Prediction of Postoperative Complications: Ventilatory Efficiency and Rest End-tidal Carbon Dioxide

Abstract

BackgroundCardiopulmonary exercise testing parameters including ventilatory efficiency (VE/VCO2 slope) are used for risk assessment of lung resection candidates. However, many patients are unable or unwilling to undergo exercise. VE/VCO2 slope is closely related to the partial pressure of end-tidal carbon dioxide (PETCO2). We hypothesized PETCO2 at rest predicts postoperative pulmonary complications. MethodsConsecutive lung resection candidates were included in this prospective multicenter study. Postoperative respiratory complications were assessed from the first 30 postoperative days or from the hospital stay. Student t test or Mann-Whitney U test was used for comparison. Multivariate stepwise logistic regression analysis was used to analyze association with the development of postoperative pulmonary complications. The De Long test was used to compare area under the curve (AUC). Data are summarized as median (interquartile range). ResultsThree hundred fifty-three patients were analyzed, of which 59 (17%) developed postoperative pulmonary complications. PETCO2 at rest was significantly lower (27 [24-30] vs 29 [26-32] mm Hg; P < .01) and VE/VCO2 slope during exercise significantly higher (35 [30-40] vs 29 [25-33]; P < .01) in patients who developed postoperative pulmonary complications. Both rest PETCO2 with odds ratio 0.90 (95% confidence interval [CI] 0.83-0.97); P = .01 and VE/VCO2 slope with odds ratio 1.10 (95% CI 1.05-1.16); P < .01 were independently associated with postoperative pulmonary complications by multivariate stepwise logistic regression analysis. There was no significant difference between AUC of both models (rest PETCO2: AUC = 0.79 (95% CI 0.74-0.85); VE/VCO2 slope: AUC = 0.81 (95% CI 0.75-0.86); P = .48). ConclusionsPETCO2 at rest has similar prognostic utility as VE/VCO2 slope, suggesting rest PETCO2 may be used for postoperative pulmonary complications prediction in lung resection candidates.