|
Affiliation
|
Department of Surgery, University of California, San Francisco, CA 94110, USA. pknudson@sfghsurg.ucsf.edu |
|
Abstract
|
OBJECTIVE: This study was undertaken to determine the current incidence of pulmonary embolism (PE) and its attributable mortality after injury.
BACKGROUND: Despite compliance with prophylactic measures, PE remains a threat to postinjury recovery. We hypothesized that the liberal use of chest computed tomography after injury has resulted in an increased rate of detection of PE but that the mortality attributable to PE has decreased over the past decade. We also postulated that the risk factors for posttraumatic PE might be different from those for deep venous thrombosis (DVT).
METHODS: We examined demographics, injury data, risk factors, and outcomes from patients with DVT and PE compiled in the recent years (2007-2009) in the National Trauma Data Bank (NTDB). For comparison, we used patient data entered into NTDB from 1994 to 2001. Statistical models were created to examine the predictors of DVT and PE and PE-related mortality.
RESULTS: Among 888,652 patients in the current NTDB cohort, there were 9398 episodes of DVT (1.06%) and 3738 of PE (0.42%). Although many risk factors overlapped, a severe chest injury (Abbreviated Injury Score ≥ 3) conferred a much higher risk of PE than DVT. When comparing results from centers that had contributed to both data sets, there was a more than 2-fold increase in PE occurrence in the current cohort (0.49% vs 0.21%, P < 0.01) but with a significant reduction in PE-adjusted mortality (odds ratio, 4.08 vs 2.42).
CONCLUSIONS: The reported incidence of PE after trauma has more than doubled in recent years, while the PE-associated mortality has significantly decreased, suggesting that we are identifying a different disease entity or stage. Chest injuries convey a substantial risk for PE, a risk not likely to be diminished by leg compression devices or vena cava filters.
Language: en |