Virtual Reality Distraction vs Nitrous Oxide for Reducing Anxiety and Injection Pain

Abstract

Objective

The purpose of this study was to compare nitrous oxide (N₂O) vs virtual reality (VR) as methods for reducing pain and anxiety during a dental injection. The primary objectives were to assess acute changes in stress responses by comparing salivary cortisol levels between the 2 groups and differences in injection pain scores.

Methods

A total of 132 female subjects serving as their own control received maxillary lateral incisor infiltration injections with the use of either N₂O or a VR headset during separate appointments spaced at least 2 weeks apart. Salivary cortisol samples were collected at 6 times throughout each appointment. Pain scores for needle insertion and solution deposition were recorded. Data were analyzed statistically using Wald and paired t tests.

Results

N₂O significantly lowered salivary cortisol concentrations and subsequent physiologic anxiety as compared with VR (P = .0089). However, no significant differences in needle insertion or solution deposition pain scores were found.

Conclusion

Although VR and N₂O may be comparable in terms of perceived pain reduction, N₂O was a more effective method than VR for physiologic analgesia and anxiolysis.

Dental pain and anxiety create significant challenges for patients and caregivers alike. Fear related to dental treatment and pain prevents many patients from seeking needed care.¹ For dental caregivers, managing a patient’s pain and anxiety can be the most difficult aspect of providing dental treatment. New and existing methods to reduce dental pain and anxiety are a constant source of interest due to their high value to both patients and providers.

One of the most widely studied and used methods for pain and anxiety management is inhalational sedation with nitrous oxide (N₂O). In patients presenting with or without dental pain (ie, symptomatic vs asymptomatic), N₂O has been demonstrated to produce analgesic, anesthetic, and anxiolytic effects.^2-5 N₂O is also easily titrated and relatively safe.⁶

An alternate source for reducing pain and anxiety is the use of a distraction technique or device. Some of the previously studied distraction methods in a dental setting involve audiovisual aids including headphones, tablets, and video games.⁷ More recently, virtual reality (VR) technology has been of high interest as a distraction method in dental settings.^8-11 VR is more immersive than other audiovisual distraction methods and may be a better option for analgesia and anxiolysis. Results from VR studies have predominantly favored VR over other audiovisual distraction methods for reducing pain and anxiety.^7-11 However, those studies tended to focus on pediatric patients, had relatively low sample sizes, and primarily used subjective forms of measurement to draw conclusions. Additionally, control groups in VR clinical trials were principally composed of either other audiovisual methods of distraction or no intervention.

Popular measurements for recording a patient’s anxiety or pain include the Corah Dental Anxiety Scale¹² and the Heft-Parker Visual Analogue Scale (HP-VAS).¹³ These scales provide a subjective recording of perceived anxiety or pain as identified by either the patient or the provider.^12,13 Although no perfect measurement of pain or anxiety exists, many researchers have incorporated readings such as blood pressure or heart rate to physiologically measure anxiety and pain.^7,14,15 One of the more promising physiologic measurements of anxiety is plasma cortisol concentration. Although cortisol is produced continuously and levels fluctuate throughout the day, the body produces higher concentrations when in a state of stress.¹⁶ One method used to approximate plasma cortisol concentrations is to collect salivary cortisol samples, a validated method that previously has been used to measure stress levels during several medical and dental studies.^17,18

The purpose of this study was to compare N₂O vs VR as methods for reducing pain and anxiety during a dental injection. The primary objectives were to assess acute changes in stress responses by comparing salivary cortisol levels between the 2 groups as well as differences in injection pain scores.

METHODS

This study was approved by The Ohio State University (OSU) Human Subjects Review Committee. Prior to starting the study, subjects were screened via a health history form and a clinical exam, signed an informed consent, and completed a Health Insurance Portability and Accountability Act research authorization form. Inclusion criteria consisted of the following: potential participants who were assigned female sex at birth and were between 18 and 65 years of age. Exclusion criteria consisted of the following: potential participants who were pregnant or nursing; those with poor health as identified by an assigned American Society of Anesthesiologists physical status classification III or greater; any clinically visible signs of injury or pathosis in the maxillary labial vestibule; contraindication to N₂O use including a history of chronic obstructive pulmonary disease, allergy, nasopharyngeal obstruction, upper respiratory infection, sinusitis, and/or recent retinal or middle ear surgery; or inability to use a VR headset (eg, epilepsy, motion sensitivity, or vision impairment). Additionally, potential participants taking corticosteroids were also excluded.

Subjects were scheduled for 2 separate appointments, during which a maxillary infiltration injection was performed while either N₂O was administered for 30 minutes or the subject watched a self-selected movie using a VR headset for 30 minutes. Appointments were scheduled at least 14 days apart for approximately the same time of day. Subjects received the same labial infiltration over either the right or left maxillary lateral incisor during each appointment. The order of the N₂O and VR interventions and the side (right or left) of the maxillary injection were assigned by use of a random number generator (www.random.org).

To avoid contamination or dilution of salivary samples or postprandial plasma cortisol alterations, subjects were instructed to avoid eating or drinking for 60 minutes prior to each appointment. Subjects were also asked to complete a Corah Dental Anxiety Scale questionnaire (Figure 1) at the start of each appointment.¹²

Figure 1.

Corah Dental Anxiety Scale

Questionnaire and rating scale for Corah Dental Anxiety Scale.

Following completion of forms and delivery of N₂O-/VR-specific instructions, the first of 6 saliva samples was collected using a Salivette Cortisol synthetic swab (Sarstedt AG & Co KG). Upon placement of the first saliva swab, a stopwatch was used to track the passage of time throughout the remainder of the appointment. The time at which the stopwatch was started was considered minute 0.

Saliva collection swabs were placed under the right side of the subject’s tongue for 2 minutes or until saliva saturation was confirmed (ie, the swab appeared dark and moist) to ensure a sufficient volume of saliva to detect cortisol concentrations during the assaying of samples. Salivary samples were collected at the following times: sample 1 at minute 0; sample 2 at 15 minutes; sample 3 at 30 minutes; sample 4 at 1 minute following delivery of local anesthetic; sample 5 at 45 minutes; and sample 6 at 60 minutes. Samples 1 to 3 were used to determine baseline readings. Immediately following collection, swab samples were placed in individual test tubes, sealed, and transferred to a deep freezer (19-21°C) until processing was completed.

Salivary cortisol concentrations collected from saliva samples were measured using a competitive immunoassay supplied by Salimetrics. The samples were assayed in duplicate on a Spectra Max 190 microplate reader (Molecular Devices Corporation). The intra-assay coefficient of variation for the assay ranged from 4% to 7%. The inter-assay coefficient of variation for the assay ranged from 3% to 11%. All samples were processed by the Lead Research Lab Technologist at the Center for Clinical Research Management at the OSU College of Medicine.

N₂O Appointment

At the start of the N₂O appointment and before the collection of sample 1, subjects were instructed on the purpose of and expectations for N₂O use as well as signs and symptoms of oversedation (eg, nausea, onset of incoherence, and sensation of imminent syncope). Following sample 1, a scented Clearview Classic Nasal Hood (Accutron, Inc) was placed over the subject’s nose. Subjects were given protective eyewear, and subject comfort was confirmed. Using a standard dental N₂O machine, N₂O was administered over a 5-minute period by titrating from 0% up to 20% to 50%. The final N₂O percentage was based on the subject achieving minimal sedation (ie, anxiolysis in which the subject was very relaxed, awake, and retained the ability to respond appropriately to verbal commands) as determined by the primary researcher (K.E.). We monitored the sedation level every 10 minutes by questioning the subject during the N₂O administration. The N₂O/O₂ sedation was maintained at that level for 30 minutes and during the local anesthetic injection. Following completion of the injection, subjects were placed on 100% oxygen for at least 10 minutes. The remaining salivary samples were collected, after which the appointment was completed and the subject dismissed if no signs of sedation were present.

VR Appointment

At the start of the VR appointment, subjects were instructed on the proper use of the provided VR headset and were given a list of films to watch during their appointment: Avengers—Infinity War, The Incredibles, The Sandlot, and Rocky. Movies were selected according to overall popularity, a PG-13 or less rating, and their appeal to different audiences. Selections also had to meet content safeguards consistent with university guidelines, which prevented the use of other media options (eg, streaming services). During sample 1 collection, a researcher prepared the VR headset (Oculus Quest 2, Meta Technologies LLC; Figure 2) including confirming orientation of the VR headset and the subject’s movie selection. This VR headset relies on 2 lenses set into the device that display overlapping images, resulting in a display that appears 3-dimensional to the viewer. The VR screen was locked, allowing the user to move their head in any direction while maintaining the display in the center of their vision. This adjustment was necessary to avoid the image dropping from the subject’s line of sight as their head orientation was altered to facilitate local anesthetic delivery. Following collection of sample 1, the subject was instructed to put on the VR headset and begin watching the movie. Over-the-ear headphones (ATH-M20X, Audio-Technica Inc) were plugged into the VR headset and placed over the subject’s ears by a researcher. Subjects continued viewing the film throughout completion of the injection. Following anesthetic delivery, the headphones and headset were removed by a researcher.

Figure 2.

Female Subject Wearing the VR Headset

A clinical picture of the virtual reality (VR) headset.

Local Anesthetic Injections and Pain Ratings

As the part of the N₂O or VR instructions, subjects were shown an HP-VAS pain rating form (Figure 3) to assess their perceived pain during needle insertion and solution deposition immediately following the local anesthetic injections. The HP-VAS was divided into 4 categories. No pain corresponded to 0 mm. Mild pain was defined as greater than 0 mm and less than or equal to 54 mm. Mild pain included the descriptors of faint, weak, and mild pain. A score greater than 54 mm and less than 114 mm indicated moderate pain and included the descriptor of moderate pain. Severe pain was defined as equal to or greater than 114 mm. Severe pain included the descriptors of strong, intense, and maximum possible pain.

Figure 3.

The Heft-Parker VAS

Values of the Heft-Parker Visual Analogue Scale (VAS). The numbers at the top of the scale were omitted on the patients’ VAS.

All infiltration injections were delivered by the primary researcher (K.E.) in a similar manner on all subjects. Topical anesthetic was not used in this study. During the injection, the maxillary lip of the assigned side was lifted using a 2 × 2 gauze pad. A dental syringe fitted with a 27-gauge, 1½-inch needle was inserted into the labial mucosa at a location best approximating the apex of the maxillary lateral incisor. Each subject received 1.8 mL of 2% lidocaine with 1:100,000 epinephrine (Xylocaine, AstraZeneca LP) delivered over 60 seconds. Following the infiltration injection, subjects were asked to rate their perceived level of pain using the HP-VAS form. Subjects were monitored for at least 30 minutes following local anesthetic delivery, during which time samples 4, 5, and 6 were collected.

Statistical Analysis

An a priori power analysis was used to determine that a minimum sample size of 100 subjects would be required based on a nondirectional α risk of .05 and assuming an SD of 33 to demonstrate a difference of ±10 points on the HP-VAS with a power of .95.^19,20

Data collected were statistically analyzed by a PhD statistician and a PhD graduate student in statistics to ensure use of the correct statistical tests. Comparison of salivary cortisol levels over time between N₂O and VR was done using a Wald test based on a mixed-effects model with a random intercept for randomized subject identification numbers. The Wald test was used to assess the significance of coefficients in regression analysis for salivary cortisol levels. Paired t tests were used to compare mean HP-VAS pain scores for needle insertion and solution deposition. The Wilcoxon rank sum test was used to compare Corah anxiety scores. All comparisons were considered significant using a P < .05.

RESULTS

A total of 132 female subjects participated in this crossover, randomized trial. The ages of subjects ranged from 19 to 57 years, with a median age of 24 years. Each subject served as their own control.

Salivary cortisol levels overall ranged from 0.014 to 1.257 µg/dL, with a mean level of 0.205 µg/dL. Mean salivary cortisol concentrations were found to be significantly lower during the N₂O appointments when compared with the VR appointments (P = .0089; Figure 4; Table 1).

Figure 4.

Average Cortisol Levels Over Time

Graph illustrates the average cortisol levels for NO and VR over 60 minutes. Overall, nitrous oxide significantly lowered cortisol concentrations and subsequent physiologic anxiety as compared with virtual reality. NO indicates nitrous oxide; VR, virtual reality.

Table 1.

Mean Cortisol Levels for Nitrous Oxide vs Virtual Reality

Sample no.	Cortisol level, mean, μg/dL
	N2O	VR
1	0.198	0.215
2	0.200	0.221
3	0.181	0.204
4	0.209	0.236
5	0.200	0.220
6	0.177	0.200

Salivary cortisol concentrations were found to be significantly lower during the N₂O appointments when compared with VR appointments for samples 3 to 5 (P = .0089).

Abbreviations: N₂O, nitrous oxide; VR, virtual reality.

No statistically significant difference between the reported VAS pain scores during the N₂O and VR appointment was found for needle insertion or solution deposition (Table 2).

Table 2.

Mean Needle Insertion and Solution Deposition Pain Scores

Injection phase	Mean N2O pain score	Mean VR pain score	P value
Needle insertion	36	39	.394
Solution deposition	43	47	.185

All mean pain values were rated as weak to mild pain on the Heft-Parker Visual Analogue Scale.

Abbreviations: N₂O, nitrous oxide; VR, virtual reality.

Most subjects listed their initial Corah rating at no/mild anxiety for the N₂O (92%) and VR (94%) appointments. The median Corah anxiety was rated as 5 (mild pain) for each group, and there was no significant difference between the groups. Most subjects (59%) selected The Incredibles (an animated movie) or Avengers—Infinity War (an action movie).

DISCUSSION

The salivary cortisol levels in this study ranged from 0.014 to 1.257 µg/dL (mean, 0.205 µg/dL), which was lower than values found in previous studies. Mean salivary cortisol values ranged from 0.76 to 1.99 µg/dL in the study by Pereira-Santos et al¹⁸ and from 0.62 to 0.83 µg/dL in the study by Shetty et al.¹¹ Those higher levels may be explained by the study population (children 5-8 years of age¹¹) and/or the third molar extractions¹⁸ or actual dental treatments used in those studies, as opposed to the simple infiltration used in this study.

Overall, salivary cortisol levels were found to be statistically (P = .0089) lower during the N₂O appointment in comparison with the VR appointment (Figure 4; Table 1). A study by Periera-Santos et al¹⁸ found that N₂O use reduced the salivary cortisol levels of subjects during a surgical appointment when compared with an appointment where subjects did not use N₂O. Although VR has also been demonstrated to lower salivary cortisol,^11,21,22 it was not as effective as N₂O in this study.

Regarding the time periods and the cortisol values, sample 1 was collected after the detailed instructions were given for either intervention but before any injection pain or any potential problems. It is possible that the anticipation of the intervention was already in place at time of sample 1 because it occurred prior to the physical use of either intervention. The cortisol values may reflect this anticipation. Sample 2 at 15 minutes showed a slight increase vs sample 1 for both N₂O and VR cortisol levels, which may reflect the patient getting comfortable with the interventions. Sample 3 at 30 minutes demonstrated lower cortisol levels for both groups, probably due to the anxiolytic actions of N₂O and VR. After sample 3 was procured, the infiltration injection was administered. Sample 4 was obtained 1 minute after the infiltration injection was completed and showed an increase in cortisol levels in both groups, likely due to pain from the injection. Samples 5 (at 45 minutes) and 6 (at 60 minutes) demonstrated decreased cortisol levels, likely reflecting the patient’s return to normal baseline without the use of N₂O or VR.

Studies have reported salivary cortisol levels falling over time during VR use^11,21,22 and N₂O use.¹⁸ We also demonstrated this effect per sample 3 (Figure 4; Table 1). These findings corroborate the theory that anxiolytic interventions help lower stress, which in turn leads to lower salivary cortisol concentrations due to the reduced cortisol production from the HPA axis.¹⁶ The present study also found a relatively sharp rise in average salivary cortisol levels immediately following the dental injections (Figure 4; Table 1). Studies by Pereira-Santos et al¹⁸ and Quartana et al²³ also reported increases in salivary cortisol levels following painful stimulation. These findings align with the expected rapid increases in cortisol levels based on an increased HPA axis response secondary to acute stress and pain from the dental injections.^16,24 This was followed by decreased stress and lower salivary cortisol levels (Figure 4; Table 1). This pattern matches the expected, rapid change in cortisol concentrations reported by Bozovic et al²⁹ and Kirschbaum et al.²⁴

Production of cortisol can be measured in the plasma between 5 and 20 minutes following exposure to stress.²⁵ Plasma cortisol transfers to the saliva in 2 to 3 minutes,²⁴ and the half-life of plasma cortisol is around 66 minutes.²⁶ We therefore would expect to be able to measure an increase in salivary cortisol within 7 to 23 minutes after injection and a decrease in salivary cortisol closer to 66 minutes postinjection. However, this study demonstrated a shorter interval between statistically significant salivary cortisol changes than has been reported in previous studies.^11,18,22,23 The decision to collect samples at 15-minute intervals gave the present study the best opportunity to capture and record potential increases or decreases in salivary cortisol concentration. The decision to collect the third and fourth samples immediately before and after the injection was made to capture any salivary cortisol concentration spikes as closely as possible.

Consideration of the stimulant effect of an action VR experience (ie, an action movie and/or a stimulating movie sequence) on cortisol levels would have been helpful in validation of the results of cortisol level rise being attributed to the injection. However, the cortisol spike from a painful injection is substantially greater than any slight variation in stress due to the intensity of the selected VR media. It would have been interesting to compare cortisol changes between types of media selected, but that analysis fell outside of the scope of this project.

The pain reported by subjects was higher on average during the VR appointment than during the N₂O appointment; however, none of the differences between the VR and N₂O pain scores were statistically significant (Table 2). The mean pain scores reported for needle insertion were 36 to 39 mm, corresponding to mild pain. The mean scores for solution deposition were 43 to 47 mm, which were higher but still corresponded to mild pain (Table 2). In previous studies from our research group by Perry et al¹⁹ and Hyde et al,²⁰ female participants from a similar population (ie, same age range, no preoperative pain, low anxiety ratings) to that used for the current study had a maxillary lateral infiltration injection administered without receiving any intervention. The mean VAS pain rating reported by those female subjects during needle insertion were higher when a male operator administered the injection¹⁹ and for subjects with light and dark eyes²⁰ (46 and 41 mm, respectively), both approaching the higher end of mild pain. Pain scores for solution deposition were markedly higher (68 and 67 mm, respectively) and into the moderate pain category for the Perry et al¹⁹ and Hyde et al²⁰ studies as opposed to the lower pain scores (43 and 47 mm) found in the current study. The lower pain ratings in this study likely demonstrated the analgesic effects of both N₂O inhalation and VR distraction.^2-5,14,15

A maxillary lateral incisor infiltration injection was used because it is a relatively simple injection to administer, and because studies have shown this injection was one of the more painful injections used in dentistry.^19,20 Using a painful injection resulted in a broader range of pain ratings to facilitate detecting any differences between treatments. Topical anesthetic was not used during the present study, to maximize measurable pain and to eliminate a potential confounding variable. Additionally, only female subjects were used in the current study because significantly higher pain scores were found during the solution deposition phase when female subjects received maxillary lateral incisor infiltrations from male operators.¹⁹ We standardized the use of a male operator delivering the injections to female subjects in order to potentially maximize injection pain.

The present study had a population with a relatively low reported level of dental anxiety, with over 120 subjects (90%) reporting no to mild anxiety (the median Corah anxiety score was 5 for both interventions) prior to starting the study. These relatively low anxiety ratings are consistent with anxiety levels seen in young, female populations in previous studies.^19,20 Incorporating a population with higher dental anxiety would be an avenue for further investigations.

Most subjects (59%) selected an animated movie (The Incredibles) or an action movie (Avengers—Infinity War). VR distraction is more effective when the media being viewed is more engaging to the viewer.^27,28 Therefore, it is possible that providing a wider array of media options may increase viewer engagement, which may lead to more effective anxiolysis and analgesia during VR use.

The monitoring and comparison of cardiovascular and respiratory vital signs were not performed in this study, assessment of these variables might have proven valuable and led to further insights. Future studies should consider including the assessment of such variables.

N₂O is commonly used to decrease injection pain and increase pulpal anesthesia.^2-5 It is thought to have several different mechanisms of action, including antagonizing N-methyl-d-aspartate glutamate receptors in the central nervous system (CNS) and acting as an agonist for γ-aminobutyric acid receptors in the CNS. It is also theorized to act on adrenergic receptors in the spinal cord and to agonize opioid receptors.^5,29 Becker et al⁶ reported that N₂O is an effective analgesic and anxiolytic drug. N₂O was selected as a comparison to VR for the present study due to its ability to decrease pain and because of its relative safety.^2-6 Oral sedation with a benzodiazepine, for example, was not used in this study because the oral route does not permit titration to the desired clinical effect.

Although a third appointment for another dental injection without VR or N₂O could have been added as a control, historical controls using similar-aged female subjects have been performed without these interventions.^19,20 Adding a third appointment would have been time and cost prohibitive for the current study. Additionally, as established in the literature,^7-11 both methods have been shown to be more effective than a control. Our purpose was not to assess if either N₂O or VR was better than a control but rather to directly compare these 2 methods.

A number of studies have reported that VR was an effective method for reducing both pain and anxiety during painful dental treatment.^7-11 However, VR is likely most effective when its full immersivity can be utilized. Some of the above studies allowed subjects to interact with the VR headset, whereas others did not. This implies that results may have differed depending on whether interactivity was permitted. However, the present study could not fully utilize the VR technology due to the requirement that subjects remain somewhat motionless during the dental injection.

These previous studies on VR also used only subjective ratings for pain and anxiety, except for Nunna et al,¹⁰ who used heart rate, and Shetty et al,¹¹ who used salivary cortisol. However, heart rate is not a reliable measurement for pain and anxiety. The study by Shetty et al¹¹ did not use subjects as their own control to account for potential intersubject differences in salivary cortisol concentrations. Additionally, all the above studies included only pediatric-aged subjects, which may have resulted in higher levels of distraction and thus higher levels of reported pain and anxiety reduction. Although a study on the analgesic and anxiolytic effects of VR using adult (ie, >18 years of age) subjects reported subjective decreases in pain and anxiety for subjects using VR in comparison with no intervention,⁷ only Glennon et al³⁰ and Yamashita et al²¹ used physiologic markers for pain and anxiety, and those measurements were limited to blood pressure and heart rate. In summary, VR has been demonstrated to be subjectively beneficial in reducing pain and anxiety, but the immersive potential of VR was not always implemented, and little physiologic evidence of analgesia or anxiolysis was present in previous studies.

Limitations and Strengths

A limitation of the current study is that it evaluated subjects with low dental anxiety even though the model used a potentially painful dental injection. Studying patients in a clinical setting with higher dental anxiety and symptomatic endodontic conditions would add further comparisons. The strengths of the study were the crossover design, the recruitment of 132 subjects, and the use of salivary cortisol samples to evaluate anxiety between the 2 treatments.

CONCLUSION

The use of N₂O significantly lowered cortisol concentrations and subsequent physiologic anxiety as compared with VR use. However, no significant differences in needle insertion or solution deposition pain were found. Although VR and N₂O may be comparable in terms of perceived pain reduction, N₂O was a more effective method than VR for physiologic analgesia and anxiolysis.

Conflict of Interest

The authors deny any conflicts of interest related to this study.