Study Objective: To evaluate the hypothesis that the bispectral index (BIS) is not affected by the hypothermia that is associated with cardiopulmonary bypass (CPB).

Design: Prospective, observational study.

Setting: Cardiac surgical operating suite of a university medical center.

Patients: 100 patients undergoing cardiac surgery requiring CPB.

Interventions: A constant effect site concentration of 2.2 ng/mL for fentanyl and 60 ng/mL for midazolam was maintained throughout surgery using a computer-assisted continuous infusion technique.

Measurements: The BIS value, percent isoflurane administered, predicted brain concentrations of midazolam and fentanyl, and nasopharyngeal temperature were recorded before CPB, at 15 minutes after the onset of CPB, at placement of the aortic cross-clamp, at start of rewarming, on separation from CPB, and 15 minutes after the end of CPB. Data were analyzed using a repeated-measures mixed-effects method, taking into account temperature, age, and predicted level of each anesthetic.

Main Results: A significant overall association between temperature and BIS was observed independent of patient age, predicted brain midazolam or fentanyl concentration, percent isoflurane administered, and surgical time point (p < 0.001). The BIS is estimated to decrease by 1.12 units for each degree Celsius decrease in body temperature.

Conclusions: Hypothermia decreases the BIS by 1.12 units per degree Celsius decline in temperature.