Predicted versus observed peak oxygen consumption after major pulmonary resection.

Abstract

The aim of this study was to verify the accuracy of predicted postoperative peak VO(2) in predicting the actual peak VO(2) after major pulmonary resection. This was a prospective longitudinal series of 110 consecutive patients who underwent lobectomy (101 patients) or pneumonectomy (9 patients), with complete preoperative and postoperative (3 months) cardiopulmonary exercise testing (CPET). Predicted postoperative peak VO(2) was calculated by subtracting from the preoperative peak VO(2) the contribution of unobstructed pulmonary segments removed during operation. Predicted postoperative peak VO(2) and actual postoperative peak VO(2) were compared by the paired sign test. The average value of preoperative peak VO(2) was 16.8 mL/kg/min or 64.1% of predicted. The actual value of postoperative peak VO(2) was 15.9 mL/kg/min or 64.4% of predicted. The actual postoperative peak VO(2) was higher than the predicted postoperative peak VO(2) (15.9 versus 13.1 mL/kg/min; p < 0.0001; 64.4% versus 50.1%; p < 0.0001). Of the 23 patients with a predicted postoperative peak VO(2) less than 10 mL/kg/min, 19 had an actual postoperative peak VO(2) greater than 10 mL/kg/min (average value 13.3 mL/kg/min). All 11 patients with a predicted postoperative peak VO(2) less than 35% of predicted had an actual postoperative peak VO(2) greater than 35% of predicted (average value, 55.8%). The prediction of postoperative peak V̇O(2) using the segmental technique was inaccurate. The use of predicted postoperative peak VO(2) for patient selection must be cautioned against; future studies are warranted to refine its estimation.