Prophylaxis
Against Venous Thromboembolism Surgical Patients
H. Brownell Wheeler,
MD,
Frederick A. Anderson, Jr., Pho,w~rcestet.
Venous thromboembolic disease is a common, potentially life-threatening, but preventable complication of surgery. Venographic studies have shown a high incidence of deep vein thrombosis in patients who do not receive prophylaxis. The long-term sequelae of deep vein thrombosis in these patients can be disabling, but the exact prevalence of postthrombotic complications has not been well documented. Fatal pulmonary embolism has been reported in 0.5% to 1% of patients after major abdominal surgery and in 2% to 6% after total hip replacement. Several methods of prophylaxis against venous thromboembolism have demonstrated efficacy, but a surprising number of patients still do not receive prophylaxis. A medical record review in 16 hospitals recently disclosed that only 32% of high-risk patients received prophylaxis, with a range from 9% in a community hospital to 56% in a major teaching hospital. The possible reasons for this comparatively low rate of utilization are discussed. Current issues and future trends in the prophylaxis of venous thromboembolism are briefly summarized.
From the Department of Surgery, University of Massachusetts Medical Center, Worcester, Massachusetts. Supported in part by grant no. ROl HL35862 from the National Heart. Lung. and Blood Institute. Requests for reprints should beaddressed to H. Brownell Wheeler, MD, Department of Surgery, University of Massachusetts Medical Center, 55 Lake Avenue North, Worcester, Massachusetts 01655. THE AMERICAN
in
Massachusetts
enous thromboembolism is widely recognized as a common, serious, and occasionally fatal complicaV tion of major surgery [I,2]. The time during which the patient is at risk is usually predictable. Methods of prophylaxis are available and have been shown to be effective [3-51, and yet, recent studies in the United States indicate that many surgical patients at risk for venous thromboembolism do not receive effective prophylaxis [6]. Some surgeons appear to have doubts about the need for prophylaxis, despite the extensive literature on this subject. Concern about possible hemorrhage in surgical patients receiving anticoagulant prophylaxis is a powerful deterrent to its use. New methods of prophylaxis may increase effectiveness, reduce complications, and encourage wider acceptance. One of the most promising is lowmolecular-weight heparin. This article provides a general overview of prophylaxis against venous thromboembolic disease in surgical patients. HISTORICAL
BACKGROUND
In 1856, Rudolf Virchow conducted meticulous autopsies on 76 patients, 11 of whom died from massive thrombi in the pulmonary arteries or right heart. In 10 of the 11 patients, he also found thrombi in the iliac and femoral veins. He concluded that the thrombi in the pulmonary arteries had in fact embolized from the pelvic and leg veins [7]. Over the next century, other autopsy series confirmed that pulmonary embolism was the immediate cause of death in 5% to 15% of patients who died in the hospital [8-101. In 1959, more than 100 years after Virchow’s original description, Sevitt and Gallagher [ 1 I] noted an extraordinarily high incidence of pulmonary embolism in patients with fractured hips. They found that pulmonary embolism accounted for 33% of all deaths in such patients and postulated that this high death rate could be diminished by prophylactic anticoagulation. In a landmark paper, they reported a reduction in mortality from pulmonary embolism from 10% to 0% among patients who received the oral anticoagulant phenindione prophylactically [ I1 1. This was the first randomized prospective trial of anticoagulant prophylaxis. Several authors had previously recommended anticoagulant prophylaxis in surgical patients on the basis of anecdotal reports and retrospective studies. The report of Sevitt and Gallagher [I I ] sparked intensive study of the incidence of deep vein thrombosis and pulmonary embolism in patients with hip fracture. In patients with hip fractures who do not receive prophylaxis, there is a 44% average incidence of venous thrombosis demonstrated by venography in nine studies conducted between 1965 and 1973. The average incidence of fatal pulmonary embolism in hip fracture patients is 5.9%, based on 13 studies conducted between 1969 and 1976
[121.
JOURNAL
OF SURGERY
VOLUME
161
APRIL
1991
507
TABLE I Average Incidence of Postoperative Venous Thrombosis In Surgical Patients Wlthout Prophylaxis Type of Surgery
No. of Studies
No. of Patients
Incidence
Elective knee surgery’ Elective hip surgery’ Hip fracture* Transvesical prostate surgery+ General surgery’ Neurosurgery+ Gynecologic surgery+ Transurethral prostate surgery+
1 5 9 13 28 6 a 7
29 335 485 a83 1,018 351 701 416
66% 52% 44% 38% 29% 29% 19% 11%
Data adapted from [ 12,26]. * Determined by venography. 1Determined by radiolabeled fibrinogen uptake.
TABLE II Average lncldence of Fatal Pulmonary Embollsm In Surgical Patlents Wlthout Prophylaxls Type of Surgery
No. of Studies
No. of Patients
Incidence
Hip fracture Elective hip surgery Abdominal trauma General surgery
13 5 11 6
1,040 249 2,071 9,390
5.9% 2.4% 1.1% 0.8%
Data adapted from
[ 121.
The high incidence of fatal pulmonary embolism in patients undergoing emergency surgery for hip fractures prompted careful investigation of patients undergoing elective hip surgery. Five studies conducted between 1974 and 1979 showed an average incidence of fatal pulmonary emboli of 2.4% after such surgery. The overall incidence of venous thrombosis in such patients proved to be 52% by venography. It is of interest that venography employed preoperatively disclosed a 7.5% incidence of venous thrombosis in these inactive patients [ 131. Extensive radiolabeled fibrinogen and venographic studies have been conducted to determine the incidence of venous thrombosis after other types of surgery. The overall incidence was much higher than surgeons had anticipated, varying from 11% of patients undergoing simple transurethral resection of the prostate to 66% of patients undergoing elective knee surgery (Table I). An excellent summary of this work and an extensive bibliography has been provided by Bergqvist in an authoritative monograph entitled, “Postoperative Thromboembolism” [IZ]. In addition to the high postoperative incidence of venous thrombosis, there also proved to be a 25.1% incidence of postoperative pulmonary embolism diagnosed by lung scan in a surgical population largely composed of patients undergoing abdominal surgery [IO]. A careful review of autopsy-verified pulmonary embolism was conducted by Bergqvist [I21 in 5,477 surgical deaths. The autopsy rate was 92%. Pulmonary emboli were demonstrated in 23.6% of all autopsies and were believed to be the primary cause of death in 6.4% of patients, constituting one of the most common sources of preventable mortality after surgery. The range of fatal pulmonary 508
THE AMERICAN
JOURNAL
OF SURGERY
VOLUME
embolism in postoperative patients who fail to receive prophylaxis is listed in Table II. After documentation of the high incidence of postoperative venous thrombosis and nonfatal pulmonary embolism, as well as the substantial incidence of fatal pulmonary embolism, many studies of thromboprophylaxis were performed in the 1970s and 1980s. An international multicenter trial of low-dose heparin prophylaxis reported reduced mortality from pulmonary emboli from 0.8% to 0.1% in general surgical patients and generated considerable enthusiasm for this method of prophylaxis [3]. A later meta-analysis of all low-dose heparin studies conducted in general surgical patients again showed a 0.8% incidence of fatal pulmonary embolism in the control group, with a 0.3% incidence of fatal pulmonary embolism in the treatment group [.5].On the basis of this metaanalysis, routine use of prophylactic low-dose heparin might be expected to save the life of approximately 1 of every 200 patients undergoing general surgery. A similar compilation of data with respect to thromboprophylaxis employing dextran 70 showed a reduction in fatal pulmonary embolism from 1.5% in the control group to 0.4% in the treatment group, or approximately 1 life saved of every 111 patients treated [12]. In addition to showing a reduction in fatal pulmonary embolism, studies showed that many prophylactic methods are effective in decreasing the incidence of venous thrombosis. Prevention of venous thrombosis may decrease the long-term incidence of postphlebitic syndrome. This may be a more common benefit of prophylaxis than prevention of life-endangering pulmonary embolism. Many years ago, Bauer [14] reported that patients with 161
APRIL 1991
PROPHYLAXIS
clinical venous thrombosis followed for more than 10 years developed a remarkably high incidence of leg swelling (91%) and ulceration (79%). Prevention of postoperative venous thrombosis clearly should be of concern to surgeons, irrespective of the compelling need to prevent fatal pulmonary embolism. Many methods of prophylaxis against venous thrombosis have been evaluated. They are generally divided into physical and pharmaceutical categories. In the United States, the most common physical modalities employed are venous gradient support stockings and intermittent pneumatic calf compression. The most common pharmacologic methods are low-dose heparin, oral anticoagulants, low-molecular-weight dextran, and antiplatelet agents such as aspirin. The scientific evidence concerning the efficacy of all these methods was reviewed at a National Institute of Health Consensus Conference in 1986. Wider prophylaxis was encouraged in both medical and surgical patients, and general guidelines were established [4]. Previously, the Council on Thrombosis of the American Heart Association had estimated that 4,000 to 8,000 lives might be saved each year if low-dose heparin prophylaxis was employed routinely in patients undergoing major general surgical procedures [15]. In short, both the medical literature and authoritative ad hoc review boards urge prophylaxis of venous thrombosis and pulmonary embolism in selected surgical patients. CONTEMPORARY INCIDENCE Despite the widespread literature supporting the efficacy of prophylaxis, and despite the lack of evidence of complications sufficient to contraindicate such treatment, several surveys have indicated that many surgeons still do not routinely prescribe prophylaxis for high-risk patients [6,16,17]. These surveys may be questioned since the information provided is subjective and nonverifiable. Surprisingly little is known about the utilization of prophylaxis in clinical practice. We have recently conducted a retrospective chart review of the management of venous thromboembolic disease in 16 participating hospitals in central Massachusetts. The study area included both rural and urban hospitals serving a population of 650,000 persons in an area of 2,500 square miles. In an 18-month period from July 1985 to December 1986, these 16 hospitals had a total of 151,349 discharges. This study has provided valuable insights about the number of patients who are candidates for prophylaxis, the number who actually receive prophylaxis, and the types of prophylaxis employed. Many patients proved to be candidates for prophylaxis. Any patient was considered at high risk of venous thrombosis and pulmonary embolism if he or she was at least 40 years of age, was hospitalized for more than 5 days, and had at least one other major risk factor, such as a major surgical procedure. By these criteria, 17% of all discharges were considered at moderate to high risk for venous thromboembolism. Recognized risk factors for venous thromboembolism proved to be common in hospitalized patients. Seventy-eight percent of all patients had one or more risk factors; 48% had two or more; 20%, three or more; 6%, four or more; and l%, five or more risk THE AMERICAN
AGAIIVST VENOUS THROMROEMBOLISM
factors. Using three or more risk factors as a criterion for prophylaxis, between 6 and 7 million patients each year might be candidates for thromboprophylaxis in the United States. This estimate is based on the fact that 20% of patients in this study had 3 or more risk factors and that there are 31 million hospital discharges in the United States each year. The number of patients who actually received prophylaxis in the central Massachusetts study proved to be considerably fewer than the number at risk. Only 32% of all high-risk patients (age greater than 40 years, length of hospital stay greater than 5 days, one or more additional major risk factors) received prophylaxis. There was a significant discrepancy in the utilization of prophylaxis in different hospitals, ranging from a low of 9% to a high of 56%. Utilization was lowest in nonteaching hospitals. Teaching hospitals had an average rate of prophylaxis use of 44%, whereas nonteaching hospitals had a rate of 19% (p
Against Venous Thromboembolism Surgical Patients
H. Brownell Wheeler,
MD,
Frederick A. Anderson, Jr., Pho,w~rcestet.
Venous thromboembolic disease is a common, potentially life-threatening, but preventable complication of surgery. Venographic studies have shown a high incidence of deep vein thrombosis in patients who do not receive prophylaxis. The long-term sequelae of deep vein thrombosis in these patients can be disabling, but the exact prevalence of postthrombotic complications has not been well documented. Fatal pulmonary embolism has been reported in 0.5% to 1% of patients after major abdominal surgery and in 2% to 6% after total hip replacement. Several methods of prophylaxis against venous thromboembolism have demonstrated efficacy, but a surprising number of patients still do not receive prophylaxis. A medical record review in 16 hospitals recently disclosed that only 32% of high-risk patients received prophylaxis, with a range from 9% in a community hospital to 56% in a major teaching hospital. The possible reasons for this comparatively low rate of utilization are discussed. Current issues and future trends in the prophylaxis of venous thromboembolism are briefly summarized.
From the Department of Surgery, University of Massachusetts Medical Center, Worcester, Massachusetts. Supported in part by grant no. ROl HL35862 from the National Heart. Lung. and Blood Institute. Requests for reprints should beaddressed to H. Brownell Wheeler, MD, Department of Surgery, University of Massachusetts Medical Center, 55 Lake Avenue North, Worcester, Massachusetts 01655. THE AMERICAN
in
Massachusetts
enous thromboembolism is widely recognized as a common, serious, and occasionally fatal complicaV tion of major surgery [I,2]. The time during which the patient is at risk is usually predictable. Methods of prophylaxis are available and have been shown to be effective [3-51, and yet, recent studies in the United States indicate that many surgical patients at risk for venous thromboembolism do not receive effective prophylaxis [6]. Some surgeons appear to have doubts about the need for prophylaxis, despite the extensive literature on this subject. Concern about possible hemorrhage in surgical patients receiving anticoagulant prophylaxis is a powerful deterrent to its use. New methods of prophylaxis may increase effectiveness, reduce complications, and encourage wider acceptance. One of the most promising is lowmolecular-weight heparin. This article provides a general overview of prophylaxis against venous thromboembolic disease in surgical patients. HISTORICAL
BACKGROUND
In 1856, Rudolf Virchow conducted meticulous autopsies on 76 patients, 11 of whom died from massive thrombi in the pulmonary arteries or right heart. In 10 of the 11 patients, he also found thrombi in the iliac and femoral veins. He concluded that the thrombi in the pulmonary arteries had in fact embolized from the pelvic and leg veins [7]. Over the next century, other autopsy series confirmed that pulmonary embolism was the immediate cause of death in 5% to 15% of patients who died in the hospital [8-101. In 1959, more than 100 years after Virchow’s original description, Sevitt and Gallagher [ 1 I] noted an extraordinarily high incidence of pulmonary embolism in patients with fractured hips. They found that pulmonary embolism accounted for 33% of all deaths in such patients and postulated that this high death rate could be diminished by prophylactic anticoagulation. In a landmark paper, they reported a reduction in mortality from pulmonary embolism from 10% to 0% among patients who received the oral anticoagulant phenindione prophylactically [ I1 1. This was the first randomized prospective trial of anticoagulant prophylaxis. Several authors had previously recommended anticoagulant prophylaxis in surgical patients on the basis of anecdotal reports and retrospective studies. The report of Sevitt and Gallagher [I I ] sparked intensive study of the incidence of deep vein thrombosis and pulmonary embolism in patients with hip fracture. In patients with hip fractures who do not receive prophylaxis, there is a 44% average incidence of venous thrombosis demonstrated by venography in nine studies conducted between 1965 and 1973. The average incidence of fatal pulmonary embolism in hip fracture patients is 5.9%, based on 13 studies conducted between 1969 and 1976
[121.
JOURNAL
OF SURGERY
VOLUME
161
APRIL
1991
507
TABLE I Average Incidence of Postoperative Venous Thrombosis In Surgical Patients Wlthout Prophylaxis Type of Surgery
No. of Studies
No. of Patients
Incidence
Elective knee surgery’ Elective hip surgery’ Hip fracture* Transvesical prostate surgery+ General surgery’ Neurosurgery+ Gynecologic surgery+ Transurethral prostate surgery+
1 5 9 13 28 6 a 7
29 335 485 a83 1,018 351 701 416
66% 52% 44% 38% 29% 29% 19% 11%
Data adapted from [ 12,26]. * Determined by venography. 1Determined by radiolabeled fibrinogen uptake.
TABLE II Average lncldence of Fatal Pulmonary Embollsm In Surgical Patlents Wlthout Prophylaxls Type of Surgery
No. of Studies
No. of Patients
Incidence
Hip fracture Elective hip surgery Abdominal trauma General surgery
13 5 11 6
1,040 249 2,071 9,390
5.9% 2.4% 1.1% 0.8%
Data adapted from
[ 121.
The high incidence of fatal pulmonary embolism in patients undergoing emergency surgery for hip fractures prompted careful investigation of patients undergoing elective hip surgery. Five studies conducted between 1974 and 1979 showed an average incidence of fatal pulmonary emboli of 2.4% after such surgery. The overall incidence of venous thrombosis in such patients proved to be 52% by venography. It is of interest that venography employed preoperatively disclosed a 7.5% incidence of venous thrombosis in these inactive patients [ 131. Extensive radiolabeled fibrinogen and venographic studies have been conducted to determine the incidence of venous thrombosis after other types of surgery. The overall incidence was much higher than surgeons had anticipated, varying from 11% of patients undergoing simple transurethral resection of the prostate to 66% of patients undergoing elective knee surgery (Table I). An excellent summary of this work and an extensive bibliography has been provided by Bergqvist in an authoritative monograph entitled, “Postoperative Thromboembolism” [IZ]. In addition to the high postoperative incidence of venous thrombosis, there also proved to be a 25.1% incidence of postoperative pulmonary embolism diagnosed by lung scan in a surgical population largely composed of patients undergoing abdominal surgery [IO]. A careful review of autopsy-verified pulmonary embolism was conducted by Bergqvist [I21 in 5,477 surgical deaths. The autopsy rate was 92%. Pulmonary emboli were demonstrated in 23.6% of all autopsies and were believed to be the primary cause of death in 6.4% of patients, constituting one of the most common sources of preventable mortality after surgery. The range of fatal pulmonary 508
THE AMERICAN
JOURNAL
OF SURGERY
VOLUME
embolism in postoperative patients who fail to receive prophylaxis is listed in Table II. After documentation of the high incidence of postoperative venous thrombosis and nonfatal pulmonary embolism, as well as the substantial incidence of fatal pulmonary embolism, many studies of thromboprophylaxis were performed in the 1970s and 1980s. An international multicenter trial of low-dose heparin prophylaxis reported reduced mortality from pulmonary emboli from 0.8% to 0.1% in general surgical patients and generated considerable enthusiasm for this method of prophylaxis [3]. A later meta-analysis of all low-dose heparin studies conducted in general surgical patients again showed a 0.8% incidence of fatal pulmonary embolism in the control group, with a 0.3% incidence of fatal pulmonary embolism in the treatment group [.5].On the basis of this metaanalysis, routine use of prophylactic low-dose heparin might be expected to save the life of approximately 1 of every 200 patients undergoing general surgery. A similar compilation of data with respect to thromboprophylaxis employing dextran 70 showed a reduction in fatal pulmonary embolism from 1.5% in the control group to 0.4% in the treatment group, or approximately 1 life saved of every 111 patients treated [12]. In addition to showing a reduction in fatal pulmonary embolism, studies showed that many prophylactic methods are effective in decreasing the incidence of venous thrombosis. Prevention of venous thrombosis may decrease the long-term incidence of postphlebitic syndrome. This may be a more common benefit of prophylaxis than prevention of life-endangering pulmonary embolism. Many years ago, Bauer [14] reported that patients with 161
APRIL 1991
PROPHYLAXIS
clinical venous thrombosis followed for more than 10 years developed a remarkably high incidence of leg swelling (91%) and ulceration (79%). Prevention of postoperative venous thrombosis clearly should be of concern to surgeons, irrespective of the compelling need to prevent fatal pulmonary embolism. Many methods of prophylaxis against venous thrombosis have been evaluated. They are generally divided into physical and pharmaceutical categories. In the United States, the most common physical modalities employed are venous gradient support stockings and intermittent pneumatic calf compression. The most common pharmacologic methods are low-dose heparin, oral anticoagulants, low-molecular-weight dextran, and antiplatelet agents such as aspirin. The scientific evidence concerning the efficacy of all these methods was reviewed at a National Institute of Health Consensus Conference in 1986. Wider prophylaxis was encouraged in both medical and surgical patients, and general guidelines were established [4]. Previously, the Council on Thrombosis of the American Heart Association had estimated that 4,000 to 8,000 lives might be saved each year if low-dose heparin prophylaxis was employed routinely in patients undergoing major general surgical procedures [15]. In short, both the medical literature and authoritative ad hoc review boards urge prophylaxis of venous thrombosis and pulmonary embolism in selected surgical patients. CONTEMPORARY INCIDENCE Despite the widespread literature supporting the efficacy of prophylaxis, and despite the lack of evidence of complications sufficient to contraindicate such treatment, several surveys have indicated that many surgeons still do not routinely prescribe prophylaxis for high-risk patients [6,16,17]. These surveys may be questioned since the information provided is subjective and nonverifiable. Surprisingly little is known about the utilization of prophylaxis in clinical practice. We have recently conducted a retrospective chart review of the management of venous thromboembolic disease in 16 participating hospitals in central Massachusetts. The study area included both rural and urban hospitals serving a population of 650,000 persons in an area of 2,500 square miles. In an 18-month period from July 1985 to December 1986, these 16 hospitals had a total of 151,349 discharges. This study has provided valuable insights about the number of patients who are candidates for prophylaxis, the number who actually receive prophylaxis, and the types of prophylaxis employed. Many patients proved to be candidates for prophylaxis. Any patient was considered at high risk of venous thrombosis and pulmonary embolism if he or she was at least 40 years of age, was hospitalized for more than 5 days, and had at least one other major risk factor, such as a major surgical procedure. By these criteria, 17% of all discharges were considered at moderate to high risk for venous thromboembolism. Recognized risk factors for venous thromboembolism proved to be common in hospitalized patients. Seventy-eight percent of all patients had one or more risk factors; 48% had two or more; 20%, three or more; 6%, four or more; and l%, five or more risk THE AMERICAN
AGAIIVST VENOUS THROMROEMBOLISM
factors. Using three or more risk factors as a criterion for prophylaxis, between 6 and 7 million patients each year might be candidates for thromboprophylaxis in the United States. This estimate is based on the fact that 20% of patients in this study had 3 or more risk factors and that there are 31 million hospital discharges in the United States each year. The number of patients who actually received prophylaxis in the central Massachusetts study proved to be considerably fewer than the number at risk. Only 32% of all high-risk patients (age greater than 40 years, length of hospital stay greater than 5 days, one or more additional major risk factors) received prophylaxis. There was a significant discrepancy in the utilization of prophylaxis in different hospitals, ranging from a low of 9% to a high of 56%. Utilization was lowest in nonteaching hospitals. Teaching hospitals had an average rate of prophylaxis use of 44%, whereas nonteaching hospitals had a rate of 19% (p
