JBJS | Symposium - Recent Advances in Venous Thromboembolic Prophylaxis During and After Total Hip Replacement

The Journal of Bone & Joint Surgery, Volume 82, Issue 2

Symposium | February 01, 2000

Symposium - Recent Advances in Venous Thromboembolic Prophylaxis During and After Total Hip Replacement

EDUARDO A. SALVATI, M.D.*, NEW YORK; VINCENT D. PELLEGRINI, JR., M.D.†, HERSHEY; NIGEL E. SHARROCK, M.B., CH.B.*, NEW YORK; PAUL A. LOTKE, M.D.‡, PHILADELPHIA, PENNSYLVANIA; DAVID W. MURRAY, F.R.C.S.§, OXFORD, ENGLAND; HOLLIS POTTER, M.D.*; GEOFFREY H. WESTRICH, M.D.*, NEW YORK, N.Y.

View Disclosures and Other Information

Presented at the Annual Meeting of the American Orthopaedic Association, Asheville, North Carolina, June 5, 1998.

The Journal of Bone & Joint Surgery. 2000; 82:252-70

5 Recommendations (Recommend) | 3 Comments | Saved by 3 Users Save Case

text A A A

Introduction

Total hip replacement is an operation that is particularly prone to thromboembolic complications with potentially life-threatening consequences. Johnson et al., in a series of 7959 total hip replacements performed between 1962 and 1973, reported that the overall prevalence of pulmonary embolism was 7.89 percent and that of fatal pulmonary embolism was 1.04 percent^51,52. Similarly, in 1974, Coventry et al. reported an overall prevalence of pulmonary embolism of 2.2 percent in a series of 2012 consecutive total hip replacements¹⁶. In a subset of sixty-two patients who had received no prophylactic anticoagulation, the prevalence of fatal pulmonary embolism was 3.4 percent. However, the average duration of the operation was 2.4 hours, the average blood loss was 1650 milliliters, and the average volume of blood transfused was 1144 milliliters. Prophylactic anticoagulation with warfarin was started five days after the operation. Patients were managed with bed rest for an average of one week before walking was allowed, and they were discharged at an average of three weeks after the operation¹⁷.

During the last three decades, substantial advances have been made in the understanding of the pathophysiology and the prevention of venous thromboembolism associated with total hip replacement. The prevalence of fatal pulmonary embolism with contemporary operative techniques in the absence of anticoagulant prophylaxis was reported to be 0.5 percent after 1162 total hip replacements performed in the United Kingdom¹²¹. In North America, antithromboembolic prophylaxis and the combined use of predonated autologous blood⁵, an expeditious operation, and early mobilization appear to have lowered the rate of fatal pulmonary embolism even further, to 0.18 percent or less^{2,63,89,91,103,126}. Importantly, an increased awareness of the intense activation of the clotting cascade, which occurs during the operation and persists for more than twenty-four hours³⁶, has recently underscored the importance of events that occur during and immediately after the operation. Specifically, it has been established that the clotting cascade is activated during the preparation of the femoral canal¹⁰⁸. The problem is compounded by stasis in the lower extremity caused by obstruction of the femoral venous flow^7,93,115, which occurs while the lower extremity is kept in an extreme position to provide adequate exposure for femoral preparation and insertion of the component. Moreover, this kinking of the femoral vein can produce endothelial injury, providing the nidus for the formation and propagation of clots.

The purposes of this symposium were to review the current status of prophylaxis and screening tests for venous thromboembolism as well as to emphasize newer measures that are available to the operative team to reduce the thrombogenic stimulus during and immediately after total hip replacement.

*Departments of Orthopaedics (E. A. S. and G. H. W.), Anesthesiology (N. E. S.), and Radiology (H. P.), The Hospital for Special Surgery, 535 East 70th Street, New York, N.Y. 10021. E-mail address for Dr. Salvati: salvatie@hss.edu. E-mail address for Dr. Westrich: westrichg@hss.edu. Please address requests for reprints to Dr. Salvati.

†Department of Orthopaedics and Rehabilitation, The Milton S. Hershey Medical Center of The Pennsylvania State University College of Medicine, 500 University Drive, Hershey, Pennsylvania 17033. E-mail address for Dr. Pellegrini: vpellegrini@psghs.edu.

‡Department of Orthopaedics, University of Pennsylvania, Silverstein Pavilion, Second Floor, 3400 Spruce Street, Philadelphia, Pennsylvania 19104.

§Nuffield Orthopaedic Centre, Windmill Road, Headington, Oxford OX3 7LD, England. E-mail address for Dr. Murray: adams@hipandkneegrp.demon.co.uk.

‡Department of Orthopaedics, University of Pennsylvania, Silverstein Pavilion, Second Floor, 3400 Spruce Street, Philadelphia, Pennsylvania 19104.

§Nuffield Orthopaedic Centre, Windmill Road, Headington, Oxford OX3 7LD, England. E-mail address for Dr. Murray: adams@hipandkneegrp.demon.co.uk.

View Larger

TABLE I: RATES OF DEEP VENOUS THROMBOSIS ASSOCIATED WITH REGIONAL BLOCK AND GENERAL ANESTHESIA ACCORDING TO A MULTICENTER STUDY BY ERIKSSON ET AL.²⁶

*The patients managed with heparain received 5000 international units three times daily, those managed with enoxaparin received forty milligrams daily, and those managed with desirudin received fifteen milligrams daily.

	Type of Prophylaxis*
Unfractionated Heparin (no. of patients)	Enoxaparin (no. of patients)	Desirudin (no. of patients)
Overall rate of deep venous^M thrombosis
Patients managed with^M regional block anesthesia	67 (24.7%) of 271	83 (19.4%) of 428	85 (12.5%) of 680
Patients managed with^M general anesthesia	46 (39.0%) of 118	112 (32.8%) of 341	100 (22.2%) of 451
Rate of proximal deep^M venous thrombosis
Patients managed with^M regional block anesthesia	42 (15.4%) of 272	22 (5.0%) of 437	28 (4.0%) of 706
Patients managed with general anesthesia	27 (22.3%) of 121	36 (10.3%) of 350	19 (4.1%) of 468

Add to My JBJS

TABLE I: RATES OF DEEP VENOUS THROMBOSIS ASSOCIATED WITH REGIONAL BLOCK AND GENERAL ANESTHESIA ACCORDING TO A MULTICENTER STUDY BY ERIKSSON ET AL.²⁶

	Type of Prophylaxis*
Unfractionated Heparin (no. of patients)	Enoxaparin (no. of patients)	Desirudin (no. of patients)
Overall rate of deep venous^M thrombosis
Patients managed with^M regional block anesthesia	67 (24.7%) of 271	83 (19.4%) of 428	85 (12.5%) of 680
Patients managed with^M general anesthesia	46 (39.0%) of 118	112 (32.8%) of 341	100 (22.2%) of 451
Rate of proximal deep^M venous thrombosis
Patients managed with^M regional block anesthesia	42 (15.4%) of 272	22 (5.0%) of 437	28 (4.0%) of 706
Patients managed with general anesthesia	27 (22.3%) of 121	36 (10.3%) of 350	19 (4.1%) of 468

View Larger

TABLE II: RATES OF DEEP VENOUS THROMBOSIS ASSOCIATED WITH USE OF INTERMITTENT PNEUMATIC COMPRESSION AS REPORTED IN THE LITERATURE

*NA = not available.

Study	No. of Patients Managed with Intermittent Pneumatic Compression	Overall Rate of Deep Venous Thrombosis (percent)	Rate of Proximal Deep Venous Thrombosis* (percent)	Rate of Pulmonary Embolism* (percent)	Additional Prophylaxis
Bradley et al.⁹, 1993	30	7	3	NA	Heparin
Francis et al.³⁴, 1992	201	27	12	NA
Grady-Benson et al.³⁷, 1994	110	18	9	1	Aspirin
Hooker et al.⁴⁵, 1999	502	5	4	1
Kaempffe et al.⁵³, 1991	48	16	12	NA
Lieberman et al.⁶¹, 1994	113	6	0	1	Aspirin
Paiement et al.⁸¹, 1987	66	17	14	0
Paiement et al.⁸⁴, 1992	66	20	17	NA
Woolson¹³⁰, 1996	322	6	6	0

Add to My JBJS

TABLE II: RATES OF DEEP VENOUS THROMBOSIS ASSOCIATED WITH USE OF INTERMITTENT PNEUMATIC COMPRESSION AS REPORTED IN THE LITERATURE

*NA = not available.

Study	No. of Patients Managed with Intermittent Pneumatic Compression	Overall Rate of Deep Venous Thrombosis (percent)	Rate of Proximal Deep Venous Thrombosis* (percent)	Rate of Pulmonary Embolism* (percent)	Additional Prophylaxis
Bradley et al.⁹, 1993	30	7	3	NA	Heparin
Francis et al.³⁴, 1992	201	27	12	NA
Grady-Benson et al.³⁷, 1994	110	18	9	1	Aspirin
Hooker et al.⁴⁵, 1999	502	5	4	1
Kaempffe et al.⁵³, 1991	48	16	12	NA
Lieberman et al.⁶¹, 1994	113	6	0	1	Aspirin
Paiement et al.⁸¹, 1987	66	17	14	0
Paiement et al.⁸⁴, 1992	66	20	17	NA
Woolson¹³⁰, 1996	322	6	6	0

View Large | Download Slide(.ppt)
Add to My JBJS

FIG1:: Fig. 1 Graph showing blood levels of fibrinopeptide A during hybrid total hip replacement. Fibrinopeptide A is a cleavage product when thrombin acts on fibrinogen to form fibrin. It has a half-life of five minutes. Note the increase in fibrinopeptide A, during insertion of the femoral component, in the group managed with saline solution and the significant suppression in both the group managed with ten units of unfractionated heparin per kilogram of body weight and the group managed with twenty units of unfractionated heparin per kilogram of body weight. The values are given as the average and the standard deviation. Event 1 = before epidural injection, event 2 = after insertion of the acetabular component, event 3 = after reaming of the femur, event 4 = after relocation of the hip, and event 5 = thirty minutes postoperatively. * indicates p = 0.0001.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG2:: Fig. 2 Maximum-intensity pixel-projection reformatted image from a time-of-flight magnetic resonance venogram of a sixty-eight-year-old patient after a recent total hip replacement. Note the superior visualization of the saphenous vein (curved arrow), the superficial femoral vein (asterisk), and the profunda femoris vein (straight arrows) when compared with conventional contrast venography.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG3:: Fig. 3 Individual axial source time-of-flight magnetic resonance image of a patient made after a recent right total hip arthroplasty, demonstrating a subtle focus of diminished signal intensity (arrow) within the external iliac vein adjacent to the acetabular component (asterisk), indicating an acute thrombosis.

The Role of Anesthesia

The use of epidural or spinal anesthesia during total hip replacement reduces the risk of postoperative deep venous thrombosis^23,69,95,106. Extension of this observation highlights the intraoperative factors involved in the development of venous thrombosis during total hip replacement¹⁰⁸. Furthermore, the recognition that deep venous thrombosis begins during, rather than after, an operation leads to the concept of intraoperative prevention rather than reliance on postoperative prophylaxis alone.

Many studies have demonstrated that epidural and spinal anesthesia reduce the risk of postoperative deep venous thrombosis by approximately 40 to 50 percent⁹⁵. The rates of proximal deep venous thrombosis may be reduced by even more than 50 percent when epidural anesthesia is used^69,106. This reduction in the rate of deep venous thrombosis is seen irrespective of the type of postoperative anticoagulant prophylaxis. In a compilation of recent multicenter studies in Europe in which deep venous thrombosis was assessed with venography, Eriksson et al. verified that epidural anesthesia reduces the risk of deep venous thrombosis in patients who receive heparin, low-molecular-weight heparin, or hirudin (Table I)²⁶. Epidural anesthesia also reduces the risk of pulmonary embolism. In a retrospective study performed at The Hospital for Special Surgery, the risk of in-hospital death due to pulmonary embolism decreased sixfold, from 0.12 percent (seven) of 5874 procedures performed with general anesthesia between 1981 and 1986 to 0.02 percent (two) of 9685 procedures performed with epidural anesthesia between 1987 and 1991 (p = 0.03)¹⁰⁷.

Carefully controlled studies have demonstrated that regional anesthesia per se has no effect on platelet function or fibrinolysis during the operation^{23,72,98,100,105,110}. Rather, epidural anesthesia increases blood flow in the lower extremity during and immediately after an operation, thereby reducing the risk of deep venous thrombosis by minimizing venous stasis^4,22. Epidural anesthesia considerably increases blood flow to the foot but not necessarily to the muscle of the leg¹³¹. However, if catecholamines with beta-agonist activity such as epinephrine are used, blood flow to the skeletal muscle is enhanced. In contrast, if beta agonists are not used, blood flow to the calf decreases with epidural anesthesia⁴. In a series of 441 consecutive total hip arthroplasties performed by the same surgeon, with the patients under epidural anesthesia, the rate of proximal deep venous thrombosis was 2.4 percent when epinephrine was used and 9.3 percent when epinephrine was not used¹⁰⁵. In that study, the rates of deep venous thrombosis, assessed with venography, were also found to be related to the duration of the operation. The overall rate of deep venous thrombosis was 10.0 percent when the duration of the operation was less than seventy minutes and 35.5 percent when it exceeded seventy minutes. Corresponding rates of proximal deep venous thrombosis were 3.3 percent and 13.3 percent, respectively.

Measurement of markers of thrombin generation (prothrombin "F1+2" and thrombin-antithrombin complexes) and fibrin formation (fibrinopeptide A) in circulating blood during total hip replacement demonstrated that activation of thrombogenesis does not begin in the initial phase of the operation (Fig. 1)^18,108. Rather, it starts during the preparation of the femur and is most pronounced with implantation of femoral components with cement rather than without cement¹⁰⁸. It is also during this phase of the operation that femoral venous occlusion occurs. Presumably, tissue thromboplastin and other intramedullary procoagulants are released from the femoral canal into the obstructed femoral vein, thereby activating femoral venous thrombosis. Such femoral venous occlusion during femoral preparation has been demonstrated both in cadaver models^7,93 and in vivo^108,115,120. In animal models of operative injury, thrombosis begins within five to ten minutes in a completely obstructed vein¹¹⁶. Thus, it behooves the surgeon to minimize the duration of femoral venous occlusion during preparation of the canal and implantation of the component.

A low dose of heparin administered intraoperatively after implantation of the acetabular cup was found to suppress fibrin formation during insertion of the femoral component¹⁰⁸. In a dose-response study, ten units of unfractionated heparin per kilogram of body weight inhibited fibrin formation to a substantial degree, whereas twenty units per kilogram of body weight completely suppressed fibrin formation¹¹² (Fig. 1). The virtue of this single intravenous dose of unfractionated heparin given approximately during the middle of the operation is that the half-life is short (thirty to forty minutes), so the risk of intraoperative and postoperative bleeding is minimized. If intraoperative hypotension is used, no additional intraoperative bleeding is evident. Perhaps the most important aspect of these observations is that they introduce the concept of targeting anticoagulation for a short period of time during the operation in order to prevent the formation of clots rather than attempting to retard the extension of existing thrombi after the operation. In a series of 212 patients who had a total hip replacement under hypotensive epidural anesthesia with epinephrine, intraoperative administration of fifteen units of heparin per kilogram of body weight before femoral preparation, an expeditious operation, and postoperative use of aspirin, the rate of deep venous thrombosis, as determined with ultrasound screening, was 6 percent (3 percent had proximal thrombosis and 3 percent had distal thrombosis)⁶⁶. There were no postoperative bleeding complications and no pulmonary emboli. This experience was expanded and at the time of writing included 1000 total hip replacements. There were five pulmonary emboli (none fatal) and no bleeding complications.

Hypotensive anesthesia may further contribute to the blunting of thrombogenesis by reducing blood loss, and the associated peripheral vasoconstriction and stasis, and by facilitating an expeditious operation^61,106,109. Excessive bleeding and intravenous administration of fluid may dilute circulating anticoagulants such as antithrombin III, altering the coagulation profile and leading to a prothrombotic state^62,76. The rate of deep venous thrombosis associated with hypotensive epidural anesthesia is 8 to 15 percent, which is substantially lower than the rate of 25 to 50 percent associated with normotensive anesthesia^106,109.

Postoperative epidural analgesia provides excellent pain relief after total hip replacement and may further reduce the risk of deep-vein thrombosis⁷⁰; however, we are not aware of any studies that have specifically addressed this question. Ultrasonic measurements after total hip replacement in patients who received epidural analgesia with use of 0.12 percent bupivacaine and fentanyl (five micrograms per milliliter) infused at ten milliliters per hour failed to demonstrate any augmentation of blood flow in the lower extremity⁶⁸. Although epidural analgesia may not directly increase femoral venous blood flow, it may reduce the rates of deep venous thrombosis by enabling patients to perform flexion and extension exercises of the foot and then facilitating early walking.

The Food and Drug Administration issued an advisory against the concurrent use of low-molecular-weight heparin and epidural anesthesia in December 1997 because of thirty-eight cases of epidural hematoma associated with a neurological deficit³². It appears that it is unsafe to administer epidural anesthesia to a patient who has received low-molecular-weight heparin preoperatively or to use low-molecular-weight heparin postoperatively before removal of an epidural catheter. Low-molecular-weight heparin is suitable for use in patients receiving general anesthesia, provided that it is administered twelve hours after the operation.

Assessment of the Risks and Benefits of Routine Thromboprophylaxis After Total Hip Replacement

Consensus conferences^27,65,79 and numerous scientific and medicolegal publications have purported that total hip replacement is associated with a high risk of venous thromboembolism and a rate of fatal pulmonary embolism ranging from 1.0 to 3.4 percent. Consequently, guidelines suggest that prophylaxis be used for all patients having a total hip replacement; such sweeping recommendations should not be issued until there is clear evidence that the benefits of thromboprophylaxis outweigh the risks.

Surrogate Outcomes

The widespread belief that the rate of fatal pulmonary embolism after total hip replacement is more than 1 percent is based on the findings of a few studies conducted mainly in the 1960s and early 1970s^16,51. Moreover, the unique characteristics of the pattern and risk factors of deep venous thrombosis after total hip replacement preclude generalization of the findings of studies of venous thrombosis after other types of procedures^77,102. A number of randomized controlled trials that subsequently were undertaken have demonstrated a substantial reduction in the prevalence of venographically evident thrombosis after total hip replacement with prophylactic anticoagulation⁵⁰. It is incorrect to assume that this reduction in the surrogate outcome of deep venous thrombosis necessarily leads to a decrease in the more important outcome of overall death rate. This assumption may be false if fatal pulmonary embolism is not very common and does not account for most deaths, if a decrease in the rate of venous thrombosis is not associated with a similar decrease in the rate of fatal pulmonary embolism, or if prophylaxis increases the risk of other causes of death. Evaluation of surrogate outcomes may lead to false conclusions.

Assessment of the Death Rate

Mortality after hip replacement may be assessed in different ways. The Report of the National Confidential Enquiry into Perioperative Deaths 1991/1992 is a large database that includes information on deaths after total hip replacement in the United Kingdom¹⁰. The overall death rate of 0.35 percent (134 of 38,000 patients)¹²⁸ was calculated on the basis of inpatient mortality within thirty days after hip replacement. Related outpatient deaths were ignored, and the death rate was potentially understated. A better way to assess the death rate is to consider events that occur each month after hip replacement as was done in the Oxford Record Linkage Study¹⁰⁴. In that study of about 12,000 patients managed between 1976 and 1985, the death rate increased threefold during the first month and an additional one and one-half-fold during the second postoperative month, which amounted to an excess death rate of about 0.5 percent. In another study, which was performed in Scotland, the death rates were compared before and after hip replacement²⁵. That study demonstrated an increased postoperative death rate for a few months, but thereafter the death rate was actually lower than it was for patients on the waiting list for total hip replacement; total mortality for the first year after hip replacement was the same as it was for the year before hip replacement. These findings suggest that hip replacement may have hastened a few deaths that would have occurred naturally during the subsequent year but did not actually increase the death rate during that year.

Meta-Analysis

An informal or opportunistic meta-analysis, not confined to randomized controlled trials⁹⁹, was undertaken in 1996 to review all of the English-language orthopaedic literature published between 1966 and 1995 that contained information about death or pulmonary embolism after total hip replacement⁷⁸. Information on 93,000 patients in 181 reports was analyzed with capture of deaths that occurred within three months after the operation. The patients were categorized according to the type of prophylaxis used: none, heparin, warfarin, aspirin, dextran, and other. The patients who were managed with no prophylaxis, placebo, or antiembolism stockings were included in the category termed none, and those managed with any form of heparin, heparinoid, or low-molecular-weight heparin, with or without stockings, were included in the heparin category. Similarly, patients who were managed with warfarin, aspirin, or dextran, with or without antiembolism stockings, were included in the categories so labeled. All patients in trials in which a combination of prophylactic regimens was used or in which the prophylactic regimen was unclear were included in the category named other. This category also included 745 patients managed with mechanical pumps, as this group was too small for a separate analysis⁸⁷.

A progressive decrease in the rate of fatal pulmonary embolism was observed over time, with a contemporary rate of 0.11 percent (confidence limits, 0.07 to 0.16 percent), which is an order of magnitude lower than that generally reported. This decrease in the rate of fatal pulmonary embolism was likely due to improvements in anesthetic techniques (such as regional anesthesia), operative techniques, and rehabilitation protocols. Because it is not always possible to determine the definitive cause of death even on postmortem examination, the rate of fatal pulmonary embolism may be inaccurate³¹. Moreover, complications of therapy for the embolism or other comorbidities may result in a fatal event. Therefore, it is important to determine the overall death rate, as it represents an upper limit for the rate of fatal pulmonary embolism.

Like the rate of fatal pulmonary embolism, the overall death rate has decreased in the last few decades⁹². The findings of the meta-analysis suggest that the current overall death rate is 0.38 percent (confidence limits, 0.29 to 0.47 percent), which is similar to the previously calculated death rate⁷⁸ of 0.35 percent (confidence limits, 0.29 to 0.41 percent). Combining both sets of data yields a better estimate of the current overall death rate after total hip replacement of 0.36 percent (confidence limits, 0.31 to 0.40 percent).

The studies in the meta-analysis were categorized according to the type of prophylaxis and the rate of fatal pulmonary embolism. Information on fatal pulmonary embolism from only 22,000 patients precluded the determination of any significant difference among the groups with respect to the type of prophylaxis. Despite an encouraging trend, no prophylactic agent significantly decreased the rate of fatal pulmonary embolism compared with the rate for the group that received no prophylaxis: four fatal emboli (0.12 percent) were observed in 3432 patients who received no prophylaxis compared with two fatal emboli (0.04 percent) in 5162 patients who received prophylaxis with warfarin (p = 0.051)⁷⁸.

It is possible that some prophylactic agents do prevent fatal pulmonary embolism after total hip replacement, but the number of patients in the meta-analysis was too small to discern such a finding. Moreover, even if it is suspected that various pharmacological agents prevent fatal pulmonary embolism, it cannot be assumed that they decrease the overall death rate because of the issue of deaths related to the use of anticoagulants²⁴. Assuming that thromboprophylaxis prevents fatal pulmonary embolism, the number of lives saved is directly related to the rate of fatal pulmonary embolism. In contrast, the risk related to anticoagulant prophylaxis is finite and independent of the rate of fatal pulmonary embolism. Therefore, when the rate of fatal pulmonary embolism is high, the benefits outweigh the risks; when the frequency of fatal pulmonary embolism is low, the risks of anticoagulation outweigh the benefits.

Meta-analyses from other fields of medicine and surgery have demonstrated that aspirin and heparin prevent a proportion of fatal pulmonary emboli^3,13. However, the rate of fatal pulmonary embolism specific to total hip replacement in the absence of chemical thromboprophylaxis was about 0.1 percent; therefore, prophylaxis cannot decrease the rate of fatal pulmonary embolism by more than about 0.05 percent in this setting. The death rate resulting directly or indirectly from the complications of anticoagulants is not known^56,122. If the death rate is 0.05 percent or more, which is not inconceivable, then the use of these agents may actually be detrimental. Prophylactic agents therefore are best assessed by consideration of the overall death rate, accounting for both risks and benefits.

Similarly, no prophylactic agent has significantly decreased the overall rate of death. Therefore, it has been suggested that recommendations in favor of the routine use of thromboprophylaxis after total hip replacement are not justified. Likewise, proponents of this philosophy advocate that it is not negligent to use no prophylactic agents for routine total hip replacement. Notwithstanding this general approach, there is a subgroup of patients who are at a particularly high risk of fatal pulmonary embolism after a total hip replacement, such as those with a history of thromboembolism and primary and secondary hypercoagulable states; other chemoprophylaxis may be appropriate for this subgroup.

Conclusions from any meta-analysis may be unreliable. Assumptions and estimates of death rates may be flawed. Likewise, comparisons between different prophylactic regimens and between management with and management without prophylaxis may be biased for a number of reasons. In the absence of such aggregate analyses, larger trials of 30,000 patients in each arm would be required to demonstrate a real decrease in the overall death rate^2,121. To date, there have not been enough reports to raise the numbers, even in a meta-analysis, to a level that would demonstrate a decrease in the overall death rate. Before the issuance of guidelines recommending the routine use of chemoprophylaxis, such large, randomized, controlled clinical trials or comparably-sized meta-analyses are necessary to demonstrate significance. However, such individual trials may be impractical to conduct.

The Risks and Benefits of Aspirin

There have been variable waves of enthusiasm, and a lack thereof, for the use of aspirin as prophylaxis against venous thromboembolism after total joint arthroplasty. The National Institutes of Health Consensus Conference on thromboembolic disease in 1986 did not include aspirin among the recommended agents⁷⁹. Nonetheless, to determine the best methods to protect patients from thromboembolic disease, it is necessary to carefully assess the risks and benefits. Until recently, the risks had not been clearly defined, and this contributed to the continued controversy. Recent studies have shown that the rate of fatal pulmonary embolism after total hip replacement is between 0.1 and 0.2 percent. This rate has been shown consistently by many investigators and appears to be unchanged by the form of prophylactic agent^{15,29,63,64,73,78,103,119,121,126}. Given that the risk of fatal pulmonary embolism is relatively low, it is important to assess the concurrent risk of hemorrhage when evaluating any antithrombotic prophylaxis because excessive bleeding is the principal anticipated adverse outcome. Indeed, over the years, the greatest attributes of aspirin have been its ease of use and safety; recommendations for its use are usually predicated on the lack of associated bleeding complications rather than on its proved efficacy. It is inexpensive, requires no monitoring, and is well tolerated in low doses.

The risks from bleeding complications vary according to the agents that are chosen. A major bleeding episode has been conventionally defined as one that causes death, intra-organ bleeding, or a reoperation³⁵. The morbidity after bleeding, to which operating surgeons are sensitive, has never been clearly defined. However, if only major bleeding episodes are considered, substantial rates of bleeding are found with some of the agents currently available for thromboembolic prophylaxis, such as enoxaparin, which is associated with a rate of 5.1 percent, and warfarin, which is associated with a rate of 2.3 percent¹¹³. Lieberman et al., in a series of 1099 total hip arthroplasties, noted that the risk of wound hematoma increased from 1.1 to 3.8 percent (p = 0.003) in a subset of 450 patients who were managed with warfarin and had a prothrombin time of more than seventeen seconds⁶³. In a meta-analysis by Imperiale and Speroff, the risks of a major bleeding episode after total hip replacement were compared among different methods of prophylaxis⁵⁰. The risk was 0.3 percent for the control group, 0.4 percent for the group managed with aspirin, 1.3 percent for patients managed with warfarin, 1.8 percent for those who received low-molecular-weight heparin, and 2.6 percent for the group managed with unfractionated heparin.

Without evidence that specific agents reduce the risk of fatal pulmonary embolism (which is between 0.1 and 0.2 percent), it is necessary to ask if the increased risk of bleeding and the subsequent morbidity are worth the potential reduction in the rate of fatal pulmonary embolism. Because of these risks, there has been a resurgence of interest in the use of aspirin. It is well known that aspirin reduces the risk of heterotopic ossification⁸⁰ and inhibits platelet function. In doses of less than 100 milligrams per day, the inhibition is dose-dependent; if more than 100 milligrams of aspirin is given per day, there is complete lifetime (approximately ten-day) suppression of platelet aggregation.

Clinical trials to assess the effectiveness of aspirin have been divided into arterial and venous studies. Although studies of the arterial effects of aspirin have demonstrated a clear efficacy in reducing mortality related to unstable angina by 60 percent and mortality related to cerebrovascular accident and transient ischemic attacks by 25 percent³, studies of the venous effects of aspirin have been less conclusive. A large meta-analysis of antiplatelet prophylaxis against venous thromboembolism was recently reported by the Anti-platelet Trialists' Collaboration³. A pooled analysis of fifty-three operative trials (only nine of which included the use of aspirin during elective orthopaedic procedures) demonstrated that fatal pulmonary embolism was reduced to 0.2 percent in the patients managed with aspirin compared with 0.9 percent in control subjects (p = 0.0001). In the studies of elective orthopaedic procedures (a total of thirteen trials with only nine involving use of aspirin), the prevalence of deep venous thrombosis was reduced to 37.5 percent in the patients who used aspirin compared with 53.2 percent in the control group. These findings may be considered to be comparable with those associated with the use of warfarin and fractionated heparin after knee arthroplasty, but they are clearly inferior to the results associated with the use of either regimen after hip arthroplasty with conventional forms of anesthesia. In one study, of 2592 total hip replacements performed with use of epidural anesthesia, between 1990 and 1994, 80 percent received aspirin postoperatively for prophylaxis¹²⁶. Warfarin was administered to the remaining 20 percent of the patients because they had additional predisposing factors for venous thromboembolism. The prevalence of venographically evident deep venous thrombosis was 10 percent, the prevalence of symptomatic pulmonary embolism was 1 percent, and the prevalence of fatal pulmonary embolism was 0.04 percent¹²⁶. Sarmiento and Goswami¹⁰³ administered aspirin after 1492 total hip arthroplasties. A fatal pulmonary embolism occurred after two arthroplasties (0.13 percent), a nonfatal pulmonary embolism was diagnosed after fourteen (0.94 percent), and clinical evidence of deep venous thrombosis developed after fifteen (1.01 percent).

In summary, one must assess risks and benefits in order to determine the appropriate chemoprophylactic agents for the prevention of fatal pulmonary embolism. As the current rates of fatal pulmonary embolism are very low, inordinately large numbers of patients are needed to scientifically prove the benefit of one regimen compared with another. Therefore, one must compare the potential risk of fatal pulmonary embolism with the potential risk of increased bleeding. Aspirin has accumulated an excellent track record on safety and its use may be indicated in conjunction with epidural or spinal anesthesia in patients with no additional predisposing risk factors for venous thromboembolism^64,103,126.

Prophylaxis with Warfarin and Fractionated Heparin

During the past three decades of total hip replacement in North America, the focus on thromboembolic disease has sharpened considerably. The recommendations of the National Institutes of Health Consensus Conference in 1986 endorsed the routine use of anticoagulant prophylaxis for thromboembolic disease after elective total joint replacement and treatment of fractures about the hip⁷⁹. An understandable preoccupation with bleeding complications was subsequently responsible for the slow acceptance of anticoagulant prophylaxis, but Paiement et al. documented the increasing use of pharmacological prophylaxis, which has had a favorable impact on thromboembolism-related mortality^39,83,85. In a recent series of 1638 patients who had a total joint replacement with routine anticoagulant prophylaxis, the overall mortality rate at six months was 0.79 percent; there were two fatal pulmonary emboli (0.12 percent), which accounted for only 15 percent of deaths from all causes⁸⁸. Despite this decrease in the mortality rate, with current methods of prophylaxis after total hip replacement the rate of venographically documented deep venous thrombosis has ranged from 15 to 26 percent for patients managed with warfarin and from 6 to 21 percent for patients managed with low-molecular-weight heparin; the rates of proximal thrombosis have been approximately 2 to 5 percent for either regimen^14,35,49,89.

Notwithstanding an apparent reduction in the prevalence of fatal thromboembolic disease with more widespread use of routine anticoagulant prophylaxis, venous thromboembolism remains the most common reason for emergency readmission after total joint replacement⁹¹, and warfarin and fractionated heparin have each had varying degrees of popularity as prophylactic agents.

Warfarin

Coumadin (warfarin) is the most commonly used single agent for prophylaxis against thromboembolic disease after total hip replacement in North America. In their landmark study concerning the rate of pulmonary embolism after 2012 hip replacements, Coventry et al. found that the occurrence of fatal pulmonary embolism was reduced from 3.4 percent in patients who did not receive anticoagulation to 0.05 percent in patients who were managed with warfarin prophylaxis¹⁶. It is noteworthy that they delayed the start of prophylactic anticoagulation until the fifth postoperative day, presumably because of a concern about bleeding problems in the early postoperative period, and still observed a rate of bleeding complications of 4.1 percent in patients managed with warfarin. In a similar report of clinically evident pulmonary embolism after total hip replacement, Amstutz et al. administered warfarin for prophylaxis for three weeks postoperatively, even when this extended beyond the date of discharge from the hospital². After 3000 total hip replacements, they noted fourteen symptomatic nonfatal pulmonary emboli (0.5 percent), confirmed by ventilation-perfusion scan or angiography. In a follow-up survey conducted by mail, they identified five more symptomatic pulmonary emboli (0.17 percent) that occurred after discharge from the hospital in patients who had received warfarin for prophylaxis for an average of 11.6 days. None of the pulmonary emboli were fatal.

Early studies with rates of bleeding episodes ranging from 8 to 12 percent have suggested that the concern about hemorrhagic complications was justified^{2,16,17,41,42}. Occasionally, the episodes were severe enough to lead to death²⁴. More recent recommendations have favored reduced-intensity anticoagulation with a prothrombin-time ratio of 1.3 to 1.5 times control, or an international normalized ratio of 2.0 to 2.5, and such protocols have been associated with lower rates of bleeding^41-43. Amstutz et al. noted an overall rate of bleeding complications of 1.5 percent (forty-four of 3000 total hip replacements), with the rate decreasing from 4.7 to 1 percent after the target level for the prothrombin time was reduced from eighteen to twenty seconds to sixteen to eighteen seconds and a bedside flow sheet was implemented to monitor daily prothrombin times and warfarin dose². Of the forty-four major bleeding complications, thirty-six were wound hematomas and eight involved bleeding in the gastrointestinal or genitourinary tract; there were no related deaths.

In 1079 consecutive patients who had a primary or revision total hip replacement, Pellegrini et al. observed a rate of bleeding complications of 1.2 percent with use of a target range of 1.3 to 1.5 for the prothrombin-time index⁸⁹. In contrast, Paiement et al. reported that 4 percent (ten) of 268 patients had a major bleeding episode (nine had a wound hematoma and one had a gastrointestinal bleeding episode) during hospitalization; the overall rate of bleeding episodes, including so-called minor bleeding episodes, during hospitalization was 6 percent⁸⁵. After discharge from the hospital and during the twelve-week period of extended prophylaxis with warfarin, an additional 5 percent (thirteen) of the 268 patients had a so-called minor bleeding episode. In a similar study with a structured program of outpatient prophylaxis with warfarin after total hip replacement, Reis et al. reported a rate of bleeding complications of 3.2 percent in 125 patients managed with warfarin for four weeks following the operation, and no patient needed to be readmitted because of bleeding⁹⁷.

Non-wound-related bleeding complications in orthopaedic patients managed with anticoagulant therapy predominantly involve the gastrointestinal and genitourinary tracts. Orthopaedic studies of a few hundred patients, by design, possess adequate statistical power to investigate venographically documented thromboembolic disease with a prevalence of 20 percent. However, the lower prevalence of bleeding complications, which ranges from 1 to 5 percent, usually renders studies of this size insufficient to determine significant conclusions relative to rates of bleeding episodes associated with various anticoagulant therapies.

Physicians in the field of internal medicine have investigated predictors of major bleeding in medical patients managed with warfarin anticoagulation in the same age-ranges as patients managed with total hip replacement. The most common site of bleeding has been the gastrointestinal tract. Landefeld et al. identified the risk predictors of a major bleeding episode in 565 outpatients being managed with warfarin in a study that emphasized comorbid medical conditions and the intensity of the anticoagulant therapy^57-59. The rate of major bleeding episodes for outpatients managed with warfarin was 3 percent in the first month after discharge, with an incremental rate of 0.8 percent per month for each subsequent month of anticoagulant therapy. These data, from study populations that were considerably larger than those commonly enrolled in orthopaedic trials, suggest a rate of major bleeding complications of 4 to 5 percent in outpatients managed with warfarin for three months. These outpatient bleeding complications often escape recognition in clinical and cost-efficacy studies, as their cost is not directly attached to the index hospitalization during which the total joint replacement was performed.

Recent evidence has suggested that a combination of continuous epidural anesthesia and postoperative epidural analgesia with warfarin prophylaxis is associated with greater efficacy in the prevention of deep venous thrombosis than either modality used alone. In 322 consecutive unilateral total hip replacements, including 210 primary and 112 revision procedures, warfarin was used to prolong the international normalized ratio to a range of 2.0 to 2.5; this represents an extension of the previously published experience²⁰. An epidural catheter was placed preoperatively and was maintained for thirty-six to forty-eight hours postoperatively for pain management, with an infusion of bupivacaine and fentanyl. With an average inpatient stay of 4.7 days, contrast venography was performed for 258 patients before discharge. The overall rate of deep venous thrombosis, as detected on venography, was 8.9 percent (twenty-three of 258 patients), with seventeen thrombi in the calf (6.6 percent) and six in the thigh (2.3 percent). Counter to historical experience, all of the proximal thrombi were contiguous popliteal extensions of thrombi in the calf; there were no segmental femoral-vein thrombi. None of the twenty-three patients with venographic evidence of deep venous thrombosis, who were managed with twelve weeks of warfarin, had additional thromboembolic disease-related events or bleeding. All of the 235 patients who had normal venograms were discharged without additional anticoagulant prophylaxis. Two (0.8 percent) of these patients were readmitted because of symptomatic thromboembolic disease; one of them had deep venous thrombosis in the thigh after a bilateral Syme amputation, and the other had a nonfatal pulmonary embolism and was admitted to an outside hospital. There were no wound hematomas necessitating a reoperation and no other morbid bleeding events. These data confirmed the findings of Westrich et al., who observed an overall prevalence of deep venous thrombosis, as determined by venography, of 10.3 percent in 2037 patients who had had a total hip replacement with hypotensive epidural anesthesia in conjunction with aspirin as prophylaxis against thromboembolic disease¹²⁶. The rate of proximal deep venous thrombosis was 4.3 percent. Late symptomatic pulmonary embolism developed in eight (0.44 percent) of 1826 patients who had had negative findings on venograms.

The prevalence of deep venous thrombosis associated with a combination of warfarin and epidural anesthesia compares favorably with the results associated with any pharmacological regimen, including fractionated heparins, but without the associated bleeding complications. Both extended epidural analgesia, which is contraindicated with use of fractionated heparins³², and early mobilization probably contribute to the low prevalence of deep venous thrombosis, and they improve on the 20 percent prevalence of deep venous thrombosis associated with warfarin prophylaxis alone. Most provocative is the observation that segmental proximal thrombi, known to account for as many as 50 percent of all deep venous thrombi after total hip replacement, did not occur with this combined prophylaxis. Warfarin, in conjunction with continuous epidural anesthesia and analgesia and early mobilization of patients after total hip replacement, is an effective strategy of prophylaxis against thromboembolic disease without the risk of bleeding associated with newer preparations of fractionated heparin. All trials with use of these newer agents should include concurrent control groups and should regulate the type of anesthesia and the duration of hospitalization, as both may affect the prevalence of thromboembolic disease.

Fractionated Heparins

The introduction of fractionated heparin with a molecular weight in the range of 5000 daltons, compared with the average 15,000-dalton molecular weight of unfractionated heparin, has provided another alternative to conventional prophylaxis against thromboembolism. With an enhanced steric affinity for antithrombin III and activated factor X, the fractionated heparin molecule provides a more active agent at an earlier point in the clotting cascade than conventional heparin. Furthermore, by virtue of less binding of plasma protein, the fractionated heparins are more bioavailable, need no monitoring, and have been associated with a lower frequency of both idiosyncratic and autoimmune-mediated thrombocytopenia⁴⁴. All of these factors, including an ambitious marketing program by commercial parties, have contributed to their initial popularity.

The findings of numerous clinical studies have proved the efficacy of fractionated heparins in reducing the frequency of thromboembolism after total hip replacement compared with placebos and with unfractionated heparins. More recently, similar controlled studies comparing fractionated heparins with warfarin have demonstrated that the former have improved efficacy in the prevention of deep venous thrombosis, with overall rates in the range of 8 to 15 percent and rates of proximal deep venous thrombosis of less than 5 percent^14,113.

Although low-molecular-weight heparins offer anticoagulant prophylaxis without the need for attendant monitoring and periodic phlebotomy (notwithstanding the need to teach self-injection for outpatient use), the fractionated heparins are associated with a frequency of bleeding complications that exceeds that associated with warfarin therapy. In a meta-analysis of methods used for the prevention of deep venous thrombosis after hip replacement, Imperiale and Speroff identified a sixfold greater risk of "clinically important" bleeding in the patients managed with low-molecular-weight heparins than in the controls and a 50 percent greater risk than in patients managed with warfarin⁵⁰. Hull et al. observed a modest reduction in deep venous thrombosis in 795 patients who had had a total hip replacement⁴⁹. Analysis of the patients for whom data were available revealed that 23.2 percent (seventy-nine) of 340 patients who received warfarin and 20.8 percent (sixty-nine) of 332 patients who received Logiparin had deep venous thrombosis, but the difference between the groups was not found to be significant. However, this reduction was offset by an increase in bleeding complications⁴⁹. Major bleeding complications were observed in six (1.5 percent) of 397 patients managed with warfarin compared with eleven (2.8 percent) of 398 patients managed with Logiparin. The prevalence of wound hematomas in the group managed with warfarin (ten of 397; 2.5 percent) was less than half that in the group managed with Logiparin (twenty-three of 398; 5.8 percent) (p = 0.03). In other studies with similar categorization of major and minor bleeding complications, investigators reported a greater frequency of bleeding events associated with use of fractionated heparins, but the groups were too small to provide sufficient statistical power to determine significant differences^15,35,123. In a prospective, controlled trial comparing dalteparin with warfarin in 580 patients who had a total hip replacement, the overall prevalence of deep venous thrombosis, as detected on venography, was 15 percent in the patients managed with dalteparin compared with 26 percent in those managed with warfarin (p = 0.006), whereas the prevalence of proximal deep venous thrombosis was 5 and 8 percent, respectively (p = 0.185)³⁵. The overall rates of major bleeding events were not found to be significantly different between the groups (2.2 and 1.4 percent, respectively). Nonetheless, more patients who received dalteparin required transfusion of red blood cells than did those who received warfarin (48 and 31 percent, respectively; p = 0.001), and bleeding complications at the operative site were more than four times more frequent in the group that received dalteparin (4.4 percent) than in the group that received warfarin (1 percent) (p = 0.03)³⁵. One patient who was managed with dalteparin had transient thrombocytopenia and needed a reoperation for evacuation of a draining hematoma.

Colwell et al., in a multicenter study of 156 centers, compared the prevalence of clinically evident venous thromboembolic disease after total hip replacement in groups receiving either adjusted-dose warfarin (1495 patients) or thirty milligrams of enoxaparin every twelve hours (1516 patients) as prophylaxis¹⁵. All clinically suspected events were confirmed by objective testing. The pharmacological prophylaxis was administered for an average of 6.5 days in each group, and venous thromboembolic events were monitored for three months after discharge from the hospital. At the time of the most recent follow-up, confirmed venous thromboembolism was noted in 3.6 percent of the patients managed with enoxaparin and 3.7 percent of those managed with warfarin. Thromboembolic events were more common during hospitalization in the group that received warfarin (1.1 percent) than in the group that received enoxaparin (0.3 percent) and were more frequent after discharge from the hospital in the group that received enoxaparin (3.4 percent) than in the group that received warfarin (2.6 percent). Clinically important bleeding occurred in eighteen (1.2 percent) of the patients managed with enoxaparin and in eight (0.5 percent) of the patients managed with warfarin. Analysis of adverse bleeding events revealed that fourteen of eighteen patients managed with enoxaparin who had a notable bleeding episode had either received the initial dose of enoxaparin within twelve hours after the operation or had had twice-daily administration of the drug rather than daily administration. Similarly, in a case-control study of 152 patients who were managed with either enoxaparin or pneumatic compression as prophylaxis against venous thromboembolism, Watson et al. demonstrated that an increased rate of bleeding episodes was associated with the use of enoxaparin¹²³. The postoperative drop in the hematocrit was significantly greater in the group managed with enoxaparin, and major bleeding events occurred in 3.3 percent of the patients managed with enoxaparin compared with 1.3 percent of the patients managed with pneumatic compression. As in the study by Colwell et al., in the study by Watson et al. the patients who received the initial dose of enoxaparin at least ten hours postoperatively had significantly fewer complications than those who received the drug within ten hours after the operation (p = 0.05).

The published literature, therefore, presents a clear dose-response relationship between increasing doses of fractionated heparin and decreasing prevalences of deep venous thrombosis, but fractionated heparin is associated with increasing hemorrhagic complications. This point was well illustrated by Spiro et al., who conducted a dose-response study of 568 patients with a hip replacement who were managed with three different regimens of enoxaparin (ten milligrams daily, forty milligrams daily, or thirty milligrams every twelve hours) for prophylaxis against deep venous thrombosis¹¹³. The prevalence of deep venous thrombosis in these groups was reduced to 25, 14, and 11 percent, respectively, as the dose of enoxaparin increased. These reductions were associated with significant increases in the overall rate of hemorrhagic events of 5, 11, and 13 percent, respectively (p < 0.05). It has become increasingly evident that bleeding complications (especially those related to the operative wound) associated with the use of fractionated heparins at a dose sufficient to substantially reduce the prevalence of deep venous thrombosis occur at considerably greater rates than those observed with a regimen of low-dose warfarin. Considering the many years necessary for surgeons to accept routine use of warfarin as prophylaxis against thromboembolic disease and its attendant rate of bleeding complications of 1 to 2 percent, it is difficult to endorse a regimen of fractionated heparin, which is associated with a substantially greater rate of bleeding episodes during hospitalization. Adoption of fractionated heparin as the standard for prophylaxis against thromboembolic disease after total hip replacement does not appear to be justified at this time.

Pneumatic Compression Devices

Intermittent pneumatic compression devices provide intermittent inflation of air-filled cuffs, which increase the velocity of venous blood flow to overcome venous stasis. Intermittent pneumatic compression devices offer many advantages over pharmaceutical methods: patients have a high tolerance for these devices, there is no associated risk of bleeding, and laboratory monitoring is not required⁶¹. The pumping mechanism of the sleeve and boots appears to stimulate endogenous fibrinolytic activity; when applied to the arm, intermittent pneumatic compression devices have been shown to decrease the development of deep venous thrombosis in the lower extremities⁵⁵. These devices should not be used on patients who have acute deep venous thrombosis; they also cannot be applied to patients who have a fracture of the lower extremity. Recently, pulsatile foot pumps have been reported to be effective in preventing venous thromboembolic disease after total hip replacement^33,46.

In patients who are at high risk for deep venous thrombosis, intermittent pneumatic compression has been shown to decrease the prevalence of deep venous thrombosis. Clinical studies^46,130 combining elastic stockings and intermittent pneumatic compression have shown a decrease in the prevalence of venous thromboembolism. However, Keith et al. were unable to show that use of a combination of graduated compression elastic stockings and intermittent pneumatic compression had an additive effect on peak velocity of venous blood flow in the superficial femoral vein⁵⁴. Studies comparing the effectiveness of these two methods when used alone for the prevention of deep venous thrombosis are lacking. A number of investigations have demonstrated that mechanical devices designed to reduce venous stasis are effective in reducing the overall rate of deep venous thrombosis after total joint replacement and traumatic injury^{9,30,33,34,37,40,45-48,53,61,81,82,84,117,127} (Table II). In at least three studies specifically concerning total hip replacement, the overall rate of deep venous thrombosis in patients managed with pneumatic compression sleeves was not found to be significantly different from that in patients managed with warfarin. Interestingly, some of those studies demonstrated a complementary decrease in distal deep venous thrombosis and a worrisome increase in proximal deep venous thrombosis after total hip arthroplasty, suggesting relative inefficacy of thigh-high pneumatic compression sleeves in the prevention of proximal thrombosis compared with warfarin prophylaxis^34,53,81.

Paiement et al. compared the effect of intermittent pneumatic compression in sixty-six patients with that of warfarin in seventy-two patients after total hip replacement^81,82. The overall prevalence of deep venous thrombosis in the groups was not found to be significantly different (p = 0.05); however, the location of the thrombi, as determined by radiographic phlebography, differed. Intermittent pneumatic compression offered better protection against thrombi in the calf, and warfarin offered better protection against thrombi in the thigh. Francis et al. reported similar results in a series of 201 patients who had had a total hip replacement; the overall prevalence of deep venous thrombosis, as determined by screening with contrast venography, was 31 percent for the patients managed with warfarin and 27 percent for those managed with pneumatic compression sleeves³⁴. Intermittent pneumatic compression provided a significant reduction in the prevalence of deep venous thrombosis in the calf, with a prevalence of 12 percent compared with 21 percent for those managed with warfarin (p = 0.021). Conversely, warfarin resulted in a significant decrease in the prevalence of proximal deep venous thrombosis (3 percent) compared with that observed with pneumatic compression (12 percent) (p = 0.012). The proximal thrombi in this trial were typically located within fifteen centimeters of the femoral head and were segmental and not continuous with thrombi in the deep veins of the calf.

In a study of intermittent pneumatic compression after total hip replacement, Wille-Jørgensen et al. noted a reduction in thromboembolism in patients managed with epidural anesthesia compared with those managed with general anesthesia¹²⁷. The prevalence of deep venous thrombosis was determined by plasmin scintigraphy with technetium-99m, and the prevalence of pulmonary embolism was assessed by pulmonary perfusion scans and ventilation scintigraphy. The prevalence of deep venous thrombosis was reduced from 31 to 9 percent, whereas the prevalence of pulmonary embolism decreased from 9 to 3 percent.¹²⁷ Similarly, Woolson, in a study of the efficacy of intermittent pneumatic compression after 289 primary or revision total hip replacements, reported an overall prevalence of deep venous thrombosis, as determined by venous ultrasonography, of 6 percent; the prevalence of proximal thrombi was 4 percent for the patients who had regional anesthesia compared with 11 percent for the patients managed with general anesthesia (p = 0.02)¹³⁰. Bradley et al. compared the combination of intermittent plantar venous compression, thigh-high antiembolism stockings, and pharmacological prophylaxis (a combination of heparin and hydroxychloroquine) in thirty patients who had a total hip replacement with the same regimen without the plantar venous compression in forty-four patients who had a total hip replacement⁹. The prevalence of deep venous thrombosis, as determined by bilateral ascending venography, was 7 percent in the former group and 27 percent in the latter (p = 0.025).

Although the optimum characteristics of these pumps for the reduction of the rate of deep venous thrombosis and pulmonary embolism are not yet known, it has been proposed that intermittent pneumatic compression is effective by means of two mechanisms. The primary mechanism is the ability to decrease venous stasis or to accelerate venous emptying, and a secondary mechanism is the ability to increase fibrinolysis^{1,9,33,40,46-48,53,67,118,127}. Sequential compression of the calf may offer advantages over single-chamber compression; sequential gradient thigh-high compression produces a 240 percent increase in the peak velocity of femoral blood flow compared with a single-chamber device, which produces a 180 percent increase in peak velocity. Sequential gradient thigh-high compression clears venographic contrast medium from the thigh eight times more quickly than single-chamber compression of the calf does. Nonetheless, in some studies these devices have been associated with increased rates of proximal thrombosis and reduced rates of distal thrombosis when compared with the use of warfarin. The available devices differ in a number of ways, including the length and location of the sleeve and the bladder, the frequency and duration of activation, the rate at which the pressure increases and the maximum pressure that is achieved, and whether the compression is simultaneous or sequential. Interest in this area has appropriately refocused attention on physiological exercises performed in bed; unassisted active movement of the ankle increases femoral venous flow by more than 75 percent compared with baseline resting values⁶⁸.

Intermittent pneumatic compression offers both theoretical and practical advantages over pharmacological prophylaxis against venous thromboembolism—namely, safety, absence of bleeding complications, and efficacy in reducing calf thrombosis. Notably, in some studies, proximal deep venous thrombosis after total hip replacement has been rather refractory to calf-thigh compression alone as a means of prophylaxis^34,53,81.

Screening Tests for Deep Venous Thrombosis

Despite negative findings for deep venous thrombosis on contrast venography and ultrasonography after total hip replacement, a small percentage of patients nonetheless sustain a pulmonary embolism. It has been postulated that the occult source of these emboli may be within the deep veins of the pelvis.

Traditional screening techniques for the detection of deep venous thrombosis include contrast venography⁸⁸ and color Doppler ultrasound. Conventional contrast venography of the lower extremity, performed by means of cannulation of a dorsal vein in the foot, is limited with regard to the assessment of pelvic thrombosis⁷⁴. In addition, the potential risk of allergy to the contrast medium is well documented, as are the risks of local soft-tissue damage created by extravasation of contrast medium at the site of puncture⁹⁶ and postvenographic thrombosis⁶. Direct femoral venipuncture affords superior visualization of the external iliac and common iliac veins of the pelvis; however, visualization of the internal iliac veins is impaired because of the fact that contrast medium injected distally must reflux against the flow in the venous direction¹¹. In addition, direct femoral venopuncture over a site of a recent hip replacement raises concerns about local hematoma formation and the potential for infection.

Color Doppler ultrasound has been very effective in the detection of symptomatic deep venous thrombosis. Application of this modality as a screening tool for the identification of nonocclusive asymptomatic thrombosis in the postoperative setting provides a greater challenge. Thrombi in the thigh have been successfully identified in several centers; however, identification of thrombi in the calf is highly variable, with sensitivities as low as 12 percent compared with venography¹². Similarly, visualization of pelvic veins is limited by the ileus in association with the administration of narcotic analgesic agents, which cause air reflection of the ultrasound beam as well as noncompressibility of the deep veins of the pelvis. In one large series in which contrast venography was compared with color Doppler ultrasound in 2000 patients who had had an orthopaedic operation, the sensitivity of ultrasound in the detection of asymptomatic proximal deep venous thrombosis was found to be only 62 percent compared with venography¹²⁴. A multicenter study including 385 patients who had had unilateral hip or knee replacement also demonstrated that color Doppler ultrasound had poor sensitivity (38 percent) in the detection of proximal venous thrombosis²¹. The advantages of color Doppler ultrasound include its noninvasive nature, relatively low cost, and portability, which allows procedures to be performed at the bedside. It is, however, highly operator-dependent, with accuracy varying widely depending on the expertise and the experience of the technician¹²⁵.

Given these limitations of both contrast venography and color Doppler ultrasound in the detection of both proximal and pelvic thrombosis, additional methods of diagnostic assessment are necessary. Magnetic resonance venography is an imaging technique providing venographic detection on the basis of the magnetic properties of soft tissue. Conventional magnetic resonance venography has successfully identified occult deep venous thrombosis following acetabular and pelvic trauma^74,75; in the initial study by Montgomery et al., contrast venography failed to detect 58 percent of the thrombi detected by magnetic resonance venography, most of which were localized in the pelvis⁷⁴. However, a high association between the findings of conventional contrast venography and those of magnetic resonance venography has been documented in preliminary studies of patients without acetabular or pelvic trauma^11,28,114.

Conventional magnetic resonance venographic techniques use a noninvasive imaging sequence, which provides increased signal based on flow characteristics, allowing for bilateral evaluation of the veins in the thigh and the pelvis with no contrast agent (Fig. 2). Differential venographic or arteriographic techniques are afforded by use of a saturation pulse, which selectively eliminates flow in either the venous or the arterial direction. High signal intensity of fully magnetized protons in flowing blood is contrasted with the low signal intensity of the stationary saturated background tissue. Individual axial images are so-called stacked by a computer algorithm into a projected image that simulates a conventional angiogram. Careful study of these individual axial sections is essential to discern diminished signal secondary to turbulent flow related to the cardiac cycle or at areas of bifurcation of vessels from fixed intraluminal saturation that is indicative of deep venous thrombosis (Fig. 3). This technique also helps to distinguish extrinsic compression, which can simulate venous occlusion on a reformatted image, from true intraluminal thrombosis¹⁰¹.

Contraindications to the use of magnetic resonance venography include the presence of a pacemaker, certain otological and penile implants, and most cerebral aneurysm clips. It is generally recommended that patients who have indwelling vascular stents and inferior vena cava filters wait six weeks before magnetic resonance imaging. In addition, claustrophobic patients may not tolerate magnetic resonance imaging and morbidly obese patients may not fit into the center of the imaging bore. The presence of orthopaedic instrumentation, including hip-replacement components, does not preclude these techniques. In a report on an early experience with magnetic resonance venography after hip replacement, occult thrombi in the pelvic veins that had not been documented on contrast venography or Doppler ultrasound were identified in three of fifteen patients⁹⁴. Nonetheless, variable segments of signal loss adjacent to the acetabular component have been observed, whereas depiction of the profunda femoris vein and the superficial femoral veins adjacent to the femoral component has been superior to that with conventional contrast venography and ultrasound. The orientation of acetabular screws, particularly in patients who had a revision, is divergent from the long axis of the magnetic field and creates greater focal disturbance and image artifact. Alterations in technique, use of fast spin-echo sequences⁹⁴, and use of gadolinium compounds as contrast agents⁶⁰ have helped to improve the visualization of the pelvic veins adjacent to the acetabulum.

Magnetic resonance venography is an evolving technique that is superior to either conventional contrast venography or color Doppler ultrasound in detecting occult thrombi in the pelvis.

Discussion

With improved operative and anesthetic techniques that afford accelerated rehabilitation, the prevalence of fatal pulmonary embolism after total hip arthroplasty has declined in recent years independent of routine anticoagulant prophylaxis. Additionally, predeposit of autologous blood is associated with a reduction in the overall prevalence of venous thromboembolic disease⁵. Notwithstanding, venous thromboembolism remains the single greatest threat to the life of a patient during the extended postoperative period after total hip arthroplasty. Although several advances in anesthetic and perioperative management have had a favorable impact on the prevalence of venous thromboembolism, controversy continues with respect to the issues of appropriateness and type of routine prophylaxis.

It has been conclusively shown that spinal anesthesia and epidural anesthesia reduce blood loss⁷¹ and the development of thromboembolism⁹⁵ compared with general anesthesia. The benefit is even more substantial if hypotension is added^61,106,109; its safety in older adults has been documented in a prospective, randomized trial¹²⁹. A surgeon who operates on a bloodless field can proceed expeditiously, without the delay caused by persistent bleeding and repeated attempts at hemostasis and with perfect visualization of the anatomical structures. The importance of an effective and expeditious operation cannot be overemphasized, as there is a direct association between the duration of the operation (particularly the duration of femoral vein occlusion), the amount of blood loss, and the prevalence of postoperative thromboembolic disease^38,106,111. Patients who are managed with spinal or epidural anesthesia do not recover active motion of the lower extremities for a few hours after the operation, which further compromises the venous return. The prompt application of mechanical devices immediately after the operation increases venous flow, preventing or minimizing the formation and propagation of clots^9,40,46,47. Once motor function has been recovered, active dorsiflexion of the foot and ankle increases femoral venous flow and such motion should be strongly encouraged. Intermittent pneumatic compression devices mimic this effect on the augmentation of venous return and exert a beneficial effect in reducing deep venous thrombosis in the calf. Use of intermittent pneumatic compression devices alone for prophylaxis against venous thromboembolism after total hip replacement, however, has been associated with an increase in the prevalence of proximal thrombi in two prospective, randomized clinical trials^34,81.

Selective heparin anticoagulation administered a few minutes before the intense activation of the clotting cascade, which occurs during femoral preparation, has been shown to reduce or obliterate the thrombogenic stimulus. The response is dose-related. Likewise, limiting the time that the lower extremity is maintained in a position that causes extreme torsion of the femoral vein reduces venous stasis and endothelial injury. Preheating of the femoral stem^8,19 or the cement, or both, can reduce the time of polymerization, without an adverse effect on their biomechanical properties⁸⁶.

Beyond these intraoperative maneuvers, the issue of routine chemical prophylaxis against venous thromboembolism remains controversial and has been challenged recently by the British orthopaedic community. The suggestion of Murray et al.⁷⁷ that the effectiveness of thromboprophylaxis should be assessed on the basis of overall rates of mortality is plausible, but the concept of Dunsmuir et al.²⁵ of so-called accelerated mortality caused by an operation with no absolute effect on overall mortality is less intuitive. Although historical death rates attributable to pulmonary embolism have ranged from 2.2 to 3.4 percent^16,51, a meta-analysis from the United Kingdom of more recent studies⁷⁸ has suggested that the rate of fatal pulmonary embolism even in the absence of chemical prophylaxis is 0.1 to 0.2 percent. The cumulative experience of five single-center series of 1079, 1099, 1492, 2592, and 3000 patients in North America demonstrated an average rate of fatal pulmonary embolism of 0.09 percent (range, 0 to 0.18 percent) or less with either warfarin or aspirin^{2,63,89,103,126}. Because the extended three to six-month follow-up period that is necessary for determining mortality rates in these studies is more reliably achieved in single-center trials, the accuracy of large numbers from multicenter pooled data is actually less than that of smaller, single-center studies. Moreover, accomplishing a prospective trial to compare pulmonary embolism-related death rates in a single center would require in excess of 50,000 patients, making it a practical impossibility. Within these limitations of comparative data from different institutions, the consensus in North America supports routine anticoagulation as prophylaxis against venous thromboembolic disease in patients having a total hip arthroplasty.

The selection of a chemoprophylactic agent remains equally controversial. Low-intensity warfarin therapy has become a time-honored method of prophylaxis, with rates of bleeding complications of less than 1 percent and a residual overall frequency of deep venous thrombosis of 15 to 20 percent in conjuction with general anesthesia and less than 10 percent with epidural or spinal anesthesia. Aspirin has demonstrated an even more favorable safety profile. Its actual efficacy in preventing venous thromboembolic events after total hip arthroplasty has been demonstrated in association with epidural and spinal anesthesia in patients who do not have additional risk factors for venous thromboembolic disease^103,126.

A review of the North American experience with low-molecular-weight heparins reveals that the risk of serious bleeding, especially that related to the operative wound, is significantly higher (p = 0.034 for wound-related bleeding complications³⁵) for patients managed with heparin than for those managed with warfarin^15,35,90,123. Moreover, serious epidural or spinal hematomas associated with spinal puncture, resulting in long-term or permanent paralysis, have been reported by the Food and Drug Administration³² and have led to warnings concerning the simultaneous use of fractionated heparins and regional anesthetic techniques. To reduce the risk of bleeding, the first dose of fractionated heparin should be delayed for at least twelve hours after the operation. As the thrombogenic stimulus¹⁰⁸ occurs during the operation, the effectiveness of such delayed anticoagulation would be decreased in the interest of safety. In view of this increased risk of bleeding complications, the popularity of fractionated heparins appears to be declining.

Imperfect as it is, warfarin has demonstrated efficacy for prophylaxis against venous thromboembolism after hip replacement, especially for patients with other risk factors for venous thromboembolic disease. Use of warfarin may begin the day before the operation because of the twenty-four to forty-eight-hour latency period before the onset of the anticoagulant effect; the combination of preoperative warfarin and epidural anesthesia has been proved to be safe in standard practice.

The most popular techniques for the assessment of deep venous thrombosis after total hip replacement have included venography (considered the so-called gold standard) and ultrasound. Ultrasound is insensitive in the identification of thrombi in the calf and variable in its ability to identify proximal thrombi; therefore, it is a poor surveillance tool for the identification of asymptomatic deep venous thrombosis in most institutions. Neither ultrasound nor conventional venography routinely visualize intrapelvic clots. Advances in magnetic resonance imaging of the venous system permit noninvasive identification of clots in both lower extremities, including the iliac veins^11,28,60. The early experience with this technique has demonstrated clots in the external iliac veins, even in patients who had negative findings on venograms and ultrasound studies. This finding may explain the occurrence of pulmonary embolism in patients who had negative findings on venography or ultrasound⁹⁴. The implications of clots in the iliac veins need to be defined, as the potential threat and morbidity associated with venous clots increase as the location becomes more proximal. A noninvasive and reliable method of surveillance for deep venous thrombosis might allow for a strategy of selective anticoagulant therapy after discharge from the hospital and for better secondary prevention of fatal embolic events, without unnecessary exposure of a larger group of patients to the risks of outpatient anticoagulation.

References

Allenby, F.; Boardman, L.; Pflug, J. J.; and Calnan, J. S.: Effects of external pneumatic intermittent compression on fibrinolysis in man. Lancet,2: 1412-1414, 1973.21412 1973 [PubMed]

Amstutz, H. C.; Friscia, D. A.; Dorey, F.; and Carney, B. T.: Warfarin prophylaxis to prevent mortality from pulmonary embolism after total hip replacement. J. Bone and Joint Surg.,71-A: 321-326, March 1989.71-A321 1989

Antiplatelet Trialists' Collaboration: Collaborative overview of randomized trials of antiplatelet therapy. III. Reduction in venous thrombosis and pulmonary embolism by antiplatelet prophylaxis among surgical and medical patients. British Med. J.,308: 235-246, 1994.308235 1994

Bading, B.; Blank, S. G.; Sculco, T. P.; Pickering, T. G.; and Sharrock, N. E.: Augmentation of calf blood flow by epinephrine infusion during lumbar epidural anesthesia. Anesth. and Analg.,78: 1119-1124, 1994.781119 1994

Bae, H.; Westrich, G. H.; Sculco, T. P.; and Salvati, E. A.: Effect of preoperative donation of autologous blood on deep venous thrombosis in total hip arthoplasty. In abstracts for the Eastern Orthopaedic Assocation. J. Bone and Joint Surg.,81-A: 10, April 1999.81-A10 1999

Bettmann, M. A.; Robbins, A.; Braun, S. D.; Wetzner, S.; Dunnick, N. K.; and Finkelstein, J.: Contrast venography of the leg: diagnostic efficacy, tolerance, and complication rates with ionic and nonionic contrast media. Radiology,165: 113-116, 1987.165113 1987 [PubMed]

Binns, M., and Pho, R.: Femoral vein occlusion during hip arthroplasty. Clin. Orthop.,255: 168-172, 1990.255168 1990 [PubMed]

Bishop, N. E.; Ferguson, S.; and Tepic, S.: Porosity reduction in bone cement at the cement-stem interface. J. Bone and Joint Surg.,78-B(3): 349-356, 1996.78-B(3)349 1996

Bradley, J. G.; Krugener, G. H.; and Jager, H. J.: The effectiveness of intermittent plantar venous compression in prevention of deep venous thrombosis after total hip arthroplasty. J. Arthroplasty,8: 57-61, 1993.857 1993 [PubMed]

Campling, E. A.; Devlin, H. B.; Hoile, R. W.; and Lunn, J. N.: The Report of the National Confidential Enquiry into Perioperative Deaths 1991/1992. London, Royal College of Surgeons, 1993.

Carpenter, J. P.; Holland, G. A.; Baum, R. A.; Owen, R. S.; Carpenter, J. T.; and Cope, C.: Magnetic resonance venography for the detection of deep venous thrombosis: comparison with contrast venography and duplex Doppler ultrasonography. J. Vasc. Surg.,18: 734-741, 1993.18734 1993 [PubMed]

Ciccone, W. J., II; Fox, P. S.; Neumyer, M.; Rubens, D.; Parrish, W. M.; and Pellegrini, V. D., Jr.: Ultrasound surveillance for asymptomatic deep venous thrombosis after total joint replacement. J. Bone and Joint Surg.,80-A: 1167-1174, Aug. 1998.80-A1167 1998

Collins, R.; Scrimgeour, A.; Yusuf, S.; and Peto, R.: Reduction in fatal pulmonary embolism and venous thrombosis by perioperative administration of subcutaneous heparin. Overview of results of randomized trials in general, orthopaedic, and urologic surgery. New England J. Med.,318: 1162-1173, 1988.3181162 1988

Colwell, C. W., Jr.; Spiro, T. E.; Trowbridge, A. A.; Morris, B. A.; Kwaan, H. C.; Blaha, J. D.; Comerota, A. J.; and Skoutakis, V. A.: Use of enoxaparin, a low-molecular-weight heparin, and unfractionated heparin for the prevention of deep venous thrombosis after elective hip replacement. A clinical trial comparing efficacy and safety. J. Bone and Joint Surg.,76-A: 3-14, Jan. 1994.76-A3 1994

Colwell, C. W., Jr.; Collis, D. K.; Paulson, R.; McCutchen, J. W.; Bigler, G. T.; Lutz, S.; and Hardwick, M. E.: Comparison of enoxaparin and warfarin for the prevention of venous thromboembolic disease after total hip arthroplasty. Evaluation during hospitalization and three months after discharge. J. Bone and Joint Surg.,81-A: 932-940, July 1999.81-A932 1999

Coventry, M. B.; Nolan, D. R.; and Beckenbaugh, R. D.: "Delayed" prophylactic anticoagulation: a study of results and complications in 2,012 total hip arthroplasties. J. Bone and Joint Surg.,55-A: 1487-1492, Oct. 1973.55-A1487 1973

Coventry, M. B.; Beckenbaugh, R. D.; Nolan, D. R.; and Ilstrup, D. M.: 2,012 total hip arthroplasties: a study of postoperative course and early complications. J. Bone and Joint Surg.,56-A: 273-284, March 1974.56-A273 1974

Dahl, O. E.: The role of the pulmonary circulation in the regulation of coagulation and fibrinolysis in relation to major surgery. J. Cardiothoracic and Vasc. Anesth.,11: 322-328, 1997.11322 1997

Dall, D. M.; Miles, A. W.; and Juby, G.: Accelerated polymerization of acrylic bone cement using preheated implants. Clin. Orthop.,211: 148-150, 1986.211148 1986 [PubMed]

Dalldorf, P. C.; Perkins, F. M.; Totterman, S.; and Pellegrini, V. D., Jr.: Deep venous thrombosis following total hip arthroplasty. Effects of prolonged postoperative epidural anesthesia. J. Arthroplasty,9: 611-616, 1994.9611 1994 [PubMed]

Davidson, B. L.; Elliott, C. G.; and Lensing, A. W.: Low accuracy of color Doppler ultrasound in the detection of proximal leg vein thrombosis in asymptomatic high-risk patients. The RD Heparin Arthroplasty Group. Ann. Internal Med.,117: 735-738, 1992.117735 1992

Davis, F. M.; Laurenson, V. G.; Gillespie, W. J.; Foate, J.; and Seagar, A. D.: Leg blood flow during total hip replacement under spinal or general anaesthesia. Anaesth. and Intens. Care,17: 136-143, 1989.17136 1989

Davis, F. M.; Laurenson, V. G.; Gillespie, W. J.; Wells, J. E.; Foate, J.; and Newman, E.: Deep vein thrombosis after total hip replacement. A comparison between spinal and general anaesthesia. J. Bone and Joint Surg.,71-B(2): 181-185, 1989.71-B(2)181 1989

Dearborn, J. T., and Harris, W. H.: Postoperative mortality after total hip arthroplasty. An analysis of deaths after two thousand seven hundred and thirty-six procedures. J. Bone and Joint Surg.,80-A: 1291-1294, Sept. 1998.80-A1291 1998

Dunsmuir, R. A.; Allan, D. B.; and Davidson, L. A.: Early post-operative mortality following primary total hip replacement. J. Royal Coll. Surgeons Edinburgh,41: 185-187, 1996.41185 1996

Eriksson, B. I.; Ekman, S.; Baur, M.; Lindbratt, S.; Bach, D.; Kälebo, P.; and Close, P.: Regional block anaesthesia versus general anaesthesia. Are different antithrombotic drugs equally effective in patients undergoing hip replacement? Retrospective analysis of 2354 patients undergoing hip replacement receiving either recombinant hirudin, unfractionated heparin or enoxaparin [abstract]. Thromb. and Haemost.,June Supplement: 487 (PS-1992), 1997.June Supplement487 1997

European Consensus Statement on the Prevention of Venous Thromboembolism. London, Med-Orion, 1992.

Evans, A. J.; Sostman, H. D.; Knelson, M. H.; Spritzer, C. E.; Newman, G. E.; Paine, S. S.; and Beam, C. A.: Detection of deep venous thrombosis: prospective comparison of MR imaging with contrast venography. AJR: Am. J. Roentgenol.,161: 131-139, 1993.161131 1993 [PubMed]

Fender, D.; Harper, W. M.; Thompson, J. R.; and Gregg, P. J.: Mortality and fatal pulmonary embolism after primary total hip replacement. Results from a regional hip register. J. Bone and Joint Surg.,79-B(6): 896-899, 1997.79-B(6)896 1997

Fisher, C. G.; Blachut, P. A.; Salvian, A. J.; Meek, R. N.; and O'Brien, P. J.: Effectiveness of pneumatic leg compression devices for the prevention of thromboembolic disease in orthopaedic trauma patients: a prospective, randomized study of compression alone versus no prophylaxis. J. Orthop. Trauma,9: 1-7, 1995.91 1995 [PubMed]

Fitts, W. T., Jr.; Lehr, H. B.; Bitner, R. L.; and Spelman, J. W.: An analysis of 950 fatal injuries. Surgery,56: 663-668, 1964.56663 1964 [PubMed]

Food and Drug Administration Public Health Advisory: Reports of Epidural or Spinal Hematomas with the Concurrent Use of Low Molecular Weight Heparin and Spinal/Epidural Anesthesia or Spinal Puncture. Rockville, Maryland, United States Department of Health and Human Resources, Dec. 15, 1997.

Fordyce, M. J. F., and Ling, R. S. M.: A venous foot pump reduces thrombosis after total hip replacement. J. Bone and Joint Surg.,74-B(1): 45-49, 1992.74-B(1)45 1992

Francis, C. W.; Pellegrini, V. D., Jr.; Marder, V. J.; Totterman, S.; Harris, C. M.; Gabriel, K. R.; Azodo, M. V.; and Leibert, K. M.: Comparison of warfarin and external pneumatic compression in prevention of venous thrombosis after total hip replacement. J. Am. Med. Assn.,267: 2911-2915, 1992.2672911 1992

Francis, C. W.; Pellegrini, V. D., Jr.; Totterman, S.; Boyd, A. D., Jr.; Marder, V. J.; Liebert, K. M.; Stulberg, B. N.; Ayers, D. C.; Rosenberg, A.; Kessler, C.; and Johanson, N. A.: Prevention of deep-vein thrombosis after total hip arthroplasty. Comparison of warfarin and dalteparin. J. Bone and Joint Surg.,79-A: 1365-1372, Sept. 1997.79-A1365 1997

Gitel, S. N.; Salvati, E. A.; Wessler, S.; Robinson, H. J., Jr.; and Worth, M. H., Jr.: The effect of total hip replacement and general surgery on antithrombin III in relation to venous thrombosis. J. Bone and Joint Surg.,61-A: 653-656, July 1979.61-A653 1979

Grady-Benson, J. C.; Oishi, C. S.; Hanson, P. B.; Colwell, C. W., Jr.; Otis, S. M.; and Walker, R. H.: Routine postoperative duplex ultrasonography screening and monitoring for the detection of deep vein thrombosis. Clin. Orthop.,307: 130-141, 1994.307130 1994 [PubMed]

Gray, D. H., and Mackie, C. E.: The effect of blood transfusion on the incidence of deep vein thrombosis. Australian and New Zealand J. Surg.,53: 439-443, 1983.53439 1983

Grossman, R. S.: Changing patterns of prophylaxis for deep venous thrombosis following elective hip replacement. Orthopedics,16: 19-22, 1993.1619 1993 [PubMed]

Hartman, J. T.; Pugh, J. L.; Smith, R. D.; Robertson, W. W., Jr.; Yost, R. P.; and Janssen, H. F.: Cyclic sequential compression of the lower limb in prevention of deep venous thrombosis. J. Bone and Joint Surg.,64-A: 1059-1062, Sept. 1982.64-A1059 1982

Hirsh, J.: Therapeutic range for the control of oral anticoagulant therapy. Arch. Internal Med.,145: 1187-1188, 1985.1451187 1985

Hirsh, J., and Levine, M. N.: The optimal intensity of oral anticoagulant therapy. J. Am. Med. Assn.,258: 2723-2726, 1987.2582723 1987

Hirsh, J.: Oral anticoagulant drugs. New England J. Med.,324: 1865-1875, 1991.3241865 1991

Hirsh, J., and Levine, M. N.: Low molecular weight heparin. Blood,79: 1-17, 1992.791 1992 [PubMed]

Hooker, J. A.; Lachiewicz, P. F.; and Kelley, S. S.: Efficacy of prophylaxis against thromboembolism with intermittent pneumatic compression after primary and revision total hip arthroplasty. J. Bone and Joint Surg.,81-A: 690-696, May 1999.81-A690 1999

Hull, R.; Delmore, T. J.; Hirsh, J.; Gent, M.; Armstrong, P.; Lofthouse, R.; MacMillan, A.; Blackstone, I.; Reed-Davis, R.; and Detwiler, R. C.: Effectiveness of intermittent pulsatile elastic stockings for the prevention of calf and thigh vein thrombosis in patients undergoing elective knee operation. Thromb. Res.,16: 37-45, 1979.1637 1979 [PubMed]

Hull, R. D., and Raskob, G. E.: Current concepts review. Prophylaxis of venous thromboembolic disease following hip and knee surgery. J. Bone and Joint Surg.,68-A: 146-150, Jan. 1986.68-A146 1986

Hull, R. D.; Raskob, G. E.; Gent, M.; McLoughlin, D.; Julian, D.; Smith, F. C.; Dale, N. I.; Reed-Davis, R.; Lofthouse, R. N.; and Anderson, C.: Effectiveness of intermittent pneumatic leg compression for preventing deep vein thrombosis after total hip replacement. J. Am. Med. Assn.,263: 2313-2317, 1990.2632313 1990

Hull, R.; Raskob, G.; Pineo, G.; Rosenbloom, D.; Evans, W.; Mallory, T.; Anquist, K.; Smith, F.; Hughes, G.; Green, D.; Elliott, G.; Panju, A.; and Brant, R.: A comparison of subcutaneous low-molecular-weight heparin with warfarin sodium for prophylaxis against deep-vein thrombosis after hip or knee implantation. New England J. Med.,329: 1370-1376, 1993.3291370 1993

Imperiale, T. F., and Speroff, T.: A meta-analysis of methods to prevent venous thromboembolism following total hip replacement. J. Am. Med. Assn.,271: 1780-1785, 1994.2711780 1994

Johnson, R.; Green, J. R.; and Charnley, J.: Pulmonary embolism and its prophylaxis following the Charnley total hip replacement. Clin. Orthop.,127: 123-132, 1977.127123 1977 [PubMed]

Johnson, R.; Carmichael, J. H. E.; Almond, H. G. A.; and Loynes, R. P.: Deep venous thrombosis following Charnley arthroplasty. Clin. Orthop.,132: 24-30, 1978.13224 1978 [PubMed]

Kaempffe, F. A.; Lifeso, R. M.; and Meinking, C.: Intermittent pneumatic compression versus Coumadin. Prevention of deep vein thrombosis in lower-extremity total joint arthroplasty. Clin. Orthop.,269: 89-97, 1991.26989 1991 [PubMed]

Keith, S. L.; McLaughlin, D. J.; Anderson, F. A., Jr.; Cardullo, P. A.; Jones, C. E.; Rohrer, M. J.; and Cutler, B. S.: Do graduated compression stockings and pneumatic boots have an additive effect on the peak velocity of venous blood flow?. Arch. Surg.,127: 727-730, 1992.127727 1992 [PubMed]

Knight, M. T., and Dawson, R.: Effect of intermittent compression of the arms on deep venous thrombosis in the legs. Lancet,2: 1265-1268, 1976.21265 1976 [PubMed]

Kong, K. C.: Differences in mortality after fracture of hip. Antithrombotic prophylaxis had little effect on overall mortality. British Med. J.,311: 572-573, 1995.311572 1995

Landefeld, C. S.; Cook, E. F.; Flatley, M.; Weisberg, M.; and Goldman, L.: Identification and preliminary validation of predictors of major bleeding in hospitalized patients starting anticoagulant therapy. Am. J. Med.,82: 703-713, 1987.82703 1987 [PubMed]

Landefeld, C. S., and Goldman, L.: Major bleeding in outpatients treated with warfarin: incidence and prediction by factors known at the start of outpatient therapy. Am. J. Med.,87: 144-152, 1989.87144 1989 [PubMed]

Landefeld, C. S.; Rosenblatt, M. W.; and Goldman, L.: Bleeding in outpatients treated with warfarin: relation to the prothrombin time and important remediable lesions. Am. J. Med.,87: 153-159, 1989.87153 1989 [PubMed]

Lebowitz, J. A.; Rofsky, N. M.; Krinsky, G. A.; and Weinreb, J. C.: Gadolinium-enhanced body MR venography with subtraction technique. AJR: Am. J. Roentgenol.,169: 755-758, 1997.169755 1997 [PubMed]

Lieberman, J. R.; Huo, M. M.; Hanway, J.; Salvati, E. A.; Sculco, T. P.; and Sharrock, N. E.: The prevalence of deep venous thrombosis after total hip arthroplasty with hypotensive epidural anesthesia. J. Bone and Joint Surg.,76-A: 341-348, March 1994.76-A341 1994

Lieberman, J. R., and Geerts, W. H.: Current concepts review. Prevention of venous thromboembolism after total hip and knee arthroplasty. J. Bone and Joint Surg.,76-A: 1239-1250, Aug. 1994.76-A1239 1994

Lieberman, J. R.; Wollaeger, J.; Dorey, F.; Thomas, B. J.; Kilgus, D. J.; Grecula, M. J.; Finerman, G. A.; and Amstutz, H. C.: The efficacy of prophylaxis with low-dose warfarin for prevention of pulmonary embolism following total hip arthroplasty. J. Bone and Joint Surg.,79-A: 319-325, March 1997.79-A319 1997

Lotke, P. A.; Palevsky, H.; Keenan, A. M.; Meranze, S.; Steinberg, M. E.; Ecker, M. L.; and Kelley, M. A.: Aspirin and warfarin for thromboembolic disease after total joint arthroplasty. Clin. Orthop.,324: 251-258, 1996.324251 1996 [PubMed]

Lowe, G. D. O.; Greer, I. A.; Cooke, T. G.; Dewar, E. P.; Evans, M. J.; Forbes, C. D.; Mollan, R. A. B.; Scurr, J. H.; and Swiet, M. D.: Risk of and prophylaxis for venous thrombo-embolism in hospital patients. Thromboembolic risk factors (THRIFT) consensus group. British Med. J.,305: 567-573, 1992.305567 1992

McCabe, J. P.; Sharrock, N. E.; Salvati, E. A.; and Sculco, T. P.: Intraoperative heparin and hypotensive epidural anesthesia in the prevention of thromboembolism following total hip arthroplasty. Orthop. Trans.,21: 124, 1997.21124 1997

McLachlin, A. D.; McLachlin, J. A.; Jory, T. A.; and Rawling, E. G.: Venous stasis in the lower extremities. Ann. Surg.,152: 678-685, 1960.152678 1960 [PubMed]

Markel, D. C.; Urquhart, B.; Derkowska, I.; Salvati, E. A.; and Sharrock, N. E.: Effect of epidural analgesia on venous blood flow after hip arthroplasty. Clin. Orthop.,334: 168-174, 1997.334168 1997 [PubMed]

Modig, J.; Borg, T.; Karlström, G.; Maripuu, E.; and Sahlstedt, B.: Thromboembolism after total hip replacement: role of epidural and general anesthesia. Anesth. and Analg.,62: 174-180, 1983.62174 1983

Modig, J.; Maripuu, E.; and Sahlstedt, B.: Thromboembolism following total hip replacement. A prospective investigation of 94 patients with emphasis on the efficacy of lumbar epidural anesthesia in prophylaxis. Reg. Anesth.,11: 72-79, 1986.1172 1986

Modig, J.: Beneficial effects on intraoperative and postoperative blood loss in total hip replacement when performed under lumbar epidural anesthesia. An explanatory study. Acta Chir. Scandinavica,Supplementum 550: 95-103, 1989.Supplementum 55095 1989

Modig, J.: Influence of regional anesthesia, local anesthetics, and sympathicomimetics on the pathophysiology of deep vein thrombosis. Acta Chir. Scandinavica,Supplementum 550: 119-127, 1989.Supplementum 550119 1989

Mohr, D. N.; Silverstein, M. D.; Ilstrup, D. M.; Heit, J. A.; and Morrey, B. F.: Venous thromboembolism associated with hip and knee arthroplasty: current prophylactic practices and outcomes. Mayo Clin. Proc.,67: 861-870, 1992.67861 1992 [PubMed]

Montgomery, K. D.; Potter, H. G.; and Helfet, D. L.: Magnetic resonance venography to evaluate the deep venous system of the pelvis in patients who have an acetabular fracture. J. Bone and Joint Surg.,77-A: 1639-1649, Nov. 1995.77-A1639 1995

Montgomery, K. D.; Potter, H. G.; and Helfet, D. L.: The detection and management of proximal deep venous thrombosis in patients with acute acetabular fractures: a follow-up report. J. Orthop. Trauma,11: 330-336, 1997.11330 1997 [PubMed]

Murray, D. J.; Pennell, B. J.; Weinstein, S. L.; and Olson, J. D.: Packed red cells in acute blood loss: dilutional coagulopathy as a cause of surgical bleeding. Anesth. and Analg.,80: 336-342, 1995.80336 1995

Murray, D. W.; Carr, A. J.; and Bulstrode, C. J. K.: Pharmacological thromboprophylaxis and total hip replacement [editorial]. J. Bone and Joint Surg.,77-B(1): 3-5, 1995.77-B(1)3 1995

Murray, D. W.; Britton, A. R.; and Bulstrode, C. J. K.: Thromboprophylaxis and death after total hip replacement. J. Bone and Joint Surg.,78-B(6): 863-870, 1996.78-B(6)863 1996

National Institutes of Health Consensus Development Conference Statement: Prevention of venous thrombosis and pulmonary embolism. J. Am. Med. Assn.,256: 744-749, 1986.256744 1986

Pagnani, M. J.; Pellicci, P. M.; and Salvati, E. A.: Effect of aspirin on heterotopic ossification after total hip arthroplasty in men who have osteoarthrosis. J. Bone and Joint Surg.,73-A: 924-929, July 1991.73-A924 1991

Paiement, G.; Wessinger, S. J.; Waltman, A. C.; and Harris, W. H.: Low-dose warfarin versus external pneumatic compression for prophylaxis against venous thromboembolism following total hip replacement. J. Arthroplasty,2: 23-26, 1987.223 1987 [PubMed]

Paiement, G. D.; Beisaw, N.; Lotke, P.; Elia, E.; Wessinger, S.; and Harris, W.: Advances in the prevention of venous thromboembolic disease after hip and knee surgery. Orthop. Rev.,18: 1-20, 1989.181 1989

Paiement, G. D.; Wessinger, S. J.; and Harris, W. H.: Cost-effectiveness of prophylaxis in total hip replacement. Am. J. Surg.,161: 519-524, 1991.161519 1991 [PubMed]

Paiement, G. D.; Schutzer, S. F.; Wessinger, S. J.; and Harris, W. H.: Influence of prophylaxis on proximal venous thrombus formation after total hip arthroplasty. J. Arthroplasty,7: 471-475, 1992.7471 1992 [PubMed]

Paiement, G. D.; Wessinger, S. J.; Hughes, R.; and Harris, W. H.: Routine use of adjusted low-dose warfarin to prevent venous thromboembolism after total hip replacement. J. Bone and Joint Surg.,75-A: 893-898, June 1993.75-A893 1993

Parks, M. L.; Walsh, H. A.; Salvati, E. A.; and Li, S.: Effect of increasing temperature on the properties of four bone cements. Clin. Orthop.,355: 238-248, 1998.355238 1998 [PubMed]

Pearson, E. S.: Confidence limits for the expectation of a Poisson variable. In Biometrika Tables for Statisticians, p. 227. Edited by E. S. Pearson and H. O. Hartley. Cambridge, University Press, 1966.

Pellegrini, V. D., Jr.; Evarts, C. M.; Clement, D.; Lush-Ehmann, C.; and Keller, G.: Venogram surveillance of deep vein thrombosis following total joint replacement. Orthop. Trans.,18: 647, 1994.18647 1994

Pellegrini, V. D., Jr.; Clement, D.; Lush-Ehmann, C.; Keller, G. S.; and Evarts, C. McC.: Natural history of thromboembolic disease after total hip arthroplasty. Clin. Orthop.,333: 27-40, 1996.33327 1996 [PubMed]

Pellegrini, V. D., Jr.; Totterman, S.; Ayers, D. C.; Johanson, N.; Stulberg, B. N.; Rosenberg, A.; Kessler, C.; and Francis, C.: Comparison of a low molecular weight heparin and warfarin in prevention of deep vein thrombosis after total hip arthroplasty. Orthop. Trans.,21: 124-125, 1997.21124 1997

Pellegrini, V. D., Jr.: Prevention of thromboembolic disease in patients after total joint arthroplasty by selective posthospitalization treatment. In Seminars in Arthroplasty, pp. 248-257. Edited by R. H. Fitzgerald. Philadelphia, W. B. Saunders, 1997.

The PIOPED Investigators: Value of the ventilation/perfusion scan in acute pulmonary embolism. Result of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). J. Am. Med. Assn.,263: 2753-2759, 1990.2632753 1990

Planès, A.; Vochelle, N.; and Fagola, M.: Total hip replacement and deep vein thrombosis. A venographic and necropsy study. J. Bone and Joint Surg.,72-B(1): 9-13, 1990.72-B(1)9 1990

Potter, H. G.; Montgomery, K. D.; Padgett, D. E.; Salvati, E. A.; and Helfet, D. L.: Magnetic resonance imaging of the pelvis. New orthopaedic applications. Clin. Orthop.,319: 223-231, 1995.319223 1995 [PubMed]

Prins, M. H., and Hirsh, J.: A comparison of general anesthesia and regional anesthesia as a risk factor for deep vein thrombosis following hip surgery: a critical review. Thromb. and Haemost.,64: 497-500, 1990.64497 1990

Rabinov, K., and Paulin, S.: Roentgen diagnosis of venous thrombosis in the leg. Arch. Surg.,104: 134-144, 1972.104134 1972 [PubMed]

Reis, S. E.; Hirsch, D. R.; Wilson, M. G.; Donovan, B. C.; and Goldhaber, S. Z.: Program for the prevention of venous thromboembolism in high-risk orthopaedic patients. J. Arthroplasty,6 (Supplement): 11-S16, 1991.6 (Supplement)11 1991

Richard, L. C.; Büller, H. R.; Bovilland, J.; and ten Cate, J. W.: Influence of anaesthesia on coagulation and fibrinolytic proteins. British J. Anaesth.,55: 869-872, 1983.55869 1983

Rosendaal, F. R.: The emergence of a new species: the professional meta-analyst. J. Clin. Epidemiol.,47: 1325-1326, 1994.471325 1994 [PubMed]

100

Rosenfeld, B. A.; Beattie, C.; Christopherson, R.; Norris, E. J.; Frank, S. M.; Breslow, M. J.; Rock, P.; Parker, S. D.; Gottlieb, S. O.; Perler, B. A.; Williams, G. M.; Seidler, A.; Bell, W.; and The Perioperative Ischemia Randomized Anesthesia Trial Study Group: The effects of different anesthetic regimens on fibrinolysis and the development of postoperative arterial thrombosis. Anesthesiology,79: 435-443, 1993.79435 1993 [PubMed]

101

Saloner, D.; van Tyen, R.; Dillon, W. P.; Jou, L. D.; and Berger, S. A.: Central intraluminal saturation stripe on MR angiograms of curved vessels: simulation, phantom, and clinical analysis. Radiology,198: 733-739, 1996.198733 1996 [PubMed]

102

Salzman, E. W., and Harris, W. H.: Prevention of venous thromboembolism in orthopaedic patients. J. Bone and Joint Surg.,58-A: 903-913, Oct. 1976.58-A903 1976

103

Sarmiento, A., and Goswami, A. D. K.: Thromboembolic prophylaxis with use of aspirin, exercise, and graded elastic stockings or intermittent compression devices in patients managed with total hip arthroplasty. J. Bone and Joint Surg.,81-A: 339-346, March 1999.81-A339 1999

104

Seagroatt, V.; Tan, H. S.; Goldacre, M.; Bulstrode, C.; Nugent, I.; and Gill, L.: Elective total hip replacement: incidence, emergency readmission rate, and postoperative mortality. British Med. J.,303: 1431-1435, 1991.3031431 1991

105

Sharrock, N. E.; Go, G.; Mineo, R.; and Harpel, P. C.: The hemodynamic and fibrinolytic response to low dose epinephrine and phenylephrine infusions during total hip replacement under epidural anesthesia. Thromb. and Haemost.,68: 436-441, 1992.68436 1992

106

Sharrock, N. E.; Ranawat, C. S.; Urquhart, B.; and Peterson, M.: Factors influencing deep vein thrombosis following total hip arthroplasty under epidural anesthesia. Anesth. and Analg.,76: 765-771, 1993.76765 1993

107

Sharrock, N. E.; Cazan, M. G.; Hargett, M. J.; Williams-Russo, P.; and Wilson, P. D., Jr.: Changes in mortality after total hip and knee arthroplasty over a ten-year period. Anesth. and Analg.,80: 242-248, 1995.80242 1995

108

Sharrock, N. E.; Go, G.; Harpel, P. C.; Ranawat, C. S.; Sculco, T. P.; and Salvati, E. A.: Thrombogenesis during total hip arthroplasty. Clin. Orthop.,319: 16-27, 1995.31916 1995 [PubMed]

109

Sharrock, N. E., and Salvati, E. A.: Hypotensive epidural anesthesia for total hip arthroplasty. A review. Acta Orthop. Scandinavica,67: 91-107, 1996.6791 1996

110

Sharrock, N. E.; Go, G.; Williams-Russo, P.; Haas, S. B.; and Harpel, P. C.: Comparison of extradural and general anaesthesia on the fibrinolytic response to total knee arthroplasty. British J. Anaesth.,79: 29-34, 1997.7929 1997

111

Sharrock, N. E.: Anesthesia. In The Adult Hip, pp. 615-624. Edited by J. J. Callaghan, A. G. Rosenberg, and H. E. Rubash. Philadelphia, Lippincott-Raven, 1998.

112

Sharrock, N. E.; Go, G.; Sculco, T. P.; Salvati, E. A.; Westrich, G. H.; and Harpel, P. C.: Dose response of intravenous heparin on markers of thrombosis during primary total hip replacement. Anesthesiology,90: 981-987, 1999.90981 1999 [PubMed]

113

Spiro, T. E.; Johnson, G. J.; Christie, M. J.; Lyons, R. M.; MacFarlane, D. E.; Blasier, R. B.; and Tremaine, M. D.: Efficacy and safety of enoxaparin to prevent deep venous thrombosis after hip replacement surgery. Enoxaparin Clinical Trial Group. Ann. Internal Med.,121: 81-89, 1994.12181 1994

114

Spritzer, C. E.; Sostman, H. D.; Wilkes, D. C.; and Coleman, R. E.: Deep venous thrombosis: experience with gradient-echo MR imaging in 66 patients. Radiology,177: 235-241, 1990.177235 1990 [PubMed]

115

Stamatakis, J. D.; Kakkar, V. V.; Sagar, S.; Lawrence, D.; Nairn, D.; and Bentley, P. G.: Femoral vein thrombosis and total hip replacement. British Med. J.,2: 223-225, 1977.2223 1977

116

Stewart, G. J.; Knight, L. C.; Arbogast, B. W.; and Stern, H. S.: Inhibition of leukocyte locomotion by tocainide, a primary amine analog of lidocaine: at study with ¹¹¹indium-labeled leukocytes and scanning electron microscopy. Lab. Invest.,42: 302-309, 1980.42302 1980 [PubMed]

117

Stranks, C. J.; MacKenzie, N. A.; Grover, M. L.; and Fail, T.: The A-V impulse system reduces deep-vein thrombosis and swelling after hemiarthroplasty for hip fracture. J. Bone and Joint Surg.,74-B(5): 775-778, 1992.74-B(5)775 1992

118

Virchow, R.: Neuer Fall von todticher Embolie der Lungenarterien. Arch. Pathol. Anat.,10: 225-228, 1856.10225 1856

119

Visuri, T.; Pulkkinen, P.; Paavolainen, P.; Koskenvuo, M.; and Turula, K. B.: Causes of death after hip arthroplasty in primary arthrosis. J. Arthroplasty,12: 397-402, 1997.12397 1997 [PubMed]

120

Warwick, D.; Martin, A. G.; Glew, D.; and Bannister, G. C.: Measurement of femoral vein blood flow during total hip replacement. Duplex ultrasound imaging with and without the use of a foot pump. J. Bone and Joint Surg.,76-B(6): 918-921, 1994.76-B(6)918 1994

121

Warwick, D.; Williams, M. H.; and Bannister, G. C.: Death and thromboembolic disease after total hip replacement. A series of 1162 cases with no routine chemical prophylaxis. J. Bone and Joint Surg.,77-B(1): 6-10, 1995.77-B(1)6 1995

122

Warwick, D., and Freeman, M.: Differences in mortality after fracture of hip. Overall survival rate is most important measure. British Med. J.,311: 572-573, 1995.311572 1995

123

Watson, B. N.; Shaieb, M. D.; and Atkinson, R. E.: Bleeding complication with enoxaparin for deep venous thrombosis prophylaxis. Orthop. Trans.,22: 1181-1182, 1998-1999.221181 1998-1999

124

Wells, P. S.; Lensing, A. W.; Davidson, B. L.; Prins, M. H.; and Hirsh, J.: Accuracy of ultrasound for the diagnosis of deep venous thrombosis in asymptomatic patients after orthopedic surgery. A meta-analysis. Ann. Internal Med.,122: 47-53, 1995.12247 1995

125

Westrich, G. H.; Schneider, R.; Ghelman, B.; Derkowska, I.; Tarantino, S. J.; Sculco, T. P.; and Ranawat, C. S.: Comparison between color Doppler imaging and ascending venography in the detection of deep venous thrombosis following total joint arthroplasty: a prospective study. Contemp. Surg.,51: 225-234, 1997.51225 1997

126

Westrich, G. H.; Farrell, C.; Bono, J. V.; Ranawat, C. S.; Salvati, E. A.; and Sculco, T. P.: The incidence of venous thromboembolism after total hip arthroplasty. A specific hypotensive epidural anesthesia protocol. J. Arthroplasty,14: 456-463, 1999.14456 1999 [PubMed]

127

Wille-Jørgensen, P.; Winter Christensen, S.; Bjerg-Nielsen, A.; Stadeager, C.; and Kjær, L.: Prevention of thromboembolism following elective hip surgery. The value of regional anesthesia and graded compression stockings. Clin. Orthop.,247: 163-167, 1989.247163 1989 [PubMed]

128

Williams, M. H.; Frankel, S. J.; Nanchahal, K.; Coast, J.; and Donovan, J. L.: Total Hip Replacement. Health Care Needs Assessment. The Epidemiologically Based Needs Assessment Reviews. Oxford, Radcliffe Medical Press, 1994.

129

Williams-Russo, P.; Sharrock, N. E.; Mattis, S.; Liguori, G. A.; Mancuso, C.; Peterson, M. G.; Hollenberg, J.; Ranawat, C.; Salvati, E.A.; and Sculco, T.: Randomized trial of hypotensive epidural anesthesia in older adults. Anesthesiology,91: 926-935, 1999.91926 1999 [PubMed]

130

Woolson, S. T.: Intermittent pneumatic compression prophylaxis for proximal deep venous thrombosis after total hip replacement. J. Bone and Joint Surg.,78-A: 1735-1740, Nov. 1996.78-A1735 1996

131

Wright, C. J., and Cousins, M. J.: Blood flow distribution in the human leg following epidural sympathetic blockade. Arch. Surg.,105: 334-337, 1972.105334 1972 [PubMed]

Topics

View Large | Download Slide(.ppt)
Add to My JBJS

View Larger

TABLE I: RATES OF DEEP VENOUS THROMBOSIS ASSOCIATED WITH REGIONAL BLOCK AND GENERAL ANESTHESIA ACCORDING TO A MULTICENTER STUDY BY ERIKSSON ET AL.²⁶

	Type of Prophylaxis*
Unfractionated Heparin (no. of patients)	Enoxaparin (no. of patients)	Desirudin (no. of patients)
Overall rate of deep venous^M thrombosis
Patients managed with^M regional block anesthesia	67 (24.7%) of 271	83 (19.4%) of 428	85 (12.5%) of 680
Patients managed with^M general anesthesia	46 (39.0%) of 118	112 (32.8%) of 341	100 (22.2%) of 451
Rate of proximal deep^M venous thrombosis
Patients managed with^M regional block anesthesia	42 (15.4%) of 272	22 (5.0%) of 437	28 (4.0%) of 706
Patients managed with general anesthesia	27 (22.3%) of 121	36 (10.3%) of 350	19 (4.1%) of 468

Add to My JBJS

TABLE I: RATES OF DEEP VENOUS THROMBOSIS ASSOCIATED WITH REGIONAL BLOCK AND GENERAL ANESTHESIA ACCORDING TO A MULTICENTER STUDY BY ERIKSSON ET AL.²⁶

	Type of Prophylaxis*
Unfractionated Heparin (no. of patients)	Enoxaparin (no. of patients)	Desirudin (no. of patients)
Overall rate of deep venous^M thrombosis
Patients managed with^M regional block anesthesia	67 (24.7%) of 271	83 (19.4%) of 428	85 (12.5%) of 680
Patients managed with^M general anesthesia	46 (39.0%) of 118	112 (32.8%) of 341	100 (22.2%) of 451
Rate of proximal deep^M venous thrombosis
Patients managed with^M regional block anesthesia	42 (15.4%) of 272	22 (5.0%) of 437	28 (4.0%) of 706
Patients managed with general anesthesia	27 (22.3%) of 121	36 (10.3%) of 350	19 (4.1%) of 468

View Larger

TABLE II: RATES OF DEEP VENOUS THROMBOSIS ASSOCIATED WITH USE OF INTERMITTENT PNEUMATIC COMPRESSION AS REPORTED IN THE LITERATURE

*NA = not available.

Study	No. of Patients Managed with Intermittent Pneumatic Compression	Overall Rate of Deep Venous Thrombosis (percent)	Rate of Proximal Deep Venous Thrombosis* (percent)	Rate of Pulmonary Embolism* (percent)	Additional Prophylaxis
Bradley et al.⁹, 1993	30	7	3	NA	Heparin
Francis et al.³⁴, 1992	201	27	12	NA
Grady-Benson et al.³⁷, 1994	110	18	9	1	Aspirin
Hooker et al.⁴⁵, 1999	502	5	4	1
Kaempffe et al.⁵³, 1991	48	16	12	NA
Lieberman et al.⁶¹, 1994	113	6	0	1	Aspirin
Paiement et al.⁸¹, 1987	66	17	14	0
Paiement et al.⁸⁴, 1992	66	20	17	NA
Woolson¹³⁰, 1996	322	6	6	0

Add to My JBJS

TABLE II: RATES OF DEEP VENOUS THROMBOSIS ASSOCIATED WITH USE OF INTERMITTENT PNEUMATIC COMPRESSION AS REPORTED IN THE LITERATURE

*NA = not available.

Study	No. of Patients Managed with Intermittent Pneumatic Compression	Overall Rate of Deep Venous Thrombosis (percent)	Rate of Proximal Deep Venous Thrombosis* (percent)	Rate of Pulmonary Embolism* (percent)	Additional Prophylaxis
Bradley et al.⁹, 1993	30	7	3	NA	Heparin
Francis et al.³⁴, 1992	201	27	12	NA
Grady-Benson et al.³⁷, 1994	110	18	9	1	Aspirin
Hooker et al.⁴⁵, 1999	502	5	4	1
Kaempffe et al.⁵³, 1991	48	16	12	NA
Lieberman et al.⁶¹, 1994	113	6	0	1	Aspirin
Paiement et al.⁸¹, 1987	66	17	14	0
Paiement et al.⁸⁴, 1992	66	20	17	NA
Woolson¹³⁰, 1996	322	6	6	0