Combined (mechanical and pharmacological) modalities for the prevention of venous thromboembolism in joint replacement surgery

Transcription

1 S. K. Kakkos, D. Warwick, A. N. Nicolaides, G. P. Stansby, I. A. Tsolakis From University of Patras, Patras, Greece S. K. Kakkos, MD, MSc, PhD, Assistant Professor of Vascular Surgery I. A. Tsolakis, MD, PhD, Professor of Vascular Surgery University of Patras, Department of Vascular Surgery, University Hospital of Patras, Hippocrates Avenue, Rio, Patras 254, Greece. D. Warwick, MD, FRCS, FRCS (Orth), Reader in Orthopaedic Surgery University Hospital Southampton NHS Foundation Trust, Department of Orthopaedic Surgery, Tremona Road, Southampton SO1 YD, UK. A. N. Nicolaides, MS, FRCS, PhD, Emeritus Professor of Vascular Surgery Imperial College, Exhibition Road, London SW7 2AZ, UK. G. P. Stansby, MChir, FRCS, Professor of Vascular Surgery Northern Vascular Unit, Freeman Hospital, Freeman Road, High Heaton, Newcastle upon Tyne NE7 7DN, UK. Correspondence should be sent to Dr S. K. Kakkos; British Editorial Society of Bone and Joint Surgery doi:1.132/31-2x.94b $2. J Bone Joint Surg Br 212;94-B: INSTRUCTIONAL REVIEW: ARTHROPLASTY Combined (mechanical and pharmacological) modalities for the prevention of venous thromboembolism in joint replacement surgery We performed a systematic review and meta-analysis to compare the efficacy of intermittent mechanical compression combined with pharmacological thromboprophylaxis, against either mechanical compression or pharmacological prophylaxis in preventing deepvein thrombosis (DVT) and pulmonary embolism in patients undergoing hip or knee replacement. A total of six randomised controlled trials, evaluating a total of 1399 patients, were identified. In knee arthroplasty, the rate of DVT was reduced from 18.7% with anticoagulation alone to 3.7% with combined modalities (risk ratio (RR).27, p =.3; number needed to treat: seven). There was moderate, albeit non-significant, heterogeneity (I 2 = 42%). In hip replacement, there was a non-significant reduction in DVT from 8.7% with mechanical compression alone to 7.2% with additional pharmacological prophylaxis (RR.84) and a significant reduction in DVT from 9.7% with anticoagulation alone to.9% with additional mechanical compression (RR.17, p <.1; number needed to treat: 12), with no heterogeneity (I 2 = %). The included studies had insufficient power to demonstrate an effect on pulmonary embolism. We conclude that the addition of intermittent mechanical leg compression augments the efficacy of anticoagulation in preventing DVT in patients undergoing both knee and hip replacement. Further research on the role of combined modalities in thromboprophylaxis in joint replacement and in other high-risk situations, such as fracture of the hip, is warranted. In recent years there has been significant progress towards more effective and practical thromboprophylaxis in joint replacement surgery, to include injectable low-molecular-weight heparins (LMWH), 1 and now oral agents (selective Factor X inhibitors and oral direct thrombin inhibitors) such as dabigatran, rivaroxaban and apixaban. 2,3 Mechanical prophylaxis is also widely used in orthopaedic surgery because of a perceived lower risk of bleeding. Mechanical methods have been used alone, in simultaneous combination or in sequence prior to delayed anticoagulation, which are referred to as stacked modalities. 4 Combined prophylactic modalities have been shown to improve-up the efficacy of single modalities in a variety of specialties. 5 Although evidence pooled from several nonorthopaedic patient populations suggests an advantage for combined mechanical and chemical prophylaxis, it is not yet clear whether combined modalities are advantageous in all types of joint replacement surgery, whether they are better than either anticoagulants or mechanical compression used alone, nor whether they can prevent pulmonary embolism (PE). The aim of the present study was to perform a systematic review and meta-analysis of the efficacy of combined modalities in joint replacement and compare them with either intermittent mechanical compression or anticoagulants when used alone. Improved efficacy would justify the use of combined modalities in orthopaedic procedures and justify additional research if the evidence was lacking or incomplete. Materials and Methods Randomised controlled trials (RCTs) of combined intermittent mechanical compression and pharmacological interventions used to prevent venous thromboembolism in patients undergoing hip or knee replacement were identified through electronic searches of MEDLINE and SCOPUS, using mechanical or pneumatic compression as the search terms. Additional studies were identified through the reference list of the relevant trials found on MEDLINE and manual searches of journals and conference proceedings. Quasi-randomised trials were included. We excluded trials in which mechanical compression was used for a short period only or the use of VOL. 94-B, No., JUNE

2 73 S. K. KAKKOS, D. WARWICK, A. N. NICOLAIDES, G. P. STANSBY, I. A. TSOLAKIS Table I. Inclusions, procedures and outcome measures of the studies included in the meta-analysis (THR, total hip replacement; TKR, total knee replacement) Authors Included conditions Procedure group * group * Outcome measure Woolson and Watt 8 Osteoarthritis, osteonecrosis, Primary or revision rheuma- THR toid arthritis, failed hip operation and other Bradley et al 9 Osteoarthritis Primary unilateral THR Silbersack et al 12 Not stated Primary unilateral THR or TKR SCD, ES and warfarin SCD and ES Asymptomatic DVT, UFH, ES and FIT UFH and ES Asymptomatic DVT LMWH enoxaparin and sequential, rapid-inflation IPC Eisele et al 11 Not stated TKR or THR LMWH certoparin, ES and rapid-inflation IPC LMWH enoxaparin and ES LMWH certoparin and ES Symptomatic and asymptomatic DVT, Symptomatic and asymptomatic DVT Edwards et al 1 Not stated TKR or THR CECT and LMWH enoxaparin LMWH enoxaparin Symptomatic and asymptomatic DVT, Yokote et al 7 Not stated Primary unilateral cementless THR Fondaparinux, ES and IPC (one group) or enoxaparin, ES and IPC (second group) ES and IPC Symptomatic and asymptomatic DVT, * SCD, sequential compression device; ES, elastic stockings; UFH, unfractionated heparin; FIT, foot impulse technology; LMWH, low-molecular-weight heparin; IPC, intermittent pneumatic compression (includes SCD and FIT); CECT, portable, continuous enhanced circulation therapy DVT, deep-vein thrombosis; PE, pulmonary embolism combined modalities was not concurrent, such as isolated intra-operative use of mechanical compression, or where different agents were used in the two study groups. We also did not consider comparisons of regimens involving different combined modalities. All forms of intermittent mechanical compression were considered. Deep-vein thrombosis (DVT) and PE were extracted as separate endpoints, while short- and long-term outcomes were extracted separately. Statistical analysis. Separate analyses for PE and DVT and also the type of surgery (hip versus knee replacement) were performed. The relative risk or risk ratio (RR) was calculated and used for assessment of dichotomous outcomes using the Mantel-Haenszel method. Chi-squared tests and inconsistency (I 2 ) statistics were used to assess statistical heterogeneity. An I 2 value of % indicates no heterogeneity, while larger values are consistent with increasing heterogeneity. Random-effect instead of fixed effect models were considered when heterogeneity was moderate or substantial (i.e. > 3%). The number needed to treat was calculated for significant findings. Review Manager v5. (RevMan; The Cochrane Collaboration, Copenhagen, Denmark) was used for statistical analysis and generation of graphs. GraphPad QuickCalcs (Graph- Pad Software Inc., La Jolla, California) was used to calculate the number needed to treat in order to prevent one negative outcome (the reciprocal of the difference between the event rate in the two groups, also known as absolute risk reduction, multiplied by 1). Cross-tabulation statistics for long-term outcomes were performed with PASW Statistics 18 (SPSS Inc., Chicago, Illinois). A p-value <.5 was considered to be statistically significant. Results A total of six RCTs, evaluating a total of 1399 patients, were identified (Table I) Of these, one had two treatment arms testing different anticoagulants, sharing the same control group, and was therefore considered as two trials. 7 All six studies evaluated patients undergoing hip replacement, and three also evaluated patients undergoing knee replacement The mean age of the patients (in the five studies that reported on 81 patients 7-1,12 ) was 5.7 years (with a range of 29 to 9 in the three studies providing this information 9,1,12 ); which broke down to a mean of 5.4 years in hip arthroplasty 7-1 and 8.4 years in knee arthroplasty. 1 The method of randomisation was unclear in four RCTs, 7,1-12 the two studies that provided this information used the sealed envelope method, 8 or a quasi-randomisation method consisting of an even or odd date of birth. 9 In two of the three studies that studied both hip and knee replacement shown in the Table I, randomisation was stratified, 1,12 and not stated as being so in the third. 11 The intervention group in all studies used combined modalities; elastic stockings were additionally used in four studies. 7-9,11 Prophylactic methods in the control group included intermittent mechanical compression in two studies 7,8 and pharmacological prophylaxis in four studies 8-12 ; elastic stockings were additionally used in five of them. 7-9,11,12 Ultrasound was the main diagnostic modality to diagnose DVT used in five trials, 7,8,1-12 while in two studies venography was employed, 8,9 and was used in combination with ultrasound in some patients in one of these. 8 These tests were performed at patient discharge and up to the twelfth post-operative day. A total of 13 patients (.93%) were excluded from the analysis. One study excluded THE JOURNAL OF BONE AND JOINT SURGERY

3 COMBINED (MECHANICAL AND PHARMACOLOGICAL) MODALITIES FOR THE PREVENTION OF VENOUS THROMBOEMBOLISM 731 Edwards et al 1 (knee) Silbersack et al 12 (knee) %.34 [.1, 8.1] Not estimable 111 % Heterogeneity: not applicable Test for overall effect: z =.7 (p =.5) % 1.%.34 [.1, 8.1] Risk reduction % Fig. 1 Forest plot showing the incidence of pulmonary embolism in the treatment group (combined modalities) and control group (anticoagulation) in total knee replacement. 1,12 A Mantel-Haenszel (M-H) method was employed with 95% confidence intervals (CI). Edwards et al 1 (knee) Silbersack et al 12 (knee) Eisele et al 11 (knee) % 15.1% 3.%.31 [.11,.91].3 [.,.5].51 [.12, 2.2] M-H, Random, 95% Cl % 7 3.7% % Heterogeneity: Tau 2 =.48; Chi 2 = 3.4, df = 2 (p =.18), l 2 = 42% Test for overall effect: z = 2.14 (p =.3).27 [.8,.89] Risk reduction 73% Fig. 2 Forest plot showing the incidence of deep-vein thrombosis (DVT) in the treatment group (combined modalities) and control group (anticoagulation) in total knee replacement The additional use of intermittent mechanical compression reduced significantly the incidence of DVT. A Mantel-Haenszel (M-H) method was employed with 95% confidence intervals (CI). eight patients because of non-compliance, confinement to bed for longer than one week, premature transfer to a different institution or re-operation or discharge from hospital without ultrasonography, 12 and another one excluded five patients because they missed their ultrasound evaluation. 7 PE was diagnosed mainly with scintigraphy 8 or CT pulmonary angiogram. 7,12 In order to assess the long-term efficacy, follow-up was prolonged in three studies for about three months. 7,1,12 Blinding of treatment allocation was not possible for studies that tested intermittent mechanical compression (on background anticoagulation), since no control (sham) compression device was used. Regarding the studies that tested anticoagulation on background compression, one used a placebo. In the entire cohort of six publications, the radiologists who read the venograms or performed the ultrasound scans were blinded to patient allocation in two studies. Short-term outcomes in knee replacement. There were no studies in the compression + anticoagulant versus compression alone category. In the compression + anticoagulant versus anticoagulant alone there were two trials exploring the effect of the additional use of compression compared with anticoagulation alone in prevention of PE. 1,12 This outcome occurred with a similar frequency of % and.89%, respectively (p =.5, test for overall effect) (Fig. 1), while in three trials exploring the effect of the additional use of compression in prevention of DVT, 1-12 this outcome was reduced from 18.7% in the control group (anticoagulation) to 3.7% in the combined modalities group (Fig. 2). There was moderate but non-significant heterogeneity (I 2 = 42%, p =.18), while the risk ratio for DVT was.27. The number needed to treat was seven (95% confidence interval (CI) 4.7 to 11.8). Short-term outcomes in hip replacement. For compression + anticoagulant versus compression only there were three trials exploring the effect of the additional use of anticoagulation, 7,8 and no PE was reported, while the rate of DVT was non-significantly reduced from 8.7% in the control group (IMC) to 7.2% in the combined modalities group (Fig. 3). There was no heterogeneity (I 2 = %, p =.92), while the risk ratio for DVT was.84. For compression + anticoagulant versus anticoagulant only there were four trials exploring the effect of the additional use of intermittent compression 9-12 ; no PE was reported, while the rate of DVT was reduced from 9.71% in the control group (anticoagulation) to.94% in the VOL. 94-B, No., JUNE 212

4 732 S. K. KAKKOS, D. WARWICK, A. N. NICOLAIDES, G. P. STANSBY, I. A. TSOLAKIS Yokote et al 7 (enoxaparin) Woolson and Watt 8 Yokote et al 7 (fondaparinux) % 41.% 29.3%.93 [.2, 2.2].73 [.28, 1.9].99 [.33, 2.94] % % Heterogeneity: Chi 2 =.1, df = 2 (p =.92), l 2 = % Test for overall effect: z =.57 (p =.57) 1.%.84 [.45, 1.55] Risk reduction 1% Fig. 3 Forest plot showing the incidence of deep-vein thrombosis (DVT) in the treatment group (combined modalities) and control group of intermittent mechanical compression (IMC) in total hip replacement. 7,8 The additional use of anticoagulation reduced the incidence of DVT but the difference was not non-significant. A Mantel-Haenszel (M-H) method was employed with 95% confidence intervals (CI). Bradley et al 9 Edwards et al 1 (hip) Silbersach et al 12 (hip) Eisele et al 11 (hip) % 8.5% 19.% 32.%.24 [., 1.1].45 [.4, 4.88].9 [.1, 1.9].5 [.,.83] % % Heterogeneity: Chi 2 = 1.84, df = 3 (p =.1), l 2 = % Test for overall effect: z = 3.51 (p =.4) 1.%.17 [.,.4] Risk reduction 83% Fig. 4 Forest plot showing the incidence of deep-vein thrombosis (DVT) in the treatment group (combined modalities) and control group (anticoagulation) in total hip replacement The additional use of intermittent mechanical compression (IMC) reduced significantly the incidence of DVT. A Mantel-Haenszel (M-H) method was employed with 95% confidence intervals (CI). combined modalities group (Fig. 4). There was no heterogeneity (I 2 = %, p =.1), while the risk ratio for DVT was.17. The number needed to treat was 12 (95% CI 7.9 to 2.3). Long-term outcomes. Because of the small number of events, no subgroup analysis between hip and knee replacement was carried out. The incidence of PE was one of 24 (.4%) with combined modalities versus of 1 (%) with single modalities (p >.999, Fisher s exact test). The incidence of DVT was 4 of 529 (.7%) for the entire cohort. After excluding one study, which used extended LMWH but failed to provide the information in which study group a DVT occurred, the incidence of DVT was 2 of 24 (.7%) with combined modalities and 1 of 1 (.3%) with single modalities (p >.999, Fisher s exact test). Discussion In the present meta-analysis we demonstrated that the additional use of intermittent compression augments the efficacy of anticoagulation in preventing DVT in both hip and knee replacement. Our findings support the recommendations of the Consensus guidelines that high-risk patients should use combined modalities. 13,14 The improved protection found with intermittent compression added to anticoagulation is consistent with observations in similar high-risk groups as shown in recent reviews, 5,15 which included cardiothoracic, 1 general surgery, 17 urology, 18,19 neurosurgery, 2 and neurology patients. 21,22 The improved efficacy of combined modalities is most likely because of an enhanced ability to ameliorate venous stasis, coagulopathy and endothelial injury, all of which promote thrombo-embolism, as first described by Virchow in ; Rosendaal 24 updated this theory a few years ago by proposing a multiple hit model. Intermittent compression reduces stasis by increasing venous velocity and flow, 25,2 stimulating endogenous fibrinolytic activity, 27,28 increasing the plasma levels of tissue factor pathway inhibitors 29 and possibly releasing nitrous oxide. 3 Unfractionated and lowmolecular weight heparin affect different aspects of Virchow s triad by inhibiting Factors II and X thus probably acting synergistically. Given the selective mode of anticoagulants, the broader action of mechanical compression explains the clinical benefit of the combined modalities. We were intrigued to find that the background efficacy of compression is not enhanced by the subsequent addition of anticoagulation, which has also been observed in general surgery. 31 This might also be explained by a selective effect although further studies and consideration are required on this. THE JOURNAL OF BONE AND JOINT SURGERY

5 COMBINED (MECHANICAL AND PHARMACOLOGICAL) MODALITIES FOR THE PREVENTION OF VENOUS THROMBOEMBOLISM 733 Also, certain operative manoeuvres like those during hip replacement cause venous obstruction as well as local venous trauma, 31 which may contribute to the high incidence of post-operative DVT; blood cell debris can be detected with ultrasound during hip replacement, and might play an active role as possibly constituting a link between blood stasis and DVT. Limitations of our study include the variety of intermittent mechanical compression devices used in the trials that were included. There are insufficient data to conclude which compression device provides the best prophylaxis in conjunction with anticoagulation, although Figures 2 and 4 suggest that sequential, mechanical leg compression is better than other forms of mechanical compression, like continuous enhanced circulation therapy (CECT) and foot compression. This reduction in protection might be the result of patient non-compliance, 32,33 inappropriate sleeve application in single centre studies, 34 or because certain types of device are not sufficiently protective, perhaps due to previous findings showing that the greater the volume of tissue compressed, the greater the fibrinolytic response. 3 Furthermore, the duration of the augmented peak venous velocity in high pressure pumps, including foot pumps, is short lasting. Additional limitations include the lack of sham devices and also placebo use in one study. Future studies should therefore have improved design. Future studies should also address cost-effectiveness, which would include looking at benefits in terms of reduced hospital stay, rehabilitation, mortality and the long term burden of the post-thrombotic syndrome. 15 The effect of adding anticoagulation to compression alone should be further studied. Far larger studies, perhaps unfeasibly big, would be needed to establish the effect of combined modalities on reducing PE. Research in other high-risk situations for which there is considerable potential for combined methods like hip fracture is limited. 11,37,38 Further research should also focus on the efficacy of combining intermittent compression with the more recent potent anticoagulants such as fondaparinux, Factor Xa inhibitors and antithrombin agents. 2,25,39-41 Finally, the use of early mechanical and subsequent chemical combinations ( stacked modalities ) requires further consideration to capture the benefits of anticoagulation yet minimise their bleeding complications. 42,43 Modern mechanical compression devices are practically noiseless, smaller and portable, with the intention to improve patient compliance. In conclusion, the additional use of intermittent mechanical compression augments the efficacy of anticoagulation in preventing DVT in both hip and knee arthroplasty, but no significant effect was seen when anticoagulation is added to compression. Further research on the role of combined modalities in the prevention of DVT and PE in patients undergoing joint arthroplasty and in other highrisk situations, such as fracture of the hip, is warranted. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. References 1. Jameson SS, Charman SC, Gregg PJ, Reed MR, van der Meulen JH. The effect of aspirin and low-molecular-weight heparin on venous thromboembolism after hip replacement: a non-randomised comparison from information in the National Joint Registry. J Bone Joint Surg [Br] 211;93-B: Eriksson BI, Kakkar AK, Turpie AG, et al. Oral rivaroxaban for the prevention of symptomatic venous thromboembolism after elective hip and knee replacement. J Bone Joint Surg [Br] 29;91-B: Raskob GE, Gallus AS, Pineo GF, et al. Apixaban versus enoxaparin for thromboprophylaxis after hip or knee replacement: pooled analysis of major venous thromboembolism and bleeding in 844 patients from the ADVANCE-2 and ADVANCE-3 trials. J Bone Joint Surg [Br] 212;94-B: Warwick D. New concepts in orthopaedic thromboprophylaxis. J Bone Joint Surg [Br] 24;8-B: Kakkos SK, Caprini JA, Geroulakos G, et al. Combined intermittent pneumatic leg compression and pharmacological prophylaxis for prevention of venous thromboembolism in high-risk patients. Cochrane Database Syst Rev 28;4:CD Roderick P, Ferris G, Wilson K, et al. Towards evidence based guidelines for the prevention of venous thromboembolism: systematic review of mechanical methods, oral anticoagulation, dextran and regional anaesthesia as thromboprophylaxis. Health Technol Assess 25;9: Yokote R, Matsubara M, Hirasawa N, et al. Is routine chemical thromboprophylaxis after total hip replacement really necessary in a Japanese population? J Bone Joint Surg [Br] 211;93-B: Woolson ST, Watt JM. Intermittent pneumatic compression to prevent proximal deep venous thrombosis during and after total hip replacement: a prospective, randomized study of compression alone, compression and aspirin, and compression and low-dose warfarin. J Bone Joint Surg [Am] 1991;73-A: Bradley JG, Krugener GH, Jager HJ. The effectiveness of intermittent plantar venous compression in prevention of deep venous thrombosis after total hip arthroplasty. J Arthroplasty 1993;8: Edwards JZ, Pulido PA, Ezzet KA, et al. Portable compression device and lowmolecular-weight heparin compared with low-molecular-weight heparin for thromboprophylaxis after total joint arthroplasty. J Arthroplasty 28;23: Eisele R, Kinzl L, Koelsch T. Rapid-inflation intermittent pneumatic compression for prevention of deep venous thrombosis. J Bone Joint Surg [Am] 27;89-A: Silbersack Y, Taute BM, Hein W, Podhaisky H. Prevention of deep-vein thrombosis after total hip and knee replacement: low-molecular-weight heparin in combination with intermittent pneumatic compression. J Bone Joint Surg [Br] 24;8- B: No authors listed. Prevention and treatment of venous thromboembolism: International Consensus Statement (guidelines according to scientific evidence). Int Angiol 2;25: Geerts WH, Bergqvist D, Pineo GF, et al. Prevention of venous thromboembolism: American college of chest physicians evidence-based clinical practice guidelines (8th Edition). Chest 28;133(Suppl): Kakkos SK, Caprini JA, Geroulakos G, et al. Can combined (mechanical and pharmacological) modalities prevent fatal VTE? Int Angiol 211;3: Ramos R, Salem BI, De Pawlikowski MP, et al. The efficacy of pneumatic compression stockings in the prevention of pulmonary embolism after cardiac surgery. Chest 199;19: Borow M, Goldson HJ. Prevention of postoperative deep venous thrombosis and pulmonary emboli with combined modalities. Am Surg 1983;49: Bigg SW, Catalona WJ. Prophylactic mini-dose heparin in patients undergoing radical retropubic prostatectomy: a prospective trial. Urology 1992;39: Sieber PR, Rommel FM, Agusta VE, et al. Is heparin contraindicated in pelvic lymphadenectomy and radical prostatectomy? J Urol 1997;158(Pt 1): Frim DM, Barker FG 2nd, Poletti CE, Hamilton AJ. Postoperative low-dose heparin decreases thromboembolic complications in neurosurgical patients. Neurosurgery 1992;3: Kamran SI, Downey D, Ruff RL. Pneumatic sequential compression reduces the risk of deep vein thrombosis in stroke patients. Neurology 1998;5: Winemiller MH, Stolp-Smith KA, Silverstein MD, Therneau TM. Prevention of venous thromboembolism in patients with spinal cord injury: effects of sequential pneumatic compression and heparin. J Spinal Cord Med 1999;22: Virchow R. Phlogose und thrombose im gefässsystem. In: Virchow R, ed. Gesammelte Adhandlungen zur Wissenschaftlichen Medicin. Frankfurt: Von Meidinger Sohn, 185:458 3 (in German). 24. Rosendaal FR. Venous thrombosis: a multicausal disease. Lancet 1999;3: Kakkos SK, Griffin M, Geroulakos G, Nicolaides AN. The efficacy of a new portable sequential compression device (SCD Express) in preventing venous stasis. J Vasc Surg 25;42: VOL. 94-B, No., JUNE 212

6 734 S. K. KAKKOS, D. WARWICK, A. N. NICOLAIDES, G. P. STANSBY, I. A. TSOLAKIS 2. Kakkos SK, Szendro G, Griffin M, Sabetai MM, Nicolaides AN. Improved hemodynamic effectiveness and associated clinical correlations of a new intermittent pneumatic compression system in patients with chronic venous insufficiency. J Vasc Surg 21;34: Comerota AJ, Chouhan V, Harada RN, et al. The fibrinolytic effects of intermittent pneumatic compression: mechanism of enhanced fibrinolysis. Ann Surg 1997;22: Kosir MA, Schmittinger L, Barno-Winarski L, et al. Prospective double-arm study of fibrinolysis in surgical patients. J Surg Res 1998;74: Chouhan VD, Comerota AJ, Sun L, et al. Inhibition of tissue factor pathway during intermittent pneumatic compression: a possible mechanism for antithrombotic effect. Arterioscler Thromb Vasc Biol 1999;19: Liu K, Chen LE, Seaber AV, Johnson GW, Urbaniak JR. Intermittent pneumatic compression of legs increases microcirculation in distant skeletal muscle. J Orthop Res 1999;17: Roberts VC, Cotton LT. Failure of low-dose heparin to improve efficacy of peroperative intermittent calf compression in preventing postoperative deep vein thrombosis. Br Med J 1975;3: Comerota AJ, Katz ML, White JV. Why does prophylaxis with external pneumatic compression for deep vein thrombosis fail? Am J Surg 1992;14: Anglen JO, Goss K, Edwards J, Huckfeldt RE. Foot pump prophylaxis for deep venous thrombosis: the rate of effective usage in trauma patients. Am J Orthop (Belle Mead NJ) 1998;27: Salvian AJ, Baker JD. Effects of intermittent pneumatic calf compression in normal and postphlebitic legs. J Cardiovasc Surg (Torino) 1988;29: Norgren L, Toksvig-Larsen S, Magyar G, Lindstrand A, Albrechtsson U. Prevention of deep vein thrombosis in knee arthroplasty: preliminary results from a randomized controlled study of low molecular weight heparin vs foot pump compression. Int Angiol 1998;17: Tarnay TJ, Rohr PR, Davidson AG, et al. Pneumatic calf compression, fibrinolysis, and the prevention of deep venous thrombosis. Surgery 198;88: Westrich GH, Rana AJ, Terry MA, et al. Thromboembolic disease prophylaxis in patients with hip fracture: a multimodal approach. J Orthop Trauma 25;19: Mehta KV, Lee HC, Loh JS. Mechanical thromboprophylaxis for patients undergoing hip fracture surgery. J Orthop Surg (Hong Kong) 21;18: Eriksson BI, Bauer KA, Lassen MR, Turpie AG. Fondaparinux compared with enoxaparin for the prevention of venous thromboembolism after hip-fracture surgery. N Engl J Med 21;345: Eriksson BI, Dahl OE, Rosencher N, et al. Oral dabigatran etexilate vs. subcutaneous enoxaparin for the prevention of venous thromboembolism after total knee replacement: the RE-MODEL randomized trial. J Thromb Haemost 27;5: Eriksson BI, Borris LC, Friedman RJ, et al. Rivaroxaban versus enoxaparin for thromboprophylaxis after hip arthroplasty. N Engl J Med 28;8: Eriksson BI, Friedman RJ, Cushner FD, Lassen MR. Potent anticoagulants are associated with a higher all-cause mortality rate after hip and knee arthroplasty. Clin Orthop Relat Res 28;4: Eikelboom JW, Quinlan DJ, O'Donnell M. Major bleeding, mortality, and efficacy of fondaparinux in venous thromboembolism prevention trials. Circulation 29;12: THE JOURNAL OF BONE AND JOINT SURGERY