The purpose of this study was to determine if echocardiography could be used to predict the presence and/or severity of cardiac tamponade in an experimental model.
Methods Animal PreparationNine mongrel dogs that weighed 15-35 kg were lightly anesthetized with a mixture of chloralose and urethane, 100 mg/kg and 1000 mg/kg, respectively. The left chest was entered through the fourth intercostal space and the pericardium was cannulated with a #8 polyethylene pigtail catheter through a stab incision. Two 00 silk ties were placed around the catheter using a pursestring suture and a cyanoacrylate ester glue was placed over the suture and surrounding areas to achieve a seal. The chest was closed, and a chest tube to an underwater seal was left in place to remove air and prevent atelectasis. Spontaneous respiration resumed after surgery was completed and the chest closed. We measured right atrial, aortic arch and left ventricular pressures using a Statham P23Db transducer at the midchest level. Cardiac output was measured by the dye-dilution principle. Indocyanine green dye (2.5 mg) was injected into the right atrium, and dye concentrations in right brachial artery blood were recorded from duplicate curves with exponential decay.
Echocardiographic StudiesEchocardiographic studies were obtained using a commercially available ultrasonoscope and a 2.25-MHz transducer focused at 7.5 cm. The dogs were examined during spontaneous respiration in the right lateral decubitus position.8 The hand-held transducer was placed in the fourth or fifth right intercostal space approximately 5 cm from the midsternal line and angled slightly inferiorly and anteriorly. An M-mode sweep was initially performed to identify the septum and posterior wall of the left ventricle just below the mitral valve for left ventricular diameter measurements. In addition, a minor sweep was initially done in order to identify the largest end-diastolic dimension measured at the R wave of the QRS complex of the simultaneous electrocardiographic tracing. After the optimal transducer position and beam angulation were initially identified, the transducer was subsequently moved as little as possible to minimize artifacts caused by changing transducer position or beam direction. Measurements were taken during expiration and inspiration in the control state, and during various levels of tamponade ( figs. 1 and 2). The phases of the respiratory cycle were identified by visual inspection of the records; this was verified in some of the dogs by simultaneously recording intrapleural pressures as well. Measurements of the ven-