Sonographically guided thoracentesis and rate of pneumothorax

PURPOSE: To assess whether thoracenteses performed with sonographic guidance are associated with a lower rater of pneumothorax and tube thoracostomy than those performed without sonographic guidance. METHODS: We reviewed the medical records of 523 subjects undergoing their initial diagnostic thoracentesis at our institution from July 1, 2001, to June 30, 2002. We excluded 73 subjects in whom no chest imaging had been performed within 5 days of thoracentesis or who had pre-existing chronic hydropneumothorax. RESULTS: Of the 450 thoracenteses performed, 305 (67.8%) were performed with sonographic guidance and 145 (32.2%) were performed without. On postthoracentesis imaging in all subjects, 30 pneumothoraces (6.7%) were found (23 inpatients, 7 out-patients). Eight patients required a tube thoracostomy for their pneumothorax. Pneumothorax occurred in 15 of 305 procedures (4.9%) performed with sonographic guidance and 15 of 145 procedures (10.3%) performed without (p < 0.05). Tube thoracostomy was performed in 0.7% of patients whose thoracentesis was performed with sonographic guidance and in 4.1% in those that were not (p < 0.05). We found no correlation between pneumothorax after thoracentesis and age, inpatient status, loculation of effusion, or volume of pleural fluid removed. CONCLUSIONS: The routine use of sonography during diagnostic thoracentesis is associated with a reduced rate of pneumothorax and tube thoracostomy.