How do you approach the risk/benefit ratio of pericardiocentesis as opposed to close observation with serial TTEs in a hemodynamically stable elderly patient on anticoagulation with a large circumferential pericardial effusion?