| ||||||||||||||
|
|
|||||||||||||
Department of Anesthesiology, Emory University School of Medicine, Childrens Healthcare of Atlanta, Atlanta, Georgia
Address correspondence and reprint requests to Keith K. Brosius, MD, Childrens Healthcare of Atlanta at Egleston, 1405 Clifton Rd. NE, Atlanta, GA 30322. Address e-mail to keith_brosius{at}emory.org
| Abstract |
|---|
|
|
|---|
IMPLICATIONS: We demonstrated a measurable sedative effect of oral midazolam in adolescents which correlated with simultaneous bispectral index (BIS) measurement. Considering the overall group, midazolam premedication did not affect intraoperative BIS, emergence times, or recovery times compared with placebo controls. Detectable preoperative sedation, and not merely midazolam administration, was predictive of prolonged emergence.
| Introduction |
|---|
|
|
|---|
| Methods |
|---|
|
|
|---|
Surgical procedures included 13 superficial plastic or general surgical procedures (scar revision, cleft lip revision, lymph node excision, foreign body removal), 8 inguinal hernia repairs, 2 central venous access port removals, 4 minor ophthalmologic procedures (examination under anesthesia and ptosis repair), 19 gastroenterology endoscopies, and 4 minor orthopedic procedures.
At the time of preanesthetic evaluation, patients were asked to report the number of hours of sleep obtained during the night before surgery. After arrival in the operating room and institution of BIS (Model A1050TM, Version 3.3; Aspect Medical Systems, Newton, MA) and standard American Society of Anesthesiologists monitoring, preinduction BIS values and OAA/S scale (13) scores were obtained and recorded by a blinded observer. Measurable sedation was defined as a score of 17 or less on the 20-point scale. Anesthesia was then induced with 8% sevoflurane in 60% N2O. IV access was established and endotracheal intubation performed after the administration of a nondepolarizing neuromuscular-blocking drug. Gas concentrations were analyzed by a Capnomac UltimaTM gas analyzer (Datex Medical Instrumentation, Inc., Helsinki, Finland). End-tidal sevoflurane and N2O concentrations were then adjusted to 3% (±0.2%) and 60% (±2%), respectively, at normocapnia (ETCO2, 3545 mm Hg) and maintained constant for 5 min before recording the associated postinduction BIS value (BIS I). Anesthesia was maintained with sevoflurane in 60% N2O, adjusted as necessary to hemodynamic variables but not titrated to BIS. At the completion of surgery, end-tidal sevoflurane and N2O concentrations were adjusted to 0.5% (±0.05%) and 60% (±2%), respectively, and again maintained constant for 5 min before recording the preemergence BIS (BIS E). Although a 5-min period of sevoflurane stabilization is insufficient to allow complete equilibration, it was sufficient to achieve stable BIS readings. BIS measurements were recorded at 5-s intervals for 1 min and averaged to obtain a single value. Blood specimens for the purpose of determining plasma midazolam levels were collected at the time BIS I and BIS E measurements were obtained. Plasma levels were determined with gas chromatography/mass spectrometry by National Medical Services, Inc. (Willow Grove, PA). Results are reported as nanograms per milliliter.
After determination of the BIS E and confirmation of the return of normal neuromuscular function, inhaled anesthetics were discontinued, and 100% oxygen was delivered through the system at a standard flow of 5 L/min. Time to awakening (as evidenced by eye opening, active facial grimace, or purposeful movement in response to a jaw thrust maneuver) was measured to the nearest 30 s. End-tidal sevoflurane concentration, end-tidal N2O concentration, and BIS at awakening were recorded. No study patient received intraoperative narcotics, but narcotics were administered immediately upon emergence for those procedures in which such a need could be reasonably anticipated. Ketorolac and local anesthetic blocks were used as appropriate.
All determinations of awakening time were made by one of two attending anesthesiologists responsible for the study. The anesthesiologist determining awakening time was blinded to grouping and monitoring and had no contact with the patient before surgery or during the induction. Monitoring and data recording were performed by the other investigator. In addition to emergence measurements, time to postanesthesia care unit (PACU) discharge readiness, as assessed on a modified (to include pulse rate and temperature) 16-point Aldrete scale (14), was recorded and compared between groups. All PACU assessments were made by a single trained observer blinded to grouping.
Statistical analysis was performed with MINITAB Statistical Software, Release 12 (Minitab, Inc., State College, PA). Demographic data, preoperative and operative time intervals, and drug concentrations were analyzed with two-tailed Students t-tests. Results are reported as the mean ± SD. Sedation scores, BIS measurements, and recovery times were compared by using the Mann-Whitney U-test. Results are reported as the median and range. The number of patients with a measurable sedative response (defined as an OAA/S score of 17 or less) in each group was compared by using Fishers exact test. The Pearson correlation was used to test for the relationships between drug levels, weights, and ages. The Spearman rank correlation was used to test for relationships involving ordinal data or nonnormally distributed data sets. A P value <0.05 was considered significant.
The relationship between plasma midazolam concentration and OAA/S score was analyzed by nonlinear regression (Fig. 1). A logistic model was applied in which it was postulated that the probability of an OAA/S score of <18 was given by the following equation: probability = C
/(C
+ C50
). Values of
and C50 were estimated by logistic regression with NONMEM statistical software (15). C is the observed plasma concentration and C50 the concentration at which 50% of subjects are expected to demonstrate a positive response.
|
| Results |
|---|
|
|
|---|
Preinduction OAA/S scores and BIS values were obtained in all patients immediately before the induction of general anesthesia. The time from the administration of midazolam to OAA/S and BIS assessment was 29.0 ± 4.8 min (range, 2340 min.). The median preinduction BIS was significantly lower (P < 0.001) in Group M (92; range, 6798) than in Group P (97; range, 8998). Ten of 25 Group M patients exhibited clinically detectable sedation, whereas only 1 of 25 placebo patients was measurably sedated (P < 0.001). BIS in the 10 midazolam patients demonstrating sedative effect was further reduced, with a median value of 82. The median preinduction OAA/S score was also significantly lower (P < 0.002) in Group M (18; range, 1220) than in Group P (20; range, 1720). There was a significant correlation between preinduction BIS and OAA/S score (r = 0.542, P < 0.006) (Fig. 2). One patient in each group reported <4 h sleep the night before surgery. The single placebo patient with detectable sedative effect (OAA/S of 17) reported 3.5 h sleep. The single midazolam patient reporting <4 h sleep had an OAA/S score of 13 and a postinduction plasma midazolam level of 49 ng/mL. The average reported hours of sleep were 6.8 ± 1.8 h and 6.5 ± 1.6 h in the Group P and Group M, respectively; these were not significantly different.
|
|
(the variable determining the steepness of the concentration-effect curve) was determined to be 3.57 (SE ± 1.24). The resultant probability curve and the distribution of the sedative response versus the corresponding plasma midazolam concentration are displayed in Figures 1 and 3.
|
In Group M, preoperative OAA/S score correlated significantly with emergence time (r = -0.566, P < 0.004) (Fig. 2), but not with time to PACU discharge readiness. Considering all patients, emergence time was 2.3 ± 1.0 min in patients with OAA/S scores of 1820 and 4.0 ± 1.1 min in patients with scores <18 (P < 0.001). Within Group M, emergence time was 2.2 ± 1.1 min and 4.0 ± 1.1 min for these same sedation ranges (P < 0.001). Neither the midazolam level at BIS E nor the BIS E value correlated with emergence time or PACU discharge time. The median time to PACU discharge readiness was 10 min (range, 582 min) in Group P and 11 min (range, 754 min) in Group M; these times were not significantly different.
Parenteral narcotics were administered after surgery as indicated on the order of the attending anesthesiologist. Fourteen midazolam- and 15 placebo-premedicated patients received narcotics in the postoperative period.
| Discussion |
|---|
|
|
|---|
Our initial midazolam levels were obtained at an average of 45 minutes after the administration of the drug. The mean plasma level at this acquisition (49 ng/mL) corresponds to that reported by Marshall et al. (17) (42 ng/mL) in pediatric patients receiving oral midazolam and achieving a similar sedative effect. This level, however, is only approximately half that reported by Malinovsky et al. (18) in pediatric patients receiving 0.5 mg/kg of oral midazolam, and it is slightly less than 58 ng/mL 45 minutes after a 0.45 mg/kg oral dose as reported by Payne et al. (19). This may in part be attributable to the smaller average per kilogram dose in our patients (0.33 mg/kg), differences in drug formulation, age-related pharmacokinetic differences, differences in sample acquisition time, or a combination of these. Because of the study design, midazolam levels were obtained simultaneously with BIS measurements in an effort to determine whether or not plasma midazolam concentration had a detectable effect on intraoperative BIS, and it did not. As a result, the plasma midazolam concentrations and OAA/S scores used to construct the concentration-effect curve (Fig. 1) were not obtained simultaneously. The sedation assessment preceded the corresponding midazolam level by approximately 15 minutes, and the curve should be interpreted accordingly. In our patients, three general categories of sedative effect and corresponding plasma level are evident, consistent with the reported adult threshold level for sedation of 40 ng/mL (20) and suggesting a similar pharmacodynamic response to that of adults. At <40 ng/mL a sedative effect did not occur. Between 40 and 80 ng/mL, the sedative response shows significant interindividual variability, with nearly equal numbers of patients in each response group. At >80 ng/mL, detectable sedation is the expected result (Fig. 3). These results suggest significant pharmacokinetic and pharmacodynamic variability in this population; this variability is also characteristic of adults receiving midazolam.
There was a weak but significant negative correlation between patient age and midazolam level which was independent of body weight. This suggests a maturational change in midazolam pharmacokinetics during the course of adolescence and suggests that their pharmacokinetic handling of midazolam is transitional between that of childhood and adulthood.
Preoperative midazolam had no influence on either of the two intraoperative BIS measurements (Table 1). It seems that at this midazolam dose, the hypnotic effect of sevoflurane/N2O overwhelms any potential contribution of midazolam to BIS. No additive or synergistic effect was detectable, even among those individuals who exhibited detectable sedation before induction. This is in contrast to the effect reported by Denman et al. (21), who demonstrated that methohexital given as premedication to pediatric surgical patients does significantly decrease BIS during sevoflurane/N2O anesthesia.
If one considers each study group as a whole, midazolam premedication does not prolong emergence time or time to PACU discharge readiness. However, the preoperative OAA/S score was significantly correlated with emergence time (Fig. 2). The emergence time in patients manifesting a detectable sedative effect after premedication was >70% longer than in those without detectable sedation. The lack of a similar effect on time to PACU discharge suggests that this residual effect is insufficient in magnitude, duration, or both to influence this measure of recovery. Neither midazolam administration nor the preinduction OAA/S score correlated with emergence concentrations of anesthetic gases.
It should be noted that our reported awakening concentration of sevoflurane is quite different than the reported MAC-awake of sevoflurane in children (22), and it was not our intention to attempt to define MAC-awake in this population. Awakening concentrations reported in this study (Table 1) were measured in combination with residual N2O, and our washout technique does not conform to either the fast or slow alveolar washout techniques often used in the determination of MAC-awake values for volatile anesthetics (23). Our range of BIS values recorded at the time of awakening (5998) is comparable to that recorded at the time of eye opening in adult patients receiving propofol/alfentanil/N2O anesthesia (24), and it is comparable to those recorded in a series of pediatric patients recovering from general anesthesia (25). We speculate that the decreased BIS at awakening in midazolam-treated patients (Table 1) may be a manifestation of residual midazolam sedative effect similar to that evident in sedate, conscious patients just before the induction (preinduction BIS). Although the hypnotic effect of sevoflurane is sufficient to overwhelm the effect of midazolam on BIS during surgery, a small but detectable effect of midazolam may again become evident when the influence of sevoflurane dissipates and consciousness is restored.
In conclusion, we demonstrated that, in a sample of healthy adolescents receiving 20 mg of oral midazolam, <50% manifested clinically detectable sedation. Preinduction BIS was significantly lower in midazolam-treated patients, was lower yet in patients who manifested a detectable sedative response, and was significantly correlated with simultaneous OAA/S score. However, intraoperative BIS measured at two fixed levels of sevoflurane administration was not affected by midazolam premedication. In aggregate, midazolam premedication did not prolong emergence from general anesthesia, did not affect the expired concentration of sevoflurane/N2O at emergence, and did not prolong time to PACU discharge readiness. Emergence time was significantly prolonged in patients who manifested clinically detectable sedation, and it is the presence of a detectable sedative effect that should be considered when anticipating possible effects of midazolam on awakening. This residual effect was insufficient to influence discharge time from the PACU.
| Acknowledgments |
|---|
The authors wish to thank James M. Bailey, MD, PhD, for his help with statistical analysis.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
R. G. Cox, U. Nemish, A. Ewen, and M.-J. Crowe Evidence-based clinical update: Does premedication with oral midazolam lead to improved behavioural outcomes in children?: [Mise a jour basee sur des donnees probantes : Ameliore-t-on le comportement des enfants par une premedication au midazolam par la bouche ?] Can J Anesth, December 1, 2006; 53(12): 1213 - 1219. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Eikermann, M. Blobner, H. Groeben, C. Rex, T. Grote, M. Neuhauser, M. Beiderlinden, and J. Peters Postoperative upper airway obstruction after recovery of the train of four ratio of the adductor pollicis muscle from neuromuscular blockade. Anesth. Analg., March 1, 2006; 102(3): 937 - 942. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. A. van den Berg and A. Abeidi Routine Inhaled Induction in Adults: A Safe Practice? Anesth. Analg., February 1, 2006; 102(2): 647 - 647. [Full Text] [PDF] |
||||
![]() |
A. J. Davidson, G. H. Huang, C. S. Rebmann, and C. Ellery Performance of entropy and Bispectral Index as measures of anaesthesia effect in children of different ages Br. J. Anaesth., November 1, 2005; 95(5): 674 - 679. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. B. McDermott, T. VanSickle, D. Motas, and R. H. Friesen Validation of the Bispectral Index Monitor During Conscious and Deep Sedation in Children Anesth. Analg., July 1, 2003; 97(1): 39 - 43. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. K. Brosius and C. F. Bannister Midazolam Premedication in Children: A Comparison of Two Oral Dosage Formulations on Sedation Score and Plasma Midazolam Levels Anesth. Analg., February 1, 2003; 96(2): 392 - 395. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. L. De Witte, C. Alegret, D. I. Sessler, and G. Cammu Preoperative Alprazolam Reduces Anxiety in Ambulatory Surgery Patients: A Comparison with Oral Midazolam Anesth. Analg., December 1, 2002; 95(6): 1601 - 1606. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|