Safety of deliberate intraoperative and postoperative hypothermia for patients undergoing coronary artery surgery: A randomized trial

Abstract

Background Hypothermia in the perioperative period is associated with adverse effects, particularly bleeding. Before termination of cardiopulmonary bypass, rewarming times and perfusion temperatures are often increased to avoid post–cardiopulmonary bypass hypothermia and the presumed complications. This practice may, however, also have adverse effects, particularly cerebral hyperthermia. We present safety outcomes from a trial in which patients undergoing coronary artery surgery were randomly assigned to normothermia or hypothermia for the entire surgical procedure. Methods Consenting patients over the age of 60 years presenting for a first, elective coronary artery surgery with cardiopulmonary bypass were randomly assigned to having their nasopharyngeal temperature maintained at either 37°C (group N; 73 patients) or 34°C (group H; 71 patients) throughout the intraoperative period, with no rewarming before arrival in the intensive care unit. All received tranexamic acid. Results There was no clinically important difference in intraoperative blood product or inotrope use. Temperatures on arrival in the intensive care unit were 36.7°C ± 0.38°C and 34.3°C ± 0.38°C in groups N and H, respectively. Blood loss during the first 12 postoperative hours was 596 ± 356 mL in group N and 666 ± 405 mL in group H (mean difference ± 95% confidence interval, 70 ± 126 mL; P = .28). There was no significant difference in blood product utilization, intubation time, time in the hospital, myocardial infarction, or mortality. The mean time in the intensive care unit was 8.4 hours less in the hypothermic group ( P = .02). Conclusions Our data support the safety of perioperative mild hypothermia in patients undergoing elective nonreoperative coronary artery surgery with cardiopulmonary bypass. These findings suggest that complete rewarming after hypothermic cardiopulmonary bypass is not necessary in all cases.