Comparison of Positive End-Expiratory Pressure of 8 versus 5 cm H2O on Outcome After Cardiac Operations.

PURPOSE: Postoperative positive end-expiratory pressure (PEEP) selection in patients who are mechanically ventilated after cardiac operations often seems random. The aim of this investigation was to compare the 2 most common postoperative initial PEEP settings at our institution, 8 and 5 cm H2O, on postoperative initial tracheal intubation time (primary outcome); cardiovascular intensive care unit (CVICU); hospital length of stay (LOS); occurrence of pneumonia; and hospital mortality (secondary outcomes).

MATERIALS AND METHODS: The electronic medical records of patients who were mechanically ventilated after isolated coronary artery bypass grafting (CABG) or combined CABG and valve operations were reviewed. Propensity score matching was used to compare patients with an initial postoperative PEEP setting of 8 cm H2O (n = 4722 [25.9%]) with those who had PEEP of 5 cm H2O (n = 13 535 [74.1%]) on the primary and secondary outcomes listed earlier.

RESULTS: There was no difference in initial postoperative intubation time between the PEEP of 8 cm H2O and the PEEP of 5 cm H2O patient groups (mean 11.9 vs 12.0 hours [median 8.2 vs 8.8 hours], P = .89). The groups did not differ on the occurrence of pneumonia (0.43% vs 0.60%, P = .25) nor on hospital mortality (0.47% vs 0.43%, P = .76). Aspiration pneumonia occurrence approached a significant difference (0.06% vs 0.21%, P value = .052), as did CVICU LOS (mean: 47.9 vs 49.8 hours [median: 28.5 vs 28.4 hours], P = .057), but were not statistically different. There was a slight but likely clinically unimportant difference in hospital LOS (7.7 vs 7.4 days, PEEP = 8 vs 5, P < .001).

CONCLUSION: Patients being mechanically ventilated after cardiac operations with an initial postoperative PEEP setting of 8 versus 5 cm H2O differed significantly only on hospital LOS but the difference was likely clinically unimportant. Thus, use of 8 cm H2O PEEP in these patients without a clinical indication, although likely not harmful, does not seem beneficial.