Session VIII - Pelvis

*Mechanical DVT Prophylaxis in Pelvic and Acetabular Fractures following Blunt Trauma

James P. Stannard, MD (a-Nutech, Kinetic Concepts, Inc.); Renee S. Riley, MD (a-Nutech, Kinetic Concepts, Inc.); Jorge E. Alonso, MD (a-Nutech, Kinetic Concepts, Inc.); David A. Volgas, MD (a-Nutech, Kinetic Concepts, Inc.); Michelle D. McClenny, BSN (a-Nutech, Kinetic Concepts, Inc.), University of Alabama, Birmingham, AL

Purpose: The hypothesis of this study is that mechanical compression provides an excellent option for prophylaxis against deep vein thrombosis (DVT) in pelvic and acetabular fracture patients without the substantial risks of bleeding that effective anticoagulation provides. A second hypothesis is that pulsatile mechanical compression is superior to standard mechanical compression in pelvic and acetabular fracture patients.

Methods: This study is designed as a prospective, randomized, blinded evaluation of two methods of mechanical DVT prophylaxis. Inclusion criteria include blunt trauma with the presence of a pelvic or acetabular fracture whose pattern requires surgical fixation. Additional inclusion criteria include the patient being at least 18 years of age, with an ability and willingness to comply with both the mechanical prophylaxis protocol and to agree to undergo the appropriate diagnostic screening studies. Exclusion criteria include a prior history of venous thromboembolic disease, initiation of mechanical compression more that 72 hours following the injury, or a stable injury that does not require surgical stabilization. All patients were randomized into one of two groups: Group A was the standard compression group and utilized thigh-high sequential compression devices (Kendall SCD, Kendall Co., Mansfield, MA). Group B was the pulsatile compression group and utilized combination sequential pumps (PlexiPulse, Nutech, San Antonio). All patients underwent two different forms of screening studies for the presence of DVT. A duplex ultrasound was obtained as was a magnetic resonance venogram (MRV) near the conclusion of the patients' inpatient hospital stay. Data collected included the presence or absence of DVT and pulmonary embolus (PE), whether DVT's were occlusive/large non-occlusive or whether it was a small (<2 cm) non-occlusive clot, injury severity score (ISS) and the number of days from injury to surgery and >from injury to screening studies.

Results: One hundred and thirty-seven patients have been entered into the study to date. One hundred and seven (107) have successfully completed the protocol. Fifty-four patients have completed the protocol in Group A, with ten patients developing DVT's, seven developing large or occlusive clots, and one patient developing a PE. Fifty-three patients have completed the protocol in Group B, with five patients developing DVT's, two developing large or occlusive clots, and no patients developing PE. The difference between the two groups in large or occlusive clots demonstrates a strong trend, but is not yet statistically significant (P = 0.16). The average ISS in Group A was 19.8, in Group B it was 16.1. Screening studies were obtained a mean of 10.8 days after injury, 6.0 days following surgical stabilization in Group A. The screening studies were obtained a mean of 10.8 days following injury, 6.5 days following surgical stabilization in Group B. None of these differences was statistically significant between the two groups. The incidence of DVT in all patients with pelvic and acetabular trauma was 14%. The incidence of PE was 0.9%. The overall incidence of DVT in Group A was 18.5%, with 13% occlusive clots and PE's in 1.9%. The incidence of DVT in Group B was 9.4%, with 3.8% occlusive clots and with 0% pulmonary emboli. Duplex ultrasound and MRV results were in conflict in 10 patients. Seven had positive MRV's and negative ultrasounds, which were thought to be false negative ultrasounds. Five of these seven patients had thrombi in pelvic veins or in the most proximal portion of the femoral vein: three external iliac veins, one iliac vein, and one proximal femoral vein. Ultrasound is far less sensitive than MRV at detecting DVT's in the pelvic veins. An additional four patients had negative MRV's and a positive duplex ultrasound. Three of these represent false positive ultrasounds while one represents a false negative MRV. Of the eleven studies where MRV and ultrasound disagreed, MRV was determined to be correct on further review in nine of the cases. One patient who developed a massive pulmonary embolus had a negative duplex ultrasound and then had pulmonary angiography confirming the PE.

Discussion: MR venography has been proposed as an excellent study to detect DVT in patients following pelvic and acetabular fractures in a study by Montgomery et al. They noted high senstivity of the MRV when compared to contrast venography. They reported an incidence of DVT of 33% in their patients using subcutaneous heparin as prophylaxis. MRV is extremely sensitive, providing a clear evaluation of the pelvic veins that can not be obtained with most other screening tests for DVT. Numerous studies have identified pelvic and acetabular trauma as presenting a very high risk for venous thromboembolic disease, but none have randomized two treatment regimens in this patient population. Most trauma studies have included patients with a wide variety of orthopaedic injuries. Mechanical compression was associated with an incidence of 14%, with pulsatile compression having only 9.4%. It is difficult to compare studies that do not use MRV as a screening test, because of the increased sensitivity in the proximal veins. Fishmann et al. noted a 9% incidence of DVT using ultrasound as a screening test and a combination of mechanical and pharmacologic prophylaxis. Webb et al. noted an incidence of PE in patients with acetabular fractures of 7% based on clinical detection. MRV provides far superior visualization of the pelvic veins when compared with ultra sound. Metal artifact created some areas of inadequate resolution but was associated with only one known false negative scan. Forty-seven percent of the thrombi detected in our study were located in pelvic veins. The ultrasound failed to detect 80% of pelvic DVT's in cases with good ultrasounds and MRV's. MRV clearly provides a far superior study of the pelvic veins when compared to ultrasound, but also has some drawbacks. MRV is far more expensive, some patients were unable to tolerate the study due to claustrophobia, metal artifact presents some problems in terms of reading, and MRV is far less conducive to serial studies.

Conclusion: Mechanical compression provided efficacious prophylaxis against deep vein thrombosis in pelvic and acetabular trauma patients screened with both MR venography and ultrasound. Pulsatile compression was associated with fewer DVT's than standard compression, with the difference representing a statistical trend with current enrollment. Magnetic resonance venography provides an excellent evaluation of the pelvic venous system.