Pericardioscopy in the etiologic diagnosis of pericardial effusion in 141 consecutive patients.

Abstract

Although previous small series have documented the utility of pericardioscopy for accurate etiologic diagnosis of pericardial effusion, this technique remains underused. The aim of our study was to assess the benefits and risks of surgical pericardioscopy in a large prospective series. One hundred forty-one consecutive patients with unexplained pericardial effusion underwent 142 pericardioscopies with a rigid mediastinoscope. For each patient, the etiologic data obtained by pericardioscopy (visualization of pericardium, guided biopsies, subxiphoid window biopsy, and fluid analysis) were compared with the results that would have been obtained with only conventional surgical drainage and biopsy (subxiphoid window biopsy and fluid analysis). After complete workup, a specific cause was found in 69 cases (48.6%); the other 73 cases were considered idiopathic effusions (51.4%). Procedural and in-hospital mortality was 8 of 141 patients (5.6%). No death was directly attributable to pericardioscopy. During long-term follow-up (median duration, 24 months; range, 6 to 96), a previously unrecognized cause was discovered in 6 patients (4%). By comparing the areas under the receiver-operating characteristic curves, the diagnostic advantage of pericardioscopy was significant for the whole series (pericardioscopy, 0.98 +/- 0.011; conventional surgical drainage, 0.89 +/- 0.029; P < .001). The increase in sensitivity was more marked for some types such as neoplastic (21%), radiation-induced (100%), or purulent (83%) effusions. Our data demonstrate that pericardioscopy increases the diagnostic sensitivity of surgical pericardial drainage and biopsy without specific risk.