Diagnosis of posttraumatic pulmonary embolism: is chest computed tomographic angiography acceptable?

Abstract

Pulmonary angiography (PA-gram) has long been the accepted criterion standard for diagnosing pulmonary embolism (PE). Computed tomographic angiography has recently been advocated as an equivalent alternative to PA-gram. CT angiography is known to be insensitive for peripheral (segmental and subsegmental) emboli. We have previously found that a significant number of posttraumatic PEs occur early. We therefore hypothesized that because of the fragmentation of these early (soft) clots, posttraumatic PEs would be found disproportionately in the lung periphery. Trauma patients with PE confirmed by PA-gram were identified from our trauma database and medical records. PA-grams and reports were re-reviewed and the location of all emboli was documented. We identified 45 patients, with an average age of 46 +/- 19 years; two thirds of the patients were men and 82% had a blunt mechanism of injury. Patients had PE diagnosed between days 0 and 57. Overall, PE was confined to segmental or smaller vessels in 27 (60%) patients and to subsegmental vessels in 7 (16%) patients. Twelve patients (27%) had a PE within the first 4 days. Furthermore, 32 patients (71%) had unilateral clot and 22 patients (48.9%) had clot confined to one region. PE frequently occurs soon after injury. The majority of PEs after trauma are found peripherally (in segmental or subsegmental vessels). Right/left pulmonary artery embolisms are likely to be found only later in a trauma patient's course. Any diagnostic study used to diagnose pulmonary embolism in trauma patients must have sufficient resolution capacity to reliably detect segmental and subsegmental clot. A diagnostic modality such as CT scanning that is insensitive to peripheral embolisms may miss a significant number of posttraumatic PEs.