Session V - Femur

The Alveolar Dead Space Fraction Predicts Respiratory Distress after Intramedullary Fixation of Femoral Shaft Fractures

W. Christopher Patton, MD; Joseph N. Rudd, Jr., BSN; Lynn Crosby, MD; Jeffrey Kline, MD; Brent Norris, MD, University of Tennessee College of Medicine, Chattanooga, TN; Carolinas Medical Center, Charlotte, NC

Purpose: To determine if alveolar dead space fraction (Vd/Vt) predicts postoperative pulmonary dysfunction in patients undergoing intramedullary nailing (IMN) of femoral shaft fractures (FSF).

Methods: Studies were conducted at two metropolitan, academic, level I trauma centers. All subjects with FSF were prospectively enrolled unless there was evidence of shock or heart failure. Three consecutive arterial blood gases (ABG's) were obtained after induction of general anesthesia. The initial ABG was obtained after steady-state anesthesia; the second and third ABG's were taken 10 and 30 minutes post-femoral canal instrumentation, respectively. The end tidal carbon dioxide (PetCO2), minute ventilation (MV), positive end-expiratory pressure (PEEP), and percent of inspired and expired inhalation agent were recorded simultaneous to the ABG. The subjects were then managed both intraoperatively and postoperatively in the routine manner as dictated by their injury. Postoperatively, non-intubated subjects were monitored for evidence of respiratory distress, i.e. need for mechanical ventilation, need for FiO2>0.4, respiratory rate > 20, use of accessory muscles. VD/VT=(PaCO2-PetCO2)/PaCO2X100%. Normal range 0-20%.

Results: Seventy-five subjects with eighty FSF completed the study. Fifty FSF had reamed IMN (62.5%) and thirty FSF had unreamed IMN (37.5%). Mean initial Vd/Vt on all subjects (n=76) was 14% (95% CI=12-16%) Twenty-one subjects (28%) had Vd/Vt>20% on the third measurement. Among the 21 subjects with Vd/Vt>20%, 16 required pulmonary support for respiratory distress; 5 subjects did not require pulmonary support. Fifteen subjects required postoperative intubation; eleven of these subjects had a Vd/Vt>20% (sensitivity for predicting respiratory distress equals 73%). The subject with the highest Vd/Vt (53%) was the only study subject who died postoperatively.

Discussion: The alveolar dead space fraction (Vd/Vt) reflects the proportion of alveoli ventilated but not perfused. This index was used to determine if intramedullary nailing of the femur produced a clinically relevant pulmonary burden secondary to marrow embolism. We hypothesized that this index could predict which patients undergoing IMN of their FSF would require postoperative intensive monitoring for respiratory dysfunction.

Conclusion: An alveolar dead space fraction of greater than twenty percent predicts the need for postoperative intensive monitoring (i.e. mechanical ventilation or clinical signs of respiratory distress). The positive predictive value of Vd/Vt is 76% when the index is measured thirty minutes after the femoral canal is instrumented. At a cutoff of 20%, Vd/Vt was 73% sensitive in detecting which patients required postoperative intubation following IMN of FSF. Our data support the utility of measuring Vd/Vt on patients undergoing IMN of their FSF. An increase in Vd/Vt of greater than 20 percent warrants postoperative intensive monitoring.