Relationship of predicted postoperative product to postpneumonectomy cardiopulmonary complications.

BACKGROUND This retrospective analytic study was to evaluate diffusing capacity of the lung for carbon monoxide (DLCO) and predicted postoperative product (PPP) as predictors of postpneumonectomy cardiopulmonary complications. METHODS One-hundred fifty-one patients underwent pneumonectomy at Vancouver General Hospital from January 1992 to December 1997. The PPP was obtained by multiplying the predicted postoperative (ppo) FEV1 by the ppo DLCO, both expressed as % of predicted. The ppo FEV1 and DLCO were derived by calculating the proportional loss of functional lung from the resected lung. We also evaluated a new index, the measured product (MP), obtained by multiplying the measured preoperative FEV1 by DLCO, both expressed as % of predicted. RESULTS Patients with complications had lower FEV1, lower DLCO, lower MP, lower ppo FEV1, lower ppo DLCO, and lower PPP than patients without complications. DLCO at 70% of predicted was the best predictor of postoperative complications, while PPP at 1400 was similar to MP at 5000 and ppo DLCO at 40% of predicted in predicting postoperative complications. The complication rate was 88% in patients with DLCO < 70% of predicted, compared with a complication rate of 19% in patients with DLCO > or = 70% of predicted (sensitivity = 68%, specificity = 93%), while the complication rate was 66% in patients with PPP < 1400, compared with a complication rate of 25% in patients with PPP > or = 1400 (sensitivity = 63%, specificity = 78%). CONCLUSIONS DLCO < 70% of predicted is associated with increased risk of complications following pneumonectomy. PPP determined preoperatively allows a patient with a critically low value (< 40% of predicted) for one variable (either ppo FEV1 or ppo DLCO) to be accepted for surgery on the basis of a good value in the other. Patients with PPP > or = 1400 have a relatively low postpneumonectomy complication rate.