Pediatric Anesthesiology: Original Clinical Research Report

Comparison of Intranasal Dexmedetomidine and Oral Pentobarbital Sedation for Transthoracic Echocardiography in Infants and Toddlers: A Prospective, Randomized, Double-Blind Trial

Miller, Jeffrey W. MD^*; Ding, Lili PhD^†; Gunter, Joel B. MD^*; Lam, Jennifer E. DO^*; Lin, Erica P. MD^*; Paquin, Joanna R. MD^*; Li, Bi Lian MD^‡; Spaeth, James P. MD^*; Kreeger, Renee N. MD^*; Divanovic, Allison MD^§; Mahmoud, Mohamed MD^*; Loepke, Andreas W. MD, PhD^‖

Author Information

From the ^*Department of Anesthesiology

^†Division of Biostatistics and Epidemiology, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio

^‡Guangzhou Women’s and Children’s Medical Center of Guangzhou Medical University, Guangzhou, China

^§Division of Cardiology, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio

^‖Division of Cardiac Anesthesiology, Department of Anesthesiology and Critical Care Medicine, The Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania.

Published ahead of print January 23, 2018.

Accepted for publication December 1, 2017.

Funding: This research was funded by a grant from the Children’s Heart Association of Cincinnati and departmental support.

The authors declare no conflicts of interest.

Clinical trial registration: NCT02250820.

LMA is a registered trademark of Teleflex Incorporated or its affiliates.

Reprints will not be available from the authors.

Address correspondence to Jeffrey W. Miller, MD, Department of Anesthesiology, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Ave, MLC 2001, Cincinnati, OH. Address e-mail to [email protected].

Anesthesia & Analgesia: June 2018 - Volume 126 - Issue 6 - p 2009-2016

doi: 10.1213/ANE.0000000000002791

Free

Abstract

KEY POINTS

Question: This prospective, randomized trial compared sedation with intranasal dexmedetomidine to oral pentobarbital in infants and toddlers with congenital heart disease undergoing transthoracic echocardiography.
Findings: Intranasal dexmedetomidine was clinically comparable to oral pentobarbital in regard to single-dose sedation success and onset time, without increasing clinically significant adverse events, while serving as an effective “rescue” sedative for failed oral pentobarbital or single-dose dexmedetomidine sedation.
Meaning: Our findings in children with congenital heart disease, a neurologically vulnerable patient population requiring repeated sedative exposures early in life, suggest that dexmedetomidine is a suitable alternative to sedatives covered by a Food and Drug Administration warning regarding their use in early childhood, such as pentobarbital, during nonpainful imaging studies.

Transthoracic echocardiography (TTEcho) is the primary noninvasive modality for initial diagnosis and periodic monitoring of congenital heart disease (CHD). Young children often require sedation during TTEcho to reduce motion artifacts, but the optimal sedative for this purpose has not been determined.^¹ Oral pentobarbital (PENT) has long been the primary sedative for infant TTEcho in many pediatric institutions, including ours.^² However, PENT is difficult to obtain, has an unpleasant taste, has a prolonged half-life, requires cooperation during oral administration, and can be associated with agitation during awakening.^³ Moreover, children over 1 year of age more frequently fail single-dose oral PENT sedation than younger infants, and may therefore require a second (“rescue”) dose.^² Airway and swallowing reflexes can be diminished after PENT administration, especially after larger doses.^³

Dexmedetomidine (DEX) is widely utilized for pediatric sedation for nonpainful procedures.^⁴ It has a short half-life, is associated with calm emergence, and maintains airway tone and airway reflexes.^⁵ Intranasal DEX is effective and painless^⁶ but can be associated with alterations in heart rate and blood pressure,^⁵ which may be problematic in children with CHD. Importantly, prolonged or repeated exposure to all commonly utilized sedatives and anesthetics, including PENT, in children <3 years of age has been the subject of a US Food and Drug Administration (FDA) warning. Because repeated echocardiographical examinations are oftentimes required for diagnosis and surveillance of CHD, in addition to surgical procedures, multiple exposures frequently occur in this patient population.^⁷ The only pharmaceuticals used in this setting that are not included in this warning, as they have not produced the same level of neurodegenerative changes in animals, are α₂-agonists, such as DEX, and opioids.^⁸

Intranasal DEX sedation has been studied for TTEcho in retrospective or observational studies^{^9–11}; however, it has not often been compared in a randomized controlled setting. Accordingly, this prospective trial compared sedation with intranasal DEX to oral PENT in infants and toddlers with CHD undergoing TTEcho. We hypothesized that sedation with intranasal DEX would be comparable to oral PENT in effectiveness and important adverse events. The primary outcome was achievement of adequate levels of sedation for TTEcho without need for rescue sedation within 30 minutes of administration. Secondary outcomes included the number of sonographer pauses, image quality in relation to sedation, and parental satisfaction. Adverse effects on heart rate, systolic blood pressure (SBP), oxygen saturation, and incidence of agitation before or after sedation were also compared between groups.

METHODS

This single-center, prospective, randomized, double-blinded trial was registered before patient enrollment at clinicaltrials.gov (NCT02250820) on September 26, 2014, with principal investigator J.W.M. DEX use is off-label in pediatric patients; this study was conducted under an Investigational New Drug (# 76–346) application. After US FDA review, the institutional review board of Cincinnati Children’s Hospital Medical Center approved this trial (# 2014–5961). Written informed consent was obtained for each subject from a legal surrogate, a parent, or legal guardian. This manuscript adheres to Consolidated Standards of Reporting Trials (CONSORT) guidelines.^¹²

Inclusion criteria included outpatient subjects 3–24 months of age with American Society of Anesthesiologists’ physical status III or less requiring a TTEcho at Cincinnati Children’s Hospital from November 19, 2014, to March 28, 2017, who were deemed too restless to obtain acceptable images without sedation. Written informed consent was obtained from parents or legal guardian.

Exclusion criteria were cardiac conduction disease, channelopathy, home ventilator use, treatment with digoxin, α- or β-adrenergic agonists or antagonists, antiarrhythmic medications, any other sedative use within 48 hours, allergies or contraindications to PENT or DEX, rhinorrhea, previous enrollment in this trial, trisomy 21, use of angiotensin-converting enzyme inhibitors on the day of the procedure, or unrepaired aortic coarctation. Former premature (<60 weeks estimated gestational age) children were not sedated for elective echocardiography, as anesthesia departmental policy required admission to the hospital. Parents requiring language interpreter services and subjects with governmental guardianship were also excluded. No other congenital cardiac diagnoses were utilized for exclusion.

Subjects were randomized 1:1 to oral PENT or intranasal DEX using a computer-generated, sealed enrollment log. Because younger age groups differ in sedative, respiratory, hemodynamic, and emergence responses to sedatives,^¹³ randomization was allocated by blocks (3–12 or 13–24 months, inclusive).

The PENT group received oral PENT 5 mg·kg⁻¹ (Oak Pharmaceuticals, Lake Forest, IL) with 2-mL cherry syrup followed by an intranasal placebo (0.5-mL atomized saline as described for DEX). The DEX group received an oral placebo with 2-mL cherry syrup followed by an intranasal DEX 2.5 µg·kg⁻¹ (100 µg/mL; Mylan, Rockford, IL) by nasal atomizer (LMA^® MAD Nasal™; Wolfe-Tory Medical, Salt Lake City, UT). Intranasal DEX was administered into 1 naris as a rapid atomized spray from a 1-mL syringe containing 0.2 mL of air as a “chaser” to ensure complete expulsion of medication. The medications and placebos were indistinguishable in appearance and volume. PENT at the studied oral dose 5 mg·kg⁻¹ has been the standard sedative for TTEcho at Cincinnati Children’s Hospital for over 20 years.^² For intranasal DEX dosing, we performed a retrospective study of our preliminary results using an initial dose of 1–3 µg·kg⁻¹ for TTEcho sedation.^⁹

During sedation, subjects were monitored by continuous pulse oximetry, automated blood pressure measurement, electrocardiogram, and continuous visual monitoring. Physical restraints were not used. If subjects were not sedated to a Ramsay score >3 after 30 minutes, “rescue sedation” with intranasal DEX (1 µg·kg⁻¹) was administered and children were considered single-dose sedation failures.

A nonblinded cardiac anesthesiologist and sedation nurse verified study group allocation and obtained medication. A sedation nurse unaware of group assignment administered sedative and placebo, monitored the subject during the TTEcho, and managed adverse effects. A research coordinator unaware of group assignment and trained in assessment of the modified Ramsay sedation scale^¹⁴ and scoring system for delirium^¹⁵ recorded data, assessed the ease of drug administration, and quantified level of sedation. Parents remained blinded to randomization until the next business day after a follow-up phone survey. The sonographer and cardiologists were similarly unaware of group assignment.

Primary outcome was adequate sedation for TTEcho within 30 minutes of drug administration, defined as eyes closed, no response to voice, but mild response to deep subxiphoid or suprasternal TTEcho probe placement, comparable to a Ramsay sedation score >3, which has been shown to be required to facilitate detailed TTEcho examination in infants.^² Times from first sedative dose to Ramsay sedation level >3 and to discharge, which required Ramsay score <3, unobstructed airway, oxygen saturation at baseline, and blood pressure and heart rate within normal ranges for age, were documented.^¹⁶

Secondary outcomes included number of sonographer pauses >2 minutes due to subject movement, image quality in relation to sedation as rated by the cardiologist and sonographer (3-point scale in long-term use in this facility—cooperative/excellent, restless/moderate, noncooperative/inadequate), and immediate and next business day parental satisfaction (10-point scale). Parents were also queried regarding prolonged drowsiness, agitation, lethargy, loss of appetite, emergency room visitation, or phone calls to doctors related to adverse events. Subject acceptance of oral and intranasal medication was documented.

Documented adverse events included agitation before achieving sedation, desaturation (peripheral capillary oxygen saturation [Spo₂] <92% or >5% decrease for chronically cyanotic subjects) or respiratory interventions, age-defined bradycardia or hypotension defined by pediatric advanced life support criteria,^¹⁶ or agitation after awakening.

Statistical Analysis

Pilot studies predicted 15% single-dose sedation failures for oral PENT and intranasal DEX.^²^,^¹⁷ Accordingly, with α =0.05 and power (1 – β) = 0.8, 140 subjects per group (280 subjects) were required to detect a clinically significant 10% difference in single-dose failure rate. Data were tested for normality (Shapiro–Wilk), and continuous variables are presented as mean ± standard deviation or median with interquartile range (IQR), as appropriate. Categorical data are presented as frequencies with percentages and were compared using χ² or Fisher exact tests, as appropriate; associations between 1 binary variable and 1 continuous variable were tested using 2-sample t test or Wilcoxon rank sum test, as appropriate. Repeated measures of heart rate, delirium score, Ramsay score, and blood pressure were analyzed using mixed-effect models with a random subject effect to account for dependence between observations from the same subject. P < .05 (2-tailed) was considered statistically significant. Exploratory analysis of factors associated with need for rescue sedation was performed. All statistical analyses were performed with SAS version 9.3 (SAS Institute, Inc, Cary, NC) or GraphPad Prism version 7.00 for Mac (GraphPad Software, La Jolla, CA).

RESULTS

A total of 280 subjects were enrolled in the trial, and their data were analyzed (CONSORT diagram; Figure 1 and Table 1). Cardiac diagnoses included 17% normal anatomy (murmur, family history of CHD), 47% one cardiac lesion (atrial septal defect, ventricular septal defect, pulmonary artery hypertension after congenital diaphragmatic hernia, pulmonary stenosis, hypoplastic aortic arch), and 36% complex cardiac lesions (transposition of the great arteries, tetralogy of Fallot, single ventricle with cavopulmonary palliation, atrioventricular canal defect). With dedicated (and blinded) research coordinators documenting the study data in real time, we had complete data capture during the sedation. For demographic data, only 1 variable had missing values (prematurity), and they were excluded from statistical analysis.

Figure 1.:
Participant flowchart. EGA indicates estimated gestational age; TTEcho, transthoracic echocardiography.

There was no difference in success rate with single-dose sedation for TTEcho (85% group PENT, 84% group DEX; P = .8697) (Table 2). Median time to Ramsay sedation level >3 allowing initiation of TTEcho was marginally shorter in group PENT versus group DEX (16.5 [IQR, 13–21] vs 18 [IQR, 16–23] minutes; P = .0095); this difference was predominantly found in the 13- to 24-month age group (Table 2). Median TTEcho scan time, median time to discharge, and total sedation to discharge time were not different between groups. Mean heart rate and SBP during sedation with single-dose intranasal DEX or oral PENT are shown in Figure 2. Factors associated with the need for rescue sedation included expelling of oral PENT in group PENT (relative risk, 2.5; 95% confidence interval [CI], 1.1–5.7; P = .0424) and agitation during onset of sedation in both groups (overall relative risk, 5.6; 95% CI, 3.0–10.3; P < .0001). Group assignment, age, weight, prematurity, nil per os time (time from last reported oral intake), and agitation on arrival to the sedation suite were not significantly associated with need for rescue sedation. All DEX primary sedation failures and 20 of 21 PENT failures (95%) achieved adequate sedation with rescue DEX; 1 PENT patient was given additional intranasal midazolam.

Table 2.:
Primary Outcome and Sedation Times After Single-Dose Sedation

Figure 2.:
Mean heart rate during sedation (Ramsay >3) was less in group DEX by a mean of 10.3/min (P < .0001; 95% CI, 7.6–13.1), which persisted into the arousal phase (P < .0001). Mean systolic blood pressure was significantly lower (mean difference, 5.1 mm Hg; P = .0009; 95% CI, 2.1–8.1) only during the arousal phase. Graphs depict mean heart rate (A) and systolic blood pressure (B) during sedation for echocardiography with single-dose intranasal DEX (2.5 µg·kg⁻¹) or oral PENT (5 mg·kg⁻¹); error bars signify standard deviation. CI indicates confidence interval; DEX, dexmedetomidine; PENT, pentobarbital.

Agitation during onset of sedation was observed in 28% of PENT and 36% of DEX subjects, without group differences (P = .1348) (Table 3). No associations between subject characteristics and agitation were observed. No adverse hemodynamic or respiratory events requiring intervention except blow-by oxygen (PENT n = 1, DEX n = 1) were reported.^¹⁸ Bradycardia <80 beats·minute⁻¹ occurred more commonly in group DEX (10/117; 8.5%) than in group PENT (1/118; 0.8%) (P = .0052; relative risk, 10.1; 95% CI, 1.3–77.5); however, no sedated heart rate below 60 beats·minute⁻¹ was observed. Hypotension (lowest SBP <70 mm Hg) during sedation occurred more commonly in group DEX (22/117 subjects; 19%) than in group PENT (11/118; 9%) (P = .0365; relative risk, 2.02; 95% CI, 1.043–3.942); no SBP <60 mm Hg was recorded. Rescue sedation did not increase the risk of bradycardia or hypotension in either group.

Table 3.:
Secondary Outcome Variables During Single-Dose Sedation

Table 4.:
Sonographer, Cardiologist, and Parental Satisfaction

Adequacy of sedation for performing the scan was rated by the sonographer as only moderate for 7 subjects (6%) in group PENT and 9 (8%) in group DEX (Table 4), without group differences (P = .5922). Cardiologist satisfaction scores were not different between groups with 8% rated as moderately adequate in each group. Immediate parental satisfaction scores were not different, with 17 parents (12.2%) in group PENT (including 9 who required rescue sedation) and 15 parents (10.7%) in group DEX (including 7 rescued) rating overall immediate satisfaction <8 of 10 (0–10 scale), due to “prolonged onset of sedation” and “agitation.” Direct, next business day parental phone follow-up was achieved in 206 cases (74%). Parents reported that agitation at home was more common in group PENT with 36 subjects (31%) than in group DEX with 11 subjects (9%) (P < .0001; relative risk, 3.2; 95% CI, 1.7–6.1).

DISCUSSION

These findings from a randomized controlled trial suggest that DEX can be used equally efficacious for sedation as PENT, a drug that has been included in the FDA warning about the effects of repeated exposures in children under 3 years of age on brain development. Importantly, children with CHD are subject to repeated and prolonged exposures to sedatives and anesthetics early in life, but are also at increased risk for hemodynamic compromise from these drugs. This trial in a limited number of patients with a wide variety of congenital heart lesions suggests that DEX may not cause exaggerated hemodynamic responses requiring interventions or admission to the hospital in this vulnerable patient population.

Our findings for PENT are consistent with those of Warden et al, where single-dose oral PENT was 87% successful and a second dose of oral PENT provided 98% successful sedation.^² Moderate bradycardia was more common in group DEX; however, heart rates <60 beats·minute⁻¹ were not observed.

The 15%–16% of subjects who failed initial sedation in our study with either technique were successfully rescued with intranasal DEX, except for 1 PENT patient who required additional intranasal midazolam. Similar use of intranasal DEX has been reported for failed chloral hydrate sedation.^¹⁹ By using DEX rescue sedation, we did not observe the 19% incidence of prolonged sedation previously reported after rescue oral PENT.^² Additional DEX also did not result in increased levels of bradycardia, hypotension, or prolonged discharge times.

Sedation practices for TTEcho differ among pediatric cardiac centers,^¹ as chloral hydrate, midazolam, PENT, DEX, ketamine, propofol, or inhalational general anesthesia have been utilized.^¹ All medications, except for DEX, have been included in the FDA warning. Chloral hydrate also has significant adverse effects, such as nausea, vomiting, prolonged sedation, and is in limited supply in the United States^²⁰^,^²¹ Intranasal midazolam has been found to only be effective as a single-dose sedation for TTEcho in 24% of children, increasing to just 80% after a second dose.^²² Intranasal DEX has been shown in an observational study to be effective for TTEcho sedation in acyanotic, relatively healthy children for brief screening echocardiography with age-dependent differences in effectiveness.^¹¹ Similar to these findings, our data showed a trend toward lower initial DEX sedation success with increasing age. The current results are consistent with our initial feasibility study,^⁹ and our previous study demonstrating that intranasal DEX is as effective as oral chloral hydrate for TTEcho sedation with similar sedation onset and recovery times and heart rate changes.^¹⁰ Accordingly, general anesthesia is rarely utilized in the echocardiology laboratory in our practice.^²³

Recent animal studies have associated many sedative agents, including PENT, with structural abnormalities in the developing brain.^²⁴ Limited studies suggest that opioids or DEX may not lead to similar degrees of neuronal degeneration as conventional anesthetics.^⁸^,^²⁵^,^²⁶ Accordingly, the US FDA has published a warning regarding repeated or lengthy exposures to general anesthetics and sedatives (including PENT but not DEX) in young children.^²⁷ Since children with CHD frequently require several anesthetic exposures and repeated diagnostic studies with sedation early in life, DEX may represent an attractive alternative sedative for less painful procedures in this population.

Intranasal DEX administration occurred undiluted via atomizer to maximize diffusion gradients and nasal mucosal surface area.^²⁸ However, undiluted DEX by incremental drop technique may have a similar onset, if subjects are cooperative.^¹³ While parental satisfaction and patient flow might improve with more rapid onset of sedation, the relatively slow onset of intranasal DEX may prevent the bradycardia and initial hypertension seen with rapid intravenous administration.^²⁹

Unlike other sedatives and anesthetics, DEX has limited effects on upper airway caliber and respiratory drive and therefore used during sleep endoscopy in subjects with obstructive sleep apnea.^³⁰^,^³¹ In contrast, respiratory effects of PENT are complex with a narrow therapeutic window.^³²^,^³³ However, our data are in agreement with a larger study from our institution showing that respiratory adverse events after oral PENT sedation are rare.^² Consequently, we did not find an advantage of intranasal DEX over PENT in this regard.

DEX causes predictable decreases in heart rate by as much as 30% from normal resting heart rate.^⁵ Treatment of DEX-associated bradycardia with anticholinergics is generally not recommended, due to the risk of concomitant hypertension.^³⁴^,^³⁵ In the present study, heart rates below 80 beats·minute⁻¹ occurred more frequently with DEX compared with PENT, but no subject had a heart rate below 60 beats·minute⁻¹. None of our patients experienced hypotension, if previously published normal reference values for systolic pressure of 50–80 mm Hg, with a lower range (−2 SDs) of 45–60 mm Hg for 3–24 months of age were utilized.^³⁶

It is unclear whether echocardiographic indices are differentially affected by different sedative techniques. However, our cardiologists were unable to detect a qualitative difference between PENT and DEX-sedated echocardiograms and did not suspect sedation-related echocardiographic effects more commonly with either medication.

There were several limitations to this study. Although outpatients with complex congenital conditions were included, potentially less stable inpatients were not included in this study; therefore, the findings cannot be extended to patients with acute illness. Subjects with trisomy 21 are at risk for anesthesia-associated bradycardia,^³⁷ and therefore excluded from participation. However, a retrospective review of our trisomy 21 patients demonstrated that DEX compared favorably to PENT or general anesthesia in the relative risk of bradycardia.^³⁸ Despite allowing oral intake up to 2 hours before sedation, many children fasted for a longer period of time, which could have affected sedation success rates and vital signs. Teleflex Medical recalled the MAD Nasal atomizer in November 2016 for technical defects in some atomizers.^³⁹ Accordingly, although we did not observe atomizer malfunction, some intranasal DEX doses may have been inadequately atomized and may have led to differences in absorption. Continuous pulse oximetry was not instituted until onset of sedation so that any transient adverse effects on heart rate or oxygen saturation may have gone unnoticed. Although the secondary outcomes are of interest to the clinician, the study was not powered for all the secondary comparisons we report. Larger studies could be required to adequately power for secondary outcomes with correction for multiple comparisons.

Intranasal DEX was comparable to oral PENT sedation for TTEcho sedation in infants and did not increase the risk of clinically important adverse events. Intranasal DEX appears to be an effective “rescue” sedative for both failed PENT and DEX sedation. DEX could be a safer option for repeated sedation in children, but further studies are needed to assess long-term consequence of repeated sedation in this high-risk population.

ACKNOWLEDGMENTS

The authors thank Pam Bernard, RN, and Elizabeth Shaw, RN, for conducting and monitoring the sedation for this study. Megan Kalin, MS, and Millicent Frimpong-Manso, BS, were the research coordinators for this study. Dr Robert Coghill advised on the study design and edited the manuscript. Maria Ashton, MS, RPh, MBA, edited the manuscript.

DISCLOSURES

Name: Jeffrey W. Miller, MD.

Contribution: This author helped with conception and design of the study, data acquisition, data analysis, drafting the article, and final approval of the version to be published.

Name: Lili Ding, PhD.

Contribution: This author helped with data analysis, drafting the article, and final approval of the version to be published.

Name: Joel B. Gunter, MD.

Contribution: This author helped with data acquisition, data analysis, drafting the article, and final approval of the version to be published.

Name: Jennifer E. Lam, DO.