Table 3: Summary of Findings From Studies on Awareness and Recall (cont).

Study (year) and location	Lehmann et al. (2003) Germany (30)	Wong et al. (2002) Canada (Ontario) (31)	Chen et al. (2002) United States (32)	Schneider et al. (2002) Germany (33)
Type of study	RCT Group 1: BIS 50 Group 2: BIS 40	RCT Group 1: BIS Group 2: SP	Prospective cohort	Prospective cohort
Quality of evidence	4-c	2	4-c	4-c
Primary purpose	To compare the hemodynamics, oxygenation, intraoperative awareness and recall, and costs at 2 different levels of BIS values (50 and 40)	To investigate the effect of BIS monitors on patients’ recovery profiles, level of postoperative cognitive dysfunction, and anesthetic drug requirement	To compare the sensitivity and specificity of BIS monitoring with that of PSI on the ability of these devices to predict loss of consciousness and emergence from anesthesia	To see if a BIS baseline score of 50 to 60 prevents awareness of getting an endotracheal intubation
Number of patients and type of surgery	62 patients undergoing first-time coronary artery bypass grafting	60 elderly patients undergoing elective orthopedic surgery	20 patients scheduled for elective laparoscopic surgical procedures	20 non-neurosurgical patients
Mean age, years (SD)	BIS 50: 63.9 (8.7) BIS 40: 65.1 (8.6)	BIS: 71 (5) SP: 70 (6)	48 (16; range, 25–72)	Responders: 40 (15) Nonresponders: 45 (17)
Female/male	BIS 50: 8/24 BIS 40: 9/21	BIS: 10/21 SP: 10/19	11/9	Responders: 3/5 Nonresponders: 3/9
Premedication	Flunitrazepam	None	Midazolam IV	Midazolam
Anesthetic agent	Midazolam and sufentanil Propofol as rescue medication for BIS values above the intended limit	Isoflurane	Induction: Propofol and fentanyl Maintenance: Desflurane and nitrous oxide	Premedication: Midazolam Anesthetic: Propofol and alfentanil
Tracheal intubation	Yes	Yes	Yes	Yes
Muscle relaxant	Yes	Yes	Yes	Yes
Methods	Patients were randomized into 2 groups. The dosage of sufentanil/midazolam was adjusted to achieve a BIS level of 45–55 (group BIS 50) in 32 patients, and a BIS level of 35–45 (group BIS 40) in 30 patients. Mild hypothermia was applied. The following data points were defined: T0: awake before the induction of anesthesia T1: at steady state after the induction of anesthesia T2: after sternotomy T3: 30 minutes after the start of CPB T4: 5 minutes after CPBT5: at the end of surgery On the third postoperative day, patients were asked to answer a standardized questionnaire to measure explicit recall.	In the SP group, the anesthesiologist was blinded to the BIS values. In the BIS group, the anesthesiologist adjusted the administration of anesthetics to maintain a BIS value of 50 to 60.	3 staff anesthesiologists administered the anesthetics and monitored depth of anesthesia using standard clinical signs. BIS monitors were positioned out of their lines of sight and a second investigator ensured proper functioning of the monitors during the operation. The third investigator recorded data at specific time intervals. Each time that electrocautery was used, the incidence of its interference with the BIS monitors was recorded. State of consciousness was assessed by the ability of patients to follow commands to open their eyes and to squeeze the investigators’ hands.	Prior to intubation, patients were tested for awareness in 1-minute intervals using the Tunstall isolated forearm technique. After at least 5 minutes of a constant BIS baseline value of 50 to 60 at unchanged infusion rates, patients were intubated. After intubation, infusion rates were kept constant for 3 minutes, then the study ended, and surgery was done. Patients were asked for recall before they left the recovery room and the next day on the ward.
Intraoperative measurements	HR, EEG, MAP, central venous pressure, mean pulmonary artery pressure, pulmonary capillary wedge pressure, cardiac output, and mixed venous oxygen saturation Cardiac index, stroke volume, left ventricular stroke work index, systemic vascular resistance, and pulmonary vascular resistance were calculated from standard formulas. Arterial and mixed venous blood gas analyses were done to calculate the index of oxygen delivery and the index of oxygen consumption according to standard formulas.	HR, systolic and diastolic BP, MAP, and mental state scores	BIS, PSI, MAP, HR, oxygen saturation, end-tidal concentration of desflurane, and nitrous oxide	ECG, HR, BP < pulse oximetry, and end-tidal carbon dioxide
Mean BIS values (SD)	BIS 50 T0: 86 (6.3) T1: 54 (4.4) T2: 51 (4.1) T3: 54 (3.1) T4: 52 (3.9) T5: 52 (4.2) BIS 40 T0: 89 (6.4) T1: 41 (7.3) T2: 39 (5.6) T3: 39 (5.8) T4: 39 (6.3) T5: 41 (4.5)	BIS: 53 (4) SP: 47 (7)	Awake: 92 (10) At induction: 89 (12) Before intubation: 38(7) Before incision: 52(11) BIS monitoring differentiated between awake and anesthetized states and was a significant predictor of unconsciousness (P <.01)	Between 50 and 60
Incidence of recall	There was no explicit recall in either group.	No patient had awareness.	No patient reported recall at the 24-hour follow-up interview.	Before intubation, no patient responded to commands, and BIS scores were between 50 and 60 according to the protocol (responders: 52; nonresponders: 54). After intubation, 8 of the 20 patients showed awareness and squeezed the investigator’s hand in response to a command. No patient had recall in 2 hours in the recovery room and in 24 hours on the ward.
Accuracy data/prediction probability	N/A	N/A	ROC curve (area under the curve): 0.79 (SD, 0.04)