Abstract
The selection of a prophylaxis regimen and its implementation have been influenced considerably by the decreased duration of hospital stays and the pressures of cost containment. The purpose of the present study was to determine the rate of symptomatic pulmonary embolism both before and after discharge, the number of days required to achieve an adequate level of anticoagulation, and the complications associated with the use of low-dose wafarin after total hip arthroplasty.Between 1987 and 1993, 1099 primary and revision total hip arthroplasties were performed in 940 patients who received low-dose warfarin for prophylaxis against thromboembolic disease. The average duration of prophylaxis was fifteen days (range, one to twenty-nine days). The target level of anticoagulation (as indicated by a prothrombin time of fourteen to seventeen seconds) was achieved an average of three days (range, one to sixteen days) after the operation. The level of anticoagulation was lower than the target range at the time of discharge after 257 total hip arthroplasties (23.4 per cent), and the target level was never achieved during the period of hospitalization after fifty-four such procedures (4.9 per cent).Twelve total hip arthroplasties were associated with a symptomatic pulmonary embolism; the over-all prevalence of this complication therefore was 1.1 per cent (95 per cent confidence interval, 0.4 to 1.9 per cent). Four pulmonary emboli were diagnosed before discharge and eight, after discharge. A fatal pulmonary embolism occurred after one procedure (0.1 per cent). Patients who had a history of symptomatic venous thromboembolic disease had a significantly increased risk of symptomatic pulmonary embolism after total hip arthroplasty (p = 0.001, Fisher exact test).A major bleeding episode occurred after thirty-two total hip arthroplasties (2.9 per cent). Patients who had a prothrombin time of more than seventeen seconds had a significantly increased risk of hematoma formation (p = 0.003, chi-square analysis). Prophylaxis with low-dose warfarin is safe and effective for the prevention of pulmonary embolism after total hip arthroplasty.
Low-dose warfarin commonly is used for prophylaxis against thromboembolic disease after total hip arthroplasty and has been associated with a low prevalence of symptomatic pulmonary embolism1,3,7,8. However, no method of prophylaxis that is currently available can eliminate the risk of thromboembolic disease completely. Pulmonary embolism has been noted to occur even after management with a meticulous regimen of prophylaxis and an intensive program of rehabilitation. The question today is how to use prophylaxis against deep-vein thrombosis in our changing health-care environment. The selection and implementation of a prophylaxis regimen have been influenced substantially by decreased hospital stays and issues of cost containment. The optimum duration of prophylaxis, concerns about thromboembolic events after discharge from the hospital2,10, the frequency of symptomatic deep-vein thrombosis after discharge, and the potential role of screening for asymptomatic deep-vein thrombosis are important issues that must be resolved2,6,10.
An analysis of the results of 2595 consecutive total hip replacements that were performed at our institution between 1974 and 1987 in patients who received warfarin for prophylaxis revealed a rate of pulmonary embolism of 0.2 per cent1. The present study includes the results of 1099 total hip arthroplasties that were performed at our institution between 1987 and 1993 in patients who received prophylaxis with low-dose warfarin. The purpose of the present study was to determine the rate of pulmonary embolism both before and after discharge, the number of days required to achieve an adequate level of anticoagulation, and the complications associated with the use of low-dose warfarin after total hip arthroplasty.
*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. No funds were received in support of this study.
†Department of Orthopaedic Surgery, University of California at Los Angeles School of Medicine, 10833 Le Conte Avenue, Los Angeles, California 90095.
‡Division of Orthopaedic Surgery, University of Texas Medical Branch, Galveston, Texas 77550.
§Joint Replacement Institute, 2400 South Flower Street, Los Angeles, California 90007.
Between 1987 and 1993, 1042 patients had a total of 1244 total hip replacements at the University of California at Los Angeles Medical Center and were entered into a computerized database. All patients who received prophylaxis with low-dose warfarin (including those who had a history of pulmonary embolism or deep-vein thrombosis) were eligible for the present study. All thromboembolic events (including pulmonary embolism and deep-vein thrombosis) and bleeding complications were entered into the database. In addition to collecting data prospectively, we reviewed all of the inpatient and outpatient medical records for these patients. Patients were followed for three months postoperatively to determine if a symptomatic pulmonary embolism or deep-vein thrombosis had occurred. Sixty patients who had not been seen at three months were contacted by telephone. The records for nine patients were obtained from other hospitals and were reviewed to determine whether any thromboembolic event or bleeding complication had occurred after discharge from our hospital. Complete data were available for all other patients. These data included the type and duration of prophylaxis, the number of days required to achieve an adequate level of anticoagulation (as indicated by a prothrombin time of fourteen to seventeen seconds), whether the prothrombin time had been less than fourteen seconds or more than seventeen seconds at any time during hospitalization, and whether the patient was receiving prophylaxis at the time of discharge. The diagnosis, the type of procedure (primary or revision), and whether the patient had received vitamin K or fresh-frozen plasma during the postoperative period because of an elevated prothrombin time also were recorded.
During the period of the study, 1164 (93.6 per cent) of the 1244 total hip arthroplasties were performed in patients who received prophylaxis with low-dose warfarin before discharge. The results of twenty-six procedures were excluded from the study because the patient had received an anti-inflammatory agent or aspirin in addition to warfarin; those of thirty-seven procedures, because the patient had received warfarin for an extended period postoperatively (usually for treatment of a cardiac problem); and those of two procedures, because the patient had had a Greenfield filter placed preoperatively. None these sixty-five procedures were associated with a symptomatic pulmonary embolism. Thus, the present study includes the results of 1099 total hip arthroplasties in 940 patients (407 men and 533 women). The average age at the time of the procedure was fifty-nine years (range, nineteen to ninety years). Six hundred and seven patients (680 hips) had a primary total hip arthroplasty, and 333 patients (419 hips) had a revision. The preoperative diagnoses included osteoarthrosis (616 hips), avascular necrosis (167 hips), rheumatoid arthritis (105 hips), post-traumatic degenerative arthritis (eighty-six hips), congenital dysplasia (sixty-four hips), ankylosing spondylitis (fifteen hips), previous infection about the hip (nine hips), psoriatic arthritis (five hips), Paget disease (five hips), osteoblastoma (two hips), and miscellaneous diagnoses (twenty-one hips); the diagnosis was unknown for the remaining four hips. Fifteen patients had a history of symptomatic deep-vein thrombosis, and five others had a history of symptomatic pulmonary embolism.
A specific dosing protocol was used for all patients. Ten milligrams of warfarin was administered orally on the night of the operation, and no warfarin was administered on the first postoperative day. On the second postoperative day, the dose was determined after the prothrombin time had been evaluated. In general, the dose of warfarin (in milligrams) was equal to twenty minus the prothrombin time. The dose of warfarin was adjusted daily in an attempt to keep the prothrombin time between fourteen and seventeen seconds.
Bleeding complications were determined by means of a careful review of medical records. A major bleeding complication was defined as gastrointestinal, retroperitoneal, or intracranial bleeding or as bleeding at the site of a major prosthetic joint. From 1987 to 1990, physical therapy was initiated on the third postoperative day; from 1991 to 1993, it was initiated on either the first or the second postoperative day.
The average duration of hospitalization was eleven days (range, four to fifty-three days), and the average duration of warfarin therapy was fifteen days (range, one to twenty-nine days). Low-dose warfarin was scheduled to be administered for at least three weeks after 337 procedures (30.7 per cent), for two to less than three weeks after 182 procedures (16.6 per cent), for one to less than two weeks after 513 procedures (46.7 per cent), and for less than one week after sixty-seven procedures (6.1 per cent). The warfarin therapy given after 483 procedures (43.9 per cent) was continued after discharge from the hospital.
The target level of anticoagulation (as indicated by a prothrombin time of fourteen to seventeen seconds) was achieved an average of three days (range, one to sixteen days) after the operation. The target level of anticoagulation was achieved by the third day after 532 procedures (48.4 per cent), by the fourth day after 717 procedures (65.2 per cent), by the fifth day after 879 procedures (80.0 per cent), and by the sixth day after 972 procedures (88.4 per cent). The target level was not achieved during the period of hospitalization after fifty-four procedures (4.9 per cent). The prothrombin time at the time of discharge was within the target range after 742 procedures (67.5 per cent), higher than the target range after 100 procedures (9.1 per cent), and lower than the target range after 257 procedures (23.4 per cent). Fresh-frozen plasma or vitamin K was administered (at the discretion of the attending surgeon) to eighteen patients because of concerns about an elevated prothrombin time of at least twenty seconds. A hematoma developed in two of these eighteen patients.
Five total hip arthroplasties (0.5 per cent) were followed by the development of symptomatic deep-vein thrombosis. Four thrombi were proximal and one was distal, and all were confirmed by venography or duplex ultrasonography. Three of the four proximal thrombi occurred during hospitalization, and two of the three were associated with a pulmonary embolism. The fourth proximal thrombus was diagnosed in a sixty-five-year-old woman on the seventeenth postoperative day and was associated with a pulmonary embolism. The distal thrombus, which was located in the calf, also occurred after discharge. The patients who had the proximal thrombi were managed with intravenous administration of heparin and oral administration of warfarin, and the patient who had the distal thrombus was managed with oral administration of warfarin. Seven patients (seven hips; 0.6 per cent) died within the three-month follow-up period: five patients died of a myocardial infarction, one died secondary to complications related to lung cancer, and one died of a presumed symptomatic pulmonary embolism on the seventeenth postoperative day.
Twelve total hip arthroplasties were associated with a symptomatic pulmonary embolism; the over-all prevalence of this complication therefore was 1.1 per cent (95 per cent confidence interval, 0.4 to 1.9 per cent). Four procedures (0.4 per cent) were associated with a symptomatic pulmonary embolism that was diagnosed during the period of hospitalization, and eight (0.7 per cent) were associated with one that was diagnosed after discharge. Ten of the twelve pulmonary emboli were diagnosed on the basis of a high-probability ventilation-perfusion scan. The eleventh pulmonary embolism was in a forty-nine-year-old man who noted shortness of breath on the sixth postoperative day. A ventilation-perfusion scan indicated a moderate probability of a pulmonary embolism, and the diagnosis was confirmed by an angiogram. The twelfth pulmonary embolism was in a seventy-five-year-old man who noted shortness of breath at home on the seventeenth postoperative day. The patient had a cardiac arrest in the emergency room of a local hospital and died before a complete evaluation could be performed; no autopsy was done. Pulmonary embolism was listed on the death certificate as the cause of death. This was the only fatal pulmonary embolism in the present study; the prevalence of this complication therefore was 0.1 per cent (95 per cent confidence interval, 0 to 0.5 per cent).
A symptomatic pulmonary embolism developed after ten (1.5 per cent) of the 680 primary total hip arthroplasties and after two (0.5 per cent) of the 419 revisions. Four emboli (three of which developed after a primary procedure and one of which developed after a revision) were diagnosed before discharge, and eight emboli (seven of which developed after a primary procedure and one of which developed after a revision) were diagnosed after discharge. There was no significant difference between patients who had a primary total hip arthroplasty and those who had a revision with regard to the prevalence of symptomatic pulmonary embolism (p = 0.47, chi-square analysis).
One of the four pulmonary emboli that were diagnosed before discharge was in an eighty-four-year-old man who had had a pulmonary embolism fifty years previously (after the operative treatment of a hernia) as well as a history of ligation of the left femoral vein. Several hours after an uneventful right total hip arthroplasty and before the initiation of warfarin therapy, the patient noted shortness of breath. The findings of a ventilation-perfusion scan were consistent with bilateral pulmonary embolism, and a duplex scan of the lower extremities revealed bilateral femoral deep-vein thrombosis. The patient was managed with intravenous administration of heparin and then received warfarin for three months. He had no additional sequelae. The pulmonary emboli in the other three patients were diagnosed on the third, fifth, and sixth postoperative days. The target level of anticoagulation was not achieved in any of these four patients before the pulmonary embolism was diagnosed.
The eight post-discharge diagnoses of pulmonary embolism were made at an average of twenty-seven days (range, fourteen to sixty-eight days) postoperatively. Two of these patients had received warfarin for eight and twelve days and were diagnosed as having a pulmonary embolism on the thirty-fifth and seventeenth postoperative days, respectively. Three patients were being managed with a three-week protocol of low-dose warfarin when a pulmonary embolism was diagnosed on the fourteenth, sixteenth, and seventeenth postoperative days. The other three patients had received warfarin for at least three weeks and were diagnosed as having a pulmonary embolism on the twenty-fifth, twenty-seventh, and sixty-eighth postoperative days (Table I).
Three of the twenty patients who had a history of a symptomatic thromboembolic event (that is, a pulmonary embolism or a deep-vein thrombosis) had a symptomatic pulmonary embolism during the study period. The first patient was the eighty-four-year-old man who had had a pulmonary embolism after the operative treatment of a hernia; this patient had a pulmonary embolism on the day of the index operation. The second patient was a thirty-eight-year-old man who had had a pulmonary embolism after a previous total hip arthroplasty; the patient still was receiving warfarin when a pulmonary embolism was diagnosed on the sixteenth day after the index operation. The third patient was the seventy-five-year-old man who died of a presumed pulmonary embolism on the seventeenth postoperative day, at which time he was still receiving warfarin; this patient had a history of deep-vein thrombosis. Patients who had a history of venous thromboembolic disease had a significantly increased risk of symptomatic pulmonary embolism (p = 0.001, Fisher exact test).
After fifty-four procedures (4.9 per cent), the target prothrombin time was not achieved either during hospitalization or before the diagnosis of a pulmonary embolism. The average duration of hospitalization (and standard deviation) for these procedures was 9 ± 2 days, and warfarin had been given for an average of 11 ± 7 days. A pulmonary embolism was diagnosed before discharge (on the day of the operation and on the third, fifth, and sixth postoperative days) after four (7 per cent) of these procedures.
After 257 procedures (23.4 per cent), a prothrombin time of fourteen to seventeen seconds usually was achieved during hospitalization but the prothrombin time was less than fourteen seconds at the time of discharge. Warfarin was administered for an average of 14 ± 6 days after these procedures. A symptomatic pulmonary embolism developed after only one of these procedures.
A major bleeding event occurred after thirty-two total hip arthroplasties (2.9 per cent). Twenty-four total hip arthroplasties (2.2 per cent) were associated with a wound hematoma. Five patients, four of whom had had a revision total hip arthroplasty, had a reoperation to drain the hematoma. Seven total hip arthroplasties (0.6 per cent) were associated with an episode of gastrointestinal bleeding; this complication was manifested by melena in five patients, a gastric ulcer in one, and a duodenal ulcer in one. One total hip arthroplasty (0.1 per cent) was associated with vaginal bleeding. Seventeen (71 per cent) of the twenty-four patients who had a wound hematoma had an elevated prothrombin time (more than seventeen seconds) during hospitalization. Over-all, a wound hematoma occurred in association with seventeen (3.8 per cent) of the 450 procedures performed in patients who had a prothrombin time of more than seventeen seconds at any time during hospitalization and seven (1.1 per cent) of the 649 procedures in patients who had a prothrombin time of seventeen seconds or less during hospitalization (p = 0.003, chi-square analysis).
The low-dose warfarin protocol used in the present study was associated with a low over-all rate of symptomatic pulmonary embolism (1.1 per cent; 95 per cent confidence interval, 0.4 to 1.9 per cent) as well as with a low rate of fatal pulmonary embolism (0.1 per cent; 95 per cent confidence interval, 0 to 0.5 per cent). The only known death that was presumed to be the result of pulmonary embolism (no autopsy was done) occurred seventeen days postoperatively. To our knowledge, this is the only fatal pulmonary embolism that has occurred after 4199 consecutive total hip arthroplasties performed at our institution since 1971 in patients managed with warfarin for prophylaxis against thromboembolic disease. The over-all rate of fatal pulmonary embolism during this period therefore is 0.02 per cent (95 per cent confidence interval, 0.0001 to 0.1 per cent).
Our data also suggest that patients who have a history of symptomatic venous thromboembolic disease have an increased risk of symptomatic pulmonary embolism after total hip arthroplasty. In such instances, it may be prudent to administer warfarin before the procedure or to select a mode of prophylaxis that is associated with a more rapid onset of action, as a sufficient level of anticoagulation may not be achieved in the immediate postoperative period. In addition, a longer duration (six weeks) of prophylaxis should be considered. If prophylaxis is going to be given for a shorter duration, however, preoperative and postoperative screening for the presence of deep-vein thrombosis should be performed for these high-risk patients.
The low-dose warfarin protocol used in the present study also was associated with a low prevalence of clinically important bleeding events (2.9 per cent). Patients who had a prothrombin time of more than seventeen seconds had a significantly increased risk of hematoma formation (p = 0.003). Five patients had a reoperation to drain a hematoma. These rates are comparable with those observed in randomized trials of patients who received warfarin for prophylaxis in association with total hip arthroplasty5,8,9.
Paiement et al.7, in a study of 268 patients who received low-dose warfarin for three months after a total hip arthroplasty, reported that no fatal pulmonary emboli occurred during the six-month study period and that no symptomatic pulmonary emboli occurred after discharge. However, because of the small sample size, the 95 per cent confidence interval indicates that the true rate of symptomatic pulmonary embolism could be as high as 1.1 per cent. Ten patients (4 per cent) had an episode of major bleeding. Balderston et al., in a study of 1392 patients who received prophylaxis with low-dose warfarin, reported that 1.1 per cent of the patients had a symptomatic pulmonary embolism and 2.4 per cent had a hematoma that necessitated operative drainage. In two randomized trials in which low-dose warfarin was compared with low-molecular-weight heparin, no symptomatic pulmonary emboli were reported in 335 patients (in the study by Hull et al.) or in 174 patients (in the study by the RD Heparin Arthroplasty Group) who had received low-dose warfarin for approximately one week. However, the patients in both studies were screened with venography and those who were found to have a proximal clot were managed with intravenous administration of heparin. Therefore, recommendations regarding the appropriate duration of prophylaxis with warfarin cannot be made on the basis of those studies. The prevalence of clinically important bleeding events in those studies was 1.5 per cent5 and 5 per cent9.
The major limitation of the present study is that it was observational and therefore the patients were not randomized to receive prophylaxis with low-dose warfarin for one of two different durations. However, a power analysis revealed that, with the low rate of symptomatic pulmonary embolism that was associated with the low-dose warfarin protocol, 10,000 patients would be necessary in each treatment group to perform such a trial adequately. In addition, the patients were not screened for evidence of proximal thrombi; therefore, we could not calculate the rate of asymptomatic proximal thrombosis.
The strengths of the present study include the large number of patients and the implementation of in-hospital protocols that reflect recent trends within the orthopaedic community regarding prophylaxis against venous thromboembolic disease. In general, as the duration of hospital stays has decreased so has the duration of prophylaxis. Another strength of the study is that it includes all patients who had a primary or revision total hip arthroplasty, even those who had a history of symptomatic pulmonary embolism or deep-vein thrombosis. Finally, we are not aware of any previous study that has included an analysis of the number of days required to achieve an adequate level of anticoagulation, the number of patients in whom the target level of anticoagulation was achieved during hospitalization, or the number of patients who had an adequate prothrombin time at the time of discharge. The results of the present study are presumed to reflect the true prevalence of symptomatic pulmonary embolism in a heterogeneous population of patients managed with low-dose warfarin for prophylaxis against thromboembolic disease after total hip arthroplasty.
One potential problem with warfarin is that it is associated with a delay in the achievement of an adequate level of anticoagulation. After 532 (48.4 per cent) of the total hip arthroplasties in the present study, the target level of anticoagulation was not achieved until the third postoperative day, and after fifty-four (4.9 per cent) of the procedures the target level was not achieved during hospitalization. The target level of anticoagulation was achieved by the fourth day after 717 procedures (65.2 per cent) and by the fifth day after 879 procedures (80.0 per cent). However, a total of three pulmonary emboli were diagnosed on the day of the operation and the third and fifth postoperative days; the target level of anticoagulation was not achieved in any of the patients in whom these emboli developed.
An analysis of the data reveals two ways in which the low-dose warfarin protocol failed. The first type of failure—the development of a symptomatic pulmonary embolism before the achievement of an adequate level of anticoagulation—occurred after four total hip arthroplasties (0.4 per cent). In addition, fifty patients (fifty hips; 4.5 per cent) in whom the target level of anticoagulation was not achieved during hospitalization presumably had an increased risk for the development of a symptomatic pulmonary embolism. The second type of failure—the development of a symptomatic pulmonary embolism despite the achievement of an adequate level of anticoagulation—occurred after eight total hip arthroplasties (0.7 per cent).
The issue of the optimum duration of prophylaxis after total hip arthroplasty has gained increased attention because of the shorter duration of hospital stays and because of concern regarding the potential for venous thromboembolism after discharge2,6,10. In general, surgeons have wanted to decrease the duration of prophylaxis because of concerns about the cost and physician time required to monitor prothrombin time, the compliance of the patient, and the potential for bleeding complications. The durations of prophylaxis reported in the literature have been quite variable, and the ideal duration remains unknown. Prophylaxis typically is started on the night before or the day of the operation and may be continued until discharge or for as long as three months5-7,9. In the present study, four pulmonary emboli were diagnosed before discharge whereas eight were diagnosed after discharge. As the duration of hospital stays after total hip arthroplasty probably will continue to decrease, the rate of pulmonary embolism after discharge probably will continue to increase; this will be especially true if short-term prophylaxis is used.
As this was not a randomized trial, it is difficult to make definitive recommendations regarding the optimum duration of prophylaxis. Six symptomatic pulmonary emboli developed in patients who had received, or who were scheduled to receive, prophylaxis with low-dose warfarin for at least three weeks. These emboli certainly would have developed if warfarin had been administered for a shorter period of time. The question that needs to be answered is whether the duration of prophylaxis can be reduced safely for patients who have had a total hip arthroplasty. In the present study, no pulmonary emboli developed after 182 total hip arthroplasties in patients for whom the warfarin protocol was two to less than three weeks. Two pulmonary emboli developed after 513 procedures in patients who received warfarin for one to less than two weeks. One cannot know whether these two symptomatic pulmonary emboli would have developed if prophylaxis with warfarin had been continued for a longer period of time. However, the slight increase in the number of symptomatic pulmonary emboli suggests that the duration of prophylaxis with warfarin may be reduced safely to two weeks; this protocol has been instituted at our hospital. In addition, as the prothrombin time was at least fourteen seconds at the time of discharge after 76.6 per cent (842) of the procedures, we do not routinely monitor the prothrombin time after discharge for patients who have received warfarin for two weeks. However, we do perform such monitoring for patients who have a prothrombin time of less than fourteen seconds at the time of discharge and for those who do not have a stable prothrombin time or international normalized ratio.
In this era of constrained health-care resources, cost-effectiveness has become a critical issue in the selection of a method of prophylaxis. Patients who receive low-dose warfarin for an extended period of time (three weeks to three months) must be monitored postoperatively, which increases the cost of the prophylaxis. Postoperative screening has received increased attention because surgeons often choose to discontinue prophylaxis at the time of discharge, but it has been recognized that even the most effective forms of prophylaxis are associated with venous thromboembolic events after discharge. One alternative is to screen all patients routinely with duplex ultrasonography before discharge from the hospital. However, the routine use of duplex ultrasonography as a screening tool remains controversial because of conflicting reports in the literature regarding the accuracy of this method for the detection of asymptomatic proximal thrombi in patients who have had a total hip arthroplasty4,11,12. The accuracy of the test also is extremely technician-dependent, and quality control may be a problem at some medical centers. In addition, a unilateral duplex scan currently costs $250 at our institution.
Prophylaxis with low-dose warfarin is safe and effective for the prevention of symptomatic pulmonary embolism after total hip arthroplasty. We continue to manage our patients with low-dose warfarin for two weeks on the basis of the findings of the present study. We do not routinely monitor the prothrombin time after discharge if the patient had an adequate and stable level of anticoagulation at the time of discharge. Patients who have a history of symptomatic venous thromboembolic disease have an increased risk for the development of a symptomatic pulmonary embolism after total hip arthroplasty; adjustments in the in-hospital regimen of prophylaxis with warfarin, an increase in the duration of prophylaxis, the use of an alternative regimen of prophylaxis, and preoperative and postoperative screening therefore should be considered for such patients.
Amstutz, H. C.; Friscia, D. A.; Dorey, F.; and |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-A321Â
1989Â
Â
Arcelus, J. I.; Caprini, J. A.; and |and |Traverso, C. I.: Venous thromboembolism after hospital discharge. Sem. Thromb. and Hemost.,19 (Supplement 1): 142-146, 1993.19 (Supplement 1)142Â
1993Â
Â
Balderston, R. A.; Graham, T. S.; Booth, R. E., Jr.; and |and |Rothman, R. H.: The prevention of pulmonary embolism in total hip arthroplasty. Evaluation of low-dose warfarin therapy. J. Arthroplasty,4: 217-221, 1989.4217Â
1989Â
[PubMed] Â
Borris, L. C.; Christiansen, H. M.; Lassen, M. R.; Olsen, A. D.; Schott, P.; and |and the|Venous Thrombosis Group: Comparison of real-time B-mode ultrasonography and bilateral ascending phlebography for detection of postoperative deep vein thrombosis following elective hip surgery. Thromb. and Haemost.,61: 363-365, 1989.61363Â
1989Â
Â
Hull, R.; Raskob, G.; Pineo, G.; Rosenbloom, D.; Evans, W.; Mallory, T.; Anquist, K.; Smith, F.; Hughes, G.; Green, D.; Elliott, C. G.; Panju, A.; and |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, 1993.3291370Â
1993Â
Â
Lieberman, J. R., and |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-A1239Â
1994Â
Â
Paiement, G. D.; Wessinger, S. J.; Hughes, R.; and |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-A893Â
1993Â
Â
Paiement, G.; Wessinger, S. J.; Waltman, A. C.; and |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, 1987.223Â
1987Â
[PubMed] Â
RD Heparin Arthroplasty Group: RD heparin compared with warfarin for prevention of venous thromboembolic disease following total hip or knee arthroplasty. J. Bone and Joint Surg.,76-A: 1174-1185, Aug. 1994.76-A1174Â
1994Â
Â
Trowbridge, A.; Boese, C. K.; Woodruff, B.; Brindley, H. H., Sr.; Lowry, W. E.; and |and |Spiro, T. E.: Incidence of posthospitalization proximal deep venous thrombosis after total hip arthroplasty. A pilot study. Clin. Orthop.,299: 203-208, 1994.299203Â
1994Â
[PubMed] Â
Wells, P. S.; Lensing, A. W.; Davidson, B. L.; Prins, M. H.; and |and |Hirsh, J.: Accuracy of ultrasound for the diagnosis of deep venous thrombosis in asymptomatic patients after orthopedic surgery. A meta-analysis. Ann. Intern. Med.,122: 47-53, 1995.12247Â
1995Â
[PubMed] Â
Woolson, S. T.; McCrory, D. W.; Walter, J. F.; Maloney, W. J.; Watt, J. M.; and |and |Cahill, P. D.: B-mode ultrasound scanning in the detection of proximal venous thrombosis after total hip replacement. J. Bone and Joint Surg.,72-A: 983-987, Aug. 1990.72-A983Â
1990Â
Â