Lethal Trauma Pulmonary Embolism Is a Black Swan Event in Patients at Risk for Deep Vein Thrombosis: An Evidence-Based Review.

Abstract We delineated the incidence of trauma patient pulmonary embolism (PE) and risk conditions by performing a systematic literature review of those at risk for deep vein thrombosis (DVT). The PE proportion was 1.4 per cent (95% confidence interval = 1.2-1.6) in at-risk patients. Of 10 conditions, PE was only associated with increased age (P < 0.01) or leg injury (P < 0.01; risk ratio = 1.6). As lower extremity DVT (LEDVT) proportions increased, mortality proportions (P = 0.02) and hospital stay (P = 0.0002) increased, but PE proportions did not (P = 0.13). LEDVT was lower with chemoprophylaxis (CP) (4.9%) than without CP (19.1%; P < 0.01). PE was lower with CP (1.0%) than without CP (2.2%; P = 0.0004). Mortality was lower with CP (6.6%) than without CP (11.6%; P = 0.002). PE was similar with (1.2%) and without (1.9%; P = 0.19) mechanical prophylaxis (MP). LEDVT was lower with MP (8.5%) than without MP (12.2%; P = 0.0005). PE proportions were similar with (1.3%) and without (1.5%; P = 0.24) LEDVT surveillance. Mortality was higher with LEDVT surveillance (7.9%) than without (4.8%; P < 0.01). A PE mortality of 19.7 per cent (95% confidence interval = 18-22) × a 1.4 per cent PE proportion yielded a 0.28 per cent lethal PE proportion. As PE proportions increased, mortality (P = 0.52) and hospital stay (P = 0.13) did not. Of 176 patients with PE, 76 per cent had no LEDVT. In trauma patients at risk for DVT, PE is infrequent, has a minimal impact on outcomes, and death is a black swan event. LEDVT surveillance did not improve outcomes. Because PE was not associated with LEDVT and most patients with PE had no LEDVT, preventing, diagnosing, and treating LEDVT may be ineffective PE prophylaxis.