CLINICAL INVESTIGATION

Sedation for Electroencephalography With Dexmedetomidine or Chloral Hydrate: A Comparative Study on the Qualitative and Quantitative Electroencephalogram Pattern Magda L. Fernandes, MD, MSc,* Welser Machado de Oliveira, MD,w Maria do Carmo Vasconcellos Santos, MD,w and Renato S. Gomez, MD, PhDz

Background: Sedation for electroencephalography in uncooperative patients is a controversial issue because majority of sedatives, hypnotics, and general anesthetics interfere with the brain’s electrical activity. Chloral hydrate (CH) is typically used for this sedation, and dexmedetomidine (DEX) was recently tested because preliminary data suggest that this drug does not affect the electroencephalogram (EEG). The aim of the present study was to compare the EEG pattern during DEX or CH sedation to test the hypothesis that both drugs exert similar effects on the EEG. Materials and Methods: A total of 17 patients underwent 2 EEGs on 2 separate occasions, one with DEX and the other with CH. The EEG qualitative variables included the phases of sleep and the background activity. The EEG quantitative analysis was performed during the first 2 minutes of the second stage of sleep. The EEG quantitative variables included density, duration, and amplitude of the sleep spindles and absolute spectral power. Results: The results showed that the qualitative analysis, density, duration, and amplitude of sleep spindles did not differ between DEX and CH sedation. The power of the slow-frequency bands (d and y) was higher with DEX, but the power of the fasterfrequency bands (a and b) was higher with CH. The total power was lower with DEX than with CH. Conclusions: The differences of DEX and CH in EEG power did not change the EEG qualitative interpretation, which was sim-

Received for publication October 25, 2013; accepted March 28, 2014. From the *Servic¸o de Anestesiologia da Santa Casa de Miserico´rdia de Belo Horizonte; wServic¸o de Neurologia da Santa Casa de Miserico´rdia de Belo Horizonte; and zDepartamento de Cirurgia da Faculdade de Medicina da Universidade Federal de Minas Gerais (UFMG), Belo Horizonte, Minas Gerais, Brazil. The English edition of the manuscript was funded by Pro´-Reitoria de Pequisa da UFMG. R.S.G. is a research fellow of CNPq. The authors have no conflicts of interest to disclose. Reprints: Renato S. Gomez, MD, PhD, Departamento de Cirurgia da Faculdade de Medicina da Universidade Federal de Minas Gerais (UFMG), Av. Alfredo Balena, 190, Sala 203, Bairro Santa Efigeˆnia, Belo Horizonte, Minas Gerais CEP 31340-300, Brazil (e-mail: [email protected]). Copyright r 2014 by Lippincott Williams & Wilkins

J Neurosurg Anesthesiol



Volume 27, Number 1, January 2015

ilar with the 2 drugs. Other studies comparing natural sleep and sleep induced by these drugs are needed to clarify the clinical relevance of the observed EEG quantitative differences. Key Words: sedation, electroencephalography, chloral hydrate, dexmedetomidine (J Neurosurg Anesthesiol 2015;27:21–25)

T

he electroencephalogram (EEG) is a test used for the diagnosis or treatment control of epilepsy and other pathologic conditions of the central nervous system. To carry out the test, sedation is necessary in individuals who are unable to communicate and cooperate with the instructions, such as children or adults with behavioral changes. The sedative drugs act on the central nervous system and could affect the EEG, leading to a great concern about the interference of these agents during the EEG recording.1 Traditionally, chloral hydrate (CH) has been utilized for sedation during EEGs,2 despite its potential anticonvulsant properties3 and its limited effectiveness as a sedative in patients with neurological disorders.4 This drug has been used for over 60 years5 and remains useful in the present days,6–8 because other drugs have not been proven to be adequate. Barbiturates might cause EEG activation, producing waves of 15 to 30 Hz.9 With the progressively increased doses of this hypnotic, slow waves of high amplitude lead to burst suppression.9 Benzodiazepines exhibit potent anticonvulsant properties and are the first choice for emergency therapy in generalized seizure crises.9 Propofol has proconvulsant and anticonvulsant effects, depending on the dose used. At low doses, it increases the number of b waves and decreases the y and a waves.10 At higher doses, the number of b waves decreases and d waves appear.10 Inhalational anesthetics such as sevoflurane might induce epileptiform electroencephalographic activity.11 Hydroxyzine has efficacy similar to CH,7,8 but its effects on the EEG have not been completely described.7,8 Dexmedetomidine (DEX), a highly selective agonist of the a2-adrenergic receptor, has been highlighted in this www.jnsa.com |

21

Fernandes et al

scenario because it does not appear to interfere with the EEG in adults or children.12,13 Previous studies have suggested that this agent induces EEG patterns similar to those of natural sleep, with characteristics suggestive of sleep phase II (N2 phase) of non–rapid eye movement, and it has a moderate increase in the power of y, a, and b bands.12–14 There are no studies comparing the effects of DEX and CH on EEG. The present study aimed to compare the pattern of the EEG in uncooperative patients with indications to conduct the examination under sedation with DEX or CH at different periods to test the hypothesis that the 2 drugs induce similar EEG patterns.

MATERIALS AND METHODS After receiving approval from the Ethics Committee of our institution, a prospective, double-blind clinical study was conducted from March 2011 to March 2013. The people responsible for the individuals were informed about the study and agreed to it by signing the informed consent form. The inclusion criteria were uncooperative patients from 12 to 40 years old for whom an EEG was indicated for the clinical management and treatment of a neurological disorder. The exclusion criteria included active respiratory disease, heart disease, kidney or liver disease, and a history of allergy to the drugs used. The subjects underwent 2 EEGs on different dates. One EEG was performed under sedation with DEX, and the other EEG was performed under sedation with CH. The patients were randomized by sealed envelopes to define the drug to be used in the first EEG; the other drug was used for the second EEG. There was no comparison of a baseline EEG with DEX or CH because all the patients underwent examination while sedated with the drugs. All subjects had fasted for 8 hours on the date of the examination, and the patients were on their usual medications, including anticonvulsants, which were not changed in the interval between the 2 examinations. The sedation was administered by the same anesthesiologist and the level of sedation was evaluated and recorded according to the following criteria as previously described: (1) awake; (2) slightly drowsy, easily alert; (3) very sleepy, tending to sleep; and (4) sleeping with minimal response to physical stimulation.15 Oxygen by a nasal cannula was administered to the subjects who had a drop in their SpO2 below 90%. For sedation with CH, the subjects received this drug orally at a dose of 50 mg/kg (CH 10%). In the absence of sedation level >2, a second dose of 25 mg/kg was administered to the patient. For sedation with DEX, venous access was obtained to begin a continuous infusion of the drug. The initial dose (bolus) was 1 mg/kg (Precedex; Laboratory Hospira, Sa˜o Paulo, Brazil) infused over 10 minutes, and the dose was maintained from 0.2 to 0.7 mg/kg/h. Initial dose adjustments or maintenance were performed to maintain a level of sedation ranging from 3 to 4 on the sedation scale. The tests were conducted in a dark and quiet room. The electrodes were placed on the scalp according to the international 10-20 system.16 The EEG recording was

22 | www.jnsa.com

J Neurosurg Anesthesiol



Volume 27, Number 1, January 2015

performed for 20 minutes in all of the patients, and no endpoint was used to finish the EEG. If there were artifacts, the EEG was paused and restarted as soon as possible to complete the examination (20 min). The heart rate (HR), mean arterial pressure, and level of sedation were recorded every 5 minutes, and the oxygen saturation (SpO2) was monitored continuously from the time of administration of the drugs to the spontaneous awakening of the patient. In the postanesthetic recovery unit, the HR and mean arterial pressure were recorded every 15 minutes, and the SpO2 was monitored continuously until the patient was discharged (a score of 10 on the modified Aldrete-Kroulik scale). We recorded the procedure time (time elapsed between the start of drug administration and the end of EEG) and the wake-up time (time between the end of EEG and spontaneous awakening of the patient). Patients in whom an intervention (administration of fluids or the use of an anticholinergic agent) was required for hypotension or bradycardia were identified.

EEG Analysis The EEG was recorded and interpreted in a Meditron MedEEG Model 420 with a software specific for brain mapping. The same neurologist expert in EEG performed the analysis of the tests, and the neurologist did not know the type of sedative drug used to ensure the blinding of the study. Two types of variables were analyzed: The qualitative variables were sleep phases (identification of the deepest phase of sleep achieved) and background activity (normal, fast activity increased, or slow activity increased). For quantitative variables, we selected the recording of the first 2-minute period of the N2 phase. Through visual recognition, the sleep spindles were located, and the density, duration, and amplitude of the sleep spindles were evaluated. The density was calculated by visually counting the number of zones present in the 2 minutes of sleep analyzed, and the result was presented as the number of spindles per minute. The duration of the spindles were obtained directly by visual counting and were expressed in seconds. The amplitude of the zones were determined in the time domain by visual scoring of the EEG and were expressed in microvolts. The absolute power (mV2/Hz) was calculated through spectral analysis based on Fourier transformation.17,18 The sample size was calculated based on a previous study that showed a difference of 0.30 in the power of the b band between DEX and the natural sleep with a SD of 0.28.13 Considering a significance level of 5% and the test power of 80%, with a 2-tailed hypothesis, we found n = 14. Reinforcing this number, a previous study evaluated the spike-wave activity before and after the administration of CH in 13 patients.3 Owing to the possibility of failure of the sedation with CH or venipuncture with DEX, we initially selected 25 patients. The patients who were referred received the EEG in our service and who met the inclusion criteria were invited to r

2014 Lippincott Williams & Wilkins

J Neurosurg Anesthesiol



Volume 27, Number 1, January 2015

participate in the study, and the informed consent form was signed. For the statistical analysis, the normality of the data was assessed by the Shapiro-Wilk test. A paired comparison was used because the data of the first examination were compared with the data of the second examination on the same patient. The tests used were: the Student paired t test (quantitative variables with normal distribution), the Wilcoxon paired test (quantitative variables without normal distribution, levels of sedation, procedure time, and wake-up time), and the McNemar test (qualitative variables). The deepest stage of sleep was compared by the w2 test. For the power analysis, there was a replacement of the values of absolute power (X) by its logarithm Y = log (X) to obtain a normal distribution.17 A value of P < 0.05 was considered to be minimally statistically significant. The statistical analysis was performed with the Statistical Package for Social Science (SPSS), version 18.0 (SPSS Inc., Chicago, IL).

RESULTS Of the 25 patients, 8 (32%) who were originally eligible for the study failed sedation with CH and were excluded from the study; however, all the subjects were sedated with DEX. A qualitative EEG analysis was performed on the 17 patients who had satisfactory sedation with the 2 treatments evaluated (DEX and CH). Three of these patients were excluded from the quantitative variable analysis because they had no period of the N2 phase with CH. Of the 17 patients evaluated, 10 were male (58.8%) and 7 were female (41.2%). The patients were aged 22.3 ± 7.6 years (mean ± SD) and had a weight of 51.3 ± 18.6 kg (mean ± SD). The neurological diseases reported in this group were: cerebral palsy (41.2%), sequela of meningitis (17.6%), genetic syndromes (17.6%), and mental retardation of unspecified cause (23.6%). All the subjects took a combination of 1 to 4 anticonvulsant drugs regularly: phenobarbital (50%), carbamazepine (37.5%), phenytoin (25%), and valproic acid (18.8%). Other psychotropic drugs used were: neuroleptics (29%), benzodiazepines (23.5%), and tricyclic antidepressants (17.6%). Four subjects (23.5%) were sedated with 50 mg/kg of CH, and 13 (76.5%) patients required an additional dose of 25 mg/kg, for a total dose of 75 mg/kg. The patients who received 2 doses of CH started the EEG recording later (median, 50 min; range, 33 to 120 min; P < 0.001, 1-way analysis of variance followed by the Tukey multiple comparison test) than the patients in the DEX group (median, 24 min; range, 15 to 50 min) and the CH group who were sedated with 1 dose (median, 33 min; range, 10 to 45 min). The initial dose of DEX (1 mg/kg) was adjusted in 9 patients (52.9%) to a value of 0.78 ± 0.35 mg/kg (mean ± SD) during the 10-minute infusion. The maintenance dose was 0.56 ± 0.89 mg/kg/h (mean ± SD). In the qualitative variable analyses, phase 3 of the non–rapid eye movement sleep (N3 phase) was the deepest phase observed on the EEG (Table 1), but there r

2014 Lippincott Williams & Wilkins

Sedation for EEG With Dexmedetomidine or Chloral Hydrate

TABLE 1. Comparison of NREM Sleep Phases and EEG Background Activity During Sedation With DEX or CH n (%) Group NREM sleep phases N1 N2 N3 Background activity Normal Fast activity increased Slow activity increased

DEX

CH

P

1 (5.8) 6 (35.3) 10 (58.9)

1 (5.8) 12 (70.6) 4 (23.6)

0.134

10 (58.9) 5 (29.4) 2 (11.7)

9 (53) 5 (29.4) 3 (17.6)

0.607

P-value (the McNemar test). % indicates percentage of patients; CH, chloral hydrate; DEX, dexmedetomidine; EEG, electroencephalogram; N1, phase 1 of NREM; N2, phase 2 of NREM; N3, phase 3 of NREM; NREM, non–rapid eye movement.

was no difference between the 2 drugs regarding the comparisons of all of the sleep phases observed (Table 1). A higher number of patients in the DEX group reach the N3 phase (10 and 4 patients for DEX and CH, respectively) compared with those who did not reach this phase (7 and 13 patients for DEX and CH, respectively) (P = 0.0365 by w2 test). There was no statistically significant difference between the drugs in relation to the background activity (Table 1). In the analysis of quantitative variables, there was no difference between the amplitude, duration, and density of the sleep spindles with the 2 drugs evaluated (Table 2). In the comparison of absolute power, there was a statistically significant difference between the drugs in all of the frequency bands, with DEX promoting higher d and y power, and lower a, b, and total power than CH (Table 3). Four and 3 patients had hypotension alone (23.5%) and bradycardia alone (17.6%), respectively, with DEX. HR values

Sedation for electroencephalography with dexmedetomidine or chloral hydrate: a comparative study on the qualitative and quantitative electroencephalogram pattern.

Sedation for electroencephalography in uncooperative patients is a controversial issue because majority of sedatives, hypnotics, and general anestheti...
132KB Sizes 2 Downloads 3 Views