Purpose: The purpose of this study was to identify the factors associated with prolonged air leak (PAL) following pulmonary lobectomy for lung cancer.
Methods: The data of 146 patients who underwent pulmonary lobectomy for lung cancer between August 2010 and July 2015 were retrospectively reviewed. Air leaks were assessed daily by a visual evaluation and were categorized as follows: forced expiratory only (Grade 1), expiratory only (Grade 2), or continuous (Grade 3). Logistic regression analyses were performed to identify the predictors of PAL (>5 days).
Results: PAL occurred in 23 patients (16%). An air leak at rest (Grade ≥ 2) was detected on postoperative day (POD) 1 in 48% of the patients with PAL and 7% of the patients without PAL. A univariate analysis demonstrated that PAL was significantly associated with male sex, a smoking history of ≥ 40 pack years, a serum albumin level of ≤4.0 mg/dL, and an air leak on POD1 (Grade ≥ 2). A multivariate analysis demonstrated that a serum albumin level of ≤4.0 mg/dL (p = 0.027) and an air leak on POD1 (p = 0.006) were independent predictors of PAL. PAL occurred in 75% of the patients with these two risk factors.
Conclusions: The preoperative serum albumin level and the presence of a visually evaluated air leak on POD1 may be useful indicators for the perioperative management of air leaks.
Keywords: Complications; Lung cancer; Nutrition; Prolonged air leak; Surgery.