Pulmonary embolism in mechanically ventilated patients requiring computed tomography: Prevalence, risk factors, and outcome

Objective: To estimate the rate of pulmonary embolism among mechanically ventilated patients and its association with deep venous thrombosis. Design: Prospective cohort study. Setting: Medical intensive care unit of a university-affiliated teaching hospital. Patients: Inclusion criteria: mechanically ventilated patients requiring a thoracic contrast-enhanced computed tomography scan for any medical reason. Exclusion criteria: a diagnosis of pulmonary embolism before intensive care unit admission, an allergy to contrast agents, and age younger than 18 yrs. Interventions: All the mechanically ventilated patients requiring a thoracic computed tomography underwent the standard imaging protocol for pulmonary embolism detection. Therapeutic anticoagulation was given immediately after pulmonary embolism diagnosis. All the included patients underwent a compression ultrasound of the four limbs within 48 hrs after the computed tomography scan to detect deep venous thrombosis. Results: Of 176 included patients, 33 (18.7%) had pulmonary embolism diagnosed by computed tomography, including 20 (61%) with no clinical suspicion of pulmonary embolism. By multiple logistic regression, independent risk factors for pulmonary embolism were male gender, high body mass index, history of cancer, past medical history of deep venous thrombosis, coma, and high platelet count. Previous prophylactic anticoagulant use was not a risk factor for pulmonary embolism. Of the 176 patients, 35 (19.9%) had deep venous thrombosis by compression ultrasonography, including 20 (57.1%) in the lower limbs and 24 (68.6%) related to central venous catheters. Of the 33 pulmonary embolisms, 11 (33.3%) were associated with deep venous thrombosis. The pulmonary embolism risk was increased by lower-limb deep venous thrombosis (odds ratio 4.0; 95% confidence interval 1.6-10) but not upper-limb deep venous thrombosis (odds ratio 0.6; 95% confidence interval 0.1-2.9). Crude comparison of patients with and without pulmonary embolism shows no difference in length of stay or mortality. Conclusions: In mechanically ventilated patients who needed a computed tomography, pulmonary embolism was more common than expected. Patients diagnosed with pulmonary embolism were all treated with therapeutic anticoagulation, and their intensive care unit or hospital mortality was not impacted by the pulmonary embolism occurrence. These results invite further research into early screening and therapeutic anticoagulation of pulmonary embolism in critically ill patients. (Crit Care Med 2012; 40:3202-3208)