thrombosis. (DVT) in 132 consecutive patients undergoing total hip replacement. After the operation, all patients had compression stockings,. 65 were treated.
PROPHYLAXIS TOTAL
HIP
COMPARISON
F. S.
From
A.
Pietro
DEEP-VEIN
THROMBOSIS
IN
REPLACEMENT
OF
SANTORI,
San
AGAINST
AND
HEPARIN
VITULLO,
M.
Fatebenefratelli
FOOT
STOPPONI,
IMPULSE
N.
Hospital,
SANTORI,
Rome,
PUMP
S.
GHERA
Italy
We performed a randomised controlled study to compare heparin with the A-V Impulse System in the prevention of deep-vein thrombosis (DVT) in 132 consecutive patients undergoing total hip replacement. After the operation, all patients had compression stockings, 65 were treated with calcium heparin and 67 with the intermittent plantar pump. DVT was diagnosed by Doppler ultrasound and thermography, followed by phlebography. There were 23 cases of DVT (35.4%) in the heparin group, with 16 m&jor and seven minor thromboses. In the impulse pump group there were nine cases (13.4%) with three major and six minor thromboses. The differences for all thromboses and for major thromboses were both significant at p < 0.005. In the heparin group there was one fatal pulmonary embolism and nine patients (13.8%) had excessive bleeding or wound haematomas, as against none in the impulse pump group.
There is a high incidence of deep-vein thrombosis (DVT) after total hip replacement (ThR), up to 70% in the absence of any prophylaxis (Bergqvist 1983). Subclinical pulmonary emboli (PE) can be detected in over 23% of cases (Harris et al 1984), with a mortality of 1% to 2% (Johnson, Green and Charnley 1977). The incidence of PE is even higher in emergency hip surgery, reaching 4% to 7% (Hull and Raskob 1986). DVT is more probable in elderly or obese patients and in those with a previous history of thromboembolism, varicose veins or a malig-
J BoneJoint
is one
Received
Surg
[Br]
28 October
1994; 1993;
76-B:579-83.
Accepted
21 December1993
nant neoplasm and et al 1970; Sikorski,
in those taking oral oestrogen (Kakkar Hampson and Staddon 1981). Many
patients who have a symptomless DVT develop clinical symptoms later, and suffer from a chronic postphlebitic syndrome (Killevich et al 1985). DVT is most commonly secondary to blood stagnation in calf veins and the valve pockets of popliteal and femoral veins (Gibbs 1957; Sevitt and Gallagher 1961). Venous stasis occurs both during anaesthetic induction and during the operation (Cotton and Roberts 1977). This of the principal
hypercoagulation
Endothelial that
the
pathogenetic due
damage femoral
to
is also a factor
veins
are
factors,
trauma twisted
and
together surgical
and it has been and
possibly
with stress.
shown
damaged
during a ThR (Planes, Vochelle and Fagola 1990). This favours the development of the proximal thrombi which have been found to cause 75% of fatal pulmonary emboli (Havig 1977). The presence of all these factors confirms that pharmacological or mechanical prophylaxis or both is justified
for all patients
When calcium ously, the incidence between 28% and Alfidi F. S. Santori, MD, Professor of Orthopaedic Surgery A. Vitullo, MD, Orthopaedic Registrar M. Stopponi, MD, Orthopaedic Registrar S. Ghera, MD, Orthopaedic Registrar Department of Orthopaedics and Traumatology, San Hospital, Via Cassia 600, 00189 Rome, Italy. N. Santori, MD, Orthopaedic 1st Cathedral Orthopaedic Piazzale Aldo Moro 5, 00185 Correspondence ©l994
British
0301-620X/94/4806
VOL
Registrar Clinic, University Rome, Italy.
should
be sent
Editorial
Society
$2.00
76-B, No. 4, JULY
1994
of
Pietro
Rome,
to Dr F. S. Santori. of Bone
and Joint
Surgery
Fatebenefratelli
‘La
Sapienza’,
1973;
Kakkar
having
a THR.
heparin is administered subcutaneof DVT has been reported to vary 73% (Harris et al 1972; Evarts and et al 1985),
but
when
the
measured
prothrombin time is maintained at between 31 .5 and 36 seconds this is reduced to 13% (Leyvraz et al 1983). Other authors have reported that a combination of calcium heparin and dihydroergotamine was associated with an incidence of DVT of 24% to 59% (Fredin, Gustafson and Rosburg 1984; Kakkar et al 1985). The more recent use of low-molecular-weight heparin (Turpie et al 1986; Lassen
et al
heparin produced
and
1991)
and
antithrombin encouraging
the
associated
III (Francis results with
use
of
calcium
et al 1990) have reduction in the 579
F. S. SANTORI,
580
Fig.
Phlebograms
to show
A. VITULLO,
venous
and
pumping
mechanism
Gardner
a physiological
foot.
Phlebographic
venous
4.9%
system
We between therefore
la
-
of the plantar R. H. Fox.)
Non-weight-bearing, venous plexus
have
showing the been ejected
respectively.
were associated with and wound haematoma
a
(Fig.
and Fox first showed
the existence
venous
pumping
mechanism
studies
indicate
that the large
la)
is rapidly
emptied
in the
when
plantar weight-
bearing flattens the plantar arch (Fig. ib). The pulsatile flow produced by walking appears to flush out the valve pockets higher in the limb. The A-V Impulse System (Fig. 2; Novamedix, Andover, UK) was designed to produce the same effect by external pressure. Early reports have shown it to be effective in reducing the incidence of proximal DVT from 32.5% to 5% after ThR (Fordyce and Ling 1992), from 23% to 0% after hemiarthroplasty for hip fracture (Stranks et al 1992) and from 18.7% to 0% in total knee replacement (Wilson et a! 1992). The combination of heparin and the A-V Impulse System has been compared with heparin alone in 74 patients undergoing hip arthroplasty (Bradley, Krugener and Jager 1993). These authors showed a 25% incidence of proximal DVT in the heparin-treated group, compared with 6.6% in the group
of the foot. Figure
Figure lb - Weight-bearing. The contents by permission of A. M. N. Gardner and
lb
10%).
In 1983, of
S. GHERA
Fig.
the physiological
incidence of DVT to 11% to 31% Both of these methods, however, considerable incidence ofbleeding and
N. SANTORI,
la
venous plexus filled with contrast medium. into the deep veins of the calf. (Photographs
(14.3%
M. STOPPONI,
treated
could heparin designed
with
find
no
heparin
and
impulse
report
of a direct
and the A-V Impulse a randomised study.
pumping.
comparison System
and
Fig. The A-V Impulse System showing inflatable pads within the slippers the plantar venous plexus.
2
the impulse to produce
generator intermittent
which rapidly compression
fills of
We excluded patients with a previous history thromboembolism, varicose veins, venous insufficiency in the legs, or the presence of a malignant neoplasm. to the
All patients gave informed study, were randomly
consent, allocated
of
and on admission to receive either
pharmacological or mechanical prophylaxis. Randomisation was by a casual numbers table, using sequentially numbered, sealed envelopes. All patients were measured for compression stockings before operation; these were put
on to both
legs
immediately
after
the
completion
of
surgery.
PATIENTS
AND
METhODS
Between June 1990 and December 1991, we studied a consecutive series of 132 patients undergoing primary THR under general anaesthesia by a direct lateral approach (Hardinge 1982). The prostheses used were the uncemented Zimmer Anatomic (Zimmer, Warsaw, Indiana), the cemented Exeter-Howmedica (Howmedica International) or a hybrid prosthesis (Harris-Galante acetabular cup, Exeter femoral stem, Zimmer), chosen according to clinical and radiographic criteria.
Pharmacological
prophylaxis
was
by
5000
IU
of
calcium heparin given subcutaneously three times a day for ten days, starting on the day before the operation. Mechanical prophylaxis was by the fitting of the A-V Impulse System to both feet immediately after the operation. The inflatable pad is placed beneath the plantar arch over the compression the retaining slipper. The
stocking and held in place by pads are then connected to the
air-impulse generator. This rapidly inflates the pads and then deflates them fully after three seconds. The cycle is repeated at approximately 20-second intervals, the time THE
JOURNAL
OF BONE
AND JOINT
SURGERY
PROPHYLAXIS
needed
to allow
the
venous
AGAINST
plexus
DEEP-VEIN
of the
foot
THROMBOSIS
to refill
IN TOTAL
differences
HIP
REPLACEMENT
581
for all thromboses
and
for major
thromboses
before the next impulse. To facilitate this the feet are kept below the level of the right atrium. Use of the A-V Impulse System continued for seven to ten days. For both
were
highly
and
In the heparin one patient
groups,
microemboli. Nine patients had either bleeding problems or wound haematoma or both. In the A-V Impulse group, three patients developed complications due to the pump; one had superficial skin abrasions on the dorsum of the
physiotherapy
bed began on began walking
with
mobilisation
of the
limb
the second postoperative day. on the fourth or fifth postoperative
in
Patients day.
The foot pump was functional during bed; impulse pumping was interrupted
the entire period in only for physio-
therapy
was
treatment
Diagnosis
or when
of DVT.
at eight to thermography whom
the
ten
All
Doppler
thrombi. group
group
not having
any
were
screened
tests
were
positive
clinical signs to determine
Phlebography and 17 of the
but
for
DVT
liquid crystal Patients in
and
all
doubtful
a negative the location
was performed impulse pump
signs We
in 31 of the group. In the
early
defined major thrombi as veins (iliac, femoral, popliteal)
were since
relationship both groups
those involving or those in the
reduce
the results
used
the chi-squared
Statistical
analysis
of
test.
The sex,
two groups of patients were well matched for age, indication for ThR, duration of operation, total blood
loss
and
Table groups
of blood
I. Details
of patients
transfused
(Table
in the heparin
and
I).
A-V
Impulse
System
Heparin
A-V Impulse
(n=65)
(n=67)
15:50
Mean
69.8
age (yr)
Indication for THR Osteoarthritis Rheumatoid arthritis Avascular necrosis Other Duration Mean
total blood
Mean
volume
(sD
6.22)
72.4
loss (ml)
transfused
(ml)
(SD
6.65)
(SD
9.89)
520
(SD
189.16)
300
(SD
267.7)
phlebographic
shown
in Table
major
DVTs,
II. In the
1 1 proximal,
proximal and distal. three major DVTs, VOL.
findings
76-B, No. 4, JULY
heparin
three
Elastic
of vessels
stimulate
fibrinolysis,
physiological randa and
mechanisms Greene 1986).
70
The discovery is activated to the
(SD
11.98)
490
(SD
195.27)
308
(SD
289.15)
in the
two
group
there
distal
and
but
(Drug
compression
and
and
is but
stockings
increase
the velocity
stimulation of leg but is painful and
they
do
not
of venous
reproduce
return
of a physiological by flattening the
development
mmHg
have used intermittent
In the A-V Impulse group two proximal and one 1994
heparins
but with applies to
methods of prophylaxis be as effective as heparin
disadvantages.
the diameter
the
(Cam,
Mi-
pump in the foot plantar arch has
of an apparatus
groups were
(Gardner
and
fluoroscopic compression
Table II. as number
Total
The
arthritis).
low-molecular-weight Bulletin 1993).
fewer
of 100
2
65
or
for
not well tolerated after operation (Browse and Negus 1970). There are a number of systems which provide intermittent pneumatic compression. These help to empty veins and also to release endothelial factors which
led
49 10 6
11 5 2 (miii)
of DVT in either or the indication
which
can
pro-
vide intermittent plantar compression. This mimics the haemodynamic effects of walking and has been shown to eject venous blood towards the heart with sufficient force to overcome a cuff around the calf inflated to a pressure
19:48
47
of operation
System
regions.
slippers, which no significant
reduce this incidence, bleeding. This also
Interest in mechanical increasing, since they may
which
Male:Female
of the found
of returning venous blood. Electrical muscles also increases return velocity
RESULTS
amount
and malleolar
fitting We
or rheumatoid
heparin can risk of major
the newer Therapeutics
with
methods.
(osteoarthritis
due to PE due to
Without prophylaxis up to 70% of patients having major hip surgery will have a DVT. There is ample evidence
assessed
same
in the heel
by poor modified.
one death problems
DISCUSSION
shown to be longer than 5 cm. Minor thrombi were recorded when calf thrombi were less than 5 cm in length. At six weeks after operation, all patients were again the
group with
