Professional Documents
Culture Documents
ORIGINAL ARTICLE
Key Words Background: The buccal mucosal bleeding time (BMBT) evaluates primary
Dogs, hemorrhage, mucosa, platelets, primary hemostasis in vivo. Three different-sized lancet devices designed for people,
hemostasis Adult (A), Junior (J), and Newborn (N), can be used to perform the BMBT
in dogs.
Correspondence
lou, Internal Medicine, Ecole
Dr. Armelle Dique
Objectives: The aim was to compare BMBT using 3 different-sized human
nationale Veterinaire de Toulouse, 23 chemin lancet devices in dogs with varying platelet counts and hematocrits.
des Capelles, BP 97614, 31076 Toulouse Methods: The BMBT was measured in 46 client-owned dogs (2 healthy,
Cedex France 44 suffering from various disorders) with varying platelet (Plt) counts and
E-mail a.diquelou@envt.fr hematocrits, using the 3 devices successively in each dog, in a randomly
determined order, over a 10- to 30-minute period. Statistical analysis
DOI:10.1111/vcp.12087 (ANOVA, Mann–Whitney U-test) was performed using commercial
software.
Results: BMBTs were significantly different between devices
(P < .00001), and shorter with devices N and J compared with device A
(P < .01). The BMBT was prolonged (> 210 s) in 10 dogs with device A and
in 7 dogs each with devices J and N, respectively. Sixteen dogs had a Plt
count < 200 9 109/L (Reference interval 200–500 9 109/L). Nine of these
dogs had prolonged BMBT with device A, and 6 dogs with device J and
device N, respectively. BMBT was longer in thrombocytopenic dogs with
devices A and J (P < .016). Anemia without thrombocytopenia did not
affect BMBT with any device.
Conclusions: The BMBT is influenced by the size of the used device, with
A resulting in the longest BMBT. Therefore, the type of device used to
obtain the BMBT has to be specified for standardized results.
Vet Clin Pathol 42/4 (2013) 451–457 ©2013 American Society for Veterinary Clinical Pathology 451
Comparison of bleeding time in dogs using 3 devices Aumann et al
to a constant pressure of 40 mmHg to provide venosta- everted to expose the buccal mucosa and a strip of
sis.4,5 In dogs and cats, cutaneous, toenail, and BMBT gauze 2 cm in width was tied around the maxilla and
have also been employed to assess primary hemosta- the mandible to hold the lip in place and to cause con-
sis.2,6,15–17 The cutaneous template BT technique used gestion of the superficial mucosal vessels. The strip of
in people has not been effective to evaluate primary gauze stayed in place until all 3 successive measure-
hemostasis in domestic animals3 because of the wide ments of the BMBT were completed.
variety in skin thickness, hair coat, and depth of subcu- Three different-sized commercially available BT
taneous fat.15 Measuring the BT in dogs after clipping devices (Surgicutt Adult, Junior and Newborn, Kitvia,
off a toenail does not differentiate adequately between Labarthe-Inard, France) containing a spring-loaded
primary and secondary hemostasis.18 The BMBT is cur- blade were used. The device was placed against the
rently the preferred method to measure the BT in dogs3 buccal mucosa in a horizontal position avoiding large
and cats.6 The technique to determine the BMBT in mucosal vessels. Activation of the trigger mechanism
dogs has been described in detail in the veterinary liter- resulted in a standardized incision with the length and
ature;2,3,6,15–17 however, standardization of the depth depending on the device used (Adult [A]
method has not been established to the same degree as 5 mm 9 1 mm, Junior [J] 3.5 mm 9 1 mm, New-
in people. Three different-sized lancet devices designed born [N] 2.5 mm 9 0.5 mm). Every 5–10 s, blood was
for human BT assessment and resulting in variable carefully removed with filter paper by blotting 1–
length and depth of the incision can be used to assess 2 mm away from the incision to avoid disruption of
the BMBT in dogs. clot formation. The stopwatch was started at the time
The aim of this study was to compare the BMBT of device activation resulting in the initiation of the
measured with 3 different-sized human lancet devices incision. It was stopped when blood no longer stained
in dogs with varying Plt count and HCT. the filter paper. If the BMBT was longer than 480 s,
the stopwatch was stopped and gauze was applied to
the incision to stop bleeding.
Material and Methods
For each dog, the 3 consecutive BMBT measure-
ments were performed by the same operator (MA, VR,
Dogs
or AD) successively over a period of 10–30 minutes.
Dogs that were presented to the Teaching Hospital of Operators varied between different dogs. The order in
the National Veterinary School of Toulouse between which devices A, J, and N were used was randomly
June 2009 and July 2010, and November–December determined (http://www.randomization.com).
2011 for routine vaccination or for evaluation of vari-
ous diseases were consecutively recruited in this study.
Statistical analysis
After informed owner consent, dogs were included in
the study if they had a CBC (Sysmex XT2000iV hema- Statistical analysis was performed using commercial
tology analyzer, Sysmex, Roissy, France) including software (Stata 9.1 software [Statacorp, College Sta-
HCT (reference interval [RI] 0.37–0.55) and Plt count tion, TX, USA]; Excel [Microsoft France, Les Ulis,
(RI 200–500 9 109/L) performed on the day of BMBT France]; and Analyze-It software Add on for Excel
assessment, and if the BMBT was performed with all 3 [Microsoft France]).
devices as described below. The Plt count was con- An ANOVA was used to compare the BMBT
firmed by evaluation of a blood smear by exclusion of among the 3 different devices. A logarithmic transfor-
platelets clumps and rough estimation of the count mation was necessary to provide normality and
(averaging the number of platelets in at least 5 oil- homogeneity of variance (homoscedasticity) of residu-
immersion fields in the monolayer of the smear, and als, and both were assessed graphically and by the
multiplying this number by 15). Anemia was defined Shapiro–Wilk, and the Breusch–Pagan and Cook–
as a HCT < 0.37 and thrombocytopenia as a Plt count < Weisberg tests, respectively. Distribution and residual
200 9 109/L based on the laboratory-specific RIs. were considered as normal and homoscedastic, respec-
tively, when P > .05 for the Shapiro–Wilk and the
Breusch–Pagan/Cook–Weisberg tests. A Bonferroni
BMBT
correction of the a-value from 0.05 to 0.016 was applied
The BMBT measurements were performed in con- to compare the different pairs of devices, using 3 paired
scious un-sedated dogs according to a standardized t-tests (level of significance P < .016). Spearman corre-
protocol previously described.15 Briefly, dogs were lation coefficients, Bland–Altman Difference plots and
restrained in lateral recumbency. The upper lip was Passing–Bablok equations were also evaluated. The
452 Vet Clin Pathol 42/4 (2013) 451–457 ©2013 American Society for Veterinary Clinical Pathology
Aumann et al Comparison of bleeding time in dogs using 3 devices
Table 1. Comparison of the buccal mucosal bleeding time (BMBT) of all dogs, dogs with and without thrombocytopenia, dogs with and without ane-
mia, and dogs with anemia but without thrombocytopenia, using 3 different-sized lancet devices: A = Adult, J = Junior, N = Newborn.
BMBT (Seconds)
Number of Dogs A J N
Vet Clin Pathol 42/4 (2013) 451–457 ©2013 American Society for Veterinary Clinical Pathology 453
Comparison of bleeding time in dogs using 3 devices Aumann et al
Table 2. Numbers of dogs with different platelet counts and with pro- with device J compared with device A. This confirms
longed buccal mucosal bleeding time (BMBT) using 3 different-sized lan- previous reports in people that both length and depth
cet devices grouped: A = Adult, J = Junior, N = Newborn.
of the incision influence the length of BT.12,13 In dogs,
BMBT > 210 Seconds the incision resulting from device N is both shorter and
9
Platelet Count (10 /L) Number of Dogs A J N more superficial than with devices A and J, and the
incision obtained with device J is of the same depth as
< 100 9 109/L (n = 10) 8 8 5 6
100 9 109/L ≤ 200 9 109/L (n = 6) 2 1 1 0
that of, but shorter than that with, device A.
≥ 200 9 109/L (n = 29) 2 1 1 1 Two of the 29 dogs with a normal Plt count and
normal HCT had a BMBT > 210 s. Surprisingly, in one
of these dogs, the BMBT was only prolonged with
device N. A technical error was suspected in this case,
but not confirmed. The second dog, with infectious
bronchitis due to S equi, had a prolonged BMBT with
devices A and J. No evidence for hemostatic abnormal-
ities was identified on clinical examination and a diag-
nostic evaluation of primary hemostasis by further
laboratory testing was declined by the owner.
The results of our study confirm previous reports
in people and dogs that the BT is prolonged in the pres-
ence of thrombocytopenia (< 100 9 109/L in people
and < 20 9 109/L in dogs).1,3,8 Specifically, 9 of 16
thrombocytopenic dogs had a prolongation of the
BMBT with the A device and 6 with the J or N device,
Figure 1. Comparison of buccal mucosal bleeding times (BMBT) mea-
sured in 45 dogs using 3 different-sized lancet devices: Adult, Junior, and respectively. All 8 dogs with a Plt count < 50 9 109/L
Newborn. The solid horizontal line indicates the median, the dotted line had a prolonged BMBT with device A, whereas only 5
indicates the published upper limit of the BMBT,7 * = P < 0.015, and 6 of these highly thrombocytopenic dogs had pro-
** = P < 0.00001. longed BMBT with J and N devices, respectively, sug-
gesting that the A device may be better suited to
identify dogs with primary defective hemostasis attrib-
Platelet count, HCT, and results of the BMBT
utable to severe thrombocytopenia. One dog with a Plt
Twenty-nine dogs had a normal Plt count. The BMBT count of 77 9 109/L had a normal BMBT with all 3
of 2 of these dogs was increased (Table 2). In one of devices. A direct relationship between the degree of
these dogs, the BMBT was only increased with the N thrombocytopenia and the degree of prolongation of
device. In the other dog, which was suffering from the BT was suggested in a study of thrombocytopenic
bronchitis due to Streptococcus equi, it was increased people;8 however, the results of this study have never
with devices A and J. Ten of the 16 thrombocytope- been confirmed.14 It has been shown that the hemostat-
nic dogs had an increased BMBT with at least one of ic effect of platelets does depend not only on the Plt
the devices (Table 2), with 4 of these dogs having an count but also on platelet size and activation efficacy.14
increase with all 3 devices. BMBT measurements For instance, people with thrombocytopenia due to
obtained in thrombocytopenic dogs were higher than immune-mediated thrombocytopenic purpura may
in nonthrombocytopenic dogs with devices A and J have normal BTs in spite of severe thrombocytopenia.19
(P < .016); significance was not reached with device Only 2 of the 6 dogs with a Plt count between 100
N (P = .0214, Table 2). Eleven of the 16 anemic dogs and 200 9 109/L had a BMBT > 210 s with one of the
had concurrent thrombocytopenia. Eight of these devices (device A or J, respectively). Both of these dogs
dogs had an increased BMBT with at least one were anemic suggesting that the effect of thrombocy-
device. None of the 5 dogs with anemia and normal topenia on the BMBT may be enhanced by concurrent
Plt count had a prolonged BMBT (Table 1). anemia. Human studies have demonstrated a relation-
ship between a decrease in HCT and a prolongation of
the BT.9 The impairment of primary hemostasis in peo-
Discussion ple with anemia may be due to altered distribution of
cells in flowing blood resulting in decreased contact of
In the current study, the BMBT was shortest with platelets with blood vessel walls and decreased release
device N compared with devices A and J, and shorter of ADP by disrupted red blood cells.20,21 In people with
454 Vet Clin Pathol 42/4 (2013) 451–457 ©2013 American Society for Veterinary Clinical Pathology
Aumann et al Comparison of bleeding time in dogs using 3 devices
Figure 2. Scatter plots, Passing Bablok equations, and Bland Altman difference plots of buccal mucosal bleeding time (BMBT) measured in 45 dogs,
comparing (A) Adult to Junior devices (A), (B) Adult to Newborn devices and (C) Junior to Newborn lancet devices.
Vet Clin Pathol 42/4 (2013) 451–457 ©2013 American Society for Veterinary Clinical Pathology 455
Comparison of bleeding time in dogs using 3 devices Aumann et al
anemia and uremia, correction of the anemia through Standardization of the length of the incision is eas-
blood transfusions improved the BT even if the Plt ier than the depth of the incision. The depth of the inci-
count remained stable.22 In a study of dogs with experi- sion does depend not only on the device but also on
mentally induced renal failure, anemia and a prolonged the operator pressure applied to the site. In people,
BMBT, the BMBT normalized temporarily after a blood variations in incision depth due to nonstandardization
transfusion increased the HCT.10 In the current study, of operator pressure may influence the BT.24 In the
7 of the 16 dogs with anemia had a prolonged BMBT, current study, variations in operator pressure may
but all of these dogs were also thrombocytopenic. have contributed to interobserver variability; how-
The median BMBT for dogs without thrombocyto- ever, it was unlikely to have contributed to intraob-
penia (Device A 88 s, device J 86 s, device N 65 s) was server variability as BMBT measurements in a
lower than the mean BMBT published previously particular dog were performed by the same operator.
(163 s),15 or (157 s);3 however, in these studies, the Variation in the degree of venostasis may have
BMBT was measured with a different device (Simplate influenced the duration of the BMBT in the dogs in our
II). The Simplate II device makes 2 incisions of 6 mm study. The degree of venostasis has been standardized
length and 1 mm depth, compared to the 5, 3.5, and in people, but not in animals.7 In people, a cuff pres-
2.5 mm length, and 1, 1, and 0.5 mm depth for the A, sure of 40 mmHg is applied to the arm where the BT is
J, and N devices, respectively. The correlation between performed.4,5 The BT tends to be shorter if venostasis is
the BMBT and plasma von Willebrand factor concen- not applied.4 Although a strip of gauze was tied around
tration using the Surgicutt device (without specifying the muzzle to provide venostasis in the dogs in our
A, J, or N device) was studied in 61 Greyhounds. The study, the degree of venostasis was difficult to stan-
mean and median BMBT were 129.5 s and 130 s, dardize and may have varied significantly among dogs.
respectively, but results of normal dogs and dogs with The 3 consecutive BMBT measurements were per-
vWD were not analyzed separately.2 formed within a period of 10–30 minutes. Even
Our study has several limitations. As each dog though the gauze was not adjusted or removed
served as its own control, potential confounding effects between the 3 measurements, the degree of venostasis
of signalment and disease states on differences in could have changed during this time period. The order
BMBT measurements using the different devices were in which the different devices were used was random-
limited. Thrombocytopenia has been associated with a ized, minimizing the influence of changes in the
prolongation of the BT in people and dogs1,3,8 and ane- degree of venostasis on the comparison of the BMBT.
mia with a prolongation of the BT in people.9 Although The number of dogs included in our study was not
our study was designed based on the methods recom- large enough to establish reference intervals for BMBT
mended by Jensen and Kjelgaard–Hansen23 and measurements with the different devices, nor was the
included 46 patients covering the entire working range study designed to calculate diagnostic sensitivity and
of the method investigated, the number of anemic dogs specificity. Reference intervals for the BMBT using the
was not large enough to assess the influence of anemia lancet device still have to be established.
on the BMBT performed with the 3 different devices. In conclusion, the results of the current study indi-
For practical and ethical reasons, the current study cate that the duration of the BMBT in dogs with and
did not assess the repeatability of the method because without thrombocytopenia is influenced by the device
this would have required a duplication of the measure- used. Determination with the A lancet device results in
ments of the BMBT with each device resulting in 6 a longer BMBT than with J and N devices. Therefore,
BMBT measurements in each dog. Intra- and interob- for standardized BMBT, the type of device has to be
server variability in BMBT measurements in dogs per- specified.
formed with the lancet device used in our study has
been published previously and the repeatability of the
Acknowledgments
measurements was low.2 Interobserver variability
could have been eliminated from our study by one and The authors thank Michelle Ingelbach for technical assis-
the same investigator performing all BMBT measure- tance and Pr Jean Pierre Braun for assisting with the revi-
ments instead of 3 main, albeit experienced, investiga- sion of the manuscript.
tors (MA, VR, and AD). We also adhered to an Disclosure: The authors have indicated that they have
established protocol as the repeatability of the BT has no affiliations or financial involvement with any organiza-
been said to improve with standardization,2,4 even tion or entity with a financial interest in, or in financial
though, in people, the beneficial effect of standardiza- competition with, the subject matter or materials discussed
in this article.
tion of the BT technique has been questioned.11
456 Vet Clin Pathol 42/4 (2013) 451–457 ©2013 American Society for Veterinary Clinical Pathology
Aumann et al Comparison of bleeding time in dogs using 3 devices
Vet Clin Pathol 42/4 (2013) 451–457 ©2013 American Society for Veterinary Clinical Pathology 457