OBJECTIVE: To determine whether currently available preoperative and intraoperative variables related to arterial oxygen tension (PaO(2)) can be used as predictors for low PaO(2) during one-lung ventilation (OLV). DESIGN: A prospective cohort study. SETTING: Primary university hospital. PARTICIPANTS: Adult patients (n = 92) undergoing thoracic surgery requiring OLV. INTERVENTIONS: Preoperative and intraoperative data, including past medical history, physical examination, and usual preoperative and intraoperative tests, were collected and used as explanatory variables for PaO(2) during OLV by univariate and multivariate analysis. A stepwise logistic regression including the same independent variables was used to identify patients who should be expected to develop arterial hypoxemia (PaO(2) <70 mmHg). Arterial blood gas samples were analyzed 15 minutes after the onset of OLV and after thoracotomy to determine the lowest PaO(2) value during OLV. MEASUREMENTS AND MAIN RESULTS: Preoperative (age, hematocrit, relative perfusion of the nondependent lung) and intraoperative (PaO(2) during 2-lung ventilation and mean arterial pressure at the lowest PaO(2)) variables were identified as independent factors affecting PaO(2) in OLV. PaO(2) during 2-lung ventilation was the only independent variable accounting for arterial hypoxemia when multivariate logistic regression was performed. CONCLUSION: The PaO(2) during OLV can be predicted using routinely available preoperative and intraoperative data. From a clinical point of view, this study failed to identify patients at risk of arterial hypoxemia when OLV is instituted because mainly intraoperative independent variables are involved in the decrease of PaO(2) in this situation. Copyright 2002, Elsevier Science (USA). All rights reserved.
OBJECTIVE: To determine whether currently available preoperative and intraoperative variables related to arterial oxygen tension (PaO(2)) can be used as predictors for low PaO(2) during one-lung ventilation (OLV). DESIGN: A prospective cohort study. SETTING: Primary university hospital. PARTICIPANTS: Adult patients (n = 92) undergoing thoracic surgery requiring OLV. INTERVENTIONS: Preoperative and intraoperative data, including past medical history, physical examination, and usual preoperative and intraoperative tests, were collected and used as explanatory variables for PaO(2) during OLV by univariate and multivariate analysis. A stepwise logistic regression including the same independent variables was used to identify patients who should be expected to develop arterial hypoxemia (PaO(2) <70 mmHg). Arterial blood gas samples were analyzed 15 minutes after the onset of OLV and after thoracotomy to determine the lowest PaO(2) value during OLV. MEASUREMENTS AND MAIN RESULTS: Preoperative (age, hematocrit, relative perfusion of the nondependent lung) and intraoperative (PaO(2) during 2-lung ventilation and mean arterial pressure at the lowest PaO(2)) variables were identified as independent factors affecting PaO(2) in OLV. PaO(2) during 2-lung ventilation was the only independent variable accounting for arterial hypoxemia when multivariate logistic regression was performed. CONCLUSION: The PaO(2) during OLV can be predicted using routinely available preoperative and intraoperative data. From a clinical point of view, this study failed to identify patients at risk of arterial hypoxemia when OLV is instituted because mainly intraoperative independent variables are involved in the decrease of PaO(2) in this situation. Copyright 2002, Elsevier Science (USA). All rights reserved.