Predicting arterial oxygenation during one-lung anaesthesia.

Eighty patients undergoing elective thoracotomy were studied to assess the possibility of predicting arterial oxygenation (PaO2) during one-lung anaesthesia (OLA). The first 50 patients were studied retrospectively. The method of multiple linear regression was used to construct a predictive equation for PaO2 during OLA. Potential predictors of PaO2 during OLA which were considered were: age, side of operation, preoperative pulmonary flow rates, preoperative and intraoperative PaO2 during two-lung ventilation. The three most significant predictors for PaO2 during OLA were: side right of operation (P < 0.05), preoperative FEV1% (P < 0.01) and intraoperative PaO2 during two-lung ventilation (P = 0.0001). The predictive equation for PaO2 after ten minutes of OLA was: PaO2 = 100 - 72 (side) - 1.86 (FEV1%) + 0.75 (two-lung) PaO2; (for side insert 0 for left-sided thoracotomy and 1 for right-sided thoracotomy). The remaining 30 patients were studied prospectively and the predicted PaO2 correlated with the observed PaO2 after ten minutes of OLA (r = 0.73, P < 0.01). Four of 30 patients had a predicted PaO2 at ten minutes of OLA < 150 mmHg. Of these, 2/4 subsequently required abandonment of OLA for pulse oximetric saturation < 85%. We conclude that although it is not possible to predict an individual patient's PaO2 during OLA with a high degree of accuracy, it is possible, before the initiation of OLA, to identify those patients whose arterial oxygenation is likely to decrease to low levels during OLA.