The accuracy of ultrasound assessment of portal vein patency has been defined by comparing it with the results of arterial portography in 115 cases. The accuracy of arterial portography was confirmed in 21 cases where orthotopic liver transplantation was performed and used as a "benchmark" against which to assess the ultrasound findings. Ultrasound correctly assessed portal vein patency in 87.5% of patients. It was more accurate in assessing patency (90%) than occlusion (68%). Ultrasound correctly assessed portal vein patency in 90% of cases of cirrhosis and hepatic malignancy. Difficulties occurred in children with biliary atresia particularly following the Kasai operation (37.5% accuracy). In the absence of previous surgery to the portal vein or biliary system, ultrasound is comparable to arterial portography and can be used as the sole means of assessment.
COMMENTS
Comparison studies of diagnostic imaging methods are fraught with methodological problems. The most common difficulties involve introduction of bias, which may occur in patient selection, through incomplete verification, through absence of a definitive reference test and through test interpretation bias, where knowledge of a test result is present (1, 2). This study suffers from these biases and some additional methodological problems. The authors retrospectively analyzed, over a 2-year period, 115 cases in whom real-time ultrasonography and arterial portography were performed for the clinical evaluation of portal venous thrombosis, for evaluation of the