Negative predictive value of computed tomography pulmonary angiography with indirect computed tomography venography in intensive care unit patients

The aim of the study was to evaluate the negative predictive value (NPV) of combined computed tomography (CT) pulmonary angiography (CTPA) and indirect CT venography (CTV) in the intensive care unit (ICU) setting. We retrospectively reviewed the records of 181 consecutive ICU patients who underwent CTPA/CTV. Radiology reports were examined to determine whether the study was positive for pulmonary embolism (PE), PE and deep venous thrombosis (DVT), or DVT alone; indeterminate; or negative. Results that were reported as negative were further evaluated for evidence of PE or DVT within 30 days by imaging, clinical evaluation, or autopsy data. The outcomes were evaluated for significance by calculating the rate ratio and 95% confidence interval. A total of 41 patients (22.7%) were diagnosed with venous thromboembolism, 29 (70.7%) with PE, 8 (19.5%) with PE and DVT, and 4 (9.8%) with DVT. Seven studies were considered nondiagnostic. Seventeen deaths occurred within 30 days of CTA/CTV, of which none was felt to be related to PE/DVT. Of the 140 studies read as negative or nondiagnostic, 4 were determined to have venous thromboembolism (3 PEs and 1 DVT) within 30 days of the initial study (NPV = 97.1%). If patients who received prophylactic anticoagulation or inferior vena cava interruption (n = 25) were excluded, NPV decreases to 96.5% A negative CTPA/CTV is reliable for the exclusion of significant venous thromboembolism in ICU patients.