Efficiency of chest computed tomography in critically ill patients with multiple traumas

Objective: The efficiency of secondary thoracic computed tomography (TCT) in critically ill patients with multiple traumas was assessed by comparison of TCT with chest radiograph findings. The subsequent therapeutic consequences based on the additional information of TCT were evaluated. Setting: A six-bed trauma intensive care unit in a university hospital. Design: Prospective, descriptive study. Patients and Interventions: One hundred one computed tomographic (CT) examinations (mean, 2.6 per patient; range, 1-10) were performed in 39 patients, fulfilling the following indications for TCT: a) sepsis with suspected pulmonary focus (n = 41); b) deterioration of pulmonary gas exchange (n = 35); c) guiding the duration of intermittent prone positioning (n = 25). The information provided by TCT was compared with corresponding chest radiographs (CXR). Therapeutic consequences drawn after TCT were compared with the additional diagnostic information of TCT. The change of therapy was documented that would not have been undertaken or may have been delayed had TCT evaluation not been used. Results: TCT was significantly superior to CXR in detecting pneumothoraces, pleural effusions, and pulmonary abscesses. Furthermore, a significantly higher accuracy regarding pulmonary densities was found. Subsequent therapeutic interventions ensued from 85 (84.2%) CT scans. After TCT, intermittent prone positioning was initiated in 31 patients, chest tubes were inserted in 16 patients, and intermittent prone positioning was terminated in 13 patients and was continued in 12 patients. Eleven thoracotomies were performed because of the TCT findings. The described therapeutic interventions were based on abnormalities seen on CT scans but were not evident in CXR in 58 patients (57.4%). Significant information that influenced therapeutic concepts was obtained in 66% (n = 23) of patients with pulmonary deterioration of gas exchange, in 61% (n = 25) of patients with sepsis, and in 40% (n = 10) of patients to guide the duration of intermittent prone positioning. Thoracotomy and specific drainage by tube thoracostomy was always dependent on the findings of TCT. Conclusion: Performed under the above displayed defined indications, TCT had an overall efficiency of 57%. It provided an increased sensitivity for intrathoracic lesions and a more comprehensive diagnosis of chest abnormalities.