Lung-diffusing capacity for carbon monoxide predicts early complications after cardiac surgery

PURPOSE: Preoperative pulmonary dysfunction has been associated with increased operative mortality and morbidity after cardiac surgery. This study aimed to determine whether values for the diffusing capacity of the lung for carbon monoxide (DL(CO)) could predict postoperative complications after cardiac surgery. METHODS: This study included 408 consecutive patients who underwent cardiac surgery between June 2008 and December 2015. DL(CO) was routinely determined in all patients. A reduced DL(CO) was clinically defined as %DL(CO) < 70%. %DL(CO) was calculated as DL(CO) divided by the predicted DL(CO). The association between %DL(CO) and in-hospital mortality was assessed, and independent predictors of complications were identified by a logistic regression analysis. RESULTS: Among the 408 patients, 338 and 70 had %DL(CO) values of ≥ 70% and < 70%, respectively. Complications were associated with in-hospital mortality (P < 0.001), but not %DL(CO) (P = 0.275). A multivariate logistic regression analysis with propensity score matching identified reduced DL(CO) as an independent predictor of complications (OR, 3.270; 95%CI, 1.356–7.882; P = 0.008). CONCLUSIONS: %DL(CO) is a powerful predictor of postoperative complications. The preoperative DL(CO) values might provide information that can be used to accurately predict the prognosis after cardiac surgery. CLINICAL TRIAL REGISTRATION NUMBER: UMIN000029985.