In pediatric anesthesia, monitoring the depth of anesthesia has received increasing attention with the advent of electroencephalogram (EEG)-derived devices.1 One of these is the bispectral index monitor (BIS®, Aspect Medical Systems, Newton, MA). The BIS is an empirically calibrated number, derived from adult EEG data, which correlates with depth of the hypnotic component of general anesthesia in adults. It is a validated monitor of depth of anesthesia for children older than 1 yr.2
In children with intellectual disability, the EEG could show abnormal activity due to the underlying cerebral pathology and/or epileptiform activity. Many intellectually disabled children also have epilepsy or cerebral palsy3,4 and thus may be treated with anticonvulsants and/or spasmolytics. As a consequence, one might question the validity and usefulness of BIS as a measure of sedation in this particular group of patients.
There is little evidence about the use of BIS in children with intellectual disability and/or those who are treated with anticonvulsants. Choudhry and Brenn5 described BIS values in 20 children with quadriplegic cerebral palsy and mental retardation changing similarly to those in 21 nonretarded children during general anesthesia. After premedication, the mean BIS value in the cerebral palsy and mental retardation group was 92 compared with 97 in the group of normal children. Saricaoglu et al.6 demonstrated that 20 children with cerebral palsy needed less propofol to obtain a BIS value between 35 and 45 during induction of general anesthesia than did 20 otherwise healthy children. Twelve of the children with cerebral palsy had a history of seizures, 10 of whom were treated with anticonvulsants.
Because of limited experience with the use of the BIS monitor in patients with a combination of neurological disorders, epilepsy, and regular anticonvulsant medication, ASPECT Medical Systems, the manufacturer of the BIS monitor, advises caution in the interpretation of BIS values in this group of patients.*
A reliable depth of anesthesia monitor would be very helpful in titrating anesthesia in children with intellectual disability for various reasons. First, they often present with many congenital anomalies, syndromes, and disorders that might require surgical intervention. Second, anesthesiologists face communication problems, difficulties in assessing level of consciousness, and unexpected consequences of pharmacodynamic interactions when anesthetizing children with intellectual disability, severe neurological disorders, or epilepsy.
The first aim of this exploratory study was to compare BIS values at different stages of anesthesia between intellectually disabled children and controls. The second aim was to investigate the discriminative properties of BIS between consciousness and unconsciousness for intellectually disabled children and for controls.
This prospective, observational study was performed at the Erasmus University Medical Center–Sophia Children’s Hospital, Rotterdam, The Netherlands, between September 2006 and September 2007.
The study protocol and data collection were approved by the local ethics review board. Written informed consent was obtained from the parents.
Children eligible for this study (aged 2–13 yr) were scheduled for elective, diagnostic gastroduodenoscopy under general anesthesia and when indicated combined with percutaneous endoscopy gastrostomy (PEG) placement. Any contraindication for the standard anesthetic regimen was an exclusion criterion. The study group included intellectually disabled children who underwent gastroduodenoscopy combined with PEG placement because of severe feeding difficulties and inability to adequately feed those children.7 The control group included children without intellectual disabilities, neurological diseases, or epilepsy, and who were not being treated with anticonvulsant medication or spasmolytics.
Diagnosis of Intellectually Disabled
The diagnosis of intellectually disabled was made by a pediatric neurologist after multiple patient clinic visits. In 16 of 17 patients, this diagnosis was further supported by magnetic resonance imaging and/or specific genetic or serum tests.
All patients received a standardized anesthetic regimen without premedication. Immediately before induction of general anesthesia, BIS electrodes (BIS Pediatric Sensor [4 sensors], Aspect Medical Systems) were placed on the patient’s forehead and temple, as recommended by the manufacturer. However, if a child resisted placement, no other attempt was made until after loss of consciousness (LOC, defined as loss of eyelash reflex). Electrodes were connected to a BIS monitor (A-2000, version 3.2; Aspect Medical Systems). Before induction of general anesthesia, 1 attempt was made to secure IV access. If successful, the patients received alfentanil (20 μg/kg IV) and propofol (2–4 mg/kg IV) to facilitate tracheal intubation. Whenever IV access could not be achieved in 1 attempt, inhaled induction with 8% sevoflurane (fraction inspired) in 100% oxygen with a fresh gas flow of 10 L/min was chosen and IV access was secured after LOC. In the case of inhaled induction, tracheal intubation was also facilitated by alfentanil (20 μg/kg) and propofol (2–4 mg/kg).
After tracheal intubation, anesthesia was maintained with isoflurane (1.5% fraction expired). The timepoint of stable intraoperative anesthesia was defined as 30 s before introduction of the gastroscope. Administration of anesthetics was then stable and a (surgical) painful stimulus was absent. All children received another dose of alfentanil (10 μg/kg) before introduction of the gastroscope. Patients’ lungs were mechanically ventilated to normocapnia (end-tidal CO2 35–40 mm Hg). Pulse oximetry, noninvasive arterial blood pressure, electrocardiogram, BIS, and concentrations of isoflurane, carbon dioxide, and oxygen were monitored throughout the procedure. At the end of the procedure, isoflurane administration was discontinued. After sufficient spontaneous breathing returned, the anesthesiologist checked for a response to verbal commands. Return of consciousness was identified on the basis of the patient’s preoperative reactions to verbal commands. Patients were tracheally extubated immediately after return of consciousness and then transferred to the postanesthesia care unit.
Data for previously defined landmark events (all medication administrations, LOC, intubation, start and end of the procedure, discontinuation of isoflurane administration, return of consciousness, extubation, and unexpected events) were recorded using Rugloop software (Demed, Temse, Belgium). Recorded data included BIS and related values (signal quality, electromyography, and suppression ratio) and standard monitoring values (pulse oximetry, noninvasive arterial blood pressure, electrocardiogram, and concentrations of isoflurane, carbon dioxide, and oxygen).
Data were analyzed using SPSS version 15.0 (SPSS, Chicago, IL) and Labgrab (Demed). A P value less than 0.05 was considered statistically significant.
Normal distribution of the data was tested by the 1-sample Kolmogorov-Smirnov test. To compare frequencies of normally distributed data, parametric tests (independent-samples t-test) were used. Nonparametric tests (Mann-Whitney U-test) served to compare the frequencies of ordinal data (e.g., all BIS values). A Fisher’s exact test was used to test differences in the distribution of nominal data. Data are given in mean (standard deviation) or median (interquartile range), as appropriate. For compensating the considerable time delay of signal processing,8 BIS values 0 and 30 s after each landmark event were analyzed.
For both groups, we used a receiver operating characteristic (ROC) curve to describe the discriminating performance of the BIS monitor between consciousness and unconsciousness. We used 4 BIS values per patient in this analysis. The awake value and the value 30 s after return of consciousness were marked as conscious values, and the value 30 s after LOC and intraoperative values were marked as unconscious values. First, we plotted the true-positive rate against (sensitivity) the false-positive rate (1-specificity) for the measured BIS values. Second, for each group, the BIS value with the highest combination of sensitivity and specificity was selected as the optimal cutoff BIS value for discrimination between conscious and unconscious state. Third, the area under the curve (AUC) was calculated. A large AUC corresponds with high discriminative properties. Fourth, we compared the ROC curve of the intellectually disabled group with the ROC curve of the control group by testing the statistical significance of the difference between the AUCs.
Seventeen patients were enrolled in the intellectually disabled group and 35 in the control group. The demographic characteristics of both groups were comparable (Table 1). None of the children in the control group was treated with anticonvulsants or spasmolytics. Ten intellectually disabled patients were treated with anticonvulsants (benzodiazepines, barbiturates, and/or antiepileptic drugs) for seizure control and 3 other patients with spasmolytics (baclofen). The etiology of underlying neurological disorders in the intellectually disabled group comprises 12 different disorders (Table 2).
Differences in BIS Values Between the Intellectually Disabled and Control Groups
BIS values (median [interquartile range]) for the intellectually disabled group were significantly lower than those for controls in the awake state (72 [48–77] vs 97 [84–98], P < 0.001), during stable intraoperative anesthesia (34 [21–45] vs 43 [33–52], P = 0.02), and during return of consciousness (59 [36–68] vs 73 [64–78], P = 0.009) (Table 3 and Fig. 1).
The Fisher’s exact test revealed that the proportion of missing awake BIS values was significantly higher in the control group: 24 of 35 vs 5 of 17 in the intellectually disabled group (P = 0.02). In the intellectually disabled group, 8 of 17 of the loss of consciousness values was missing versus 25 of 35 in the control group (P = 0.13). None of the intraoperative values was missing in the intellectually disabled group versus 3 of 35 in the control group (P = 0.54). On the return of consciousness values, 3 of 17 was missing in the intellectually disabled group versus 5 of 35 in the control group (P = 1.00).
The discriminative properties of the BIS monitor for the state of consciousness were comparable between the groups. The AUC for the control group was not statistically significantly larger than the AUC for the intellectually disabled group (0.92 vs 0.78, P = 0.07). In Figure 2, the ROC curve for each group is displayed and optimal cutoff BIS values have been marked.
For the control group, the optimal cutoff BIS value was 65 (sensitivity = 0.81 and specificity = 0.93) and for the study group it was 47 (sensitivity = 0.73 and specificity = 0.81).
BIS values observed in the awake state, during stable intraoperative anesthesia, and during return of consciousness were significantly lower for intellectually disabled children compared with controls. The differences between the 2 groups ranged from 9 to 25. BIS distinguished the conscious and unconscious state equally well in the 2 groups. Nevertheless, the optimal cutoff BIS value for discrimination between the conscious and unconscious state was 28 points lower for the intellectually disabled group.
On the BIS scale from 100 (fully awake) to 0 (isoelectric EEG), differences of these magnitudes could easily give rise to misinterpretation of the patient’s state of consciousness.
Comparison with Previous Studies
These data confirm and supplement the observations of Choudhry and Brenn5 and Saricaoglu et al.6 Both groups reported lower BIS values in intellectually disabled children. However, contrary to our findings, in those previous studies, the BIS values of intellectually disabled children remained within the defined ranges for the different stages of anesthesia.
Still, we should be aware of essential differences between these 2 studies and this study. First, anesthesia management was substantially different. We administered both IV and inhaled anesthetics, whereas Saricaoglu et al. administered only propofol. Patients in the study by Choudhry and Brenn were premedicated with midazolam, received inhaled induction with sevoflurane, and tracheal intubation was facilitated with rocuronium. Second, children in the study groups of both previous studies had been diagnosed with cerebral palsy and intellectual disability, but details on the underlying causes of cerebral palsy or the severity of intellectual disability are lacking.
Possible Explanations for Lower BIS Values in Intellectually Disabled Children
The small sample size and the heterogeneity of the group of intellectually disabled children in this study do not allow establishing causal relationships that might explain the lower BIS values in this group. Alternatively, we suggest 3 possible explanations based on findings from the literature.
- A review by Dahaba9 summarizes a large variety of conditions (anesthetic drugs, clinical conditions, and electric device interference) that could result in the BIS indicating an incorrect hypnotic state. The effect of most of those conditions on the EEG signal is clear. Dahaba gives no definite explanation for lower BIS values in intellectually disabled children. We suggest that the underlying cerebral pathology could cause epileptiform and nonepileptiform EEG abnormalities and thus affect BIS values. For example, nonepileptiform EEG abnormalities were observed in the majority of a cohort of children with tetraplegia/diplegia.10
- Epileptiform activity during general anesthesia has been reported to affect BIS values. Verma and Radtke11 reviewed studies about ictal and interictal EEG activity. Interictal activity in partial epilepsy concerns several patterns of δ waves. The frequency band for δ waves is between 1 and 4 Hz. Those low-frequency waves are generated during natural sleep but are also seen during general anesthesia as an effect of hypnotic drugs.9,12 These interictal δ waves could artificially decrease BIS values in children with epilepsy.
- Not only epileptiform activity but also anticonvulsants might influence BIS values. Most anticonvulsants exert their effect by facilitating γ-aminobutyric acid (GABA)-mediated inhibition via allosteric interaction with neuronal postsynaptic GABAA receptors. Propofol and volatile anesthetics also act on the GABAA receptor in the central nervous system. In this way, anticonvulsants may directly influence depth of anesthesia. Ten of the intellectually disabled children in our study were treated with anticonvulsants versus none of the control children.
Limitations of This Study
A considerable number of awake BIS values for the control group is missing. However, the obtained BIS values in the control children were comparable to those reported in other pediatric studies.13,14 We assume that the awake BIS values of the other control children in our study would have been comparable to those reported by Denman et al.13 and by Blussé et al.14 The crux of the matter is that the awake BIS values of intellectually disabled children were much lower than the ones measured in the control group. Furthermore, for logistical reasons and the fact that one-quarter of the patients received inhaled induction, a larger proportion of the LOC data is missing for both groups. Therefore, the LOC data should be interpreted with caution.
Heterogeneity in neurological diagnoses (Table 2) is a realistic reflection of the intellectually disabled patient population in need of a PEG but makes it difficult to generalize the results of the study to all intellectually disabled patients.
We advise anesthesiologists to be alert to possible lower BIS values in children who are intellectually disabled. There is a risk that they will inadvertently misinterpret the state of consciousness in these children.
Given the variety and rarity of underlying neurological disorders, only large multicenter trials might provide decisive information about BIS monitoring in intellectually disabled children. In addition, the effects of epileptiform EEG activity and anticonvulsant therapy on BIS should be studied. It would also be advisable to study other depth of anesthesia monitors in an attempt to find the optimal manner of evaluating (un)consciousness in intellectually disabled patients with documented and confirmed specific etiologies of their intellectual disability.
1. Davidson AJ. Monitoring the anaesthetic depth in children—an update. Curr Opin Anaesthesiol 2007;20:236–43
2. Sadhasivam S, Ganesh A, Robison A, Kaye R, Watcha MF. Validation of the bispectral index monitor for measuring the depth of sedation in children. Anesth Analg 2006;102:383–8
3. Terstegen C, Koot HM, de Boer JB, Tibboel D. Measuring pain in children with cognitive impairment: pain response to surgical procedures. Pain 2003;103:187–98
4. Airaksinen EM, Matilainen R, Mononen T, Mustonen K, Partanen J, Jokela V, Halonen P. A population-based study on epilepsy in mentally retarded children. Epilepsia 2000;41: 1214–20
5. Choudhry DK, Brenn BR. Bispectral index monitoring: a comparison between normal children and children with quadriplegic cerebral palsy. Anesth Analg 2002;95:1582–5
6. Saricaoglu F, Celebi N, Celik M, Aypar U. The evaluation of propofol dosage for anesthesia induction in children with cerebral palsy with bispectral index (BIS) monitoring. Paediatr Anaesth 2005;15:1048–52
7. Sullivan PB. Gastrointestinal disorders in children with neurodevelopmental disabilities. Dev Disabil Res Rev 2008;14:128–36
8. Pilge S, Zanner R, Schneider G, Blum J, Kreuzer M, Kochs EF. Time delay of index calculation: analysis of cerebral state, bispectral, and narcotrend indices. Anesthesiology 2006;104: 488–94
9. Dahaba AA. Different conditions that could result in the bispectral index indicating an incorrect hypnotic state. Anesth Analg 2005;101:765–73
10. Kulak W, Sobaniec W, Smigielska-Kuzia J, Kubas B, Walecki J. A comparison of spastic diplegic and tetraplegic cerebral palsy. Pediatr Neurol 2005;32:311–7
11. Verma A, Radtke R. EEG of partial seizures. J Clin Neurophysiol 2006;23:333–9
12. Amzica F, Steriade M. Electrophysiological correlates of sleep delta waves. Electroencephalogr Clin Neurophysiol 1998;107: 69–83
13. Denman WT, Swanson EL, Rosow D, Ezbicki K, Connors PD, Rosow CE. Pediatric evaluation of the bispectral index (BIS) monitor and correlation of BIS with end-tidal sevoflurane concentration in infants and children. Anesth Analg 2000; 90:872–7
14. Blussé van Oud-Alblas HJ, Peters JW, de Leeuw TG, Tibboel D, Klein J, Weber F. Comparison of bispectral index and composite auditory evoked potential index for monitoring depth of hypnosis in children. Anesthesiology 2008;108:851–7
*Available at: http://www.biseducation.com/ViewAsset.aspx?aaid=38&lang=en-US. Accessed July 1, 2009.