RESUMO
BACKGROUND: Air emboli released from incompletely deaired cardiac chambers may cause neurocognitive decline after open heart surgery. Carbon dioxide (CO2) field flooding is reported to reduce residual intracavital air during cardiac surgery. A protective effect of carbon dioxide insufflation on postoperative brain function remains unproven in clinical trials. METHODS: Eighty patients undergoing heart valve operations by median sternotomy were randomly assigned to either CO2 insufflation (group I, n = 39) or unprotected controls (group II, n = 41). Preoperative evaluation included neurocognitive test batteries consisting of six different tests, and objective measurements of brain function by means of P300 wave auditory-evoked potentials (peak latencies, ms). Neurocognitive testing and P300 measurements were repeated on postoperative day 5. Neurocognitive deficit (ND) was defined as a 20% decrement in two or more tests. RESULTS: Preoperatively, P300 peak latencies did not differ between groups (374 +/- 75 vs 366 +/- 72 ms, not significant [n.s.]). Five days after surgery, P300 peak latencies were significantly shorter with CO2 protection as compared with the unprotected control group (group I: 390 +/- 68 ms, group II: 429 +/- 75 ms, p = 0.02). Clinical outcome was comparable as for mortality (group I: 1 patient; group II: 2 patients) and cerebrovascular events or confusional syndromes (group I: 5 patients; group II: 4 patients) or other clinical variables as intubation time or hospital stay. Neurocognitive test batteries did not reveal differences between groups. CONCLUSIONS: Shorter P300 peak latencies after surgery indicate less brain damage in patients who underwent heart valve operations with CO2 flooding of the thoracic cavity. Even if these findings were not supported by clinical results or neurocognitive test batteries in our cohort, carbon dioxide field flooding has proven efficiency and should be advocated for all patients undergoing open heart surgery.