thrombosis was 68.7% in patients receiving no prophylaxis and 50% in those using the device. The difference was not significant. There was, however, a ...
THROMBO-EMBOLIC TOTAL
PROPHYLAXIS
KNEE
EVALUATION
OF
N. V. WILSON,
S. K. DAS,
J. G. SMIBERT,
From
King’s
IN
REPLACEMENT THE
A-V
IMPULSE
V. V. KAKKAR,
E. M. THOMAS,
College
Hospital,
SYSTEM
H. D. MAURICE,
J. E. NIXON
London
We performed a prospective randomised controlled trial of a new mechanical method of prophylaxis for venous thrombo-embolism in 60 patients undergoing knee replacement surgery. The method uses the A-V Impulse System to produce cyclical compression of the venous reservoir of the foot. The overall incidence of deep-vein thrombosis was 68.7% in patients receiving no prophylaxis and 50% in those using the device. The difference was not significant. There was, however, a reduction of the extent of thrombosis in the treated group. There were 13 major calf-vein thrombi and six proximal-vein thrombi in the control group compared with only five major calf-vein thrombi in the treated group. This difference was significant (j = 0.014). No patient developed clinical features of a pulmonary embolism. Venous thrombo-embolism is a common complication particularly of hip and knee surgery. The incidence of deep-vein thrombosis after elective knee replacement is between 40% and 84% (Cohen et al 1973 ; Stulberg et al 1984; Lynch et a! 1988 ; Stringer et a! 1989). Pulmonary embolism occurs in approximately 13% of cases (Stulberg et a! 1984; Stringer et al 1989). Sevitt and Gallagher (1959) first demonstrated the advantages of prophylaxis in postoperative thromboembolism. Since then there have been many reports confirming its benefits, particularly using low-dose heparin. Despite suggestions to the contrary, a recent overview of more than 70 randomised trials of heparin prophylaxis in general, orthopaedic and urological surgery has verified the efficacy oflow-dose heparin in all groups of patients (Collins et al 1988). A previous survey has indicated that only 5% of orthopaedic surgeons routinely use prophylactic measures (Morris 1980). Of
N. V. Wilson,
FRCS,
Research
Fellow
PATIENTS
S. K. Das, FRCS, Research Fellow V. V. Kakkar, FRCS, Professor ofSurgical H. D. Maurice, J. G. Smibert,
FRCS, FRCS,
Orthopaedic
Science
Senior
Registrar
Orthopaedic Senior Registrar E. M. Thomas, FRCS, Consultant Orthopaedic Surgeon J. E. Nixon, FRCS, Consultant Orthopaedic Surgeon The Thrombosis Research Institute and the DepartmentofOrthopaedic Surgery, King’s College Hospital, Denmark Hill, London SE5 England. Correspondence
©
1992 British 0301-620X/92/l
J Bone Joint
50
should
be sent
to Professor
Editorial Society ofBone 357 $2.00 Surg [Br] 1992; 74-B :50-2.
particular concern in orthopaedic surgery is the potential risk of haemorrhagic complications. Although this has not been a significant hazard in hip replacement surgery, there has been a reluctance to use heparin in total knee replacement (Kakkar and Murray 1985). Mechanical methods of prophylaxis do not have this risk and may therefore be of benefit. In 1983, Gardner and Fox described a previously unrecognised physiological venous pump mechanism in the sole of the foot which is activated by the flattening of the plantar archon weight-bearing. The resultant forceful ejection of venous blood into the calf veins can overcome an occlusive tourniquet inflated to a pressure of 100 mmHg. Artificial activation of this pump during surgery may therefore abolish venous stasis and prevent the development of venous thrombosis (Gardner and Fox 1989). The arteriovenous impulse system foot pump (A-V Impulse System) is such a device which we have evaluated in this prospective study.
V. V. Kakkar. and
Joint
Surgery
9RS,
AND
METHODS
We studied 59 consecutive patients undergoing 60 elective total knee replacements with Biomet AGC 2500 (Biomet, Glamorgan, Wales) or Insall-Burstein (Zimmer, Swindon, UK) prostheses and a standard technique. Alllimbs were exsanguinated and an occlusive tourniquet was used. The indication for surgery and the duration of the operation were recorded. In addition, measurement ofcalfcircumference, 25 cm above the medial malleolus, was made in both legs. A history of previous deep-vein THE JOURNAL
OF BONE
AND
JOINT
SURGERY
THROMBO-EMBOLIC Table
I. Details
ofpatients
PROPHYLAXIS
total knee replacement
undergoing Nopump
Pflp:.
(n=31)
(a=28
Meanage(.yr)
70.1 (sD±6.8)
Sex(male:female)
10:21*
Osteoarthritis
26
20
6
8
IN TOTAL
KNEE
51
REPLACEMENT
seconds. On inflation, the pad flattens the plantar arch, simulating weight-bearing, and thus empties the venous plexus of the foot. This cycle is repeated every 20 Ieconds; thereby allowing the venous plexus to refill. three
71.1 (sn±6.7)
5:23
RESULTS Rheumatoid Duration
arthritis ofoperation
Calf-circumference
(per cent increase #{149}one patient
(mm)
132.1
measurements in operated leg)
had bilateral
Table II. replacements
(sD±
l39.2(sD±
32.4)
4.91 (sD±5.1)
4.16
Venographic
findings
in 60 knee
Pump (n=28)
Normal
10
14
MajorcalfDVT
13
5
6
0
DVT
x2 =8.508, DVT,
df=2, deep-vein
(sD±4.2)
operations
No pump (n=32)
Proximal
34.9)
p=O.O14 thrombosis
thrombosis and pulmonary embolism and the presence of varicose veins and obesity were not recorded as we have previously shown that they are not significant predisposing factors (Stringert al 1989). All patients gave informed consent before operation and were randomly assigned to receive either no prophylaxis or the A-V Impulse System. On completion of surgery, those in the treatment group had the device fitted to the operated limb. The pump was used continuously(exceptduring mobilisation) until ascending ipsilateral venography (Kakkar 1969) was performed on the 9th or 10th postoperative days. Calf-vein thrombi less than 5 cm long were classified as minor, and more extensive calf-vein thrombi and popliteal and femoral-vein thrombi as major deep-vein thrombosis. Isolated popliteal and femoral thrombi (excluding an isolated cusp thrombus) were described as proximal deep-vein thrombosis. Clinical suspicion of pulmonary embolism required confirmation by ventilation perfusion lung scanning. All venograms were independently assessed, blind. We performed statistical analysis using the chisquare test and the Student’s 1-test. The A-V Impulse System (Novamedix, Andover, England). The device is an electrically driven air compressor with reservoir that intermittently inflates a pneumatic pad applied over stockinette to the sole of the foot and held in place by a slipper. The compressor rapidly inflates the pad (0.4 sec) and then deflates it after a period of VOL.
74-B, No. 1, JANUARY
1992
In our series there were 15 men and 44 women ranging in age from 55 to 81 years. The clinical details are shown in Table I. Owing to a technical problem in one patient the device was fitted for only two days but this patient was included in the analysis on an intention-to-treat basis. The venographic findings in the 60 limbs are shown inTable II. Theoverall incidence ofdeep-vein thrombosis in the untreated group was 68.7% compared with 50% in those using the pump. There was a significantly higher incidence ofmajor deep-vein thrombosis in the untreated patients(59.4%)compared with the treated group (17.8%) (p = 0.014, x2 = 8.508). This was largely due to the reduction ofproximaldeep-vein thrombosis in the treated group. There was no relationship between the incidence of deep-vein thrombosis and operating time, osteoarthritis or rheumatoid arthritis and calf-circumference measurements. At the time of discharge, no patient had any clinical evidence of pulmonary embolism. Two patients developed complications associated with the use of the device, namely, areas of blistering and superficial skin abrasions on the dorsum of the foot. Both lesions had healed by the time of discharge. The cause was related to the fitting ofthe slipper, and this has since been rectified.
DISCUSSION An incidence of 68.7% of deep-vein series confirms the high risk after unprotected patients and is similar Stulberg et al (1984) and slightly previous study (Stringer Ct al 1989). finding of a higher incidence in arthritis (McKenna et al 1976) we in the incidence
with with
ofvenous
thrombosis in our knee replacement in to that reported by higher than in our Despite a previous patients with osteofound no difference
thrombo-embolism
in patients
rheumatoid arthritis or osteoarthritis. This agrees previous findings and those of Stulberg et al (1984). The significant reduction of deep-vein thrombosis by the A-V Impulse System compares favourably with other methods of prophylaxis. In Stulberg’s series of 468 patients receiving a variety of medications (aspirin, lowdose heparin, supplemented low-dose heparin and warfarm), the incidence was 57%. The regimens in this study were not randomised and these data should therefore be interpreted with caution. High-dose aspirin (1.3 g tds) has been reported as being beneficial (McKenna et al 1980) but, recently, Lotke et al (1984) reported an incidence of 72% in patients receiving this treatment. In another randomised prospective study of patients
N. V. WILSON,
52
S. K. DAS,
undergoing hip and knee replacements, Dextran 40 (Pharmacia, Sweden) was of no benefit but low-dose warfarin gave effective protection with an incidence of deep-vein thrombosis of 21% (Francis et al 1983). Warfarin therapy was associated, however, with bleeding complications in 4% of patients. Since therapeutic anticoagulation may be accompanied by an unacceptable risk of bleeding into the joint (Lotke et al 1984), there is a definite need for safer effective methods of prophylaxis. Mechanical methods using pneumatic compression stockings and calf- and thigh-compression devices have produced encouraging results in small numbers of patients undergoing total knee replacement (Hull et al 1979 ; McKenna et al 1980), but the A-V Impulse System is much
simpler
to use
and
has
shown
promising
results.
The ability of the system to reduce the incidence of major deep-vein thrombosis is particularly encouraging since it is well known that this is the source of pulmonary embolism (Kakkar et al 1969). It is extremely unlikely that small calf-vein thromboses cause pulmonary embolism 2%,
(Moser however,
Impulse thousand
and
LeMoine 1981). Its incidence is less than and to show a reduction using the A-V
System would patients.
require
the participation
of several
V. V. KAKKAR,
ET AL
Prospective randomised controlled trials are needed to improve thrombo-embolic prophylaxis in knee replacement surgery. Pharmacological methods may prove superior to mechanical devices and a comparison of these in knee
replacement
surgery
is required.
Of
particular
interest is the introduction of low-molecular-weight heparin which has proved very effective in general surgery (Kakkar and Murray 1985). Such an agent alone or in combination with dihydroergotamine needs assessment, particularly since this approach has already been found to be effective in hip surgery (Turpie et al 1986; Lassen Ct al 1988). We have shown, however, that the A-V Impulse System is an effective means of prophylaxis for deep-vein thrombosis against which pharmacological methods should be evaluated.
We
wish
to thank Mr A. M. N. Gardner for introducing us to the of the venous foot pump and the potential benefit of the A-V Impulse System. It was with his help that this study was initiated. We are also grateful to the radiographers at Dulwich North Hospital, who helped us with the venographic studies.
concept
No benefits a commercial
from this
in any form have been received or will be received party related directly or indirectly to the subject of
article.
REFERENCES
Cohen
SH,
Ehrlich
following
Kauffman
GE,
knee
surgery.
MB,
J Bone
Joint
Cope C. Thrombophlebitis Surg [Am] 1973; 5-A:
embolism
and
Peto thrombosis
venous
ofsubcutaneous
heparin.
Marder
N Eng/J
VJ, EvartsCM,
therapy : prevention excessive bleeding.
Lynch
R. Reduction by perioperative
in fatal pulmonary administration 1988 ; 18:1162-73.
Med
YaukoolbodiS.
Two-step
McKenna
elastic
stockings
thrombosis Res 1979;
Kakkar
for
in patients 16:37-45.
VV. Hunterian
The problems
England,
20 March,
Lecture
the
undergoing
of
elective
Surg [Br]
WJG,et
thrombo-embolism 1985 ; 67-B :538-42.
Kakkar
VV, Murray WJG. Efficacy heparin (CY216) in preventing embolism: a co-operative study. Bords
LC,
tamine for venous dose heparin and BrJSurg
Cbristiaoum
thrombosis low
surgery.
College
VV, Howe Cl, Flanc C, Clarke MB. postoperative deep-vein thrombosis. Lancet
MR,
knee
thigh
Morris
molecular
1988; 75 :686-9.
of the leg: of
hip
Natural
1969;
history
of
2:230-2.
the knee
A. Deep-
after
total
knee
1988; 70-A :11-4.
SP, Ga1*IIte total
for total
Donald
RN, motion
knee
JO.
replacement.
Thromboembolic J Bone Joint
calfand
Prevention used
necks
F, et *1. Prevention of venous knee replacement by high-dose thigh compression. Br MedJ 1980;
of venous
by orthopaedic
thromboembolism and
Is embolic Ann
Intern
general
: a survey of Lancet 1980;
surgeons.
risk conditioned by location Med 1981 ; 94:439-44.
Preventionofvenousthrombosisandpulmonary patients: a trial of anticoagulant in
offemur.
middle-aged and elderly Lancet 1959; ii :981-9.
Stringer MD, Steadman CA, Hedges AR, et al. Deep after elective knee surgery : an incidence study JBoneJoint Surg[Br] 1989; 71-B :492-7.
of
prophylaxis patients
with
vein thrombosis in 312 patients.
BN, Insall JN, Williams GW, Gbelman B. Deep-vein thrombosis following total knee replacement : an analysis of six hundred and thirty-eight arthroplasties. J Bone Joint Surg [Am]
Stulberg
and
safety of low-molecular-weight postoperative venous thromboBrJSurg 1985; 72:786-91.
HM, et al. Heparin/dihydroergoprophylaxis : comparison of lowweight heparin in hip surgery.
[Am]
J, Bachmaiui after total
KM, LeMOlne JR. deep vein thrombosis?
fractured
1984; Rankin
following 66-A :202-8.
1976; 58-A :928-32.
embolism in injured with phenindione
arthroplasty.
Swg
Indications
H.
passive
F, Kaushal
SevittS,GallagherNG.
45:257-76.
and dihydroergotamafter
GK. methods ii :572-4.
CH,
undergoing
aspirin or intermittent 280:514-7.
vein Thromb
of Surgeons
Engl 1969;
*1. Heparin
Kakkar
Lassen
and
in the deep veins
delivered at the Royal 1969. Ann R CollSurg
VV, Fok PJ, Murray inc prophylaxis against
calf
Berkowitz thrombosis
continuous
J BoneJoint
McKenaa R, Galante thromboembolism
Moser of thrombosis
Kakkar
J BoneJoint
prevention
and
in patients
Surg [Am]
Fox RH. The venous pump of the human foot: report. BristolMed Chir J 1983; 98:109-14. Gardner AMN, Fox RH. The return ofblood to the heart: venous pumps in health anddisease. London, etc : John Libbey and Co, 1989. Hull R, DelmoreTJ, HirshJ,et *1. Effectivenessof intermittent pulsatile
Rorabeck
R, Bachmanii
disease
Gardner AMN, preliminary
RB,
thrombosis
arthroplasty.
warfarin
Alavi A, venous
J Bone Joint Surg [Am]
AF, Bonnie vein
of postoperative venous thrombosis without JAMA 1983; 249:374-8, 2021 (correction).
ML, deep
of
replacement.
YusufS,
R, ScrlmgeourA,
Fr*DcISCW,
Ecker
treatment
106-12.
Collins
PA,
Lotke
1984; 66-A:194-201.
Turple
AGG, Levine MN, Hlrsh J, et *1. A randomized of a low-molecular-weight heparin (Enoxaparin) vein thrombosis in patients NEnglJMed 1986; 315:925-9.
THE JOURNAL
undergoing
elective
controlled to prevent hip
trial deepsurgery.
OF BONE AND JOINT SURGERY
