Session I - Pelvis

Deep Vein Thrombosis in Patients with Pelvic or Acetabular Fractures: A Review of 486 Patients

Drake S. Borer, MD; Ashutosh V. Rao, MD; Adam J. Starr, MD; Joan Walker, RN; Shelly Whitlock, CCRA; Charles M. Reinert, MD; University of Texas Southwestern Medical Center, Dallas, TX

Purpose: The purpose of this study was to document the rate of deep venous thrombosis (DVT) in 486 patients with fractures of the pelvis and acetabulum, and to examine the efficacy of screening for DVT in these patients.

Methods: Data were obtained from a prospectively compiled trauma database and were retrospectively analyzed for purposes of this study. Findings were confirmed by examining appropriate radiological reports. Patients admitted to our center with the diagnosis of a pelvic or acetabular fracture, or both, between November 1, 1997 and November 30, 1999 were treated according to a standard protocol for DVT prophylaxis and screening. Beginning on the day of admission, patients were treated with subcutaneous unfractionated heparin, 5000 units 3 times per day. In addition, patients underwent mechanical prophylaxis against DVT using either pneumatic foot pumps or sequential compression devices (SCDs). Screening for DVT was accomplished with bilateral thigh ultrasound and pelvic magnetic resonance venography (MRV). Our goal was to obtain these studies within 3 days of injury. If screening studies showed a DVT, either systemic heparinization was begun or a filter was placed in the inferior vena cava.

Results: Data from 486 patients with fractures of the acetabulum or pelvis (or both) were compiled during the study period.

Two hundred thirty-seven patients underwent study by MRV; 13 MRVs were read as positive for DVT within the pelvis and 19 MRVs were read as inconclusive or indeterminate. Two hundred five MRVs were read as negative for DVT within the pelvis, and 249 patients did not undergo a MRV. Of these, 36 died prior to discharge, and 153 were discharged prior to MRV. All 153 patients were seen at least once as an outpatient for follow-up.

Three hundred two patients underwent study by Duplex Doppler (sonogram). Eight sonograms were read as positive, 293 sonograms were negative, and 1 sonogram was read as inconclusive. Two hundred twenty-three patients were studied with both sonogram and MRV, and two of the patients with a positive sonogram also had a positive MRV.

Ten patients were diagnosed as having a pulmonary embolism (PE). This diagnosis was confirmed by pulmonary arteriography (PA-gram) in 8 patients, high probability ventilation-perfusion nuclear medicine scan (V/Q scan) in 1 patient, and by autopsy in 1 patient. PE was determined to be the cause of death in 3 patients. Seven of the 10 patients with a PE underwent study by MRV. None of these studies was positive, 1 of these studies was read as indeterminate, and 6 of these studies were read as negative.

Nine of the 10 patients with a PE underwent study by sonogram and were read as negative for DVT.

The total number of patients with DVT was calculated by adding the number of patients with a PE to the number of patients with a positive result of either MRV or sonogram. Patients who had an overlap of any of these 3 variables were only counted once. Twenty-nine patients had a DVT. The DVT rate for the 486 patients was 6%, and the rate of PE was 2%.

Discussion: Patients with pelvic or acetabular fractures are at high risk for development of DVT if no prophylaxis is used (1). The best method to screen for DVT remains controversial. MRV is noted to be highly sensitive, but its specificity has been reported to be poor (2,3). Thus, it may be that several of the patients with a positive MRV in our study had, in fact, no DVT. It is possible that these patients had an IVCF placed unnecessarily. For this reason, our current practice is to confirm positive findings on MRV with a selective venogram. Early results of this practice have revealed that MRV is, in fact, often falsely positive. But, we have too little experience to date to state the actual incidence of falsely positive MRV's. Furthermore, in our study, MRV did not prove to be protective against the clinically relevant result of a DVT, pulmonary embolism. None of the patients diagnosed with a pulmonary embolism had a DVT detected by MRV. This study clearly demonstrates that the prophylactic regimen employed is effective in preventing DVT and PE.

References:

1. Geertz WH, Code KI, Jay RM, et al.: A prospective study of venous thromboembolism after major trauma. N Engl J Med 1994; 331: 1601-1606.

2. Montgomery KD, Potter HG, Helfet D: Magnetic resonance venography to evaluate the deep venous system of the pelvis in patients who have an acetabular fracture. J Bone Joint Surg 77-A: 1639-1649, 1995.

3. Stover MD, Morgan SJ, Bosse MJ, et al.: Prospective comparison of contrast enhanced computed tomography vs magnetic resonance venography in the detection of occult deep pelvic vein thrombosis in patients with pelvic and acetabular fractures. Presented at the annual meeting of the Orthopaedic Trauma Association, Vancouver, B.C., October 8-10, 1998.