Prediction of recurrent spontaneous pneumothorax: CT scan findings versus management features

Objective: To assess the association between primary spontaneous pneumothorax (PSP) recurrence and pulmonary CT scan findings, and other variables pertaining to clinical presentation and management. Methods: Consecutive patients hospitalized for the first episode of PSP and treated by various strategies including chest tube or pteurocatheter drainage and, in persistent pneumothorax, by chemical or surgical pleurodesis. All patients had a pulmonary CT scan examination in the week following hospital discharge in order to calculate a score combining distribution, number and size of dystrophic pulmonary abnormalities. This score as well as other pertinent clinical and therapeutic parameters were compared between patients who ultimately experienced PSP recurrence and those who did not. Results: Eighty patients (mean age: 27 +/- 7 yr) were admitted for PSP and had a chest drainage with either a drain or pleurocatheter. Chest drainage and hospitalization durations were 4.7 +/- 3.2 and 6.2 +/- 3.5 days, respectively. Sixteen patients required chemical pleurodesis. Dystrophic bullae were present in CTscans in 72.5% patients. After a mean follow up of 34 +/- 20 months, 15 out of the 80 patients (19%) had a PSP recurrence. Multivariate statistical analysis disclosed the use of pleurocatheter (OR = 5; 95% Cl: 1.4-20; P = 0.02) and of chemical pleurodesis (OR = 8; 95% Cl: 1. 5-47; P = 0.015) as independent predictors of PSP recurrence. The severity of dystrophic lesions inferred from the dystrophic score was not statistically associated with the risk of recurrence. Conclusion: Dystrophic lesions are frequently present in PSR They are not associated with an increased risk of recurrence. (c) 2006 Elsevier Ltd. All rights reserved.