Secondary Logo

Journal Logo

Patient Safety: Research Report

A Pilot Study for a Prospective, Randomized, Double-Blind Trial of the Influence of Anesthetic Depth on Long-Term Outcome

Short, Timothy G. MB ChB, MD, FANZCA*; Leslie, Kate MB BS, MD, M Epi, FANZCA; Campbell, Douglas BM, FRCA, FANZCA*; Chan, Matthew T. V. MB BS, FANZCA; Corcoran, Tomas MB BCh, BAO, MRCPI, FCARCSCI, MD, FCICM§‖; O’Loughlin, Edward MB BS, FANZCA; Frampton, Chris BSc, PhD#; Myles, Paul MB BS, MPH, MD, FCARCSI, FANZCA**††

Author Information
doi: 10.1213/ANE.0000000000000209
  • Free


Observational studies have demonstrated an association between deep anesthesia, as measured by processed electroencephalographic (EEG) indices such as Bispectral Index (BIS), and postoperative mortality.1–6 Conducted in a variety of settings, these studies have generally reported a relative risk of increased mortality in the deep anesthesia groups of approximately 20%. Some of these studies reported mean volatile anesthetic concentrations that were lower in patients with lower BIS values implying increased sensitivity to anesthesia. This finding highlights the possibility that it is anesthetic sensitivity, perhaps a marker of frailty, and not deep anesthesia per se, that is the true risk factor for a poor outcome.6–13 An additional source of confounding is that relative anesthetic overdose, as well as contributing to low processed EEG values, may also result in organ hypoperfusion and dysfunction.8,14

The most robust way of determining whether the relationship between low processed EEG values and poor outcomes is causal is to perform a prospective randomized controlled trial of deep versus light anesthesia in a high-risk population.8,10 Recent studies that include outcomes after major surgery identify that patients aged >60 years, ASA physical status III or IV, having general anesthesia for surgery lasting more than 2 hours, and an expected hospital stay of ≥2 days have a 1-year mortality of about 10%.15–18 There is a wide range of practice with regard to acceptable anesthetic depth, with BIS or spectral entropy (SE) values of 30 to 60 being within the range of normal clinical practice.2–6,18–20 A randomized controlled trial comparing 2 depths of anesthesia in an at-risk group would be large, requiring 6500 patients assuming a 1-year mortality rate of 10% and a 20% relative risk reduction for mortality in the “light” group. We performed a prospective, multicenter, randomized, controlled pilot study to determine whether accurate BIS or SE titration to targets of 35 or 50 in a high-risk population was feasible, to test and refine the protocol, to establish accurate recruitment targets and trial costs, and to confirm the sample size calculation.


The study protocol was approved by the ethics review boards at each participating site. The study was registered with the Australian and New Zealand Clinical Trials Registry (ACTRN 12610000668000). Written informed consent was obtained from each patient before randomization. There were 5 participating sites in New Zealand, Australia, and Hong Kong.


Patients aged ≥60 years, ASA physical status III or IV, scheduled for surgery under general anesthesia lasting ≥2 hours, and with expected postoperative hospital stay of ≥2 days were recruited. Patients were excluded if they were having cardiac surgery, surgery requiring a wake-up test, if BIS or SE could not be monitored (e.g., cranial or intracranial surgery), or if the patients were not expected to be contactable in 1 year.


Before randomization, attending anesthesiologists were asked to define and record an appropriate mean arterial blood pressure (MAP) range that they would maintain for each patient during the procedure. This was done to ensure differences in MAP between the groups due to differences in anesthetic depth would not become a confounding variable when interpreting the results. All patients were monitored with BIS (Covidien, Mansfield, MA) or SE (GE Healthcare, Helsinki, Finland). Patients were then randomly assigned to either the “low” target (BIS/SE = 35) or “high” target (BIS/SE = 50) groups using sequential randomization envelopes. Randomization was concealed until after consent was obtained. The aim for the attending anesthesiologist was to maintain the EEG index value within 5 U of the target during the maintenance phase of anesthesia.

All patients received standard anesthetic care and monitoring. Anesthesia was maintained with either a volatile anesthetic drug or propofol infusion. Regional anesthesia combined with general anesthesia was permitted. Choice of anesthetic drugs and IV fluids was at the discretion of the attending anesthesiologist with the exception of nitrous oxide. Nitrous oxide was excluded from the anesthetic regimen because of its confounding effect on EEG monitoring.20,21 Group identity was concealed from the surgeon, patient, and research staff collecting postoperative and outcome data. At the end of the procedure, the intraoperative case report form and record of group assignment were placed in a sealed envelope to ensure blinding.


Patient demographic details, medical, and surgical history were recorded, and the Charlson comorbidity score was calculated.22 Intraoperative drug doses and physiological measurements including MAP, BIS or SE, and end-tidal volatile anesthetic concentration were recorded electronically, and if not feasible, manually at a minimum of 5-minute intervals. In the postanesthesia care unit (PACU), pain scores were assessed by a verbal rating of 0 to 10 on a numeric scale (0 = pain free and 10 = worst imaginable pain). The PACU staff recorded the time of eligibility for PACU discharge as defined by local practice and a modified Aldrete score of ≥9.23

Patients were followed up on days 1, 2, and 3 after surgery to assess quality of recovery after surgery using a 9-item quality of recovery (QoR-9) score24 and to detect postoperative complications. They were then seen before hospital discharge to detect postoperative complications and record intensive care unit and hospital length of stay. The following postoperative complications were actively searched for pneumonia, myocardial infarction, stroke, pulmonary embolism, heart failure, and wound infection. Patients were contacted by phone at 30 days and 1 year to assess for additional outcomes. If appropriate, hospital records were accessed, the patient’s doctors were approached, and/or national mortality databases were contacted to establish presence of cancer recurrence and to determine date of death.

Statistical Analysis

All patients randomly assigned to the low or high BIS targets were considered to be an intention-to-treat population for all primary and secondary analyses. Normally distributed data were summarized using mean and 95% confidence intervals (CI95), skewed data were summarized using median and interquartile range (or range), and categorical data were summarized using number and percent.

The accuracy of BIS, minimum alveolar concentration (MAC), and MAP recordings was assessed by calculating the median and interquartile range for each patient and then using the mean and CI95 of the medians for the population statistics. The differences between groups were tested using independent samples tests, χ2 tests, and Fisher exact tests as appropriate. BIS was assessed from the first decrease in BIS <60 after induction, until the last BIS <60 after switch off of the anesthetic. All records were also visually inspected for artifact. MAP was assessed for the same epoch as the BIS data. End-tidal volatile anesthetic concentration was assessed from the time of first decrease in BIS to <60 until switch off at the end of the anesthetic, defined as an inspired anesthetic concentration <0.1 MAC. The within-patient variation in BIS, MAC, and MAP was examined by comparing the individual variances using a Mann-Whitney U test and the between-patient variation using Levenes variance test.

Primary and secondary end points as well as differences in anesthetic procedures and recovery characteristics were assessed with t tests, Wilcoxon rank sum tests, χ2 tests, or Fisher exact probability tests as appropriate. All reported P values were 2 sided and not adjusted for multiple comparisons. The statistical analysis was performed in R (R Development Core Team, 2013, R: A Language and Environment for Statistical Computing, R Foundation for Statistical Computing, Vienna, Austria). The postoperative complications of pneumonia, myocardial infarction, stroke, pulmonary embolism, heart failure, and death were grouped into a composite adverse outcome score. Because the occurrence of >1 of these events was expected to be uncommon, the outcome was analyzed as the presence or absence of any of these composite adverse events and comparisons of the groups summarized as risk differences with CI95.

A sample size calculation was not performed because this was a pilot study. The primary purpose of the study was to determine whether BIS or SE titration was possible in a high-risk population, that adequate separation of depth targets was achievable, and to confirm the sample size calculation for a larger trial. It was felt these goals could be achieved with 25 patients at each of 5 study centers, giving a total sample size of 125.


The 5 participating centers recruited subjects between July 2010 and June 2011. One hundred twenty-five patients were enrolled (61 in the low group and 64 in the high group), and data collection was completed on all patients. Demographic and medical characteristics were similar in both patient groups (Table 1).

Table 1
Table 1:
Demographic Data

Mean (BIS or SE values in the low group and high group were significantly different: 39 [CI95, 38 to 41] vs 48 [CI95, 46 to 49], P = 0.0001; Fig. 1). There were no differences in the within-patient or between-patient individual variances between the groups (P = 0.19 and P = 0.47, respectively). The mean difference in BIS between the groups was 8.0 (CI95, 5.7 to 10.4). The difference in mean volatile anesthetic or propofol dosing to achieve the 2 depth targets was also significant. There was a 37% relative reduction in volatile anesthetic dosing; mean MAC = 0.98 (CI95, 0.92 to 1.05) and 0.64 (CI95, 0.59 to 0.69), P = 0.0001), the mean difference in MAC between groups being −0.35 (CI95, −0.44 to −0.26). There was a significant difference in the within-patient and the between-patient individual variances for volatile use (P = 0.0001 and P = 0.011, respectively). There was a 23% relative reduction in propofol dosing (mean 4.0 [CI95, 3.7 to 4.3] μg/mL and 3.1 (CI95, 2.7 to 3.5] μg/mL, P = 0.004) in the low group compared with the high group (Table 2). The difference in propofol targets was −0.74 (−1.2 to 0.32) μg/mL. There were inadequate data to examine the variances for the propofol infusion group. The mean difference in propofol targets was −0.8 (CI95, −1.2 to −0.3). The mean MAP was 85 (CI95, 81 to 88) mm Hg in the low group and 87 (CI95, 84 to 90) mm Hg in the high group (P = 0.86). The mean difference in MAP between the groups was 2 (CI95, −2 to 6) mm Hg. There were no differences in the within-patient or between-patient individual variances between the groups (P = 0.45 and P = 0.59, respectively). The chosen target range and number of patients tracking accurately within this range were similar in both groups. More regional anesthesia, predominantly epidural analgesia, was used in the high group (P = 0.012), but with wide CIs. Postoperative pain scores (P = 0.67), QoR-9 scores (P = 0.67), and hospital lengths of stay (P = 0.15) were similar in the 2 groups.

Table 2
Table 2:
Comparison of Intraoperative and Recovery Characteristics
Figure 1
Figure 1:
Mean of median (interquartile range [IQR]) Bispectral Index (BIS) or spectral entropy (SE) for each of the 125 patients. The solid line indicates the group each patient was assigned to. The dashed lines are the mean (95% confidence intervals) BIS/SE targets achieved for the group.

Postoperatively, there was a higher incidence of wound infection in the low group than the high group (13% vs 3%, risk difference −10 [CI95, −21 to −0.1], P = 0.04; Table 3). There were no differences in macroscopic cancer recurrence (P = 0.86), any major complications at 1 month (P = 0.56) and 1 year (P = 0.15), or all-cause mortality at 1 year (P = 0.70). Two patients in each group had >1 nonfatal adverse event.

Table 3
Table 3:
Postoperative Outcomes


This pilot study demonstrated that BIS and SE titration to assigned low and high targets was possible. Although there was a centralizing tendency, we consider the separation in mean BIS or SE of 9 U is adequate in terms of a clinically relevant difference in pharmacodynamic effect. Also, there was no overlap of the CI95. BIS values change very little over wide ranges of end-tidal volatile anesthetic concentration in some patients, and our data reinforce this finding.25 BIS or SE titration can be problematic in individual patients as demonstrated by the large interquartile ranges of EEG index values in individual patients. Recent publications highlight these issues. In the dexamethasone, light anesthesia and tight glucose control (DeLiT) randomized controlled trial, a 3-fold factorial design study, 2 of the assigned groups were BIS targets of 35 and 55, yet the targets achieved were 44 and 50 showing a strong centralizing tendency and a BIS separation of only 6 U.26 A study by Radtke et al.27 comparing BIS-guided anesthesia in the range of 40 to 60 with BIS concealed found no difference in mean BIS, which was 39 in both groups. In contrast, we achieved a clinically relevant and statistically significant separation in depth targeting with a 37% difference in mean end-tidal volatile anesthetic concentrations and 23% difference in mean propofol target concentrations but no difference in MAPs. Hence, 2 of the stated goals of the pilot trial were achieved without a confounding difference in MAP. There was increased within- and between-patient variance in the MAC for the “light” group. This was presumably because the BIS is more responsive to changes in surgical stimulus at light levels of anesthesia, leading to more frequent changes in volatile drug dose to maintain target BIS.

More epidurals were performed in the high group (8% vs 23%, P = 0.012). It is unclear whether this was due to a random allocation difference between groups or bias introduced by more regional anesthesia being performed after randomization in the high group. This potential confounder will need to be accounted for in any definitive trial. There were no other clinically significant differences in intraoperative variables between the 2 groups.

Early recovery variables including time to eligibility of PACU discharge, PACU pain score, postoperative day 1 QoR-9 score, and hospital length of stay were similar in the 2 groups. The 1-month and 1-year composite outcome of major complications, cancer recurrence, and 1-year all-cause mortality also showed no statistically significant differences. These findings were expected because this trial was not powered for outcome differences of this nature. The overall 1-year all-cause mortality was 10.4%, and overall incidence of composite end points was 22.4%, which suggests we have identified a suitable high-risk population for a large trial. The statistically significant difference in wound infection between groups could be a type 2 error. However, the study by Chan et al.18 in 2013 also found a 28% increase in infection in a BIS-concealed group, who had a similar level of anesthesia to the (“deep”) group in this study. This finding warrants further investigation in a larger trial.

In conclusion, this pilot study showed that BIS or SE titration is possible in a high-risk patient population and a clinically important separation of depth targets can be achieved. The outcome data in this pilot study suggest that an appropriate at-risk population has been identified and confirm the complication rates observed in observational studies of this patient group. An accurate sample size calculation for a definitive randomized controlled study can be made. In conclusion, we have demonstrated the feasibility of a large, multicenter, randomized, controlled, safety, and efficacy trial of the influence of anesthetic depth on long-term outcome.


This study was supported by a pilot study grant from the Australian and New Zealand College of Anaesthetists and a project grant from the Auckland District Health Board Research Trust.


Name: Timothy G. Short, MB ChB, MD, FANZCA.

Contribution: This author helped design the study, conduct the study, collect the data, analyze the data, and prepare the manuscript.

Attestation: This author reviewed the original study data and data analysis and attests to the data integrity and analysis as reported in the manuscript. This author is the archival author.

Name: Kate Leslie, MB BS, MD, M Epi, FANZCA.

Contribution: This author helped design the study, conduct the study, collect the data, and prepare the manuscript.

Attestation: This author reviewed the analysis of the data and approved the final manuscript.

Name: Douglas Campbell, BM, FRCA, FANZCA.

Contribution: This author helped design the study, conduct the study, collect the data, analyze the data, and prepare the manuscript.

Attestation: This author reviewed the original study data and data analysis and attests to the data integrity and analysis as reported in the manuscript.

Name: Matthew T. V. Chan, MB BS, FANZCA.

Contribution: This author helped design the study, conduct the study, collect the data, and prepare the manuscript.

Attestation: This author reviewed the analysis of the data and approved the final manuscript.

Name: Tomas Corcoran, MB BCh, BAO, MRCPI, FCARCSCI, MD, FCICM.

Contribution: This author helped conduct the study, collect the data, and prepare the manuscript.

Attestation: This author reviewed the analysis of the data and approved the final manuscript.

Name: Edward O’Loughlin, MB BS, FANZCA.

Contribution: This author helped conduct the study, collect the data, and prepare the manuscript.

Attestation: This author reviewed the analysis of the data and approved the final manuscript.

Name: Chris Frampton, BSc, PhD.

Contribution: This author helped design the study, analyze the data, and prepare the manuscript.

Attestation: This author reviewed the analysis of the data and approved the final manuscript.

Name: Paul Myles, MB BS, MPH, MD, FCARCSI, FANZCA.

Contribution: This author helped design the study, conduct the study, collect the data, and prepare the manuscript.

Attestation: This author reviewed the analysis of the data and approved the final manuscript.

This manuscript was handled by: Sorin J. Brull, MD, FCARCSI (Hon).


1. Monk TG, Saini V, Weldon BC, Sigl JC. Anesthetic management and one-year mortality after noncardiac surgery. Anesth Analg. 2005;100:4–10
2. Lindholm ML, Träff S, Granath F, Greenwald SD, Ekbom A, Lennmarken C, Sandin RH. Mortality within 2 years after surgery in relation to low intraoperative bispectral index values and preexisting malignant disease. Anesth Analg. 2009;108:508–12
3. Leslie K, Myles PS, Forbes A, Chan MT. The effect of bispectral index monitoring on long-term survival in the B-aware trial. Anesth Analg. 2010;110:816–22
4. Kertai MD, Pal N, Palanca BJ, Lin N, Searleman SA, Zhang L, Burnside BA, Finkel KJ, Avidan MSB-Unaware Study Group. . Association of perioperative risk factors and cumulative duration of low bispectral index with intermediate-term mortality after cardiac surgery in the B-Unaware Trial. Anesthesiology. 2010;112:1116–27
    5. Kertai MD, Palanca BJ, Pal N, Burnside BA, Zhang L, Sadiq F, Finkel KJ, Avidan MSB-Unaware Study Group. . Bispectral index monitoring, duration of bispectral index below 45, patient risk factors, and intermediate-term mortality after noncardiac surgery in the B-Unaware Trial. Anesthesiology. 2011;114:545–56
      6. Sessler DI, Sigl JC, Kelley SD, Chamoun NG, Manberg PJ, Saager L, Kurz A, Greenwald S. Hospital stay and mortality are increased in patients having a “triple low” of low blood pressure, low bispectral index, and low minimum alveolar concentration of volatile anesthesia. Anesthesiology. 2012;116:1195–203
      7. Monk TG, Weldon BC. Anesthetic depth is a predictor of mortality: it’s time to take the next step. Anesthesiology. 2010;112:1070–2
      8. Leslie K, Short TG. Low bispectral index values and death: the unresolved causality dilemma. Anesth Analg. 2011;113:660–3
      9. Cohen NH. Anesthetic depth is not (yet) a predictor of mortality! Anesth Analg. 2005;100:1–3
      10. Kurata J. Deep hypnosis as a sign of “imbalance” in balanced anesthesia. Anesth Analg. 2010;110:663–5
      11. Kalkman CJ, Peelen LM, Moons KG. Pick up the pieces: depth of anesthesia and long-term mortality. Anesthesiology. 2011;114:485–7
      12. Arbous MS, Grobbee DE, van Kleef JW, de Lange JJ, Spoormans HH, Touw P, Werner FM, Meursing AE. Mortality associated with anaesthesia: a qualitative analysis to identify risk factors. Anaesthesia. 2001;56:1141–53
      13. Sigurdsson GH, McAteer E. Morbidity and mortality associated with anaesthesia. Acta Anaesthesiol Scand. 1996;40:1057–63
      14. Ballard C, Jones E, Gauge N, Aarsland D, Nilsen OB, Saxby BK, Lowery D, Corbett A, Wesnes K, Katsaiti E, Arden J, Amoako D, Amaoko D, Prophet N, Purushothaman B, Green D. Optimised anaesthesia to reduce post operative cognitive decline (POCD) in older patients undergoing elective surgery, a randomised controlled trial. PLoS One. 2012;7:e37410
      15. Rigg JR, Jamrozik K, Myles PS, Silbert BS, Peyton PJ, Parsons RW, Collins KSMASTER Anaethesia Trial Study Group. . Epidural anaesthesia and Anaesthesia and outcome of major surgery: a randomised trial. Lancet. 2002;359:1276–82
      16. Myles PS, Leslie K, Chan MT, Forbes A, Paech MJ, Peyton P, Silbert BS, Pascoe EENIGMA Trial Group. . Avoidance of nitrous oxide for patients undergoing major surgery: a randomized controlled trial. Anesthesiology. 2007;107:221–31
        17. Story DA, Leslie K, Myles PS, Fink M, Poustie SJ, Forbes A, Yap S, Beavis V, Kerridge RREASON Investigators, Australian and New Zealand College of Anaesthetists Trials Group. . Complications and mortality in older surgical patients in Australia and New Zealand (the REASON study): a multicentre, prospective, observational study. Anaesthesia. 2010;65:1022–30
          18. Chan MT, Cheng BC, Lee TM, Gin TCODA Trial Group. . BIS-guided anesthesia decreases postoperative delirium and cognitive decline. J Neurosurg Anesthesiol. 2013;25:33–42
          19. Myles PS, Leslie K, McNeil J, Forbes A, Chan MT. Bispectral index monitoring to prevent awareness during anaesthesia: the B-Aware randomised controlled trial. Lancet. 2004;363:1757–63
          20. Ozcan MS, Ozcan MD, Khan QS, Thompson DM, Chetty PK. Does nitrous oxide affect bispectral index and state entropy when added to a propofol versus sevoflurane anesthetic? J Neurosurg Anesthesiol. 2010;22:309–15
          21. Liley DT, Leslie K., Sinclair NC, Feckie M. Dissociating the effects of nitrous oxide on brain electrical activity using fixed order time series modeling. Comput Biol Med. 2008;38:1121–30
          22. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373–83
          23. Aldrete JA. The post-anesthesia recovery score revisited. J Clin Anesth. 1995;7:89–91
          24. Myles PS, Hunt JO, Nightingale CE, Fletcher H, Beh T, Tanil D, Nagy A, Rubinstein A, Ponsford JL. Development and psychometric testing of a quality of recovery score after general anesthesia and surgery in adults. Anesth Analg. 1999;88:83–90
          25. Whitlock EL, Villafranca AJ, Lin N, Palanca BJ, Jacobsohn E, Finkel KJ, Zhang L, Burnside BA, Kaiser HA, Evers AS, Avidan MS. Relationship between bispectral index values and volatile anesthetic concentrations during the maintenance phase of anesthesia in the B-Unaware trial. Anesthesiology. 2011;115:1209–18
          26. Abdelmalak BB, Bonilla A, Mascha EJ, Maheshwari A, Tang WH, You J, Ramachandran M, Kirkova Y, Clair D, Walsh RM, Kurz A, Sessler DI. Dexamethasone, light anaesthesia, and tight glucose control (DeLiT) randomized controlled trial. Br J Anaesth. 2013;111:209–21
          27. Radtke FM, Franck M, Lendner J, Krüger S, Wernecke KD, Spies CD. Monitoring depth of anaesthesia in a randomized trial decreases the rate of postoperative delirium but not postoperative cognitive dysfunction. Br J Anaesth. 2013;110(Suppl 1):i98–105
          © 2014 International Anesthesia Research Society