Estimating hospital costs attributable to prolonged air leak in pulmonary lobectomy.

OBJECTIVE Prolonged air leak (PAL) after pulmonary surgery is a frequent occurrence and is reported to cause increased length of stay (LOS) and hospital costs although the costs directly attributable to PAL have never been published. The present study aims to compare the prevalence of pulmonary complications (atelectasis, pneumonia and pleural empyema) in patients with or without PAL and to quantify economic costs directly incurred by PAL in a series of pulmonary lobectomies. METHODS A series of 238 patients scheduled for pulmonary lobectomy (January 2001-December 2003) have been reviewed. PAL was defined as air leakage which prevented hospital discharge for 5 postoperative days or over. Hospital costs (excluding operating room) for pulmonary lobectomies have been obtained and calculated as mean daily costs. Age, body mass index, diagnosis, Charlson co-morbidity index, ppoFEV1 and major post-operative cardio-pulmonary morbidity have been used to construct a Cox-regression model for hospital stay, considering deaths as censored cases. Individual risk function has been used as a new variable and expected LOS calculated for each case. This data has been used to estimate total excess hospital stay and costs incurred by cases with PAL. RESULTS Prevalence of PAL was 23 cases (9, 7%). Mean daily hospital cost for lobectomy was 632.49. For the whole series, mean hospital stay was 5 days (10 days for patients with PAL). PAL cases had more postoperative pulmonary morbidity (risk-ratio: 2.78). Variables showing independent influence on stay were: diagnosis of non-malignant disease (P=0.001); FEV1ppo (P=0.032) and cardio-respiratory morbidity (P<0.001). Calculated total excess stay for PAL patients was 62 days. A total expense of 39,437.39 (38,724.96 hospital and 712.43 pharmacy charges) were estimated to result from postoperative air-leak. CONCLUSIONS PAL patients are prone to developing major postoperative morbidity. PAL calculated costs are over 13,000 per year. This data is useful for designing technical cost-effective strategies to avoid post-lobectomy PAL.