Respiratory failure and intraoperative oxygenation following oesophagectomy

Respiratory complications are a major cause of morbidity andmortality following oesophagectomy.¹ We hypothesized that patientswho develop postoperative respiratory failure would have anunstable intraoperative course. Two stage oesophagectomy requiresa period of one lung ventilation (OLV) and one measure of intraoperativeinstability is desaturation during OLV. We therefore studiedthe relationship between Sa_O2 during OLV and postoperative course.Previous work has focused on preoperative factors associatedwith postoperative complications.² There is little informationon intraoperative factors that may influence outcome. We performed a retrospective analysis of all elective oesophagectomiesperformed between January 1998 and August 1999. Seventy-sevenpatients had a 2-stage sub-total oesophagectomy requiring OLV.Oxygen saturation recorded on the anaesthetic chart during OLVwas plotted on a fixed linear scale of 0–100% againsttime for each patient. The area above the curve was calculated,as a measure of intraoperative oxygenation during OLV. Thiswas standardized by dividing the area by the OLV duration, thusobtaining the Area Per Unit Time (APUT). The postoperative courseof each patient was recorded including outcome, total durationof ITU stay and the worst Pa_O2/F_O2 ratioattained. ARDS was defined using the American-European ConsensusConference on ARDS criteria. Twenty-one patients had a prolongedITU stay (>48 h) and all had a persistent Pa_O2/F_O2<27. Thirteen patients (17% of the total) fulfilled the ConsensusConference criteria for ARDS and seven of these died in hospital. Total OLV time was not significantly different between the groupsbut patients who developed ARDS had significantly greater intraoperativehypoxaemia as measured by the APUT (Table 17). We conclude thatintraoperative hypoxaemia is associated with respiratory failurefollowing oesophagectomy.