Clinical Outcomes Associated With Intranasal Dexmedetomidine Sedation in Children Free

We conducted a multicenter prospective cross-sectional study of children receiving IN dexmedetomidine for sedation undergoing a nonpainful procedure. We enrolled children between September 2018 to January 2020 presenting to 1 of 7 member institutions of the Pediatric Sedation Research Consortium. The Pediatric Sedation Research Consortium is a collaborative multidisciplinary group of over 50 institutions dedicated to improving the practice of pediatric sedation medicine through the sharing of prospective observational outcome data on procedural encounters.¹²  The 7 participating institutions were self-selected and all used IN dexmedetomidine as an established part of their practice before participating in this study. The institutional review board for each participating site approved this study with a waiver of informed consent. This study was registered at clinicaltrials.gov (NCT03220880).

We enrolled children aged 2 months to 17 years (ie, before their 18th birthday) undergoing either an MRI, ABR, ECHO, EEG, or CT scan. Exclusion criteria included patients with known allergy to dexmedetomidine, abnormal renal or hepatic function, concurrent use of digoxin or β-blockers, or cardiac conduction abnormality or heart block. Patients with known structural or congenital cardiac abnormalities or history of cardiac surgery were not excluded unless they had a cardiac conduction abnormality or heart block. A consecutive sample was enrolled for patients undergoing all procedures except for ECHO, which was a convenience sample. A review of missed eligible patients was performed for the convenience sample.

The dose of IN dexmedetomidine administered and use of adjunct sedatives were as per the treating clinician’s discretion. All patients received the same concentration of dexmedetomidine (100 mcg/mL), had IN medications administered using a mucosal atomization device (Wolfe Tory Medical, Inc.), and received a maximum volume of administration per nostril of 1 mL. Total volumes >1 mL were dispensed by alternating between nostrils until the total volume was administered. The maximum total single dose was 200 mcg at 5 sites and 100 mcg at 2 sites (Supplemental Table 7). There was no blinding of medications administered.

Outcomes were assessed by a trained clinician in real-time during the sedation and recorded using a standardized web-based data collection tool. All patients were monitored during the sedation using at least continuous pulse oximetry and blood pressure monitoring. Use of cardiorespiratory and end tidal carbon dioxide monitoring varied between sites and procedure types. The decisions to abort a procedure, administer an intravenous sedative, or perform an intervention in response to an event were made by the treating clinician. Patients were monitored after procedure completion by a nurse or advanced practice provider until patients were determined ready for discharge based on both study and institutional criteria.

The primary outcome for determining quality of sedation was the Pediatric Sedation State Scale (PSSS), a scale scored from 0 to 5 with strong validity and reliability in children for assessing the quality of sedation and adverse events (Table 1).¹³  The quality of sedation was assessed by determining the proportion of patients who experienced an acceptable sedation state, oversedation, and undersedation. Acceptable sedation state was a procedure-specific outcome defined as the range of PSSS scores represented during the entirety of a procedure that was considered to be acceptable for conducting and completing a specific procedure and not requiring administration of an intravenous sedative because of inadequate sedation. The range of scores representing an acceptable sedation state for each procedure was chosen by consensus by the nine study investigators. The PSSS scores chosen to represent acceptable sedation state were 2 or 3 for MRI and CT, 2 for ABR, and 2, 3, or 4 for echocardiograms and EEG. Oversedation was defined as having a PSSS score of 0 or 1 at any time during the procedure, indicating that there were abnormal vital signs requiring an intervention. Undersedation was defined as having a PSSS score of 5 at any time during the procedure, indicating that a patient required forceful immobilization. Oversedation and undersedation were not reported in patients who received an intravenous sedative because of inadequate sedation. The PSSS was assessed starting from time of procedure start until time of procedure completion. A PSSS score best representing each consecutive 5-minute block was documented, with the occurrence of scores representing oversedation and undersedation taking precedence.

For our secondary outcomes, we assessed procedure completion, defined as having obtained interpretable results without the use of an intravenous sedative because of inadequate sedation. The interpretability of test results was assessed by asking the provider performing the procedure (eg, sonographer for ECHO or audiologist for ABR tests) or the physician interpreting the study (eg, MRI, CT, or EEG) to classify the results in 1 of 3 categories: (1) adequate, no limitations; (2) minor limitations, still interpretable; or (3) major limitations, unable to interpret. Time-based outcomes included time of onset of sedation, time to procedure start, procedure duration, and total sedation time. Onset of sedation was assessed using the University of Michigan Sedation Scale (UMSS), a 5-point scale that has strong validity and reliability for identifying depth of sedation in children.¹⁴  The UMSS progresses from a score of 0, representing awake and alert, to a score of 4, representing unarousable. For this current study, adequate depth of sedation was defined as a UMSS score of 3 or greater. The UMSS was assessed starting from the time of dexmedetomidine administration until procedure completion and documented at least every 5 minutes, or whenever a change in depth of sedation occurred. Time to procedure start was defined as time from dexmedetomidine administration to when testing was initiated. Procedure duration was defined as time from procedure start to when the test ended and data collection ceased. Total sedation time was assessed using the Modified Aldrete Scoring System (MASS), which is scored from 0 to 10, with higher numbers representing greater readiness for discharge.¹⁵  Total sedation time was defined as time from dexmedetomidine administration until a MASS score of ≥9 (which represented readiness for discharge), with a MASS score assessed starting from time of procedure completion and documented at least every 15 minutes, or whenever a change in MASS score occurred. Events and interventions (Supplemental Table 8) were assessed starting from time of dexmedetomidine administration until a patient achieved a MASS score of ≥9.

Descriptive statistics were used to describe the characteristics of enrolled patients. Means and 95% confidence interval (CI) were determined for continuous outcomes (eg, time-based outcomes). Proportions and 95% CI were obtained for nominal outcomes (eg, acceptable sedation state). Outcomes were described in subgroups based on IN dexmedetomidine dose and regimens and procedure type. Time-based outcomes and events associated with different doses and regimens were reported across all procedures and within each procedure type. Predictors of acceptable sedation state were identified using a multivariable logistic regression model. Predictors were selected a priori based on clinically plausible relevance and included age (years), procedure type (MRI, ABR, ECHO, EEG), IN dexmedetomidine dose (2–2.9 mcg/kg, 3–3.9 mcg/kg, ≥4 mcg/kg), use of adjunct sedative (no adjunct sedative, oral midazolam, IN midazolam), and patient history of autism spectrum disorder. Patients who underwent CT scans were removed from the regression analysis because of insufficient number of patients required for analysis. Statistical analyses were performed using SPSS version 26 (IBM).

We enrolled 578 patients whose characteristics are shown in Table 2. The median age was 2.5 years (IQR 1.6–3), with more than 60% of patients <3 years of age. There were 10 (7.6%), 53 (16.9%), 4 (4.5%), and 10 (29.4%) patients with autism spectrum disorder that underwent MRI, ABR, ECHO, and EEG, respectively. The most common procedures were ABR testing (54.3%) and MRI (22.8%). The most frequently-used dose of IN dexmedetomidine was 3 to 3.9 mcg/kg (55%). Adjunct sedatives were administered in 39.3% of cases, all of which were either oral or intranasal midazolam, with the doses of adjunct sedatives administered shown in Supplemental Table 9. The missed eligible patients who underwent ECHO were similar to those enrolled with respect to age, weight, sex, American Society of Anesthesiologists status, dose of IN dexmedetomidine administered, adjunct sedative use, procedure completion, intravenous sedative use because of inadequate sedation, and interpretability of results. There was a greater proportion of patients with no developmental issues in the missed eligible compared with enrolled cohort (73.3% and 62.9%, respectively) (Supplemental Table 10).

The quality of sedation achieved by different IN dexmedetomidine doses and regimens are shown in Table 3, Fig 1, and Supplemental Tables 11 to 13. Rates of procedure completion are shown in Table 3 and Supplemental Table 14 and were 100% across all dose and regimen categories for ECHO, EEG, and CT scans. Among all patients who had successful procedure completion, only 88.3% achieved an acceptable sedation state for the entirety of the procedure; 7% experienced undersedation and required forcible immobilization during the procedure. Supplemental Tables 15 to 17 show the proportion of patients who had interpretable results, required an intravenous sedative because of inadequate sedation, and required intravenous sedation to initiate the procedure. There were no predictors of acceptable sedation state identified in the regression analysis (Table 4).

Table 5, Fig 2, and Supplemental Tables 18 and 19 show the time-based outcomes associated with different IN dexmedetomidine doses and regimens. Events and interventions associated with IN dexmedetomidine sedation are shown in Table 6. All events and interventions occurred in patients who only received IN dexmedetomidine with or without an adjunct sedative; none occurred in patients who received an IV sedative. The most common event overall was bradycardia (1.9%), with 0.7% of patients experiencing hypotension. However, no cardiovascular interventions were administered in response to any episodes of bradycardia or hypotension. Similarly, there were few airway or breathing events overall (2.3%), with all episodes of apnea (0.2%) and complete airway obstruction (0.2%) managed with airway repositioning and/or supplemental oxygen only. There were no patients who required an oral or nasopharyngeal airway, positive pressure ventilation, or endotracheal intubation in response to these events.

In this prospective study, we described the quality of sedation achieved in children undergoing sedation for nonpainful procedures when using different doses of IN dexmedetomidine and regimens with and without midazolam. The dose of IN dexmedetomidine and use of adjunct sedatives were not associated with the odds of achieving an acceptable sedation state. There were no events that prompted a serious airway, breathing, or cardiovascular intervention. Rates of acceptable sedation state were lower than rates of procedure completion, with a notable proportion of patients who successfully completed procedures but still experienced unacceptable sedation states or required forceful immobilization.

The range of doses of IN dexmedetomidine represented in our study are the same as those previously reported for sedating children for nonpainful procedures. Three studies that compared different doses of IN dexmedetomidine in children undergoing the same type of procedures described in our study showed that higher doses provided adequate sedation in more patients than lower doses.^7,16,17  We did not power our study to identify differences in acceptable sedation state between different doses. However, the proportion of patients who achieved an acceptable sedation state and procedure completion appeared to be comparable across each dose category and there was no difference in odds of achieving an acceptable sedation state based on dose. In addition, patients who received ≥4 mcg/kg of IN dexmedetomidine without midazolam appeared to have a longer time to onset of sedation compared with lower doses, which would seem counterintuitive. These observations may have been due in part to dexmedetomidine potentially having a wider range of pharmacodynamic effect on children compared with other sedatives. Future studies comparing different doses may be warranted to identify the minimum effective dose needed and better describe time-based outcomes associated with each dose.

The use of midazolam as an adjunct sedative with IN dexmedetomidine has been previously described in children undergoing MRI, ABR, ECHO, and CT scans. Although these studies did not report on the quality of sedation, the rates of successful procedure completion when using midazolam with IN dexmedetomidine for procedures, such as MRI, ABR, ECHO and CT have ranged from 42.1% to 98.4%.^9,10,18–20  Our regression analysis showed that the use of oral or IN midazolam did not predict acceptable sedation state. However, this finding differs from prior work that demonstrated a higher rate of procedure completion when using dexmedetomidine with midazolam compared with dexmedetomidine alone.⁹  Conversely, we observed that there appeared to be fewer patients who received oral midazolam that achieved an acceptable sedation state and procedure completion. These findings raise the question of whether the use of adjunct sedatives, particularly oral midazolam, with IN dexmedetomidine may be of benefit or potential detriment. Future study is warranted to evaluate the adjunct use of midazolam with IN dexmedetomidine, as well as any differences based on route of midazolam administration.

Cardiovascular events are commonly associated with dexmedetomidine sedation, with bradycardia and hypotension occurring in 1.9% and 0.7% of our cohort, respectively.¹  However, none of these events we observed required a cardiovascular intervention (eg, isotonic fluid bolus, vasoactive medication, chest compressions). The few events that did occur and required an intervention were related to airway and breathing and limited to airway repositioning and supplemental oxygen only; no patients required positive pressure ventilation or endotracheal intubation. Although our sample size was too small to identify more serious events, such as pulmonary aspiration or death, our rates of airway, breathing, and cardiovascular events are similar to that reported in prior studies of IN dexmedetomidine-based regimens.²¹  In addition, our study provides useful information not only regarding the occurrence of airway, breathing or cardiovascular events, but greater detail regarding the clinical significance of these events based on whether the event prompted a response in the form of an intervention and the specific nature of that intervention.

An important aspect of our study was our choice of quality of sedation as our primary outcome, rather than procedure completion. Procedure completion alone does not adequately reflect what a patient or clinician may experience during the course of a sedation. A patient’s sedation state may change during a sedation: the change may not be to the degree of warranting aborting the procedure or administering a parenteral sedative, but enough to instigate an interruption or pause in the procedure or a brief moment of forceful immobilization until the patient returns to an acceptable sedation state. Similarly, forceful immobilization may be used instead of administering additional sedatives to complete a procedure if the procedure is nearly complete or the ability to obtain interpretable results is less susceptible to patient movement. The occurrence of such events may not prevent the successful completion of a procedure, but the quality of the sedation would be considered suboptimal for both the patient and the sedation provider. Our study demonstrated that procedure completion was not synonymous with achieving an acceptable sedation state and reinforces the fact that the gap between acceptable sedation state and procedure completion exists and likely often goes unmeasured. Therefore, we believe that future studies of procedural sedation should measure not only procedure completion but quality of sedation to truly be able to discern clinically meaningful differences between different sedation regimens.

Our study had limitations. Because of the observational nature of our study, we did not enroll sufficient number of patients within every subgroup of IN dexmedetomidine dose, adjunct use, and procedure necessary for comparisons. There were few patients in certain subgroups (eg, only 1 patient who received 2 to 2.9 mcg/kg of IN dexmedetomidine with oral midazolam for an ECHO), and not all doses or regimens were represented across all procedures (eg, no patients who underwent ABR testing received IN midazolam). We enrolled a convenience sample for patients who underwent ECHOs, which could have introduced bias. However, missed eligible patients had a smaller proportion of patients with developmental issues and were similar to those enrolled across all other clinically-relevant demographics and clinical outcomes. The number of events, particularly those requiring an intervention, were too few to allow for comparison between different doses of IN dexmedetomidine or adjunct sedative regimens.

Intranasal dexmedetomidine-based regimens can achieve acceptable sedation states and high rates of procedure completion in children undergoing sedation for nonpainful procedures with a low incidence of adverse events requiring serious airway, breathing, or cardiovascular interventions. Further study is needed to identify the optimal dose of IN dexmedetomidine and delineate the effect of adjunct sedatives on clinically meaningful outcomes such as quality of sedation.