 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1999 The Journal of Bone and Joint Surgery, Inc. Efficacy of Prophylaxis Against Thromboembolism with Intermittent Pneumatic Compression After Primary and Revision Total Hip Arthroplasty* , CHAPEL HILL, NORTH CAROLINAInvestigation performed at the Department of Orthopaedics, University of North Carolina, Chapel Hill
Background: Thromboembolism is a common and important complication after total hip arthroplasty. A variety of pharmacological and mechanical measures have been proposed for prophylaxis. The purpose of the present study was to evaluate the efficacy of intermittent pneumatic compression as prophylaxis against thromboembolism following total hip arthroplasty. Methods: The prospective study involved a consecutive series of 425 patients in whom a total of 502 (324 primary and 178 revision) total hip arthroplasties had been performed by two surgeons. The patients were managed intraoperatively and postoperatively with use of thigh-high elastic compression stockings and thigh-high intermittent pneumatic compression sleeves. Experienced vascular technologists performed venous duplex ultrasonography on both lower extremities of all patients at a mean of six days (range, two to fifteen days) postoperatively. All patients were followed for at least one year in order to detect late thromboembolism. Results: An asymptomatic deep-vein thrombosis was noted on the scans made after twenty-three (4.6 percent) of the 502 procedures. Nineteen (3.8 percent) of the arthroplasties were followed by the development of a proximal thrombosis and four (0.8 percent), a distal thrombosis. Nineteen of the thromboses were ipsilateral (eighteen were proximal and one, distal), and four were contralateral (one was proximal and three, distal). No symptomatic deep-vein thrombosis developed in the hospital. In addition, three (two proximal and one distal) symptomatic ipsilateral deep-vein thromboses (a prevalence of 0.6 percent) developed three to twenty-three weeks after postoperative scans revealed negative findings and the patients were discharged from the hospital. Three symptomatic pulmonary embolisms (a prevalence of 0.6 percent) were confirmed by ventilation-perfusion scanning while the patients were in the hospital. There were no symptomatic pulmonary embolisms after discharge, and there were no fatal pulmonary embolisms. With the numbers available, we were unable to detect an association between deep-vein thrombosis and age (p = 0.76), gender (p = 0.13), body-mass index (p = 0.12), type of arthroplasty (primary or revision) (p = 0.12), operative approach (p = 0.37), duration of the operation (p = 0.21), type of anesthesia (general or regional) (p = 0.51), units of blood transfused (autologous, p = 0.79; homologous, p = 0.57), blood type (p = 0.18), or the presence of a so-called classic risk factor for the development of thrombosis (p = 0.22). Five arthroplasties (1.0 percent) were followed by the development of a wound hematoma, but only one hematoma necessitated operative drainage. Conclusions: The use of intraoperative and postoperative thigh-high intermittent pneumatic compression, combined with duplex ultrasonography performed by experienced vascular technologists, is effective for prophylaxis against thromboembolism after both primary and revision total hip arthroplasties. The low prevalence of deep-vein thrombosis (4.6 percent) and symptomatic pulmonary embolism (0.6 percent) is comparable with that associated with pharmacological prophylaxis.
Patients who are managed with total hip arthroplasty have an increased risk of postoperative thromboembolic complications. In series ranging in size from fifty to 7959 procedures, 45 to 75 percent of patients who received no prophylaxis had a deep-vein thrombosis and 1 to 3 percent had a fatal pulmonary embolism2,4,8,9. These events are the most potentially lethal complications in the early postoperative period10. Proximal thrombi are believed to be the precipitating factor in the development of most of these pulmonary emboli; thus, effective prophylaxis against deep-vein thrombosis is essential in the management of all patients who are managed with hip arthroplasty11. Recent studies have demonstrated lower rates of thromboembolism after total hip arthroplasty without any chemical or mechanical prophylaxis. In a series of 104 hips treated without routine prophylaxis, Clarke et al. reported a prevalence of deep-vein thrombosis of 33 percent (thirty-four hips), as demonstrated by routine postoperative venography, and no deaths due to pulmonary embolism1. Warwick et al. studied the results after 1162 consecutive total hip replacements in patients who were managed with antithromboembolism stockings and early mobilization (but not routine anticoagulation therapy)18. No screening was performed, and so-called high-risk patients were excluded and given anticoagulation therapy. The rate of fatal pulmonary embolism was only 0.3 percent (95 percent confidence interval, 0.09 to 0.88 percent). Several prophylactic methods have been shown to reduce the prevalence of deep-vein thrombosis. In a meta-analysis of fifty-six trials involving various methods to prevent venous thromboembolism after total hip arthroplasty in 8000 patients, Imperiale and Speroff reported a rate of proximal deep-vein thrombosis of 5 percent in ten studies that included 864 patients managed with warfarin, a rate of 6 percent in twenty studies that included 2065 patients managed with low-molecular-weight heparin, and a rate of 13 percent in six studies that included 391 patients managed with either intermittent pneumatic compression or graded stockings6. (The latter six studies were grouped into one category for the purposes of the meta-analysis.) Woolson showed that intraoperative initiation of intermittent pneumatic compression reduced the risk of proximal deep-vein thrombosis to 6 percent (twenty of 322 procedures)20. However, in that study, patients who had a risk factor for the development of thrombosis, such as a history of thromboembolism or varicose veins, were managed with intermittent pneumatic compression and low-dose warfarin. The major advantages of prophylaxis with intermittent pneumatic compression are the presumed decrease in the prevalence of bleeding complications and the lack of necessity for the monitoring required with oral anticoagulation therapy. Both warfarin and low-molecular-weight heparin have been associated with bleeding complicatios5,6. A prophylactic method that is not associated with such risks would clearly deserve serious consideration for more universal use if it provided a similar level of protection against thromboembolism. In this prospective study, we examined the efficacy of prophylaxis against thromboembolism with intermittent pneumatic compression in a consecutive non-selected series of patients who were managed with either primary or revision total hip arthroplasty.
Five hundred and fifteen primary or revision total hip arthroplasties were performed in 433 patients by two of us (P. F. L. and S. S. K.) at one institution between June 1991 and June 1997. All patients, except those who were already receiving warfarin for cardiovascular problems but including those who had a known history of so-called classic risk factors for thromboembolism, were considered for entry into the prospective study. Fourteen patients (twenty-eight hips) had a bilateral staged total hip arthroplasty during a single period of hospitalization. These procedures were included in the study as separate events. No patient had two total hip arthroplasties under the same anesthesia. Thirteen hips were excluded from the study. Five hips were excluded because the patients were receiving chronic warfarin therapy; one of these patients (one hip) was being so managed because a prosthetic heart valve had been inserted and two patients (four hips), because of recurrent thromboembolism. Seven hips, in four patients who were managed with intermittent pneumatic compression, were excluded because the patients had been inadvertently discharged from the hospital without a venous duplex scan having been performed. One hip, also in a patient who was managed with intermittent pneumatic compression, was excluded because the postoperative scan had shown chronic thrombotic changes, precluding a determination of a recent postoperative deep-vein thrombosis. Preoperative duplex scans were not routinely performed in this study12. All patients were instructed to stop taking aspirin and other nonsteroidal anti-inflammatory medication one to two weeks before the operation. The 425 patients (502 hips) included in the study and the four patients (seven hips) who had not had a postoperative scan were managed both intraoperatively and postoperatively with intermittent pneumatic compression. All patients wore bilateral thigh-high antithromboembolic stockings and intermittent pneumatic compression devices (both manufactured by Kendall, Boston, Massachusetts) from the time of the operation until the venous duplex scan was performed, at a mean of six days (range, two to fifteen days) postoperatively. The patients were permitted to remove the intermittent pneumatic compression devices while bathing or participating in physical therapy. The devices consisted of a sleeve with four transverse calf chambers and two transverse thigh chambers. The chambers inflated sequentially for eleven seconds each, beginning distally and continuing proximally. There was a sixty-second deflation period between each cycle. A nonsterile stocking and intermittent pneumatic compression sleeve were applied to the uninvolved limb before the patient arrived in the operating room. After the induction of anesthesia and sterile draping of the involved limb, a sterile thigh-high intermittent pneumatic compression sleeve was applied over the sterile stockinette, and compression was begun before the incision. It was not difficult to maintain sterility with use of this device, as the long plastic hose was attached at the knee with an adhesive drape before it was dropped off the operating table for attachment to the pump. Two hundred and ninety-nine of the arthroplasties were performed in women, and 203 were done in men. The mean age was 61.5 years (range, eighteen to ninety-one years). The mean height was 168 centimeters (range, 140 to 196 centimeters), and the mean weight was seventy-five kilograms (range, thirty-six to 142 kilograms). The mean body-mass index was 26.6 kilograms per square meter (range, 14.9 to 49.1 kilograms per square meter). Three hundred and twenty-four procedures were primary total hip arthroplasties. The preoperative diagnosis was osteoarthritis in 179 hips (55 percent), avascular necrosis in sixty-one hips (19 percent), rheumatoid arthritis in thirty-nine hips (12 percent), arthritis secondary to a pediatric or adolescent hip disorder in thirteen hips (4 percent), fracture in ten hips (3 percent), posttraumatic arthritis in seven hips (2 percent), and a miscellaneous condition in fifteen hips (5 percent) that had a primary arthroplasty. Two hundred and twenty-nine primary procedures were so-called hybrid total hip arthroplasties (performed with insertion of the acetabular component without cement and the femoral component with cement), ninety-four were performed without cement, and one was performed with cement. The remaining 178 procedures were revision total hip arthroplasties. Eighty-two revisions involved both the acetabular and the femoral component; sixty-two of these procedures were hybrid total hip arthroplasties, and twenty were performed without cement. Fifty-two revisions were of the acetabular component only, twenty-six consisted of exchange of the acetabular liner and of the femoral component, eleven involved the femoral component only, and seven were revisions of the acetabular liner and the femoral head only. The posterior approach was used in 372 (74 percent) of the 502 hips; the anterior approach, in seventy-eight (16 percent); and the transtrochanteric approach, in fifty-two (10 percent). All patients were strongly encouraged to have regional anesthesia, if possible, because previous studies have shown a lower prevalence of deep-vein thrombosis with the use of that type of anesthesia11,13. Regional anesthesia was used either alone or in combination with general anesthesia in 440 procedures (88 percent). General anesthesia alone was used in sixty-two procedures (12 percent). The type of anesthesia was not chosen randomly as regional anesthesia was refused by some patients and was not technically possible in others. The distributions of patients according to gender, age, and type of operation (primary or revision) were not equal between the anesthesia groups (Table I).
The patients were assessed preoperatively for the presence of so-called classic risk factors (a history of thromboembolism, obesity, a history of cancer, congestive heart failure, atherosclerotic-hypertensive cardiovascular disease, and diabetes) for the development of deep-vein thrombosis. Patients who had more than one procedure during the course of the study were assessed separately during each hospitalization. Two hundred and thirty-one procedures (46 percent) were performed in patients who had atherosclerotic-hypertensive cardiovascular disease; 105 (21 percent), in obese patients (obesity was defined as a body-mass index of more than thirty kilograms per square meter); fifty-two (10 percent), in patients who had a history of cancer; thirty-eight (8 percent), in diabetic patients; twenty-seven (5 percent), in patients who had a history of thromboembolic events; and twenty (4 percent), in patients who had congestive heart failure. Three hundred and twelve (62 percent) of the procedures were done in patients who had at least one of these risk factors and 128 (25 percent), in those who had at least two. At the beginning of the study period, patients routinely began walking in the physical therapy department on the third or fourth postoperative day, depending on the level of pain. Duplex scans were generally performed on the seventh to tenth postoperative day. Because of economic constraints, in 1994 the postoperative protocol was changed so that walking in the physical therapy department was begun on the first or second postoperative day, depending on the level of pain, and duplex scans were generally performed on the fourth postoperative day. All patients were encouraged to perform exercises in bed immediately after the operation. Vascular technologists with a mean of fifteen years of experience performed bilateral venous duplex ultrasonography on all patients with use of an Acuson XP-5 Color Flow Duplex Scanner (Mountain View, California). Ultrasound examinations for venous thromboembolism have been performed at our institution for twenty-five years and have been previously validated7,15. In a study from our institution, in which 244 patients were examined both with an ultrasonic Doppler technique and with venography, the Doppler examination was found to have an accuracy of 94 percent, a specificity of 96 percent, and a sensitivity of 86 percent7. In another study from our institution, in which a Doppler examination, iodine-15-fibrinogen, and venography were used after twenty-eight total hip arthroplasties in patients who had been given a variety of pharmacological agents, the mean duration until the Doppler examination showed positive findings was three days (range, one to ten day)15. Real-time imaging information was obtained from the external iliac vein through the proximal portions of the calf vessels. A scan was considered negative if it showed complete coaptation of the walls of the vessels. In the present study, patients who were seen to have an asymptomatic proximal or distal deep-vein thrombosis on a duplex scan were routinely managed for three months with low-dose warfarin alone as part of the protocol, as described previously12. All other patients were advised to take 650 milligrams of aspirin daily after discharge, but there was no method to determine whether the patients complied. All patients included in this study were clinically evaluated for the symptoms of deep-vein thrombosis and pulmonary embolism for at least one year after discharge.
It was our overall subjective impression that the nursing staff's compliance with the administration of the intermittent pneumatic compression prophylaxis and the patients' tolerance of the treatment were excellent. No patient had a symptomatic deep-vein thrombosis while in the hospital. An asymptomatic deep-vein thrombosis was identified on the duplex scans made after twenty-three arthroplasties (4.6 percent). There were nineteen (eighteen ipsilateral and one contralateral) proximal deep-vein thromboses (3.8 percent) and four (one ipsilateral and three contralateral) distal deep-vein thromboses (0.8 percent). With the numbers available, we were unable to detect significant differences between patients with and without deep-vein thrombosis with respect to mean age (p = 0.76), gender (p = 0.13), body-mass index (p = 0.12), type of arthroplasty (primary or revision) (p = 0.12), operative approach (p = 0.37), duration of the operation (p = 0.21), type of anesthesia (general or regional) (p = 0.51), units of blood transfused (autologous, p = 0.79; homologous, p = 0.57), blood type (p = 0.18), or the presence of a so-called classic risk factor (p = 0.22) (Table II). Patients who were seventy-five years of age or older actually had a lower prevalence of deep-vein thrombosis (two of 100 procedures; 2 percent) than those who were less than seventy-five years of age (twenty-one of 402 procedures; 5 percent), but this difference was not found to be significant with the numbers available (p = 0.28, chi-square test; 1 - ß = 0.37). Deep-vein thrombosis developed in four (6 percent) of the sixty-two hips in patients who had received general anesthesia alone and in nineteen (4 percent) of the 440 hips in patients who had received regional anesthesia (p = 0.51, Fisher's exact test; 1 - ß = 0.12). Deep-vein thrombosis developed after seven (4 percent) of the 190 procedures in patients who did not have a classic risk factor and after sixteen (5 percent) of the 312 procedures in patients who had at least one risk factor (p = 0.22, chi-square test; 1 - ß = 0.25). There was no difference in the prevalence of deep-vein thrombosis between the patients who had duplex ultrasonography on the second through fifth postoperative days (eleven of 251) and those who had the test on or after the sixth postoperative day (twelve of 251). A power analysis was performed, and it was estimated that between 1500 and 2000 patients would be required for a study to determine a significant association between these variables and the prevalence of deep-vein thrombosis.
A symptomatic pulmonary embolism developed in three patients (a prevalence of 0.6 percent), all women, and was confirmed with a ventilation-perfusion lung scan while the patients were in the hospital. These patients were managed with heparin followed by warfarin for six months, and there were no more complications. Two of the women, who were sixty-six and seventy-five years of age, had negative duplex ultrasonography before chest pain developed, on the tenth and fourth postoperative days, respectively. The third woman, who was sixty-eight years of age, had chest pain on the sixth postoperative day; she had had a positive scan that showed an ipsilateral femoral-vein thrombosis. There were no other differences between these patients and the rest of the study group. A wound hematoma, defined as a large amount of swelling of the buttock and proximal aspect of the thigh associated with a change in the hemoglobin level, developed after five (1.0 percent) of the procedures. Four of the hematomas drained spontaneously, and the incision healed without additional sequelae. The fifth hematoma, which was large and involved the buttock and thigh, was treated with operative drainage. There were no other bleeding complications. As mentioned, all patients were clinically evaluated for symptoms of deep-vein thrombosis and pulmonary embolism for at least one year after discharge from the hospital. Three symptomatic ipsilateral deep-vein thromboses (a prevalence of 0.6 percent) developed after the patients were discharged from the hospital. Two of these patients, a seventy-seven-year-old man and a forty-two-year-old man who had both had regional anesthesia, had a proximal deep-vein thrombosis, with symptoms developing twenty-three and 158 days postoperatively, respectively. The third patient, a sixty-five-year-old woman who also had had spinal anesthesia, had a distal deep-vein thrombosis, with the symptoms developing sixty-two days postoperatively. No patient had a symptomatic pulmonary embolism after discharge from the hospital, and there were no fatal pulmonary embolisms. There were no symptomatic deep-vein thromboses or pulmonary embolisms after the arthroplasties in the thirteen hips that were excluded from the study.
In this prospective study, deep-vein thrombosis was detected by postoperative duplex ultrasonography after 4.6 percent (twenty-three) of 502 total hip arthroplasties in patients managed with thigh-high antithromboembolic stockings and thigh-high intermittent pneumatic compression devices. Proximal deep-vein thrombosis developed after 3.8 percent (nineteen) of the 502 procedures. It is important to note that the mechanical prophylaxis was begun on both limbs before the incision was made. All patients were evaluated for the presence of thrombi by experienced vascular technologists using duplex ultrasonography, at a mean of six days postoperatively. A symptomatic pulmonary embolism developed in three patients (a prevalence of 0.6 percent) while they were in the hospital, but no patient had a fatal pulmonary embolism. The 0.6 percent prevalence of symptomatic, nonfatal pulmonary embolism compares favorably with the rates of 0.7 percent for patients managed with low-molecular-weight heparin (in a 1994 meta-analysis of 2065 patients in twenty studies6) and 2.7 percent for those managed with warfarin (in a 1994 meta-analysis of 864 patients in ten studies6). The prevalence of thromboembolism with the mechanical prophylaxis used in our study is similar to those reported for pharmacological prophylaxis with either warfarin or low-molecular-weight heparin. As mentioned, Imperiale and Speroff found a 5 percent rate of proximal deep-vein thrombosis in studies of patients managed with warfarin and a 6 percent rate in studies of patients managed with low-molecular-weight heparin6. The 3.8 percent rate of proximal deep-vein thrombosis in the present study is also comparable with the 4 percent rate of proximal deep-vein thrombosis (ten of 233) reported by Woolson after a similar prophylactic protocol was carried out in association with procedures performed with regional anesthesia20. These results should also be viewed with the knowledge that our study included patients with the so-called classic risk factors for the development of deep-vein thrombosis, such as a history of cancer or thromboembolism, atherosclerotic-hypertensive cardiovascular disease, diabetes, obesity, and congestive heart failure, as well as patients who had a revision arthroplasty. The low prevalence of deep-vein thrombosis in our study suggests that intermittent thigh-high pneumatic compression devices, in conjunction with duplex ultrasonography surveillance, may be sufficient prophylaxis even in high-risk groups. Intermittent pneumatic compression has not been associated with a high rate of bleeding complications after primary or revision hip arthroplasty. In a randomized study, major bleeding was a complication in 1.5 percent (six) of 397 patients who had been given warfarin and 2.8 percent (eleven) of 398 patients who had been given low-molecular-weight heparin5. A wound hematoma occurred in 2.5 percent (ten) of 397 patients who had been managed with warfarin and 5.8 percent (twenty-three) of 398 patients who had been managed with low-molecular-weight heparin. In the present study, although a postoperative wound hematoma developed in five patients, only one hematoma necessitated operative evacuation. No patient had any other bleeding complication associated with the arthroplasty or the prophylaxis. During the course of the study, the day on which the duplex scan was generally performed was changed, as a result of an altered economic environment, from the seventh to tenth postoperative day to the fourth postoperative day. One study suggested that a longer delay in testing after total hip arthroplasty might increase the prevalence of positive duplex scans, presumably as a result of the propagation of smaller thrombi20. However, we found no difference in the prevalence of deep-vein thrombosis between the patients who had duplex ultrasonography two to five days postoperatively and those who had the test on or after the sixth postoperative day. This finding suggests that the asymptomatic deep-vein thromboses that were detected on the postoperative duplex scans had most likely developed intraoperatively or in the early postoperative period. In a previous study from our institution, in which twenty-eight patients had a Doppler examination every other day after the operation, deep-vein thrombosis developed at a mean of three days (range, one to ten days) postoperatively15. Several studies have demonstrated kinking and subsequent occlusion of the common femoral vein during total hip arthroplasty14,17,21. Venous stasis obviously plays a large role in the development of deep-vein thrombosis during total hip arthroplasty. Wuh et al. showed that intraoperative intermittent pneumatic compression can help to maintain blood flow through the kinked vein, thereby decreasing the prevalence of deep-vein thrombosis21. The intermittent pneumatic compression prophylaxis in the present study was initiated before the incision was made, and this factor may be related to the low prevalence of deep-vein thrombosis. Stamatakis et al. specifically associated kinking of the common femoral vein with posterior and transtrochanteric approaches to the hip17. However, in the present study, no association between the operative approach (anterior, posterior, or transtrochanteric) and the development of deep-vein thrombosis was detected (p = 0.37). We also found no association between the type of operation (primary or revision) (p = 0.12) or the duration of the operation (p = 0.21) and the development of deep-vein thrombosis. A power analysis showed that between 1540 and 2910 hip arthroplasties would be required for a study to determine a significant difference. It is possible that, with a larger number of patients, an association between the prevalence of deep-vein thrombosis and the approach, type, and duration of the operation could become apparent. Several studies have demonstrated a decreased prevalence of deep-vein thrombosis after use of spinal or epidural anesthesia during total hip arthroplasty3,11,13,16. In the present study, however, although there was a trend suggesting that the use of regional anesthesia might be protective, we could not detect a significant difference between general and regional anesthesia with regard to the development of deep-vein thrombosis (p = 0.51, Fisher's exact test; 1 - ß = 0.12). As 88 percent (440) of the 502 procedures were performed with regional anesthesia, the study did not have sufficient power to detect such small variations (8100 procedures would have been required). Three symptomatic deep-vein thromboses (a prevalence of 0.6 percent) developed after a postoperative duplex scan revealed negative findings and the patients were discharged from the hospital. Woolson postulated that such late thrombi might develop through propagation from a small calf thrombus that was not detected at the time of the original scan20. This finding emphasizes the importance of educating patients about seeking prompt medical attention if symptoms of deep-vein thrombosis occur after discharge. In the present study, both lower extremities were evaluated for the presence of deep-vein thrombosis with duplex scans. Only four thromboses (a prevalence of 0.8 percent) developed in the contralateral extremity. The low prevalence of contralateral thrombosis is similar to that reported in the study by Woolson et al., in which a contralateral proximal deep-vein thrombosis developed in only one (0.7 percent) of 152 patients19. The present study showed the efficacy of prophylaxis against thromboembolism with intermittent pneumatic compression after either primary or revision total hip arthroplasty. The intermittent pneumatic compression devices used intraoperatively and postoperatively reduced the rates of both deep-vein thrombosis and pulmonary embolism compared with those for historical controls who had received no prophylaxis. The overall prevalence of thromboembolism was similar to that following prophylaxis with either warfarin or low-molecular-weight heparin, but there were fewer serious bleeding complications. In this prospective study of 502 consecutive total hip arthroplasties, intermittent pneumatic compression combined with duplex ultrasonography performed by experienced vascular technologists was found to be a safe and effective method of preventing both deep-vein thrombosis and pulmonary embolism, regardless of a patient's risk factors for these complications.
*One or more of the authors has received or will receive benefits for personal or professional use from a commercial party related directly or indirectly to the subject of this article. In addition, benefits have been or will be directed to a research fund, foundation, educational institution, or other nonprofit organization with which one or more of the authors is associated. Funds were received in total or partial support of the research or clinical study presented in this article. The funding source was the Kendall Corporation.
  CiteULike  Connotea  Del.icio.us  Facebook  Technorati  Twitter What's this?
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |