Background: Prolonged air leak (longer than three days) was hypothesized to be the primary cause of extended hospital stays following pulmonary resection. Its effect on length of stay (LOS) was compared with that of suboptimal pain control, nausea and vomiting, and other causes. Predictors of prolonged LOS and of prolonged air leaks were investigated.
Design: Retrospective review of 91 patients. Primary reasons for prolonged hospitalization were determined. Patient characteristics (demographic information, pulmonary function test results, body habitus measurements, smoking history), operative factors (procedure performed, duration of operation, complications) and postoperative factors (time of chest tube removal) were considered. Student's t test and chi2 analysis were used to compare continuous and ratio data, respectively, and linear regression analysis was used to define the equation relating two variables.
Results: The mean postoperative LOS was 6.4 days. Only prolonged air leak was predictive of increased LOS (9.4 days versus 5.4 days, P<0.001). Forced expiratory volume in 1 s less than 1.5 L/min, carbon monoxide diffusing capacity less than 80% predicted and the detection of a pneumothorax were all predictive of prolonged air leak. A strong correlation between the time of chest tube removal and LOS was found (r=0.937, P<0.001). Linear regression analysis showed postoperative LOS and duration of thoracostomy tube insertion to be related by the equation y = 0.88x + 2.49 days.
Conclusion: These results suggest that increased LOS following pulmonary resection is due primarily to prolonged air leaks. Furthermore, patients who have their chest tubes removed sooner are discharged sooner.