Abstract
Introduction
Pediatric perioperative anxiety affects over 60% of children presenting for anesthesia.1,2 Due to a sense of powerlessness in an unfamiliar environment, perioperative anxiety may lead to short- and long-term consequences. 3 Short-term effects include a higher risk of emergence delirium (ED), increased pain perception, longer recovery stays, and decreased parental satisfaction.2,4–6 Long-term sequelae include maladaptive behavioral changes such as enuresis, sleep disturbances, changes in eating patterns, and aggression toward authority.2,7,8
Mask induction of anesthesia in the pediatric population poses unique considerations. The introduction of the anesthesia mask during inhalational induction of anesthesia elicits high levels of anxiety. 9 Pediatric anesthetic inductions differ from that of consenting adults since uncooperative children may require coercion or even physical restraint. 10 Additionally, up to 15% of children undergoing anesthesia experience postoperative ED, with symptoms increasing with higher preoperative anxiety levels. 2 Enhancement of the mask induction experience maximizes the likelihood of cooperation and minimizes anxiety, reducing negative effects on children's well-being. 11
Typical approaches to reducing pediatric perioperative anxiety and optimizing cooperation include parent-present inductions of anesthesia (PPIA) and/or anxiolytic medication such as oral midazolam. 12 Although these traditional techniques can be effective, disadvantages to PPIA include adverse parent reactions and high states of parental anxiety potentiating a child's anxiety.12,13 Drawbacks of oral midazolam include timing the administration for optimal effect, paradoxical reactions in 3-18% of children, and compliance with consumption due to its bitter taste.14–17 Given these limitations, enhanced strategies are needed for managing pediatric perioperative anxiety and cooperation during mask induction.
Nonpharmacological interventions, such as immersive technology using extended reality headsets, are promising adjuncts for pediatric perioperative anxiety without the side effects associated with pharmacological agents.18,19 Various levels of immersion, from smartphones to virtual reality (VR) provide distracting images, sounds, and interactive content to the patient.19,20 Although VR reduces pain and anxiety through complete immersion during minor procedures in hospitalized children, disadvantages include the weight and size of VR headsets on pediatric patients, the risk of motion sickness, and the inability to interact with the real world, including their caregivers.21–23 VR can decrease anxiety during inhalation induction, however disadvantages to this technique include difficulty achieving a tight face mask seal due to the headset and obstruction of the patient's eyes from the anesthesia provider. 23
Compared to VR, augmented reality (AR) uses head-mounted displays (HMD) to depict holograms superimposed on the real world. 24 Some HMDs utilize a visor, which allows full access to the face for proper face mask application, while also allowing visualization of the patient's eyes. Because the risk of motion sickness is lower with AR than VR, many providers feel more comfortable combining AR with midazolam premedication.25,26
Given the limitations of VR and the prevalence of perioperative anxiety leading to uncooperative mask inductions, we hypothesized that AR would be an effective tool for improving the patient experience during mask induction. Our primary aim was to investigate mask acceptance during inhalational induction utilizing AR. The secondary aims explored whether AR improved cooperation during mask induction and reduced the incidence of ED in the recovery room.
Methods
Design
This was a prospective, matched case-control study with historical controls. Our institutional Internal Review Board approved a waiver of consent for participation (eProtocol No. 62835, 10/01/2021). The manuscript adheres to the STROBE guidelines for case-control studies. 27
Setting
We conducted this study at a freestanding, academic, quaternary care children's hospital in Northern California. The hospital contains 365 beds, 16 operating rooms, and 10 preoperative bays. Patients have preoperative access to certified Child Life Specialists, an assortment of preoperative anxiolytic medications including midazolam and dexmedetomidine, and technology-based adjuncts such as tablets and VR. Anesthetic providers include board-certified pediatric anesthesiologists, pediatric anesthesiology fellows, and anesthesiology residents. AR participants were collected prospectively from 30 August 2021 to 24 June 2022.
Participants
Patients greater than 4 years and undergoing elective procedures requiring general anesthesia with inhalational mask induction of anesthesia were eligible. At our institution, inhalational induction is offered as a default for patients unless contraindicated by the patient medical history or if the patient tolerates preoperative PIV placement. Patients with facial abnormalities or injuries prohibiting comfortable use of the headsets, seizures, severe developmental delay, vertigo, or nausea or postoperative nausea and vomiting were excluded. Those patients undergoing ophthalmologic procedures and those receiving intravenous induction of anesthesia were also excluded. Additionally, patients less than 4 years of age were excluded due to the size and fit limitations of the AR headset.
Each study participant was matched to 3 historical controls elicited from the electronic medical record (EMR, Epic Systems, Verona, WI, USA). The controls had identical American Society of Anesthesiologists (ASAs) class, use of premedication or no premedication, similar age, and same gender of each matched AR participant. If more than 3 control patients were matched, 3 control patients were randomly selected using a random number generator (Microsoft Excel for Microsoft 365. Version 16.72). No control patients used VR during mask induction because it is only used preoperatively during intravenous placement at our institution.
Intervention
Several AR HMD are commercially available, ranging from several hundred to several thousand dollars. 24 In order to increase applicability and usability, we selected the Dream Glass (Dream Glass, Fremont, CA, USA). Advantages of the Dream Glass include its lower cost compared to higher-end AR HMDs and its corded controller that allows providers to easily control content. 28 The Dream Glass projects holograms onto a glass visor that sits approximately 2 inches from the face. The images are superimposed onto the real environment, creating the equivalent of a 200-inch screen with 90° field of vision optics.26,28 We introduced the Dream Glass into our clinical practice as a technological adjunct for improving patients’ anxiety and compliance during mask induction.
Preoperatively, Child Life Specialists offered AR to interested patients who met eligibility criteria. After a brief orientation to the headset, the patient selected a commercially available video from 1 of 3 streaming applications. The headset was adjusted for optimal fit, and the patient was transported from the preoperative area to the procedure room while viewing their holographic video.
All patients received a sevoflurane inhalational induction with the headset in place (Figure 1). The headset was removed once the patient's eyes remained closed after the induction of anesthesia. Anesthetic care was conducted independent of Dream Glass use and an intravenous catheter was placed after induction. The patient's anesthesiologists completed the induction assessment form in the EMR per routine, which included assessments that described the patient's response to mask induction.

Dreamglass AR headset on patient during mask induction of anesthesia.
Outcome Variables
The primary outcome compared mask acceptance during inhalational induction utilizing AR compared to historical case-matched controls receiving standard of care (SOC), which included oral premedication, child life specialists and/or parent present induction of anesthesia. Patients in the AR group were also eligible for oral midazolam.
The first secondary outcome explored whether AR improved cooperation during mask induction. The final secondary outcome investigated the incidence of ED in the recovery room between those utilizing AR during mask induction compared to the historical controls.
Data Sources and Measurements
Matched case-control patient data were collected via chart review from the EM. The primary outcome was assessed using the Mask Acceptance Scale (MAS), which is a commonly used tool for determining mask acceptance during inhalational induction of anesthesia. 29 The MAS is scored on a 4-point scale: 1 = unafraid, cooperative, accepts mask readily, 2 = slight fear of mask, easily reassured, 3 = moderate fear of mask, calmed with reassurance, 4 = terrified, crying or combative. We compared patients with MAS scores of 1 to those with scores of 2 or greater between the AR and SOC groups. The scores were collected from the EMR's induction assessment form completed by the anesthesiologists.
The first secondary outcome, cooperation, was determined by the anesthesiologist and defined by a binary answer of yes or no. Regarding the final secondary outcome, the Watcha scale is used to assess ED in the recovery room. The Watcha scale is a 4-point scale that ranges from “calm” to “agitation with thrashing around.” 30 Scores greater than 3 are consistent with a diagnosis of ED. Watcha scores were recorded by the patient's recovery room nurse.
Bias
Although patient blinding was not possible due to the nature of the intervention, the anesthesiologists caring for the patient were blinded to the aims of the study. In addition, the statistician was blinded to the case and SOC groups to minimize any potential biases arising from the analyses. Finally, the anesthesiologists of record for the historically matched case-control data were unaware that data would be utilized, further reducing the risk of reporting bias.
Study Size
The primary outcome was based on the incidence of ideal mask acceptance (defined as MAS scores of 1). We based our sample size calculation on institutional pilot data of 20 patients that demonstrated an incidence of poor mask acceptance in the AR group compared to SOC to be 7% and 30%, respectively. Applying a 1 to 3 intervention to matched control sampling approach, we estimated that we required 37 patients in the AR group and 111 in the SOC to have 80% power to be able to detect a > 20% improvement in mask acceptance. An effect size of around 20% was chosen based on previous studies on interventions that reduce preoperative anxiety. 31 To account for potential incomplete data, we collected 50 patients in the AR group and 150 in the SOC group.
Statistical Methods
Demographic and baseline characteristics were reported as mean ± interquartile range (IQR) as appropriate for continuous variables. Categorical variables were presented as percentages.
For the primary outcome, the relative risk (RR) and its 95% confidence interval (CI) were calculated. Categorical secondary outcomes such as cooperation and ED were compared between 2 groups using chi-square analysis or Fisher's exact test as appropriate. The continuous variables were compared using a t-test or the Wilcoxon Rank Sum test as appropriate. Statistical analyses were performed using Stata version 14.0 (StataCorp., College Station, TX, USA).
Results
Demographics
Fifty patients, ages 4-12 years (M = 7.5 years, [IQR = 6-9 years]), used AR for inhalational induction. Within the AR group, 66% were male and 34% were female, and the majority were ASA class 1 and 2. Within this group, half the AR patients also received premedication (Table 1). There were 150 patients in the SOC group with matched demographics.
Demographics.
ASA, American Society of Anesthesiologists.
Primary Outcome
MAS scores ranged from 1 to 4 (M = 1, [IQR = 1-1]). Scores of 2, 3, and 4 accounted for 2 of 50 (4%, 95% CI [0, 9.4%]) in the AR group and 29 of 150 (19.3%, 95% CI [13%, 25.7%]) in the SOC group (RR 0.21, 95% CI [0.05, 0.84], P = .027) (Table 2).
Comparison of Dream Glass Headset Versus Case-Matched Controls.
Secondary Outcomes
Regarding cooperation during induction, 49 of 50 in the AR group were considered cooperative (98%, 95% CI [94.1%, 100%]) compared to 137 of 150 (93.8%, 95% CI [86.8%, 95.8%]) in the SOC group (P = .457) (Table 2). The incidence of ED during recovery was analyzed and there were 0 patients in the AR group with Watcha scores greater than 3 and only 1 of 150 (0.7%) in the SOC group (P = 1.000) (Table 2).
Discussion
The primary aim was to investigate mask acceptance with AR utilization during inhalational mask induction of children compared to those receiving SOC. Our results indicate an association between the use of AR and an ideal MAS score. No relationship was observed between cooperation and the incidence of ED to AR use.
The improvement of mask acceptance has important implications for the use of nonpharmacologic techniques to enhance a child's experience during mask induction. Several other nonpharmacologic strategies, including preoperative education workshops, cognitive behavioral therapy, clowns, music therapy, and acupuncture, have reduced preoperative anxiety in pediatric patients.32,33 However, these nonpharmacologic techniques are limited due to time constraints, need for trained personnel, and cost.32,33
HMDs with various levels of immersion have been used for anxiolysis and pain reduction.19,20,24,26 Similar to these findings, one study demonstrated that specialized glasses which displayed distracting videos during mask induction reduced anxiety and were noninferior to midazolam for preoperative anxiolysis. 34 Although those investigators assessed anxiety, there were no measures of mask acceptance or cooperation, which are important components to a successful mask induction. 34 A child can still have difficulty accepting the mask while exhibiting low levels of distress. Additionally, these video glasses from 10 years ago offered a 50-inch virtual screen, in contrast to the Dream Glasses which offer a 200-inch virtual screen. 34 The larger virtual screen increases the level of immersion, increasing the likelihood of distraction and engagement to improve anxiolysis. 35 Overall, these findings support the use of AR as a promising and affordable nonpharmacologic approach to improve mask acceptance in pediatric patients.
Regarding cooperation, no differences were observed between patients who used AR compared to the SOC group. This is likely due to the oversimplification of cooperation to a binary scale on the EMR induction assessment form at our institution. Although this truncated scale improves clinical efficiency, typical research-based scales that measure cooperation incorporate various positive and negative behaviors. 12 In this study, the majority were assessed as cooperative with half the patients receiving a premedication. Future studies are needed to distinguish potential differences in cooperation and AR using robust research-based scales as well as controlling for the use of premedication.
The incidence of ED and induction facilitated by AR was also investigated since high anxiety increases the risk of ED postoperatively. 36 No differences in the rate of ED were observed between the AR and SOC groups. Similar to cooperation, the rate of ED was very low in both groups, likely due to our local routine practice of deep extubations with propofol and occasionally dexmedetomidine.37,38 Future investigations with larger sample sizes are needed to adequately power studies that explore the relationship between ED and MAS within the setting of AR during induction.
This study had several limitations. First, although the MAS is a commonly used scale for assessing mask acceptance and is integrated into the standard EMR form, the scale itself has never been appropriately validated. 29 Second, the MAS does not measure anxiety directly. The scale incorporates behavior and cooperation into categories, making it difficult to differentiate nuances between a child that is cooperative but still fearful at mask induction. Behavior, affect, and cooperation are separate constructs that are commonly assessed independently. Additional studies are required to further delineate the mask induction experience with AR by utilizing well-known, validated research scales that measure affect and cooperation. 39 Third, we only utilized a single measure of ED, which is the one used in our EMR at our institution. The addition of another scale, such as the Pediatric Anesthesia Emergence Delirium scale, may have provided a more robust correlation measure to detect potential effects on the incidence of ED. Fourth, the results are not applicable to patients less than 4 years due to the fit of the HMD. Younger children tend to have high levels of anxiety, and it is unknown if an HMD designed specifically for young children would also benefit from AR during mask induction. 40 Fifth, due to the nature of the AR HMD, the anesthesiologist was not blinded to the intervention, which may have altered their behavior in an unpredictable direction. Sixth, as with all observational, case-matched studies, the results may have been subjected to bias which may have been mitigated with a prospective, randomized controlled study. Finally, although AR HMDs are substantially lower in cost than most AR HMDs, they still rely on trained staff and resources that may not be translatable to variable resource settings, limiting their widespread appeal. However, similar to other technologies, over time these devices will continue to have improved optics with reduced costs, increasing accessibility.
Managing pediatric anxiety and optimizing cooperation by providing an ideal mask acceptance experience is essential to the perioperative experience.1–8 AR during mask induction improved mask acceptance compared to SOC. Future research will investigate the integration of this technology into routine clinical practice as a nonpharmacologic intervention with randomized trial designs.
Footnotes
Assistance
The authors gratefully acknowledge the assistance of Ahtziri Fonseca and Maria Menendez
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethics Statement
This study was approved by the Stanford University Internal Review Board and provided a waiver of consent for participation.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
