Prospective Analysis of the Predictive Value of Sonographic Pleural Fluid Echogenicity for the Diagnosis of Exudative Effusion.

ABSTRACT

BACKGROUND: Pleural effusion echogenicity on ultrasound has previously been suggested to allow identification of exudates. A case series suggested that homogenously echogenic effusions are always exudates. With modern imaging techniques and more advanced ultrasound technology, this may no longer be true. OBJECTIVES: This study aims to prospectively assess the predictive value of echogenicity in the identification of exudates. METHOD: Patients undergoing thoracic ultrasound before pleural fluid sampling were analysed prospectively (n = 140). Pleural fluid was classified as an exudate if both fluid total protein (TP) > 29 g/L and fluid lactate dehydrogenase (LDH) > 2/3 upper limit of normal serum LDH (which is 255 IU/L in females and 235 IU/L in males) were present. If only one of these criteria was met, the effusion was considered to have discordant biochemistry. RESULTS: Fifty-five (39%) patients had non-echogenic and 85 (61%) had echogenic effusions. Six (7.1%) patients with echogenic effusions had transudates; the median fluid TP for this group was 18.5 g/L (IQR 9.75) and median LDH 63.0 IU/L (IQR 40.3). The specificity of echogenicity identifying exudates from transudates, excluding patients with discordant biochemistry, was 57.1%, positive predictive value (PPV) 90.3%, sensitivity 65.1%, and negative predictive value (NPV) 21.0%. The specificity of echogenicity identifying exudates (including discordant biochemistry) from transudates was 57.1%, PPV 92.9%, sensitivity 62.7%, and NPV 14.5%. CONCLUSIONS: Echogenicity of a pleural effusion has a low specificity for identifying an underlying exudate, and the echogenic qualities of the fluid should not influence clinical decision-making.