The Effect of Music on Preoperative Anxiety in an Operating Room: a Single-Blind Randomized Controlled Trial

Abstract

Objective:

This study aimed to determine the effect of music as an intervention on relieving preoperative anxiety in patients with dental fear in an outpatient operating room (OR) before intravenous sedation (IVS).

Methods:

Sixty adult patients with dental fear undergoing dental surgery under IVS were divided into 2 groups (music and nonmusic). The music group listened to music in the waiting room until immediately before the initiation of IVS whereas the nonmusic group did not. Patient anxiety was objectively measured using heart rate variability (HRV) analysis to assess the low-frequency/high-frequency ratio as an indication of sympathetic or parasympathetic nervous system activity. Subjective preoperative anxiety was evaluated with a visual analog scale (VAS).

Results:

Heart rate variability analysis failed to demonstrate any significant difference between the 2 groups from baseline to start of IVS. There were also no significant differences between the 2 groups regarding changes in VAS scores.

Conclusion:

Music intervention was not found to reduce preoperative anxiety in patients with dental fear before IVS in the dental outpatient OR as determined by HRV analysis or VAS scores.

Intravenous sedation (IVS) is useful for the reduction of psychological stress that could lead to perioperative complications like vasovagal syncope, especially in patients with dental fear.^1,2 However, IVS does not reduce anxiety prior to the time it is initiated. One suggested intervention for relieving anxiety during this period is music.³ Music intervention is a psychological therapy that has many advantages when used in the outpatient setting.^3,4 As a bridging intervention, music may enable stress management upon arrival of the patient until IVS is started. However, it is difficult to ensure subject blinding when examining the effectiveness of music therapy, and many investigations utilize subjective measures that might compromise the quality of evidence.^4–6

Heart rate variability (HRV) analysis enables the intensity of sympathetic nervous system (SNS) and parasympathetic nervous system (PNS) activity to be analyzed separately,^7–12 thereby providing a more detailed picture of autonomic nervous system (ANS) activity. Heart rate variability is the millisecond order fluctuation of the interval between consecutive R waves on an electrocardiogram. The frequency domain index can be obtained as the power spectrum density by Fourier transformation of the sampled RR interval data. The frequency domain exceeding 0.15 Hz is the high-frequency (HF) component and that between 0.05 and 0.15 Hz is the low-frequency (LF) component. Positive changes in the LF/HF ratio indicate increased SNS activity whereas negative changes indicate increased PNS activity.^11,13 Emotional stress reportedly increases SNS activity as reflected by an increase in LF and a decrease in HF, and consequently an increase in LF/HF.^8–10 There are some reports recognizing the stress-relieving effect of preoperative music intervention as demonstrated via HRV analysis.^13–15

Our previous randomized controlled trial (RCT)¹⁶ revealed that music intervention was useful for relieving preoperative anxiety in patients with dental fear before entering the dental outpatient operating room (OR) as determined by HRV analysis. However, that study did not investigate the period after participants entered the OR, nor were the investigator and analyst blinded. Therefore, we performed a single-blind RCT to evaluate the effect of music intervention on preoperative anxiety for patients with dental fear upon entering the OR until initiation of IVS using HRV analysis to evaluate ANS activity and a visual analog scale (VAS) to evaluate subjective fear.

METHODS

This study included 60 participants aged 18 years or older with dental fear scheduled to undergo oral surgery or dental treatment with or without local anesthesia under IVS using propofol in an outpatient setting in the Department of Dental Anesthesiology at Hokkaido University Hospital between May 2017 and November 2018. Participants were randomly allocated to either the music or nonmusic group. Concealment of the group assignment was ensured by employing a centralized allocation of the randomization sequence. Written informed consent was obtained from all participants. The study protocol was approved by the local ethics committee (Institutional Review Board of Hokkaido University Hospital, clinical study code 016-0256) and conducted in accordance with the Helsinki Declaration guidelines. This manuscript adheres to the applicable Statement of Consolidated Standards of Reporting Trials.

To determine eligibility, candidates for participation were asked the following dental fear question the day of their preanesthetic examination: “How frightened are you of dental treatment? Please choose the option that best describes your feelings from among the 4 options mentioned here: very afraid, afraid, neutral, and not afraid.” Only patients who chose “very afraid” or “afraid” were enrolled. Additional exclusion criteria included an American Society of Anesthesiology physical status ≥III, patients with hearing disorders (ie, unsuitable for music intervention), arrhythmias, diabetes mellitus, psychiatric disorders that affect PNS activity (eg, schizophrenia or depression),^17,18 and medications that impact anxiety or ANS activity.

Study Protocol

This single-blind RCT used the protocol detailed in Figure 1. Upon arrival to the outpatient clinic, patients were instructed to lie supine on a bed in a recovery room and rest for 5 minutes, after which HRV was measured for 5 minutes and VAS scores were obtained (baseline period). Participants in the music group chose their favorite piece from 4 composed musical options to listen to via wireless earphones while lying on a stretcher in the operating waiting room for at least 10 minutes before entering the OR. The selection of music was done according to a previous study.¹⁶ In short, we composed 4 healing tunes in advance, and after listening to the first 30 seconds of each, participants selected their favorite. Sound volume was adjusted so participants were able to concurrently listen to the music and hear directions from health providers. The nonmusic group participants followed this same procedure, wearing earphones without any music to ensure blinding to the investigator responsible for data measurement and analysis. All participants were transported to the OR on a stretcher and moved to the operating chair without any postural change. Heart rate variability measurements were repeated for 5 minutes (pre-IVS period) while standard anesthesia monitors were placed, the peripheral IV was established, and another VAS score was recorded. The music/nonmusic intervention was stopped after completion of the repeat HRV measurement. All HRV measurements were performed by a single investigator.

Figure 1. .

Study protocol. (a) Changing into a surgical gown; (b) choosing a favorite from 4 prepared music pieces; (c) lying down on a stretcher; (d) moving to the operating room on the stretcher; (e) moving to an operating table without postural change; (f) “surgical timeout” to inhibit wrong side/site surgery; (g) attaching monitors and establishing intravenous access.

HRV Analysis and VAS.

The stress-relieving effect of music intervention was evaluated using an HRV monitor (Relax Meijin, Crosswell Co) in accordance with a previous study.¹⁶ The ratio of LF/HF was used as an index of increased SNS activity.^7–10,13 The coefficient of component variance for HF (CCVHF) was used as an index for HF evaluation.¹⁹ The coefficient of variation of R-R intervals (CVRR) was considered to reflect total ANS activity.¹¹

A VAS was used to evaluate subject fear²⁰ using a 100-mm line with 0 mm representing no anxiety and 100 mm as the highest level of fear one has ever experienced in life. Participants were asked to mark their appropriate level of fear on the line, which the investigator later measured.

Statistical and Power Analyses

In our previous study,¹⁶ the difference in LF/HF from baseline to waiting in a private room before entering the operating waiting room (the primary outcome) was −1.45 ± 1.88 (mean ± SD) for the music group and 1.05 ± 4.08 for the nonmusic group, producing an effect size of −2.5 and a pooled SD of 3.2. Therefore, in this study, we regarded an increase of >78% (2.5/3.2) over the difference value of 3.2 as a significant change. An a priori power analysis was performed to calculate the required sample size using an α = .05 and β = .2 by a 2-sample t test with equal variance, indicating that 27 participants were needed in each study group. The total number of participants was set at 60 to account for participant dropouts.

The primary study outcome was LF/HF and secondary outcomes were CCVHF, CVRR, heart rate, and VAS scores. The mean value of HRV analysis for each period was regarded as the representative value for that period. The difference between the representative values for the pre-IVS and baseline periods was used to make intergroup comparisons. Specifically, the difference between pre-IVS period and baseline period was compared as Δ(pre-IVS period − baseline period).

JMP Pro14 (SAS Institute Inc) was used for statistical analysis. The Shapiro-Wilk W test was used for the evaluation of normal distribution. An unpaired Student t test was used for differences in patients' age, height, weight, and body mass index, and the χ² test was used for sex differences and degree of fear. The Mann-Whitney U test and unpaired Student t test were used for intergroup comparisons of primary and secondary outcomes. The Wilcoxon signed-rank test and paired Student t test were used for intragroup comparisons. Level of significance was set at P < .05. Data values are presented as mean (SD), median (25th to 75th percentile), or No. (%).

RESULTS

A flowchart illustrating the data for participant enrollment, allocation, and analysis is detailed in Figure 2. Of the 28 participants whose data were analyzed in the music group, 6 responded “very afraid” and 22 “afraid.” Of the 27 participants in the nonmusic group, 8 responded “very afraid” and 19 “afraid.” There were no significant differences in the initial self-reported degree of fear between the 2 study groups (P = .48). The participants' demographic characteristics are presented in Table 1. There were no significant differences noted between the 2 groups.

Figure 2. .

Flowchart of enrollment, allocation, and data analysis. ¹ Change to another sedative because of not following preoperative fasting period; ² intravenous (IV) access failure (twice; n = 1); ³ IV access failure (twice; n = 1) and trouble with heart rate variability (HRV) measurement equipment (n = 2).

Table 1. .

Participant Demographics

Characteristics	Music Group (n = 28)	Nonmusic Group (n = 27)	P Value
Sex, No. (%)			.67
Men	3 (10.7)	2 (7.4)
Women	25 (89.3)	25 (92.6)
Age, mean (SD), y	37.6 (14.5)	36.4 (12.1)	.75
Height, mean (SD), m	1.61 (0.06)	1.60 (0.06)	.37
Weight, mean (SD), kg	57.3 (9.9)	54.9 (10.1)	.38
Body mass index, mean (SD), kg/m2	22.1 (3.5)	21.5 (3.0)	.49

Intergroup Comparisons

Although HR increased from baseline to pre-IVS in both groups, there was a significantly higher increase in the music group compared to the nonmusic group (5.47 vs 1.36 bpm; P < .001). Differences between the 2 groups for the HRV parameters or VAS scores all lacked statistical significance (Table 2).

Table 2.

Intergroup Comparisons Between Preintervention and Postintervention for Primary and Secondary Outcomes*

Variable	Δ (Pre-IVS Period − Baseline Period)
	Music Group (n = 28)	Nonmusic Group (n = 27)	P Value
LF/HF, median (25th to 75th percentile)	1.24 (−0.3 to 2.59)	0.66 (−0.39 to 1.70)	.36
CCVHF, mean (SD)	0.09 (0.72)	0.04 (0.48)	.79
CVRR, median (25th to 75th percentile)	0.50 (−0.14 to 1.49)	0.58 (−0.04 to 1.49)	.92
HR, median (25th to 75th percentile), bpm	5.47 (2.46 to 9.36)	1.36 (−1.00 to 4.61)	<.001
VAS, mean (SD), mm	6.54 (21.7)	13.56 (15.85)	.18

^*CCVHF indicates coefficient of component variance for high frequency; CVRR, coefficient of variation of R-R intervals; HR, heart rate; IVS, intravenous sedation; LF/HF, low-frequency component/high-frequency component; and VAS, visual analog scale.

Intragroup Comparisons

Comparing intragroup data from baseline to the pre-IVS period, both groups had significantly increased LF/HF and CVRR values, whereas changes in CCVHF were insignificant (Table 3). The HR from baseline to pre-IVS increased in both groups, although only the increase in the music group (+6.8 bpm vs +1.4 bpm) was significant (P < .001). Although VAS scores increased for both groups, only the nonmusic group had a significant change from baseline (+13.6 mm; P < .001).

Table 3.

Intragroup Comparisons of Post–Music Intervention Changes for Primary and Secondary Outcomes*

Variable	Music Group (n = 28)			Nonmusic Group (n = 27)
	Baseline	Pre-IVS Period	P Value	Baseline	Pre-IVS Period	P Value
LF/HF	2.14 (1.37–4.56)	3.39 (2.63–4.40)	.009	1.62 (1.22–3.36)	2.35 (1.42–3.71)	.026
CCVHF	1.70 (1.09–2.24)	1.64 (1.08–2.73)	.55	1.75 (0.70)	1.79 (0.68)	.67
CVRR	3.83 (2.93–5.26)	4.41 (3.23–6.76)	.001	3.90 (1.26)	4.62 (1.42)	.007
HR, bpm	70.2 (11.0)	77.0 (12.5)	<.001	66.9 (8.9)	68.3 (10.1)	.25
VAS, mm	65 (31.5–74)	66.5 (38–86.75)	.17	58.1 (22.0)	71.7 (16.6)	<.001

^*CCVHF indicates coefficient of component variance for high frequency; CVRR, coefficient of variation of R-R intervals; HR, heart rate; IVS, intravenous sedation; LF/HF, low-frequency component/high-frequency component; and VAS, visual analog scale.

Neither hyperventilation nor hypoventilation was observed while measuring HRV in any of the participants. Additionally, there were no adverse physical effects derived from music intervention noted with any of the participants.

DISCUSSION

With intragroup comparison, significant increases in the LF/HF ratio and CVRR were observed in both groups in this study. That is, the increased SNS activity of patients with dental fear was recognized, followed by an increase in ANS activity upon entering the OR irrespective of music intervention. There were no significant intergroup differences in Δ values (pre-IVS period − baseline period) of the LF/HF ratio and CVRR between the 2 groups. Music intervention did not contribute to relieving anxiety in patients with dental fear before IVS in the OR. Subjective fear (VAS scores) data also failed to demonstrate a significant difference between the music and nonmusic groups.

We selected 5 studies from an exhaustive list of similar RCTs in the literature that investigated the use of music to relieve preoperative anxiety while patients were conscious^{3,14,16,21,22} and compared them with this study. Notably, only Chiu et al¹⁴ and this study considered preoperative anxiety after entering an OR. Both studies evaluated the effect of music intervention via HRV, which is applied in clinical settings because it is a simple, noninvasive, and objective indicator for assessing mental stress.^8–10 Chiu et al¹⁴ found that the HF component increased significantly and LF/HF decreased significantly with music, while not found in this study. This difference between the 2 studies may derived from the differences in operating circumstances, comparison method (intra- or intercomparisons), and the abscence or presence of participant's surgical anxiety, rendering direct comparisons difficult.

With respect to where the effect of music intervention was evaluated, this study included measurements and interventions occurring inside an OR, whereas an operating waiting room was used in our previous study.¹⁶ The effect of music on reducing preoperative anxiety in the OR might be inadequate because of the patient being in an environment with unfamiliar visual and auditory stimuli (eg, OR equipment, anesthesia monitor sounds), especially for patients with substantial dental fear.^23,24 Moreover, participants had to respond to questions from medical personnel (the “surgical timeout”) and hear instructions about the operation. All these factors might have prevented them from concentrating on listening to the music.

Focusing on the connection between the effect of music and evaluation using subjective and/or objective parameters, 4 studies^3,16,21,22 found that music intervention was effective in reducing anxiety according to subjective parameters. However, as participants cannot be blinded to a music intervention, a decrease in the quality of evidence cannot be avoided.^4–6 In terms of objective parameters, an effect of music was found in 2 studies.^14,16 The investigator and analyst were blinded only in the study by Jeppesen et al²² and in this study. In both studies, an anxiety-relieving effect of music intervention was not demonstrated by objective parameters. It is likely that a single-blind RCT on the stress-relieving effect of music using objective parameters will require accumulated knowledge derived from high-evidence papers hereafter.

With respect to the time spent listening to music, the music intervention lasted 15 minutes or longer in this study, although the mean length of the music intervention was approximately 30 minutes in our previous study.¹⁶ The rationale for using 15 minutes in this study was to minimize the impact on fearful nonmusic participants and avoid prolong waiting time until starting IVS. Furthermore, a recent study²⁵ reported that there was no significant difference between 15 and 30 minutes of music intervention for stress relief.

In the music group, the increase in HR was significantly higher as compared with the nonmusic group. Previous studies^26,27 found that music intervention decreased heart rate and interpreted that to be an expected finding consistent with stress relief. However, to date there are no other reports of healing music intervention increasing heart rate as found here. This study was designed as an RCT to avoid bias. Moreover, none of the corresponding confounding factors described in biomedical research²⁸ were present in this study. Therefore, the reason for this finding is unclear.

Music intervention on its own was not found to reduce anxiety in patients with dental fear before initiating IVS in the dental outpatient OR. The effect of shortening the duration from entering the OR until starting IVS should be further investigated. Lehrner et al²⁹ reported that exposure to ambient odor of orange in the dental waiting room has a relaxant effect. Further investigations on the effect of relieving preoperative stress with a combination of music and odor interventions may be indicated.

This study has several limitations. First, as for the location where baseline measurements were performed, a quiet private room separated from the operating waiting room was used in our previous study,¹⁶ whereas a recovery room near the operating waiting room was used in this study. This occurred as we were unable to obtain another quiet private room because of relocation of our outpatient facility while this study was being conducted. Second, prolonged waiting time until starting IVS in the OR is a significant problem for apprehensive patients. Heart rate variability analysis can be affected by postural change.¹⁹ As per our preliminary data, the influence of postural change on HRV lasts for 2 minutes. Therefore, the protocol mandated that participants enter the OR in the supine position so that we could start measuring HRV immediately upon entry. However, the duration of 1 to 2 minutes, which was the HRV measurement time in the OR (5 minutes) minus the time required for monitoring preparation and venous access, was considered as extra waiting time caused by participation in this study. Third, the timeout system that interrupted the patients' concentration on listening to music was beyond the scope of assumptions when planning this study protocol.

CONCLUSION

This experimental research showed that music intervention failed to reduce preoperative anxiety in patients with dental fear in the dental outpatient OR before IVS as determined by ANS activity using HRV analysis and subjective VAS scores.