
50

50

100 150 200 250

100
Ristocetin Cofactor (%)

Figure 1 Bland-Altman plots comparing ACL TOP with reference method results. Bland-Altman plots show no bias between
the 2 factor VIII methods (A) or the 2 von Willebrand factor (vWF) antigen methods (B). The ristocetin cofactor result tended to
be higher than the ACL TOP automated activity result, indicating that the automated assay tends to underestimate the vWF
activity (C). MDA ACL, result from the MDA-180 analyzer minus that from the ACL TOP analyzer; Risto ACL, ristocetin
cofactor result minus the ACL TOP result. For proprietary and other information about the methods, see Table 1.

732
732

Am J Clin Pathol 2007;127:730-735

DOI: 10.1309/CEPND3LFHQ87XU4D

American Society for Clinical Pathology

Coagulation and Transfusion Medicine / ORIGINAL ARTICLE

that the ACL TOP activity assay tends to underestimate the

value, or in other words, overestimate the likelihood for vWF
deficiency (a potentially acceptable characteristic for a
screening assay).
Comparison of Clinical Diagnoses Determined by the
Assays
Table 2 compares the clinical results of the reference
method (ristocetin cofactor to measure vWF activity, with
factor VIII and vWF antigen on the MDA-180) with the ACL
TOP method (automated vWF activity, factor VIII and vWF
antigen on the ACL TOP). There were no cases of a reference
method abnormality with normal ACL TOP results (no ACL
TOP false-negative results). Thus, the ACL TOP had 100%
sensitivity for detecting a vWF abnormality (defined as a
nondiagnostic abnormality or vWD identified by the reference method).
Of the 54 cases, 42 were normal by the reference method.
As shown in Table 2, there were 6 cases in which an ACL TOP
nondiagnostic abnormality was present with normal reference
method results (vWF assays >60% but abnormal ratio in 4
cases; vWF activity normal for blood type but 60% in 2
cases). Thus, with the ACL TOP, there were 6 (11%) of 54
false positive results, and the specificity was 86%.
Of the 54 cases, 7 had nondiagnostic abnormalities as
determined by the reference method. With the ACL TOP, an
abnormality was present in all 7 cases, although the type of
abnormality was slightly different from that identified by the

reference method in 2 of 7 cases, and in a third case among the

7, the ACL TOP result met the criteria for vWD (Table 2). All
7 of these cases are considered in clinical agreement between
the 2 methods because any abnormal result in the ACL TOP
method would lead to reflex testing with ristocetin cofactor
(reference method) for confirmation of the abnormality.
The remaining 5 cases (of the 54) were classified as vWD
by the reference method. The ACL TOP results were abnormal in all 5 cases, although in 2 of 5 cases, the ACL TOP diagnosis or subtype classification was different from that of the
reference method (Table 2). Again, all 5 cases are considered
in clinical agreement between the 2 methods because any
abnormal result in the ACL TOP method would lead to reflex
testing with ristocetin cofactor (reference method) for confirmation of the abnormality. Collectively, the ACL TOP was
100% sensitive for detecting a vWF abnormality. Reflex testing with the ristocetin cofactor assay when an ACL TOP
abnormality is present would be necessary to reach a conclusion regarding the nature of the abnormality (ie, to confirm
whether it is vWD and to determine the subtype of vWD)
Figure 2.
Of the 54 specimens, 36 had normal vWF activity (and
antigen) by the reference method and the ACL TOP method.
Thus, in 67% of specimens (36/54), the ACL TOP automated
vWF activity assay would have successfully eliminated the
need for a ristocetin cofactor activity assay. The other 18 specimens had at least 1 abnormality in the ACL TOP results, as
described earlier and as shown in Table 2.

Table 2
Clinical Classification of Results With ACL TOP vs Reference Method in 54 Cases*
Results
Reference Method
Result Category
Normal (n = 42)

Nondiagnostic
abnormality (n = 7)

von Willebrand
disease (n = 5)

No. of
Cases

Reference
Method

ACL TOP

36

Normal

Normal

Normal

NDR

2
4

Normal
ND60

ND60
ND60

O, 21 (58%); A,13 (36%);

B, 2 (6%); AB, (0%)
A, 3 (75%);
AB, 1 (25%)
O, 2 (100%)
O, 3 (75%); A, 1 (25%)

1
1
1
1

ND60
ND60
ND60, NDR
Type 1

ND60, NDR
Type 2
ND60
Type 2

O, 1 (100%)
O, 1 (100%)
O, 1 (100%)
A, 1 (100%)

Type 2

Type 2

O, 1 (100%)

1
1
1

Type 2
Type 2
Type 1

ND60
Type 2
Type 1

O, 1 (100%)
B, 1 (100%)
A, 1 (100%)

Comment/Details

Blood Type

vWF >100% by both methods; reference ratios, 0.9-1.1;

ACL TOP ratios, 0.28-0.65
vWF, 72% and 54% in one case; 67% and 50% in the other
vWF for both methods within 10% of each other in all
4 cases
vWF, 50% and 42%; ratios, 0.76 and 0.68, respectively
vWF, 47% and 24%; ratios, 0.76 and 0.44, respectively
vWF, 41% and 48%; ratios, 0.67 and 0.76, respectively
vWF, 43% and 29%; ratios, 0.83 and 0.67, respectively
vWF, 55% and 61%; ratios, 0.47 and 0.41, respectively
(previously diagnosed; receiving vWF concentrate)
vWF, 35% and 42%; ratios, 0.61 and 0.73, respectively
vWF, 23% and 36%; ratios, 0.26 and 0.43, respectively
vWF, 35% and 19%, respectively

ND60, nondiagnostic abnormality consisting of vWF of 60% but not low for blood type; NDR, nondiagnostic abnormality consisting of a ratio of <0.7; vWF, von Willebrand
factor activity.
* Normal was defined as vWF activity of >60% and a ratio of >0.7; and von Willebrand disease as follows: type 1, vWF low for blood type with a normal ratio; type 2, vWF low
for blood type with an abnormal ratio. For proprietary and other information about the methods, see Table 1.
Respectively refers to the reference method and the ACL TOP.
There were 6 (11%) total false-positive results.

Am J Clin Pathol 2007;127:730-735

American Society for Clinical Pathology

733

DOI: 10.1309/CEPND3LFHQ87XU4D

733
733

Salem and Van Cott / AUTOMATED VON WILLEBRAND ACTIVITY ASSAY

In general, factor VIII results were similar to vWF results,

as expected. One of the 61 specimens was from a male patient
with mild hemophilia A with normal vWF activity and antigen
results by all methods but 12% factor VIII on the MDA-180
and 15% on the ACL TOP (results for factors IX, XI, and XII
and the prothrombin time were normal). His factor VIII
response to desmopressin lasted for a normal duration, making vWD type 2N less likely.
The clinical classification of results shown in Table 2
was identical regardless of which of the 2 definitions of blood
typespecific lower limits of normal was used. One definition
for the lower limit of normal was 36% for type O, 48% for
type A, 57% for type B, and 64% for type AB.4 The other definition was based on ABO-specific ranges determined on an
ACL TOP (lower limit of normal, 40.3% for type O and
48.8% for nontype O).

Discussion
Screening assays should have high sensitivity, even at the
expense of specificity, so that no cases of disease are missed.
The presence of disease can then be confirmed or excluded
by a confirmatory assay, such as the ristocetin cofactor assay.
In this study, the automated vWF activity assay performed
very well as a screening assay, with 100% sensitivity and
86% specificity. Use of the automated vWF activity assay
would have successfully eliminated the need for 67% of the
ristocetin cofactor assays. Many clinical laboratories send out
ristocetin cofactor assays rather than perform the assay in
house owing to the skilled labor and specialized equipment
required and the labor-intensive nature of the assay. As an
alternative, the results of the present study suggest that laboratories could perform the automated vWF assay in house
(along with a vWF antigen and factor VIII assay), reducing
the need for ristocetin cofactor send outs by 67%. To detect
all possible cases of vWD, we recommend that a ristocetin
cofactor assay be performed if the automated vWF activity
result is 60% or less and/or if the vWF activity/vWF antigen
ratio is less than 0.7.
Laboratories that currently perform ristocetin cofactor
assays in house often perform the assay relatively infrequently in batches. The automated vWF activity assay is easier to
perform, and, therefore, it would be easier to offer on a more
frequent or even a stat basis, and it could also reduce the volume of ristocetin cofactor assays.
Although the present study involved 61 specimens with a
broad range of results, including type 1 and type 2 vWD, nondiagnostic abnormalities, normal results, and elevated values,
further study should be performed on an even larger number
of specimens to confirm the findings. The automated assay
assesses the ability of vWF to bind to platelet glycoprotein Ib.
734
734

Am J Clin Pathol 2007;127:730-735

DOI: 10.1309/CEPND3LFHQ87XU4D

Perform automated assays in house:

Factor VIII, vWF antigen, vWF activity
vWF activity 60 or
ratio <0.7

vWF activity >60 and

ratio >0.7

Perform ristocetin
cofactor activity

No further testing
at this time

Ristocetin cofactor >60

and ratio >0.7
Ristocetin cofactor 60
or ratio <0.7, consider repeated
testing in future or specialized
tests if indicated (eg, multimers)

Figure 2 Algorithm for possible use of the automated von

Willebrand factor (vWF) activity assay. It is proposed that
laboratories could perform the automated vWF activity assay
(HemosIL, Instrumentation Laboratories, Lexington, MA)
along with factor VIII and vWF antigen in house. Specimens
with normal results (vWF activity values >60% and ratios
>0.7) would not need a ristocetin activity assay performed (if
illness, injury, stress, pregnancy, or estrogen use are present,
testing should be repeated at a later date despite the normal
results). Specimens with low-normal or abnormal results
(vWF activity values 60 or ratios <0.7) would need a
ristocetin cofactor activity assay to further assess for von
Willebrand disease.

Type 2B vWD reportedly is also detected by this assay

because the level of vWF activity in the plasma is low.5
Other commercially available vWF activity tests include
an enzyme-linked immunosorbent assay (ELISA), which
requires manual labor and usually batch testing. In addition, it
has not performed as well as the ristocetin cofactor activity
assay in some studies.6 The collagen-binding activity assay is
another type of ELISA-based assay for which several methods
are available. Some of the collagen-binding methods have performed well in some studies, depending on the method used.6,7
However, collagen-binding assays in an ELISA format typically still require batch testing and manual labor.
After this manuscript was submitted, another study examining the HemosIL vWF activity assay was published, which
also reports 100% sensitivity if a cutoff of 60% is used in comparison with ristocetin cofactor activity.8 The specificity was
87.6%. In this study, the new activity assay was performed on
a STAC coagulation analyzer (Diagnostica Stago), whereas in
the present study, the assay was performed on an ACL TOP.
The automated ACL TOP method had 100% sensitivity
for detecting vWD and possible vWD, if any abnormal result
American Society for Clinical Pathology

Coagulation and Transfusion Medicine / ORIGINAL ARTICLE

leads to reflex testing with a ristocetin cofactor activity assay

for further assessment. The assay functioned well as a screening assay for vWD. Further study should be performed on
more patients with vWD to confirm these findings.
From the 1Division of Laboratory Medicine, Department of
Pathology, Massachusetts General Hospital and 2Harvard Medical
School, Boston.
Address reprint requests to Dr Van Cott: Clinical
Laboratories, Gray 235, Massachusetts General Hospital, 55 Fruit
St, Boston, MA 02114.
Instrumentation Laboratories (IL) provided the IL reagents
used in this study and reimbursed medical technologist labor costs
incurred in performing the IL assays.

References
1. Rodeghiero F, Castman G, Dini E. Epidemiological
investigation of the prevalence of von Willebrands disease.
Blood. 1987;69:454-459.
2. National Committee for Clinical Laboratory Standards.
Collection, transport, and processing of blood specimens for
testing plasma-based coagulation assays. 4th ed. Wayne, PA:
National Committee for Clinical Laboratory Standards; 2003.
Approved Guideline H21-A4.

3. National Committee for Clinical Laboratory Standards.

Assays of von Willebrand factor antigen and ristocetin
cofactor activity. Wayne, PA: National Committee for
Clinical Laboratory Standards; 2002. Approved Guideline
H51-A.
4. Gill JC, Endres-Brooks J, Bauer PJ, et al. The effect of ABO
blood group on the diagnosis of von Willebrand disease. Blood.
1987;69:1691-1695.
5. Pinol M, Costa M, Sales M, et al. New automated von
Willebrand factor activity assay to distinguish type 1 and type
2 von Willebrand disease [abstract]. J Thromb Haemost.
2003;suppl 1:1679.
6. Favaloro EJ. Detection of von Willebrand disorder and
identification of qualitative von Willebrand factor defects:
direct comparison of commercial ELISA-based von
Willebrand factor activity options. Am J Clin Pathol.
2000;114:608-618.
7. Paczuski R. Determination of von Willebrand factor activity
with collagen-binding assay and diagnosis of von Willebrand
disease: effect of collagen source and coating conditions. J Lab
Clin Med. 2002;140:250-254.
8. de Vleeschauwer A, Devreese K. Comparison of a new
automated von Willebrand factor activity assay with an
aggregation von Willebrand ristocetin cofactor activity assay
for the diagnosis of von Willebrand disease. Blood Coagul
Fibrinolysis. 2006;17:353-358.

Am J Clin Pathol 2007;127:730-735

American Society for Clinical Pathology

735

DOI: 10.1309/CEPND3LFHQ87XU4D

735
735