Falsely elevated bispectral index during deep hypothermic circulatory arrest

G. Mychaskiw*, B. J. Heath and J. H. Eichhorn

University of Mississippi School of Medicine, Cardiac Anesthesiology, University of Mississippi Medical Center, 2500 North State Street, Jackson, MI 39216-4505, USA

Accepted for publication: May 27, 2000


    Abstract
 Top
 Abstract
 Introduction
 Case report
 Discussion
 References
 
A 2-month-old infant underwent repair of a ventricular septal defect under deep hypothermic circulatory arrest. Bispectral index and EEG suppression ratio were evaluated using an Aspect BIS monitor. Erroneous readings from the monitor could have led to a potentially dangerous alteration in surgical and anaesthetic management.

Br J Anaesth 2000; 85: 798–800

Keywords: monitoring, electroencephalography


    Introduction
 Top
 Abstract
 Introduction
 Case report
 Discussion
 References
 
The bispectral index (BIS) is rapidly gaining acceptance as a measure of hypnosis under anaesthesia. Additionally, the BIS monitor unit can be used as a readily available and easily interpreted tool to ensure EEG burst suppression during deep hypothermic circulatory arrest. This is displayed as a suppression ratio, which is the percentage of time during the preceding 60 s in which EEG burst suppression occurred. We report a case where a BIS monitor (Aspect Medical Systems, Spotsylvania, VA, USA) demonstrated erroneous readings during deep hypothermic circulatory arrest that could have altered the anaesthetic and surgical management of the patient.


    Case report
 Top
 Abstract
 Introduction
 Case report
 Discussion
 References
 
A 2-month-old, 3 kg male patient presented to the cardiac surgery service for repair of a ventricular septal defect (VSD). Cardiac catheterization revealed a large muscular VSD with a pulmonary to systemic blood flow ratio of greater than 4 to 1. The patient was medically managed with digoxin and furosemide and had no clinical evidence of congestive heart failure at the time of surgery.

After placement of standard, non-invasive monitors, the patient was anaesthetized with fentanyl, 10 µg kg–1, vecuronium, 0.1 mg kg–1, 30% oxygen in nitrous oxide, and isoflurane, 0.5–2.0 vol%. Anaesthesia was maintained with the volatile agent and occasionally supplemented with fentanyl and vecuronium. Additional monitors placed after induction included a femoral arterial catheter, internal jugular triple lumen catheter, Aspect model A-2000 BIS monitor (system revision 0.41) and cerebral venous oximeter (Somanetics Corporation, Troy, MI, USA). In preparation for deep hypothermic circulatory arrest, methylprednisolone, 10 mg kg–1 was administered and the head was packed in ice. Following anticoagulation with heparin, aortic and right atrial cannulae were placed and cardiopulmonary bypass (CPB) was started. At the start of CPB the BIS was in the 40s. The patient was cooled to a core temperature of 18°C and deep hypothermic circulatory arrest was instituted after 15 min of cooling time on CPB. At the time of deep hypothermic circulatory arrest the BIS was zero with a suppression ratio of 100. The cerebral mixed venous oxygen saturation (SrO2) had increased from baseline by more than 100% during cooling.

Within the first 5 min of deep hypothermic circulatory arrest, the SrO2 decreased to baseline. By 15 min of deep hypothermic circulatory arrest the SrO2 had decreased over 50% from baseline and was below the limits of measurement. The patient’s core temperature had started to slowly drift upward and by this time had reached 22°C. At 16 min of deep hypothermic circulatory arrest the BIS abruptly increased to 98–100, with a suppression ratio of 0–10 (Fig. 1). The raw EEG tracing on the BIS monitor appeared flat and unchanged. The EMG readout on the monitor throughout CPB was negligible. The surgeon was advised of the monitor readings and was asked to resume CPB as soon as possible. The repair was expedited and the patient returned to CPB within 5 min. Immediately on return to CPB, the SrO2 again increased to over 100% of baseline. After approximately 10 min of rewarming the BIS returned to its pre-deep hypothermic circulatory arrest level in the 40s. The patient was weaned from CPB and the procedure concluded without further incident. He was extubated the next morning having suffered no apparent neurological sequelae.



View larger version (10K):
[in this window]
[in a new window]
 
Fig 1 Printout of BIS values during case. Note BIS in 40s at start of CPB, 09.35 h. BIS decreased to 00 by start of deep hypothermic circulatory arrest, 09.50 h. BIS abruptly increased to 98–100 at 10.06 h, with return to baseline starting at 10.16 h as rewarming continued.

 

    Discussion
 Top
 Abstract
 Introduction
 Case report
 Discussion
 References
 
The bispectral index is an artificial variable arrived at by computer processing of, most commonly, a one-channel EEG pattern via a proprietary algorithm of the Aspect Corporation. This value, scaled from 0 to 100 reflects a patient’s level of hypnosis, and potentially, probability of recall.1 This has become an important, if somewhat emotionally charged, anaesthetic safety issue in the last few years. Some have even speculated that BIS monitoring will become the next ‘standard of care’.2 This has led to a rapid proliferation in the number of BIS monitors available in operating rooms across the USA. The Aspect monitors, the only presently available, are easy to use and interpret, requiring very little time to fully orient and train most anaesthesia practitioners. An added advantage of the Aspect monitors is their ability to display real-time EEG tracings, accompanied by the suppression ratio. With the caveat that this offers a very limited picture of the total brain EEG activity, the BIS monitors are thus a readily available and easily used EEG processing system for the detection of gross global changes.

EEG monitoring during deep hypothermic circulatory arrest is of controversial benefit. It has been suggested that outcome is improved when EEG burst suppression is achieved by cooling.3 Burst suppression has been used as an indicator of adequate brain protection during deep hypothermic circulatory arrest.4 There are other studies, however, which do not demonstrate improved outcome from use of EEG monitoring.5 As a result of these issues, coupled with the logistical and economic pressures of equipment and personnel availability, EEG monitoring during deep hypothermic circulatory arrest is not applied universally and varies greatly among institutions. The wide availability and ease of use of the Aspect BIS monitors obviates some of these concerns. Although its benefit has not been demonstrated in peer-reviewed studies, the suppression ratio detected by BIS monitors may be of value in helping to assess optimal brain protection prior to deep hypothermic circulatory arrest.

In this case the brain was cooled for 15 min prior to deep hypothermic circulatory arrest. The period of cooling was somewhat shorter than the usual practice at this institution, and was dictated by the surgeon as he felt the arrest time would be very brief for anatomic reasons. This appeared to have the desired result of metabolic suppression, as evidenced by the significant increase in SrO2. Appropri ately, the BIS decreased while the suppression ratio increased, correlating with EEG suppression induced by cooling. The rapid decrease in SrO2 following deep hypothermic circulatory arrest, however, was disturbing and suggested a possible depletion of oxygen stores in the brain, which may be seen following heterogeneous cooling, especially following an abbreviated cooling time. As the temperature drifted upward, the BIS increased abruptly to high levels at 23°C after 16 min of deep hypothermic circulatory arrest. The most worrisome thought was that this could possibly represent the monitor’s interpretation of seizure EEG secondary to anoxic insult. The surgeon was advised of these concerns, but this was tempered by the real-time EEG waveform, which appeared to be highly suppressed. It was our impression that this was most likely an artefactual product of the algorithm’s misinterpretation of EMG activity or other radiofrequency interference. The ice surrounding the patient’s head was wrapped in plastic and towelling, thus increased electrode impedance was not thought to be a factor. The surgeon quickly completed the repair and the patient was returned to CPB. When the brain had been rewarmed, the BIS returned to pre-cooling levels, further reinforcing our belief that this was an artefact, although amelioration of seizure activity by improved perfusion remained a possibility. In approximately 30 deep hypothermic circulatory arrest cases prior to the one reported, the BIS had not displayed this activity, but rather remained highly suppressed until the brain had been rewarmed. It has also not appeared in the 10 cases subsequent to the case reported. In the postoperative period the patient demonstrated no evidence of neurologic insult and was discharged home after an uneventful recovery.

Both EMG activity and radiofrequency interference from electrical equipment in the operating theatre influence the BIS algorithm. Under normal conditions, the EEG is of sufficient power to override these other electrical sources and is thus weighted appropriately by the algorithm. In circumstances of extreme EEG suppression, as during deep hypothermic circulatory arrest, it is possible that electrical interference from either EMG or radiofrequency noise may be interpreted by the algorithm as EEG activity and assigned a high BIS value. When EEG activity resumes, as during rewarming, the monitor then re-interprets it appropriately, discounting the EMG and radiofrequency activity in the algorithm (personal communication, David Zaraket, Aspect Corporation). This appears to have happened in our case, as there was no demonstrable evidence of neurologic insult, such as seizure activity or choreoathetosis, seen postoperatively.

In conclusion, we present a case where electrical interference was misinterpreted as EEG activity by an Aspect A-2000 BIS monitor during a period of intense EEG suppression. This misinterpretation was displayed as a high BIS value with a low suppression ratio. Observation of the real-time EEG waveform suggested that this was an artefact, but was not absolutely conclusive. During periods of significant EEG suppression, BIS monitors may misinterpret electrical interference as EEG and could possibly lead to unnecessary therapeutic interventions. Observation of the real-time EEG waveform may aid in the diagnosis of this artefact, which, in our experience, occurs infrequently. Further studies are necessary to determine if monitoring suppression ratio and BIS during deep hypothermic circulatory arrest is beneficial and cost effective.


    Acknowledgement
 
All work funded through the Department of Anesthesiology, University of Mississippi School of Medicine.


    Footnotes
 
* Corresponding author Back


    References
 Top
 Abstract
 Introduction
 Case report
 Discussion
 References
 
1 Sigl JC, Chamoun NC. An introduction to bispectral index analysis for the EEG. J Clin Mon 1994; 10: 392–404[ISI]

2 Mychaskiw G, Eichhorn JH. Patient safety and cost benefits of monitoring. In: Lake C, Hines RL, Blitt C, eds. Clinical Monitoring: Practical Applications. New York, NY: W.B. Saunders (in press).

3 Nakashima K, Todd MM, Warner DS. The relation between cerebral metabolic rate and ischemic depolarization. A comparison of the effects of hypothermia, pentobarbital and isoflurane. Anesthesiology 1995; 82: 1199–1208[ISI][Medline]

4 Sebel PS. Central nervous system monitoring during open heart surgery: an update. J Cardiothorac Vasc Anesth 1998; 12: 3–8

5 Roach GW, Newman MF, Murkin JM, et al. Ineffectiveness of burst suppression therapy in mitigating perioperative cerebro vascular dysfunction. Anesthesiology 1999; 90: 1255–64[ISI][Medline]