The clinician's perspective on pneumothorax management.

OBJECTIVE: We sought to determine the current practice habits among clinicians treating spontaneous pneumothorax and bronchopleural fistula.

METHODS: Practice habits were determined by a randomized postal survey of 3,000 American College of Chest Physicians members. Group comparisons are performed by chi2 analysis with p<0.05 being significant.

RESULTS: Four hundred nine respondents (13.6%) included 176 practicing pulmonologists (43.0%), 67 academic pulmonologists (16.4%), 102 thoracic surgeons (25.0%), and 64 others (15.6%). More than 50% of respondents treat a first small primary spontaneous pneumothorax (PSP) by simple observation, a first small secondary spontaneous pneumothorax (SSP) by chest tube, persistent air leak in both PSP and SSP with chest tube+video-assisted thoracoscopy, and use a 20 to 24F chest tube in mechanically ventilated ARDS-related tension pneumothorax. First recurrences of PSP and SSP were treated by a variety of interventions that included simple observation (PSP=14%, SSP=4%), chest tube (22%/17%), chest tube+sclerosis (20%/16%), chest tube+video-assisted thoracoscopy (36%/48%), and chest tube+thoracotomy (5%/12%). The most popular sclerosing agents are doxycycline (48%), talc slurry (24%), and talc poudrage (19%). More than 75% of physicians intervened in a persistent air leak between 5 and 10 days. Chest tubes are initially placed to suction by 48% of respondents in PSP and removed >24 h after air leak ceases in 79%. Chest tube clamping prior to removal is employed by 67% of respondents. Significant differences exist between thoracic surgeons and pulmonologists with surgeons placing more chest tubes for first-time PSP and performing chest tube+video-assisted thoracoscopy for first recurrences of PSP more often than pulmonologists. Thoracic surgeons seldom use sclerosis in spontaneous pneumothorax compared to pulmonologists.

CONCLUSIONS: Marked practice variation exists in clinicians' approaches to the management of spontaneous pneumothorax and bronchopleural fistulas that is partially explained by differences between pulmonologists and thoracic surgeons. A national consensus statement is needed to guide randomized studies in pneumothorax management.