Title Computed tomography scoring system for discriminating between parapneumonic effusions eventually drained and those cured only with antibiotics
Author(s) Porcel JM, Pardina M, Alemán C, Pallisa E, Light RW, Bielsa S.
Source Respirology. 2017 Aug;22(6):1199-1204. doi: 10.1111/resp.13040. Epub 2017 Mar 30.
Abstract BACKGROUND AND OBJECTIVE: Due to limited data, we aimed to develop and validate a computed tomography (CT)-based scoring system for identifying those parapneumonic effusions (PPEs) requiring drainage.
METHODS: A retrospective review of all patients with PPE who underwent thoracentesis and a chest CT scan before any attempt to place a tube thoracostomy, if applicable, over an 8-year period was conducted. Eleven chest CT characteristics were compared between 90 patients with complicated PPEs (CPPEs), defined as those which eventually required chest drainage, and 60 with non-complicated effusions (derivation sample). A scoring system was devised with those CT findings identified as independent predictors of CPPE in a logistic regression analysis, and further validated in an independent population of 59 PPE patients.
RESULTS: CT scores predicting CPPE were pleural contrast enhancement (3 points), pleural microbubbles, increased extrapleural fat attenuation and fluid volume ≥400 mL (1 point each). A sum score of ≥4 yielded 84% sensitivity (95% CI: 62-85%), 75% specificity (95% CI: 62-85%), 81% diagnostic accuracy (95% CI: 73-86%), likelihood ratio (LR) positive of 3.4 (95% CI: 2.1-5.4), LR negative of 0.22 (95% CI: 0.13-0.36) and area under the receiver operating characteristic curve (AUC) of 0.829 (95% CI: 0.754-0.904) for labelling CPPE in the derivation set. These results were reproduced in the validation sample. The CT grading scale also exhibited a fair ability to identify patients who needed surgery or would die from the pleural infection (AUC: 0.76, 95% CI: 0.61-0.9).
CONCLUSION: A novel CT scoring system for adults with PPE may allow clinicians to predict the need for chest tube drainage with good accuracy.