W observed 2 patients with large pleural effusions and small to moderate pericardial effusions who had echocardiographic signs of cardiac tamponade that resolved completely after pleural tap. Because many diseases that produce pleural effusions also cause pericardial effusions, the occurrence of both a pericardial and a pleural effusion in the same patient is not uncommon. In this study, we wished to discover whether pleural effusions produced clinical and echocardiographic signs of cardiac tamponade, and if so, to assess the size of pleural effusion that could cause signs of tamponade. • • • Our study was approved by the institutional review board; all subjects signed an informed consent before the study began. Procedures were in accord with the guidelines set forth by our institution. Thirty-seven patients with pleural effusions and no clinical or echocardiographic evidence of organic heart disease comprised the study group. The etiologies of pleural effusions were: malignancy in 18 patients, chronic renal failure in 2 patients, human immunodeficiency virus related in 3 patients, pneumonia in 3 patients, tuberculosis in 2 patients, empyema in 2 patients, systemic lupus erythematosus in 2 patients, and unknown etiologies in 5 patients. The patients were divided into 2 groups. Group 1 consisted of 10 patients (7 men and 3 women, age range 33 to 88 years [mean 69]) with pleural effusions occupying less than a hemithorax on chest x-ray. Group 2 consisted of 27 patients (13 men and 14 women; age range 20 to 74 years [mean 49]) with effusions greater than a hemithorax. The sizes and distributions of pleural effusions in group 1 were as follows: 3 patients had right pleural effusions, 5 had left pleural effusions, and 2 patients had small bilateral effusions. Two subjects had small pericardial effusions. In group 2, 11 patients had left pleural effusions (Table 1). Eleven patients had effusions occupying 1 lung. Seven patients also had small pericardial effusions and 4 had moderate pericardial effusions. All patients underwent an evaluation of jugular venous pressure (JVP) and blood pressure measurement to assess the presence of pulsus paradoxus. The size of the pleural effusion was determined via posterior-anterior chest x-ray, then confirmed by lateral decubitus chest x-ray. A large pleural effusion is assumed when at least half a lung is involved. The same criteria are used when bilateral effusions are involved. Pulsus paradoxus, JVP, lung and cardiac auscultation, respiration rate, blood pressure, and the site and size of the pleural effusion were noted. Two-dimensional echocardiography was performed in all standard views. Apical 4-chamber, 2-dimensional echocardiograms and pulse-wave Doppler of the mitral inflow signals during quiet respiration were obtained by placing the sample volume at the tips of the mitral leaflets with a 2-MHz transducer. After echocardiograms were performed, patients were referred back to their primary care physicians to proceed with the indicated thoracentesis. The amount of fluid removed ranged from 900 to 1,500 ml (mean 1,267 ml) in 9 patients without chest tube placement. In the 11 patients who did have chest tubes placed, we cannot report exact volumes of fluid removed due to the unknown excess amount versus normal drainage of pleural fluids. Volumes removed were estimated to range from 2,000 to 4,000 ml (mean 3,050). Clinical examinations and echocardiograms were repeated within 24 hours after thoracentesis. Post-thoracentesis echocardiograms were analyzed for resolution of abnormalities noted in the previous study in 20 patients from group 2. Follow-up data were not available for 7 patients in group 2 due to death (n 2) or refusal to continue the study (n 5). The echocardiograms were reviewed. Right ventricular (RV) diastolic collapse was considered present if there was an inward motion of the right ventricle in diastole. The minimum and maximum velocities of E waves during quiet respiration were measured, and the percent change in velocity was determined. A change of 25% was considered abnormal respiratory variation in transmitral flow or flow velocity paradoxus. Group 1 patients had no clinical or echocardiographic signs of cardiac tamponade. In group 2, 8 subjects had elevated JVP ( 5 cm above the clavicle in the 45° position), 8 had pulsus paradoxus, 6 had RV From the Division of Cardiology, LAC & USC Medical Center, Los Angeles, California. Dr. Chandraratna’s address is: Long Beach VA Medical Center, Section of Cardiology (111C), 5901 E 7th Street, Long Beach, California 90822. E-mail: premindra.chandraratna @med.va.gov. Manuscript received June 6, 2001; revised manuscript received and accepted September 11, 2001. TABLE 1 Location and Size of Pleural Effusions in Group 2