Feasibility and safety of ultrasound-aided thoracentesis in mechanically ventilated patients

Objective: Thoracentesis in a ventilated patient is rarely performed because of the risk of pneumothorax. We have evaluated the safety of this procedure when aided by ultrasound. Design: Prospective study. Setting: Medical intensive care unit, university-affiliated hospital. Patients: 45 procedures were performed in 40 consecutive patients with ultrasound signs of pleural effusion, all mechanically ventilated. Interventions: Pleural effusion was defined on ultrasound as a collection of fluid between parietal and visceral pleura leading to variations in interpleural distance during breathing. When the interpleural distance was greater than or equal to 15 mm and visible over three intercostal spaces, a needle (16 or 21 G) was inserted after ultrasound localization in a patient in either dorsal or lateral decubitus. Results: No complication occurred in the 45 thoracenteses. Fluid was obtained in 44 of 45 procedures, thus confirming the diagnosis of pleural effusion. The procedure was immediate (less than 10 s) in 40 of 45 cases. It was easy (i.e., keeping the patient supine) in 22 of 45 procedures. In 44 cases where fluid was obtained, only 27 bedside radiographs revealed signs of effusion, whereas 17 showed absence of a visible effusion. Ultrasound thus appeared more efficient than bedside X-ray in detecting pleural effusion. Conclusions: If basic rules are followed, ultrasound localization makes thoracentesis a safe, easy and simple procedure in patients on mechanical ventilation.