Abstract Rationale Noninvasive differentiation of exudative from transudative pleural effusions may assist clinical triage while laboratory results are pending. Computed tomography (CT) attenuation, expressed in Hounsfield units (HU), reflects fluid composition and may support effusion classification. This study quantified attenuation differences between exudative and transudative pleural effusions. Methods A single-center retrospective cohort was conducted at Wyckoff Heights Medical Center (Brooklyn, NY). Adults who underwent thoracentesis between January 2022 and December 2024 were included if they had (1) complete paired serum protein and lactate dehydrogenase (LDH) values with pleural fluid chemistries to calculate Light’s criteria, and (2) a chest CT obtained prior to thoracentesis. Effusions were classified by Light’s criteria. On axial CT images, four circular regions of interest (ROIs) were placed within the largest dependent fluid pocket, and their mean defined the pleural fluid HU average. Because data were non-normally distributed (Shapiro-Wilk test), group comparisons were performed using the Mann-Whitney U test. Results A total of 124 pleural effusions were analyzed (79 exudates, 45 transudates). The mean patient age was 66.1 ± 16.1 years, and 62% were male. Although exudates were more frequent in males (70.1%) than in females (53.2%) and tended to occur in younger adults, sex distribution, smoking history, and age did not differ significantly between groups. Mean HU attenuation was higher in exudates (14.07 ± 7.31) than in transudates (9.78 ± 4.71) (Mann-Whitney p 0.001). Conclusions Pleural fluid attenuation on noncontrast chest CT was approximately 4.3 HU higher in exudative than in transudative effusions. Elevated HU values were associated with exudates and may help prioritize diagnostic or therapeutic thoracentesis and early management when exudate is suspected. However, CT attenuation should not replace thoracentesis or Light’s criteria, nor delay indicated procedures. Hemorrhagic, purulent, or contrast-contaminated effusions may increase attenuation and should be interpreted cautiously. Prospective multicenter validation and integration of attenuation into multivariable clinical models are warranted. This abstract is funded by: None
Gache et al. (Fri,) studied this question.