Deep vein thrombosis after reconstructive shoulder arthroplasty: A prospective observational study

J Shoulder Elbow Surg (2009) 18, 100-106

www.elsevier.com/locate/ymse

Deep vein thrombosis after reconstructive shoulder arthroplasty: A prospective observational study Andrew A. Willis, MDa, Russell F. Warren, MDb,*, Edward V. Craig, MDb, Ronald S. Adler, MD, PhDb, Frank A. Cordasco, MDb, Stephen Lyman, PhDb, Stephen Fealy, MDb a b

Tri-County Orthopedic & Sports Medicine, P.A., Morristown, NJ Sports Medicine and Shoulder Service, Hospital for Special Surgery, Weill Cornell Medical Center, New York, NY Summary This clinical study was performed to document the prevalence of deep vein thrombosis (DVT) after prosthetic shoulder replacement surgery. We prospectively followed 100 consecutive shoulder arthroplasty procedures (total shoulder replacement in 73 and hemiarthroplasty in 27) in 44 male and 56 female patients for 12 weeks (mean age, 67 years; range, 17-88 years). Risk factors for venous thromboembolic disease were assessed preoperatively and postoperatively. A 4-limb surveillance color flow Doppler ultrasound was performed at 2 days (100 patients) and 12 weeks (50 patients randomly selected) after surgery, and the presence and location of DVT were recorded. Postoperative symptomatic or fatal pulmonary emboli (PE) were also recorded. The overall prevalence of DVT was 13.0%, consisting of 13 DVTs in 12 patients. These included 6 ipsilateral and no contralateral upper extremity DVTs and 5 ipsilateral and 2 contralateral lower extremity DVTs. The prevalence of DVT was 10.0% (10/100) at day 2 after surgery and 6.0% (3/50) at week 12 after surgery. The incidence of symptomatic nonfatal PE was 2.0% (2/100), and that of fatal PE was 1.0% (1/100). Risk factors associated with venous thromboembolic disease did not reach statistical significance because of the small study population sample size. At our institution, the prevalence of DVT after reconstructive shoulder arthroplasty was 13.0%, a rate comparable to that after hip arthroplasty (10.3%) but lower than that after knee arthroplasty (27.2%). Shoulder arthroplasty surgeons should be aware of the potential risk of perioperative thromboembolic complications in both the acute and subacute postoperative periods. Ó 2009 Journal of Shoulder and Elbow Surgery Board of Trustees.

Prosthetic shoulder replacement has gained widespread use in the treatment of a variety of shoulder disorders since its introduction by Neer.16 Though performed far less frequently than hip or knee replacement surgery, shoulder arthroplasty is not uncommon. In 1998, there were 17,468 This study received the Charles S. Neer Award at the annual meeting of the American Shoulder and Elbow Surgeons; March 2006; Chicago, IL. Grant support was provided by the Sports Medicine Institute, Hospital for Special Surgery. *Reprint requests: Russell F. Warren, MD, Hospital for Special Surgery, 535 E 70th St, New York, NY 10021. E-mail address: [email protected] (R.F. Warren).

shoulder replacements performed in the United States.10 Unlike arthroplasty of the hip or knee, the prevalence of deep vein thrombosis (DVT) after shoulder arthroplasty is unknown. The frequency and potential morbidity of venous thromboembolic (VTE) disease in this patient population have not been clearly defined. Given the paucity of literature, clinical guidelines for managing VTE risk in patients after shoulder arthroplasty are not currently available. The purpose of this study was to document prospectively the prevalence of DVT after shoulder arthroplasty in 100 consecutive patients at a single center. This investigation details the presence, location, frequency, timing, and

1058-2746/2009/$36.00 - see front matter Ó 2009 Journal of Shoulder and Elbow Surgery Board of Trustees. doi:10.1016/j.jse.2008.07.011

Deep vein thrombosis severity of DVT. To our knowledge, this is the first report providing exact data about these specific issues. We hypothesized that the prevalence of DVT in patients after shoulder arthroplasty would be greater than the published prevalence of DVT in the general population. We also hypothesized that the prevalence of DVT in patients after shoulder arthroplasty would be less than the published prevalence of DVT in patients after hip and knee arthroplasty at our institution. Additional goals of this study were to (1) describe the frequency and morbidity of associated thromboembolic events occurring in a series of patients undergoing shoulder replacement during the acute and subacute perioperative period and (2) characterize individuals who may be at high risk of the development of VTE complications after shoulder arthroplasty based on historical and operative factors.

Materials and methods A prospective, observational study was designed to (1) document the prevalence of DVT after shoulder arthroplasty in 100 consecutive patients at our center, (2) describe associated VTE complications in this patient population, and (3) identify and characterize potential high-risk individuals based on specific historical and operative factors. This study was approved by the orthopaedic research committee as well as the institutional review board at our center. All patients undergoing total shoulder replacement or hemiarthroplasty of the shoulder at our institution were eligible for inclusion in this study. To minimize potential adverse and confounding effects associated with low surgeon volume, we included only patients of surgeons who perform more than 5 shoulder arthroplasties per year.13 Patients receiving routine preoperative anticoagulation, such as heparin or warfarin; patients with active thromboembolic disease; and patients who declined to participate in the study were excluded. Between August 2003 and August 2004, 124 consecutive shoulder arthroplasties were performed at our institution. Of these, 100 met the study criteria, and all 100 were voluntarily enrolled into the study after informed written consent was obtained. There were 73 total shoulder replacements and 27 hemiarthroplasties performed in 44 male patients and 56 female patients at a mean age of 67 years (range, 17-88 years). The most common indication for surgery was osteoarthritis, and the majority of cases were primary procedures (Table I). All procedures were performed by use of a standard deltopectoral approach with the patient in the semi-reclined, beachchair position under interscalene block regional anesthesia. The implant design and method of fixation were based on surgeon preference. No perioperative DVT prophylaxis such as antiembolic stockings, pneumatic compression devices, or intraoperative heparin administration was used in this series. After surgery, DVT prophylaxis was initiated in all patients and consisted of enteric coated aspirin (325 mg twice a day), unless medically contraindicated (25 patients); pneumatic compression foot pumps; and early ambulation. In light of our past experience, as well as a recent report by Rockwood et al20 detailing significant thromboembolic complications in patients

101 Table I

Demographics of study patient population Overall

Total shoulder Shoulder replacement hemiarthroplasty

Age (y) [mean 67 (17-88) 68 (26-85) (range)] Male/female 44:56 31:42 Diagnosis OA 62 54 RA 9 7 AVN 1 1 Post-traumatic 7 6 Acute trauma 9 0 Other 12 5 Primary surgery 94 70 Revision surgery 6 3

65 (17-88) 13:14 8 2 0 1 9 7 24 3

OA, Osteoarthritis; RA, rheumatoid arthritis; AVN, avascular necrosis.

after reconstructive shoulder arthroplasty, we did not believe it was ethical to withhold postoperative DVT prophylaxis in this study. The prophylactic measures used were specifically selected to match those in previously published studies documenting the prevalence of DVT after hip and knee arthroplasty.29,31 This was done so that the clinical significance of our findings could be assessed and interpreted relative to the well-established VTE risk associated with hip and knee replacement surgery. A 4-limb surveillance color flow Doppler ultrasound was used to determine the prevalence of DVT. All tests were performed and interpreted by a single board-certified musculoskeletal radiologist specializing in ultrasonography. Color flow Doppler ultrasound has been shown to be an accurate and effective screening modality for both upper and lower extremity DVT.3,12 Although most thromboembolic events after elective orthopaedic surgery develop in the first postoperative week, DVTand related thromboembolic complications have been reported up to 12 weeks after lower extremity implant arthroplasty.22,31 As a result, a surveillance color flow Doppler ultrasound of both upper and lower extremities was performed in all patients on the second day and, because of cost restrictions, in 50 randomly selected patients at 12 weeks after surgery. The presence and location of DVT in both the acute and subacute postoperative period were assessed and recorded. A screening surveillance Doppler ultrasound was not performed before surgery. Preoperatively, all patients were interviewed and examined on the day of surgery. Numerous potential risk factors associated with VTE disease were assessed and recorded in each patient (Table II).1,4,11,20 Operative time, additional procedures, and surgical complications were also recorded. Postoperatively, time to ambulation, length of stay, and any associated VTE complication, such as symptomatic or fatal pulmonary embolism occurring during the index hospitalization, were noted. All patients were contacted by telephone 12 weeks after surgery, and any subacute thromboembolic complication occurring after hospital discharge was investigated and documented.

Statistical analysis All statistical analyses were performed with SPSS computer software (SPSS, Chicago, IL) by use of c2 analysis, Fisher exact

102

A.A. Willis et al.

Table II Potential preoperative, operative, and postoperative VTE disease risk factors assessed in study population Potential VTE risk factors

No. of study patients/result

Preoperative assessment Age <50 y 9 50-59 y 14 60-69 y 23 70-79 y 39 80 y 15 Race White 92 Black 4 Hispanic 4 Other 0 Body mass index <18.5 2 18.5-24.9 23 25-29.9 35 30 40 Medical comorbidities Hypertension 51 Cardiovascular disease 16 Diabetes 11 Venous stasis disease 5 Active cancer 2 (prostate) History of cancer 19 History of stroke 1 Liver disease 1 Kidney disease 3 Inflammatory bowel disease 8 History of recent long bone 1 fracture History of DVT 8 History of pulmonary embolism 2 History of thrombophilia 1 (lupus anticoagulant) Limited independent mobility 2 Family history of VTE disease 7 Active smoker 8 History of recent prolonged or 14 distant travel History of recent hospitalization 2 (diverticulitis) History of recent major surgery (within 5 y) Neurologic 0 Pelvic 4 Orthopaedic 22 Gynecologic or obstetric 3 Risk factor assessment score of Caprini et al4 Low (0-1) 14 Moderate (2-4) 63 High (>4) 23 Operative and postoperative assessment Mean operative time (min) 127.5 No. of patients requiring 34 additional procedures (continued)

Table II (continued ) Potential VTE risk factors No. of intraoperative surgical complications Mean time to ambulation after surgery (d) Mean length of hospital stay (d)

No. of study patients/result 2 1.3 3.3

test, or Student t test. An a value was set at .05. No corrections were made for multiple comparisons.

Results The prevalence of DVT after shoulder arthroplasty was 13% in this series (Figure 1). We documented 10 DVTs (prevalence, 10%) in 9 patients by surveillance Doppler ultrasound on the second postoperative day. An additional 3 new DVTs (incidence, 6%) were documented in 3 patients 12 weeks after surgery. DVT was documented in both the upper and lower extremities and was detected in both the acute and subacute postoperative periods (Figure 2). All upper extremity DVTs were found in the operative extremity (Figure 3). Five DVTs involved the axillary vein and extended proximally to the subclavian and internal jugular veins, whereas one upper extremity DVT extended distally and involved the brachial vein. There were 5 ipsilateral and 2 contralateral lower extremity DVTs documented. Two DVTs involved the popliteal vein and extended proximally, whereas five DVTs involved the posterior tibial and peroneal veins of the calf. Pulmonary embolism occurred in 3 patients (Figure 4). Symptomatic nonfatal pulmonary emboli were documented by spiral computed tomography (CT) scan in 2 patients. One presented with complaints of acute-onset dyspnea and chest discomfort on the second day after a total shoulder replacement for osteoarthritis. A spiral CT scan showed bilateral massive pulmonary emboli. A 4-limb surveillance Doppler ultrasound performed when the patient was in stable condition on the third postoperative day was significant for a posterior tibial DVT, the presumed source of the VTE event. Another patient presented with a fluctuating tachyarrhythmia beginning on the first postoperative day after a hemiarthroplasty and glenoid biologic resurfacing procedure for rheumatoid arthritis. Because of a persistent tachyarrhythmia and low oxygen saturation on room air, a spiral CT scan was performed on the fourth postoperative day and showed right upper lobe and right lower lobe segmental pulmonary emboli. A 4-limb surveillance Doppler ultrasound performed at that time showed a large subclavian vein thrombosis, the presumed source of the VTE event. Both patients were transferred to the intensive care unit, treated with systemic

Deep vein thrombosis

103

13/100

14 12

Number of DVTs

10/100

10 8 6 3/50

4 2 0 Overall

POD #2

Week 12

Figure 1 Prevalence of DVT after reconstructive shoulder arthroplasty. POD, Postoperative day.

population risk of DVT, and the prevalence of DVT after hip and knee replacement surgery at our institution. The 13% prevalence of DVT in this study was statistically greater than the 0.05% overall general population prevalence reported by Fowkes et al8 in a recent metaanalysis of the literature (P < .0001) (Figure 5). Our prevalence of DVT was also statistically greater than the prevalence of DVT in age-matched population control subjects (Figure 6).17 It was not significantly different than the 10.3% prevalence of DVT after hip replacement surgery at our institution (P ¼ .4) (Figure 7).29 However, the 27.0% prevalence of DVT after knee replacement surgery at our institution was significantly greater than that found after shoulder replacement (P ¼ .02) (Figure 7).31

Discussion 8

Number of DVTs

7

7 6

6

6 5

4

4 3

2

2

1

1 0 Overall

POD #2

Upper Extremity

Week 12

Lower Extremity

Figure 2 Presence, location, and timing of DVT after reconstructive shoulder arthroplasty. POD, Postoperative day.

anticoagulation, and ultimately discharged in stable condition 14 days and 10 days, respectively, after surgery. Sudden respiratory distress developed after a Valsalva maneuver in a third patient while in the bathroom approximately 7 weeks after a shoulder hemiarthroplasty for a traumatic 4-part proximal humeral fracture. The patient’s wife noted that he quickly became pale, dyspneic, diaphoretic, and unresponsive. All resuscitation efforts failed, and he died rapidly. His death was attributed to cardiovascular collapse from a massive pulmonary embolism. Autopsy was declined by the family. The initial 4-limb color flow surveillance Doppler ultrasound performed on postoperative day 2 was negative for DVT. Because of the small study population size, we were unable to characterize high-risk individuals based on specific historical or operative factors. Increasing patient age, weight, prolonged operative time, and previous history of VTE disease trended toward significance. To assess and interpret the clinical relevance of our findings in this study, we compared our results with previously published data in the literature detailing the prevalence of DVT in the general population, the age-related

VTE disease is a common medical condition and a growing public health concern, with a mean annual incidence of more than 1 case per 1,000 person-years.23 It can be potentially lethal as a result of pulmonary emboli, one quarter of which present as sudden death.11 Secondary complications, such as postphlebitic syndrome and pulmonary hypertension, contribute to long-term morbidity and death. Venous thrombosis is one of the most frequent complications after elective orthopaedic surgery. Patients who undergo major lower extremity orthopaedic surgery, particularly hip and knee arthroplasty, are among those considered to be at greatest risk for venous thromboembolism.9 Numerous studies have described an increased incidence of DVT and pulmonary embolism in this patient population as well as defined risk factors for and mortality rate from venous thromboembolism after hip and knee arthroplasty.5,6,26,28,30 Given the silent nature of VTE disease, its widespread prevalence, its rapid onset, and its potentially catastrophic consequences, prevention rather than treatment is considered necessary in patients undergoing hip and knee arthroplasty. Postoperative thromboprophylaxis with pharmacologic agents is widely advocated to improve survival rates and reduce health care costs.7,9 Although extensive research has been done on VTE disease after hip and knee arthroplasty, there is very little information concerning DVT and pulmonary embolism after shoulder arthroplasty.2,14,20,21,24 Currently, objective clinical guidelines to help assess and manage VTE risk after shoulder replacement do not exist. This is the first study to report the prevalence of DVT after reconstructive shoulder arthroplasty. Overall, the 13% prevalence in this study was higher than anticipated and statistically greater than the general population prevalence of DVT as reported by Fowkes et al8 in a recent metaanalysis of the literature. We postulated that this higher prevalence of DVT may be because of several factors. First, the patient population undergoing this procedure is generally older. Age has been shown to be a significant risk

104

A.A. Willis et al.

Figure 3 Color flow Doppler ultrasound image of a patent jugular vein in a study patient on postoperative day 2 (left) and at 12 weeks after reconstructive shoulder arthroplasty with DVT now present (right).

Incidence PE = 3.0% Number of VTE Complications

3.5 3/100

Nonfatal PE:

3

POD#2: POD#4:

(Presumed Source) UE DVT LE DVT

2.5 2/100

Mortality: PE @ 7 weeks No DVT @ POD#2

2 1.5

1/100

1 0.5 0 Pulmonary Embolism

Figure 4

14

Fatal PE

VTE complications after reconstructive shoulder arthroplasty. PE, Pulmonary embolism; POD, postoperative day.

30

13%

12

p < 0.0001

10 8 6 4

26.7%

25 21.7%

Percentage

Percentage

Symptomatic PE

20 p=0.001

p<0.0001

p<0.0001

p=0.002

15 10

7.7%

7.1%

5 0.29%

2

0.05%

0 Shoulder Arthroplasty Study Population

General Population

Figure 5 Prevalence of DVT after reconstructive shoulder arthroplasty compared with general population prevalence of DVT.

factor for VTE disease.1 Second, the axillary vein may be traumatized or kinked when the humerus is rotated repeatedly or positioned during the procedure, increasing the chance of intimal damage to the vessel. Third, these patients may undergo a physiologic stress response resulting in an increase in coagulation factors in response to intramedullary instrumentation and use of acrylic bone cement.18,25 Fourth, shoulder arthroplasty requires the patient to be in an upright, nearly seated position for an extended period of time, which may result in significant

0.33%

0.56%

0.82%

0 50-59

60-69

70-79

> 79

(Age Range) Shoulder Arthroplasty Study Population

General Population

Figure 6 Prevalence of DVT after reconstructive shoulder arthroplasty compared with age-matched population control subjects.

pooling of blood in the lower extremities because of the lack of muscular pumping. Long surgical procedures have also been associated with hematologic alterations in the clotting cascade and the systemic release of thrombogenic factors.15 To determine whether the increased prevalence of DVT was simply a factor of the older age of this patient population, we calculated the prevalence of DVT by decade in our series and compared these findings with age-matched population control subjects.17 We found that age alone

Deep vein thrombosis

105

30.0

p=0.02

25.0 27.2% (22/81)

Percentage

p=0.40

20.0 15.0

13.0% (13/100)

10.3% (210/2037)

10.0 5.0 0.0 Shoulder Arthroplasty

Hip Arthroplasty

Knee Arthroplasty

Figure 7 Prevalence of DVT after reconstructive shoulder arthroplasty compared with published prevalence of DVT after total hip and knee arthroplasty at our institution.

could not account for the higher prevalence of DVT documented in this study. To assess the clinical significance of our findings in a relevant orthopaedic context, we compared our findings with the published prevalence of DVT after hip and knee arthroplasty at our institution.29,31 Surprisingly, we found that the prevalence of DVT after shoulder arthroplasty is not statistically different than that after hip replacement surgery at our institution, using a similar thromboprophylaxis protocol of aspirin, pneumatic compression foot pumps, and early ambulation. The prevalence of DVT after shoulder arthroplasty is, however, significantly lower than that after knee replacement surgery at our institution. This study suggests that DVT after reconstructive arthroplasty is not uncommon. DVT can occur in the operative upper extremity or either of the lower extremities. The greatest risk for DVT is in the acute perioperative period, although it can occur up to 12 weeks postoperatively. Because preoperative Doppler ultrasound evaluation was not performed, we cannot state whether perioperative DVTs documented in this study were new or preexisting. Nevertheless, we believe that the presence or absence of DVT postoperatively, not the time of genesis, is most relevant clinically. Upper extremity DVTs typically involve the axillary vein. Significant pain and swelling in the postoperative extremity may indicate DVT rather than a superficial phlebitis of the cephalic vein. Lower extremity DVTs typically involve the distal veins of the calf, although they can develop proximally as well. DVT-related thromboembolic complications can and do occur after shoulder arthroplasty, including symptomatic and fatal pulmonary embolism. Surgeons should have a high degree of suspicion for pulmonary embolism when respiratory difficulty or a cardiac tachyarrhythmia develops in a patient after shoulder replacement surgery. This is the first study to investigate VTE disease prospectively in this patient population, and as a result, there are several limitations in its design. Given the paucity of existing literature to support our initial clinical concern that VTE events in patients undergoing shoulder

arthroplasty may occur more frequently than commonly believed, we chose not to perform preoperative screening Doppler ultrasounds because of cost restrictions. For the same reason, we chose to select half of our study patients randomly for follow-up 12-week Doppler ultrasound. Because of these limitations, coupled with the fact that our patient sample size is relatively small, we recognize that it is possible that our findings may not accurately reflect the true incidence of DVT after shoulder arthroplasty. Nevertheless, we believe that the findings from this study are still important and will serve to alert physicians to the possibility of VTE in this patient population as well as to stimulate further research in this area. Anticoagulation therapy in postsurgical patients carries inherent risks and may result in potential major bleeding complications at the site of surgery or in a critical organ. Such adverse events have been implicated as a cause of wound complications, infection, the need for reoperation, or even death.7,19,27,29-31 At this time, we are not aware of any ideal VTE disease prophylaxis available that is both completely safe and effective. On the basis of our findings, we now use perioperative compression stockings and intraoperative pneumatic compression boots, as well as postoperative aspirin, as prophylaxis against VTE disease. We recommend discussing the risk of thromboembolic complications preoperatively with patients. In patients with significant medical comorbidity or a history of prior VTE disease or cancer, we routinely perform screening postoperatively for DVT using Doppler ultrasound.

Acknowledgments We thank Tara Holmes, PA-C, and Maureen Doyle, PA-C, Sports Medicine and Shoulder Service, Hospital for Special Surgery, New York, NY, for their assistance in collecting and processing data for this study.

References 1. Anderson FA Jr, Wheeler HB, Goldberg RJ, Hosmer DW, Patwardhan NA, Jovanovic B, et al. A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism. The Worcester DVT Study. Arch Intern Med 1991;151:933-8. 2. Arcand M, Burkhead WZ Jr, Zeman C. Pulmonary embolism caused by thrombosis of the axillary vein after shoulder arthroplasty. J Shoulder Elbow Surg 1997;6:486-90. 3. Baxter GM, Kincaid W, Jeffrey RF, Millar GM, Porteous C, Morley P. Comparison of colour Doppler ultrasound with venography in the diagnosis of axillary and subclavian vein thrombosis. Br J Radiol 1991;64:777-81. 4. Caprini JA, Arcelus JI, Hasty JH, Tamhane AC, Fabrega F. Clinical assessment of venous thromboembolic risk in surgical patients. Semin Thromb Hemost 1991;17(Suppl 3):304-12. 5. Clagett GP, Anderson FA Jr, Geerts W, Heit JA, Knudson M, Lieberman JR, et al. Prevention of venous thromboembolism. Chest 1998;114(Suppl):531S-60.

106 6. Clarke MT, Green JS, Harper WM, Gregg PJ. Screening for deepvenous thrombosis after hip and knee replacement without prophylaxis. J Bone Joint Surg Br 1997;79:787-91. 7. Colwell CW Jr. Managing thromboembolic risk in hip and knee arthroplasty: state of the art. Orthopedics 2003;26(Suppl):s231-6. 8. Fowkes FJ, Price JF, Fowkes FG. Incidence of diagnosed deep vein thrombosis in the general population: systematic review. Eur J Vasc Endovasc Surg 2003;25:1-5. 9. Geerts WH, Heit JA, Clagett GP, Pineo GF, Colwell CW, Anderson FA Jr, et al. Prevention of venous thromboembolism. Chest 2001;119(Suppl):132S-75. 10. Hasan SS, Leith JM, Smith KL, Matsen FA III. The distribution of shoulder replacement among surgeons and hospitals is significantly different than that of hip or knee replacement. J Shoulder Elbow Surg 2003;12:164-9. 11. Heit JA, Silverstein MD, Mohr DN, Petterson TM, O’Fallon WM, Melton LJ III. Predictors of survival after deep vein thrombosis and pulmonary embolism: a population-based, cohort study. Arch Intern Med 1999;159:445-53. 12. Kassai B, Boissel JP, Cucherat M, Sonie S, Shah NR, Leizorovicz A. A systematic review of the accuracy of ultrasound in the diagnosis of deep venous thrombosis in asymptomatic patients. Thromb Haemost 2004;91:655-66. 13. Kreder HJ, Deyo RA, Koepsell T, Swiontkowski MF, Kreuter W. Relationship between the volume of total hip replacements performed by providers and the rates of postoperative complications in the state of Washington. J Bone Joint Surg Am 1997;79:485-94. 14. Lyman S, Sherman S, Carter TI, Bach PB, Mandl LA, Marx RG. Prevalence and risk factors for symptomatic thromboembolic events after shoulder arthroplasty. Clin Orthop Relat Res 2006;448:152-6. 15. Mansfield AO. Alteration in fibrinolysis associated with surgery and venous thrombosis. Br J Surg 1972;59:754-7. 16. Neer CS II. Articular replacement for the humeral head. J Bone Joint Surg Am 1955;37:215-28. 17. Nordstrom M, Lindblad B, Bergqvist D, Kjellstrom T. A prospective study of the incidence of deep-vein thrombosis within a defined urban population. J Intern Med 1992;232:155-60. 18. Paramo JA, Rocha E. Changes in coagulation and fibrinolysis after total hip replacement and their relations with deep vein thrombosis. Haemostasis 1985;15:345-52.

A.A. Willis et al. 19. Parvizi J, Ghanem E, Joshi A, Sharkey PF, Hozack WJ, Rothman RH. Does ‘‘excessive’’ anticoagulation predispose to periprosthetic infection? J Arthroplasty 2007;22(Suppl 2):24-8. 20. Rockwood CA Jr, Wirth MA, Blair S. Warning: pulmonary embolism can occur after elective shoulder surgerydreport of two cases and survey of the members of the American Shoulder and Elbow Surgeons. J Shoulder Elbow Surg 2003;12:628-30. 21. Saleem A, Markel DC. Fatal pulmonary embolus after shoulder arthroplasty. J Arthroplasty 2001;16:400-3. 22. Sikorski JM, Hampson WG, Staddon GE. The natural history and aetiology of deep vein thrombosis after total hip replacement. J Bone Joint Surg Br 1981;63:171-7. 23. Silverstein MD, Heit JA, Mohr DN, Petterson TM, O’Fallon WM, Melton LJ III. Trends in the incidence of deep vein thrombosis and pulmonary embolism: a 25-year population-based study. Arch Intern Med 1998;158:585-93. 24. Sperling JW, Cofield RH. Pulmonary embolism following shoulder arthroplasty. J Bone Joint Surg Am 2002;84:1939-41. 25. Stern SH, Insall JN. Hematologic effects of total knee arthroplasty. A prospective evaluation. Clin Orthop Relat Res 1993:10-4. 26. Stulberg BN, Insall JN, Williams GW, Ghelman B. Deep-vein thrombosis following total knee replacement. An analysis of six hundred and thirty-eight arthroplasties. J Bone Joint Surg Am 1984; 66:194-201. 27. Torn M, Bollen WL, van der Meer FJ, van der Wall EE, Rosendaal FR. Risks of oral anticoagulant therapy with increasing age. Arch Intern Med 2005;165:1527-32. 28. Warwick D, Williams MH, Bannister GC. Death and thromboembolic disease after total hip replacement. A series of 1162 cases with no routine chemical prophylaxis. J Bone Joint Surg Br 1995;77:6-10. 29. Westrich GH, Farrell C, Bono JV, Ranawat CS, Salvati EA, Sculco TP. The incidence of venous thromboembolism after total hip arthroplasty: a specific hypotensive epidural anesthesia protocol. J Arthroplasty 1999;14:456-63. 30. Westrich GH, Sanchez PM. Prevention and treatment of thromboembolic disease: an overview. Instr Course Lect 2002;51:471-80. 31. Westrich GH, Sculco TP. Prophylaxis against deep venous thrombosis after total knee arthroplasty. Pneumatic plantar compression and aspirin compared with aspirin alone. J Bone Joint Surg Am 1996;78: 826-34.