The Incidence of Pulmonary Embolism After Combat Related Traumatic Amputation Is Higher Than We Thought: An Analysis of 366 Combat Casualties

Matthew D Tadlock, MD1, Matthew Hannon, MD2, Ted Melcer, PhD3, Jay Walker, BA3, Jesse Bandle, MD2, Kameran Nieses3, Michael Galarneau, MS, NREMT3. 1Naval Hospital Camp Pendleton, 2Naval Medical Center San Diego, 3Naval Health Research Center

Objective: Between 2001-2013, U.S. Soldiers, Sailors and Marines have suffered 1,558 major limb amputations during combat operations in Iraq and Afghanistan. Since 2009, 44% of these occurred in Afghanistan usually from a complex dismounted blast injury. Studies have shown that the incidence of pulmonary embolism (PE) after combat related extremity amputation is nearly ten-fold higher than in civilian traumatic amputees. Our goal was to determine the incidence and examine the risk factors for PE after combat related traumatic amputation.

Methods: Traumatic amputation (TA) was defined as surgical amputation within 48 hours of injury. The Expeditionary Medical Encounter Database and chart review were used to evaluate all patients suffering TA proximal to the wrist or ankle presenting to a Navy Role 2 or 3 facility from January 2009-December 2011.  Outcomes (PE and DVT) were followed from injury through 12 months.  PE risk factors were identified utilizing multivariable logistic regression.

Results: During the 3-year study period, 426 suffered a TA. Of the 366 with adequate records for review, 99.5% were male, 97.5% suffered a blast injury and 94.5% were injured in Afghanistan. Mean age was 24.3, median Injury Severity Score 21 and 86.9% received chemical prophylaxis.  PE incidence was 16% (59), of which only 25% (15) had a concomitant DVT identified. The DVT rate was 16% and a PE and/or DVT found in 28%. The overall mortality was low at 2.5%; PE was not found to be a contributing factor in any mortality. Those with at least 1 above knee amputation had a higher PE rate compared to those with a lower amputation level (21% vs.12.3%, p<0.04). The median ISS was higher in amputees with DVT compared to those without DVT (29 vs. 19, p<0.02) and in amputees with PE compared to those with no PE (26 vs. 21, p<0.03). The majority of patients, 318 (86.9%), received chemical VTE prophylaxis.  Massive transfusion (> 10 units) with packed red blood cells and/or fresh whole blood (PRBC/FWB) occurred in 64.2%.  As units of transfused PRBC/FWB and fresh frozen plasma increased, so did the PE incidence. Upon multivariable analysis, only increasing ventilator days (Odds Ratio [OR]1.97, 1.16-3.37 95% Confidence Interval [CI]) and units of PRBC+FWB transfused ( OR 1.72, 1.11-2.68 95% CI) were independently associated with the identification of a post-injury DVT. Only increasing units of FFP transfused was independently associated the identification of a PE (OR 1.31, 1.10-1.55, 95% CI). Receiving chemical VTE prophylaxis was independently associated with no post injury DVT (OR 0.37, 018-0.78, 95% CI) or PE (OR 0.34, 0.17-0.70, 95% CI) identification.

Conclusion: The incidence of post-injury PE in this large cohort of combat amputees is significantly higher than previously described and increased as the level of lower extremity amputation increased.  However, only increasing units of FFP transfused was independently associated with PE identification. Further, in this group of patients with severe and often multiple amputations, the incidence of PE without DVT was high, challenging the traditional DVT to PE sequence.

« Return to SAGES 2015 abstract archive