External validation of the CAPAD score to predict the occurrence of clinically significant pneumothorax after CT-guided percutaneous transthoracic lung biopsy

Résumé fourni par la source

BACKGROUND: The most common complication preventing outpatient management after percutaneous CT-guided lung biopsy (PTLB) is the occurrence of pneumothorax requiring drainage. We previously developed a pragmatic score to predict the risk of clinically relevant pneumothorax after PTLB (hereafter called the CAPAD score), to guide patient management. OBJECTIVES: To externally validate the CAPAD score in an independent cohort, with a specific focus on its ability to rule out pneumothorax requiring drainage after PTLB. METHOD: We retrospectively analysed a pre-existing institutional database of consecutive adult patients undergoing CT-guided PTLB at Bichat-Claude-Bernard University Hospital (Paris, France) between 1 January 2015 and 31 December 2017. The CAPAD score combines five variables: COPD, Anterior PTLB approach, more than one Pleural crossing, pneumothorax ≥10 mm on the end-of-procedure CT (post-procedural immediate pneumothorax), and skin-pleura Distance ≤30 mm, with a pre-specified cut-off of 73 points. The primary endpoint was the diagnostic performance of the score for post-PTLB pneumothorax on the 4-6 h control CXR and secondary endpoint was its performance for pneumothorax requiring drainage. RESULTS: The CAPAD score was computable in 313 of 474 patients. Pneumothorax occurred in 95 patients (30.4%), and 23 patients (7.3%) required drainage. At the pre-specified cut-off of 73, the score had a C-index of 0.77, a sensitivity of 75%, a specificity of 56%, a negative predictive value (NPV) of 84% and a positive predictive value of 43% for any post-PTLB pneumothorax. For pneumothorax requiring drainage, a score < 73 was associated with a 2.0% risk (3/147), corresponding to a NPV of 98%. CONCLUSION: Our pragmatic CAPAD score demonstrated a good prognostic performance for predicting the risk of post-procedural pneumothorax, achieve a NPV of 98% for pneumothorax requiring drainage.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.