Pulmonary embolism after abdominal flap breast reconstruction: prediction and prevention.

BACKGROUND Symptomatic pulmonary embolism constitutes a significant risk following abdominal flap breast reconstruction. Reported rates vary from 0 to 6 percent. The authors assessed risk factors associated with symptomatic pulmonary embolism and constructed a prediction model to identify high-risk patients. METHODS Patients undergoing deep inferior epigastric perforator or transverse rectus abdominis musculocutaneous flap breast reconstructions at two academic centers from January of 2005 through January of 2011 were included. Thromboprophylaxis measures included early ambulation, low-molecular-weight heparin, elastic stockings, A-V Impulse System foot pumps, and pneumatic stockings. Risk factors for symptomatic pulmonary embolism were analyzed and weights were assigned to these risk factors. Sensitivity and specificity were maximized using receiver operating characteristic curves. RESULTS Of 430 consecutive patients, symptomatic pulmonary embolism occurred in 17 cases (4.0 percent). Two independent predictors for symptomatic pulmonary embolism were found, body mass index higher than 25, additionally higher than 28, and the BRCA gene mutation. Operation duration and bilaterality of reconstructions were dependent on the BRCA mutation and both indirect predictors for symptomatic pulmonary embolism. Optimization of sensitivity and specificity resulted in a prediction model. No significant differences in efficacy were found between the different thromboprophylaxis measures. CONCLUSIONS The rate of symptomatic pulmonary embolism was 4.0 percent, despite standard thromboprophylaxis. Body mass index and BRCA were significant predictors for symptomatic pulmonary embolism. The authors integrated these factors into a prediction model, which provides a useful tool for identification of high-risk patients. This latter category may benefit from a more aggressive thromboprophylaxis approach. CLINICAL QUESTION/LEVEL OF EVIDENCE Risk, III.