Real-Time Ultrasound-Guided Paracentesis by Radiologists: Near Zero Risk of Hemorrhage without Correction of Coagulopathy.

PURPOSE: To evaluate the rate and risk factors for hemorrhage in patients undergoing real-time, ultrasound-guided paracentesis by radiologists without correction of coagulopathy. MATERIALS AND METHODS: This was a retrospective study of all patients who underwent real-time, ultrasound-guided paracentesis at a single institution over a 2-year period. In total, 3116 paracentesis procedures were performed: 757 (24%) inpatients and 2,359 (76%) outpatients. Ninety-five percent of patients had a diagnosis of cirrhosis. Mean patient age was 56.6 years. Mean international normalized ratio (INR) was 1.6; INR was > 2 in 437 (14%) of cases. Mean platelet count was 122 x 103/µL; platelet count was < 50 x 103/µL in 368 (12%) of patients. Seven hundred seven (23%) patients were dialysis dependent. Patients were followed for 2 weeks after paracentesis to assess for hemorrhage requiring transfusion or rescue angiogram/embolization. Univariate analysis was performed to determine risk factors for hemorrhage. Blood product and cost saving analysis were performed. RESULTS: Significant post-paracentesis hemorrhage occurred in 6 (0.19%) patients, and only 1 patient required an angiogram with embolization. No predictors of post-procedure bleeding were found, including INR and platelet count. Transfusion of 1125 units of fresh frozen plasma and 366 units of platelets were avoided, for a transfusion-associated cost savings of $816,000. CONCLUSIONS: Without correction of coagulation abnormalities with prophylactic blood product transfusion, post-procedural hemorrhage is very rare when paracentesis is performed with real-time ultrasound guidance by radiologists.

Hepatic arterioportal fistula presenting as gastric variceal hemorrhage.

Hepatic arterioportal fistulae (APF) are abnormal communications between the hepatic artery and the portal vein. In this report, we present the second case in the literature of a symptomatic APF presenting as a gastric variceal bleeding. A 55-year-old female presented to our facility with hematemesis. Upper endoscopy revealed a bleeding gastric varix. A computed tomography scan identified a large left hepatic lobe APF between the left hepatic artery and the left portal vein. Through angiography coil embolization was performed and with resultant loss of arterial flow, the APF was decompressed. On hospital day 3, the patient developed new melena. Portovenogram was performed and a TIPS stent was deployed. The patient subsequently did well. Hepatic arterioportal fistulae can result in portal hypertension secondary to arterial blood flowing directly into the portal vein bypassing the hepatic sinusoids. Iatrogenic causes (e.g. percutaneous liver biopsy) represent more than 50% of published cases of APFs. Most APFs resolve spontaneously as they are small and peripherally located. In rare instances, when APFs are centrally located, clinical symptoms develop. There have been 30 reported cases of symptomatic intrahepatic APFs following percutaneous liver biopsy. Of those, only one case presented as a gastric variceal bleed. Digital subtraction angiography is the gold standard in the diagnosis and treatment of APFs. In addition to initial embolization, we elected to treat the patient with TIPS due to the magnitude of her bleed. Although rare, intrahepatic APF should be kept on the differential of a patient presenting with isolated gastric varices.