SCIENTIFIC PAPERS
Venous Thromboembolism in Patients with Major Trauma Steven R. Shackford, MD, FACS, James W. Davis, MD, Peggy Hollingsworth-Fridlund, RN, Nancy S. Brewer, RCP, David B. Hoyt, MD, FACS, Robert C. Mackersie, MD, San Diego,California
The risk of venous thromboembolism after trauma is thought to be high, but the specific risk factors and the incidence of venous thromboembolism in the trauma population are poorly defined. Between October 1, 1987, and March 1, 1988, 719 patients were evaluated; 5 4 2 had no risk factors and 177 had at least 1 risk factor. No venous thromboembolism occurred in any of the 5 4 2 patients without a risk factor, whereas 12 of 177 patients ( 7 % ) with at least 1 risk factor had a venous thromboembolism. Pneumatic compression hose was the most common form of prophylaxis used, but it could not be applied to 35% of limbs because of plaster immobilizers, external fixators, complex wounds, or traction. In the high-risk group, 25 patients ( 1 4 % ) received no prophylaxis because of a physical impediment to application of these hose and a contraindication to antieoagulation. Age greater than 45 years was the only risk factor predictive of venous thromboembolism by logistic regression analysis. Patients with more than one risk factor had a significantly higher incidence of venous thromboembolism than those with only one risk factor. We conclude that a selected subgroup of trauma patients appears to be at risk of venous thromboembolism and should receive prophylaxis. Approximately one in seven high-risk patients cannot receive anticoagulant or mechanical prophylaxis because of their injuries.
he riskofvenous thromboembolism (VTE) in patients with a major traumatic injury is thought to be high The reported incidence of venous thrombosis in patients with trauma varies between 20% and 90%, whereas the incidence of pulmonary embolism is reported to be between 4% and 22% [2]. As a result, prophylaxis is recommended [3-5]. However, these incidence data are skewed by the inclusion of elderly patients with hip fractures and patients with paraplegia, groups known to be at high risk. The true rate of VTE in young, healthy trauma patients with multiple major injuries is unknown. Knowledge of the relative risk of VTE is important in making decisions regarding the use of prophylaxis. We undertook this prospective study to determine which patients with major injury were at risk for VTE, how frequently conventional prophylactic measures were impractical or inadvisable, and the frequency of complications associated with prophylaxis in patients with major trauma.
T [1,2].
PATIENTS AND M E T H O D S Between October 1, 1987, and March 1, 1988, all major trauma patients triaged to the University of California, San Diego, Trauma Center according to San Diego County Regional Trauma system guidelines [6] were evaluated by one of the authors (SRS, JWD, RCM, or DBH) for factors thought to increase the risk of VTE (Table I). If none of the factors was present, the patient was assigned to the low-risk group. The low-risk group was followed clinically for signs and symptoms of VTE during hospitalization and follow-up. Clinical suspicion of venous thrombosis or pulmonary embolism was an indication for objective evaluation (see following). When patients in the low-risk group were transferred to another service (Orthopedics or Neurosurgery, among others), they were followed by a trauma nurse coordinator (PH) and one of the authors (SRS). Patients with one or more risk factors were assigned to the high-risk group. Highrisk patients were screened with impedance plethysmography or two-dimensional ultrasound before the use of any type of prophylaxis. All noninvasive studies were performed by one of the authors (NSB) and three vascular technicians. Patients with negative results were studied at approximately weekly intervals during prophylaxis. Equivocal studies were repeated the following day. Patients with positive and repeatedly equivocal examinations underwent venography. All patients suspected of having deep venous thrombosis of the lower extremity From the Divisionof Trauma, Departmentof Surgery,Universityof based on clinical findings (swelling, pain, or calf tenderCalifornia,San DiegoMedical Center,San Diego,California. ness) or a positive or equivocal impedance plethysmograRequests for reprints shouldbe addressedto StevenR. Shackford, phy underwent venography. The diagnosis of pulmonary MD, FACS, Departmentof Surgery,Universityof Vermont,The Giv- embolism was pursued on the basis of clinical suspicion en Building,Burlington,Vermont05405. Manuscript submitted May 16, 1989, revisedJuly 25, 1989, and (tachycardia, tachypnea, or chest pain, among other symptoms). Pulmonary embolism was confirmed by acceptedAugust 2, 1989. THE AMERICAN JOURNALOF SURGERY VOLUME159 APRIL 1990 365
SHACKFORD ET AL
TABLE II Comparison of High- and Low-Risk Groups
TABLE I Risk Factors 1. 2. 3. 4. 5. 6. 7. 8. 9.
Age >45 years and enforced bed rest >3 days Previous history of venous thromboembolism Spine fracture without neurologic deficit Coma (Glascow Coma Scale
Venous Thromboembolism in Patients with Major Trauma Steven R. Shackford, MD, FACS, James W. Davis, MD, Peggy Hollingsworth-Fridlund, RN, Nancy S. Brewer, RCP, David B. Hoyt, MD, FACS, Robert C. Mackersie, MD, San Diego,California
The risk of venous thromboembolism after trauma is thought to be high, but the specific risk factors and the incidence of venous thromboembolism in the trauma population are poorly defined. Between October 1, 1987, and March 1, 1988, 719 patients were evaluated; 5 4 2 had no risk factors and 177 had at least 1 risk factor. No venous thromboembolism occurred in any of the 5 4 2 patients without a risk factor, whereas 12 of 177 patients ( 7 % ) with at least 1 risk factor had a venous thromboembolism. Pneumatic compression hose was the most common form of prophylaxis used, but it could not be applied to 35% of limbs because of plaster immobilizers, external fixators, complex wounds, or traction. In the high-risk group, 25 patients ( 1 4 % ) received no prophylaxis because of a physical impediment to application of these hose and a contraindication to antieoagulation. Age greater than 45 years was the only risk factor predictive of venous thromboembolism by logistic regression analysis. Patients with more than one risk factor had a significantly higher incidence of venous thromboembolism than those with only one risk factor. We conclude that a selected subgroup of trauma patients appears to be at risk of venous thromboembolism and should receive prophylaxis. Approximately one in seven high-risk patients cannot receive anticoagulant or mechanical prophylaxis because of their injuries.
he riskofvenous thromboembolism (VTE) in patients with a major traumatic injury is thought to be high The reported incidence of venous thrombosis in patients with trauma varies between 20% and 90%, whereas the incidence of pulmonary embolism is reported to be between 4% and 22% [2]. As a result, prophylaxis is recommended [3-5]. However, these incidence data are skewed by the inclusion of elderly patients with hip fractures and patients with paraplegia, groups known to be at high risk. The true rate of VTE in young, healthy trauma patients with multiple major injuries is unknown. Knowledge of the relative risk of VTE is important in making decisions regarding the use of prophylaxis. We undertook this prospective study to determine which patients with major injury were at risk for VTE, how frequently conventional prophylactic measures were impractical or inadvisable, and the frequency of complications associated with prophylaxis in patients with major trauma.
T [1,2].
PATIENTS AND M E T H O D S Between October 1, 1987, and March 1, 1988, all major trauma patients triaged to the University of California, San Diego, Trauma Center according to San Diego County Regional Trauma system guidelines [6] were evaluated by one of the authors (SRS, JWD, RCM, or DBH) for factors thought to increase the risk of VTE (Table I). If none of the factors was present, the patient was assigned to the low-risk group. The low-risk group was followed clinically for signs and symptoms of VTE during hospitalization and follow-up. Clinical suspicion of venous thrombosis or pulmonary embolism was an indication for objective evaluation (see following). When patients in the low-risk group were transferred to another service (Orthopedics or Neurosurgery, among others), they were followed by a trauma nurse coordinator (PH) and one of the authors (SRS). Patients with one or more risk factors were assigned to the high-risk group. Highrisk patients were screened with impedance plethysmography or two-dimensional ultrasound before the use of any type of prophylaxis. All noninvasive studies were performed by one of the authors (NSB) and three vascular technicians. Patients with negative results were studied at approximately weekly intervals during prophylaxis. Equivocal studies were repeated the following day. Patients with positive and repeatedly equivocal examinations underwent venography. All patients suspected of having deep venous thrombosis of the lower extremity From the Divisionof Trauma, Departmentof Surgery,Universityof based on clinical findings (swelling, pain, or calf tenderCalifornia,San DiegoMedical Center,San Diego,California. ness) or a positive or equivocal impedance plethysmograRequests for reprints shouldbe addressedto StevenR. Shackford, phy underwent venography. The diagnosis of pulmonary MD, FACS, Departmentof Surgery,Universityof Vermont,The Giv- embolism was pursued on the basis of clinical suspicion en Building,Burlington,Vermont05405. Manuscript submitted May 16, 1989, revisedJuly 25, 1989, and (tachycardia, tachypnea, or chest pain, among other symptoms). Pulmonary embolism was confirmed by acceptedAugust 2, 1989. THE AMERICAN JOURNALOF SURGERY VOLUME159 APRIL 1990 365
SHACKFORD ET AL
TABLE II Comparison of High- and Low-Risk Groups
TABLE I Risk Factors 1. 2. 3. 4. 5. 6. 7. 8. 9.
Age >45 years and enforced bed rest >3 days Previous history of venous thromboembolism Spine fracture without neurologic deficit Coma (Glascow Coma Scale
