Patient-selected music reduces propofol requirements during combined upper gastrointestinal endoscopy and colonoscopy: A prospective single-center study (MUSICENDO)

Abstract

BACKGROUND

Gastrointestinal (GI) endoscopy is frequently performed under propofol sedation. Music is a lowrisk sensory intervention that may reduce anxiety and sedative requirements; however, its objectively measured sedativesparing effects during combined gastroscopy and colonoscopy remain uncertain.

AIM

To compare propofol consumption rates (mg/kg/minute) between patients receiving patient-selected music and those receiving standard care alone.

METHODS

An openlabel, singlecenter, parallelgroup randomized controlled trial involving adult outpatients scheduled for combined diagnostic gastroscopy and colonoscopy was conducted. Participants were randomized 1:1 to receive either patientselected music, initiated at least 10 minutes before anesthesia induction and continued throughout the procedure, in addition to standard care, or standard care alone. Propofol was the sole hypnotic agent via targetcontrolled infusion using the Schnider model. Among 89 randomized patients, 88 were included in the modified-intentiontotreat analysis and 82 in the perprotocol analysis.

RESULTS

The mean propofol consumption rate was significantly lower in the music group than in the standard-care group (0.18 ± 0.05 mg/kg/minute vs 0.26 ± 0.07 mg/kg/minute in the modified-intentiontotreat analysis; P < 0.0001), with consistent findings in the per-protocol analysis (0.19 ± 0.05 mg/kg/minute vs 0.26 ± 0.07 mg/kg/minute; P = 0.0009). The music group also received a lower total propofol dose (258.24 ± 57.52 mg vs 344.88 ± 86.86 mg; P = 0.0009) and required fewer propofol vials per procedure (0.52 ± 0.12 vs 0.69 ± 0.17; P = 0.0009). No between-group differences were observed in examination duration, total ambulatory stay, hemodynamic parameters, or preprocedural anxiety scores. No unplanned hospital admissions occurred.

CONCLUSION

Patientselected music administered before and during combined GI endoscopy significantly reduced propofol consumption under a standardized sedation protocol. These findings warrant further investigation of music as a simple, non-pharmacological adjunct to procedural sedation in larger, multicenter randomized controlled trials.

Key Words: Gastrointestinal endoscopy; Colonoscopy; Propofol; Target-controlled infusion; Music; Ambulatory care

Core Tip: This trial investigated whether patient-selected music could reduce propofol requirements during combined gastroscopy and colonoscopy under target-controlled infusion. Compared with standard care alone, patient-selected music substantially reduced propofol consumption without increasing procedure duration or delaying discharge. No significant between-group differences were observed in hemodynamic parameters or adverse-event rates. The reduction in propofol requirements was also reflected in a practical measure of resource utilization, with fewer propofol vials required per procedure in the music group.

INTRODUCTION

Gastrointestinal (GI) endoscopy is among the most frequently performed medical procedures[1]. Many of these procedures are performed in ambulatory settings, with patient eligibility determined by individual characteristics and periprocedural risk. Optimizing peri-procedural care is therefore essential to ensuring patient safety, improving procedural efficiency, and enhancing the patient experience in ambulatory settings[2]. Pre-procedural anxiety is common, often arising from concerns about anesthesia, pain, and potential adverse outcomes. When inadequately addressed, it may increase anesthetic and analgesic requirements[3-6]. Contemporary perioperative optimization programs therefore emphasize both organizational measures and practical non-pharmacological interventions that facilitate early recovery and support efficient ambulatory care[7,8].

Music listening is a low-risk, patient-centered sensory intervention that can be readily integrated into clinical workflows with minimal disruption. In adult surgical settings, randomized trials and meta-analyses have demonstrated that perioperative music reduces anxiety and improves pain-related outcomes compared with usual care or silence[9-11]. In digestive endoscopy, music and related sensory interventions have been associated with improved patient experience, reduced anxiety, and lower pain scores in several studies[12-15]. Importantly, several trials involving GI endoscopy have reported reduced sedative requirements when music is played during the procedure[16,17].

Sedation practices in endoscopy remain heterogeneous, and combined gastroscopy and colonoscopy (“double indication”) may require deeper or more prolonged sedation than either procedure performed alone. Most prior endoscopy studies have focused on patient-reported outcomes, with objective sedative consumption assessed as the primary endpoint less frequently, particularly in studies using standardized hypnotic protocols across GI endoscopies procedures[16-18]. Given that anxiolysis and attentional modulation may mediate the effects of music, initiating patient-selected music before anesthesia induction and continuing it throughout the procedure could enhance its potential sedative-sparing effects.

We therefore conducted a randomized controlled trial in adult outpatients undergoing combined diagnostic gastroscopy and colonoscopy to determine whether patient-selected music, added to standard care, reduces propofol requirements without compromising procedural efficiency or patient safety. The primary endpoint was the propofol consumption rate (mg/kg/minute); secondary endpoints included examination duration, total ambulatory stay, hemodynamic parameters, anxiety levels, serious adverse events (SAEs), and practical measures of propofol utilization.

MATERIALS AND METHODS

Study design and oversight

This single-center, parallel-group randomized controlled trial was conducted at Princess Grace Hospital, Monaco. Consecutive adult outpatients scheduled for GI endoscopic procedures were screened over a 5-month period (20 March-16 July 2024). The study was reported in accordance with the CONSORT guidelines; the flow diagram is presented in Figure 1 and corresponding checklist is included in the Supplementary materials. All participants provided written informed consent. The study received no external funding.

Figure 1 CONSORT flow diagram of participant screening and analysis populations. ITT: Intention-to-treat; PP: Pulse pressure; TCI: Target-controlled infusion.

Participants

Eligible participants were adults (≥ 18 years) with American Society of Anesthesiologists physical status I-III, scheduled for GI endoscopic procedures in an ambulatory setting. Exclusion criteria included legal incapacity; a planned hospital stay exceeding 24 hours; a psychiatric disorder; alcohol or other substance use; severe renal impairment (creatinine clearance < 30 mL/minute); pregnancy or breastfeeding; hearing impairment severe enough to prevent music exposure; urgent procedures; and lack of social insurance. For patients undergoing multiple eligible procedures during the study period, only the first procedure was included in the analysis.

Randomization and masking

Participants were randomized in a 1:1 ratio to receive either patient-selected music plus standard care or standard care alone. Randomization was performed using a computer-generated schedule with fixed permuted blocks, prepared by an independent statistician. Allocation concealment was ensured through sequentially numbered, opaque, sealed envelopes, which were opened after enrollment and baseline assessment. Because the intervention was audible, participants, clinicians, and the endoscopy team could not be blinded. Group allocation was disclosed during the pre-anesthesia consultation to enable patients assigned to the music group to select their preferred music.

Interventions

Patients assigned to the music group selected their preferred music during the pre-anesthesia consultation, several days before the procedure, and confirmed their selections upon arrival at the ambulatory unit. They could select from predefined playlists (jazz, classical, rock, or relaxing music) or provide an indication of their preferred music to be played. No particular musical style was recommended, and selection was based entirely on individual patient preference. Music was initiated by nursing staff upon the patient’s arrival at the ambulatory surgery unit and continued during transfer to the endoscopy room. For all patients, the audio device was placed in a dedicated holder attached to the upper right side of the stretcher, following a standardized protocol. Headphones were not used because of patient positioning during gastroscopy.

Music was played continuously from admission to the ambulatory surgery unit until the patient’s return to the unit following the endoscopic procedure. Before sedation, the volume was adjusted to the patient’s preference and subsequently maintained at a comfortable level, without interfering with routine patient care or communication among healthcare professionals. Music playback was uninterrupted throughout the study, ensuring that all patients in the intervention group received continuous exposure for the entire planned intervention period. In the control group, patients received standard care without music intervention, and ambient noise was not modified.

The sedation strategy and predefined clinical sedation targets were standardized prospectively across both groups. Propofol was used as the sole hypnotic agent, and no sedative premedication was administered. Propofol was administered via a Fresenius Kabi Base Primea Orchestra target-controlled infusion (TCI) pump (Fresenius Kabi AG, Bad Homburg, Germany) using the Schneider pharmacokinetic model. Propofol 1% (10 mg/mL) was supplied in 50-mL vials. The initial effect-site target concentration was set at 6.5 μg/mL and subsequently adjusted according to the patient’s clinical response, as assessed using the Ramsay Sedation Scale. A Ramsay score of 6 was targeted during gastroscopy to account for the greater reflexogenic stimulation associated with upper GI endoscopy, whereas a Ramsay score of 5 was targeted during colonoscopy. Following gastroscopy, the propofol target concentration was progressively reduced and titrated to achieve the predefined Ramsay sedation target for colonoscopy. The same sedation targets and titration strategy were applied in both randomized groups. Neither a maximum target concentration nor a standardized rescue-dose protocol was predefined. Instead, the anesthesiologist continuously adjusted propofol administration according to the specific sedation targets and each patient’s response. The Ramsay Sedation Scale was routinely used to guide propofol titration; however, individual scores were not prospectively documented in the case report form.

Processed electroencephalogram (EEG) monitoring was performed using a Masimo Root® monitor (model RDS7A; Masimo Corp., Irvine, CA, United States) with the SedLine Brain function activated to measure the Patient State index (PSi). PSi was used as a supplementary measure of sedation depth rather than the primary parameter guiding propofol titration. A PSi range of 25-50 served as a reference for identifying potentially excessive or inadequate sedation, while clinical assessment using the Ramsay Sedation Scale remained the primary basis for dose adjustment. PSi values were recorded at 5-minute intervals throughout the procedure. At the end of the examination, the mean PSi was calculated from all recorded values, documented in the case report form and included in the analysis.

All participants underwent the same combination of procedures: Gastroscopy followed by colonoscopy. Endoscopists were assigned according to routine clinical practice, independently of treatment allocation, resulting in their random distribution across the two study groups.

Anesthesia and monitoring

Propofol was used as the sole hypnotic agent in both groups and was administered via effect-site TCI using the Schnider pharmacokinetic model[19]. The attending anesthesiologist adjusted the propofol effect-site target to achieve sufficient sedation for endoscopy while maintaining spontaneous ventilation. Supplemental oxygen was routinely administered via nasal cannula. Standard monitoring included continuous electrocardiography, noninvasive blood pressure measurements at 3-5 minutes intervals, pulse oximetry, and respiratory rate. No sedative premedication was administered. Pre-procedure anxiety was assessed using the Amsterdam Preoperative Anxiety and Information Scale (APAIS) during the pre-anesthesia consultation, upon arrival at the ambulatory unit, and immediately before the procedure[20,21].

Gastroscopy and colonoscopy

The indication for combined gastroscopy and colonoscopy was established during the initial gastroenterology consultation, after which pre-anesthesia consultation and endoscopic procedures were scheduled. All combined upper GI endoscopy and colonoscopy procedures were performed for diagnostic or screening purposes.

On the day of the procedure, the indications for both gastroscopy and colonoscopy were reconfirmed before anesthesia induction. A standardized preprocedural checklist, based on the recommendations of the French Society of Digestive Endoscopy, was completed. This included verification of bowel preparation on the preceding day, fasting status, anticoagulant management, allergies, removal of contact lens and removable dental appliances, suitability for ambulatory care (including the availability of an accompanying person), and readiness of the equipment and clinical team. Anesthesia and endoscopy proceeded only after all checklist requirements had been satisfied.

The bite block was inserted immediately before the initiation of propofol administration began. Gastroscopy was performed first and commenced only after the predefined clinical sedation target had been achieved. The examination included the recommended photographic documentation and biopsies when indicated.

Colonoscopy commenced immediately after completion of gastroscopy, with the sedation target adjusted accordingly. The examination included recommended photographic documentation and, when indicated, biopsies, polypectomies, or endoscopic mucosal resections compatible with ambulatory management.

Propofol administration was discontinued upon completion of the colonoscopy. Once sufficiently awake and recovered, each patient received both verbal and written endoscopy results from the endoscopist in the ambulatory unit before discharge.

Examination duration was defined as the interval from the initiation of anesthesia induction to completion of the colonoscopy. Total ambulatory stay was defined as the interval from the initiation of anesthesia induction to actual discharge from the ambulatory surgery unit.

Patients were discharged in accordance with the institutional ambulatory surgery protocol once they achieved a modified Post-Anesthetic Discharge Scoring System (PADSS) score of ≥ 9. Discharge criteria included stable vital signs, adequate spontaneous ventilation, satisfactory pain and nausea control, the ability to ambulate when appropriate, and no evidence of clinically significant bleeding. PADSS assessments were performed by ambulatory care nurses as part of routine clinical practice, rather than at prespecified study intervals. Consequently, the earliest time at which each patient met the discharge criteria was not prospectively determined. Accordingly, the study reports the time to actual discharge rather than the time to discharge readiness.

Outcomes

The primary endpoint was the propofol consumption rate (mg/kg/minute), calculated by dividing the total administered propofol dose (mg) by the patient’s body weight (kg) and by the duration of sedation (minutes from anesthesia induction to completion of endoscopy). Propofol dosage data were obtained directly from the infusion device records.

Secondary endpoints included: (1) Total propofol dose (mg); (2) Examination duration (minutes from anesthesia induction to completion of endoscopy); (3) Total ambulatory stay (minutes from anesthesia induction to actual discharge, with discharge assessed using the modified Aldrete score according to local practice); (4) Anxiety (assessed using APAIS scores); (5) Hemodynamic parameters, including systolic and diastolic blood pressures and pulse pressure, measured immediately before induction and at the end of the procedure, together with changes between these time points; (6) Serious peri-procedural adverse events; and (7) The number of propofol vials used per procedure.

Safety assessment

Safety data collection was performed according to International Council for Harmonization of Good Clinical Practice, in order to identify SAEs. Additionally, we used the Common Terminology Criteria for Adverse Events (CTCAE) version 5.0 to grade the SAEs. The assessment focused on clinically significant respiratory adverse events occurring during anesthesia and endoscopy, including bronchospasm and laryngospasm requiring therapeutic intervention. Heart rate, (noninvasive) blood pressure, and oxygen saturation were continuously or periodically monitored, as appropriate, throughout anesthesia in accordance with routine clinical practice. Minor, transient physiological fluctuations, were not prospectively recorded as adverse events. For respiratory adverse events, the clinical presentation and required management (due to life-threatening status) were documented.

Samplesize calculation

The sample size was calculated based on the primary outcome. Assuming a mean propofol consumption rate of 0.185 mg/kg/minute in the standard-care group (SD: 0.025) and an anticipated reduction to 0.170 mg/kg/minute in the music group, a total of 44 participants per group was estimated to provide 80% power at a two-sided significance level (α) of 0.05[22]. Accordingly, the planned enrollment was 88 participants.

Statistical analysis

Statistical analyses were performed using R (version 4.4.1; R Foundation for Statistical Computing, Vienna, Austria)[23]. Continuous variables are expressed as mean ± SD or median (interquartile range), as appropriate, and categorical variables are presented as n (%). The primary efficacy analysis was prespecified for the intention-to-treat population, comprising randomized participants who received their allocated intervention, underwent protocol-specific propofol TCI, and had available primary-outcome data. A per-protocol sensitivity analysis was also conducted. Of the 89 randomized participants, 1 withdrew consent after randomization and did not authorize the use of their data. Consequently, 88 participants were included in the modified-intention-to-treat analysis of the primary outcome, as a full intention-to-treat analysis was not feasible. The per-protocol analysis included 82 participants. Six additional participants were excluded from this population because of protocol deviations or missing secondary-outcome data: 3 did not receive the allocated music intervention as planned, 1 required tracheal intubation following laryngospasm, 1 did not receive propofol via TCI, and 1 had missing secondary-outcome data. The participant exclusions are summarized in Figure 1. The primary outcome was compared between groups using an independent-samples t-test, with the treatment effect expressed as the mean difference and its 95%CI. A prespecified adjusted analysis was performed using linear regression with treatment group, age, sex, and body mass index (BMI) as covariates. An additional post hoc analysis incorporated intervention duration into the prespecified regression model. Categorical outcomes were compared using Fisher’s exact test. Hochberg’s step-up procedure was used to control the family-wise error rate across secondary outcomes[24]. Two-sided P values below 0.05 were considered statistically significant. No imputation was performed.

RESULTS

Participants

Between 20 March and 16 July 2024, 434 patients scheduled for endoscopic procedures identified by the French Common Classification of Medical Procedures (Classification Commune des Actes Médicaux) codes HEQE002 (gastroscopy) and HHQE002 or HHQE005 (colonoscopy) were assessed for eligibility, and 89 participants were randomized[25]. One participant withdrew consent after randomization and did not authorize the use of their data, leaving 88 participants in the modified intention-to-treat population for analysis of the primary outcome. The per-protocol analysis included 82 participants after 6 additional patients were excluded because they did not meet the prespecified criteria, as detailed in Figure 1. Baseline demographic and clinical characteristics, including APAIS anxiety scores and pre-induction blood pressure measurements, were comparable between groups (Table 1).

Table 1 Baseline characteristics of participants in a modified intention-to-treat population, n (%).

Characteristic	n	Music protocol (n = 44)	Standard care (n = 44)
Sex	82
Female		21 (51.2)	21 (51.2)
Male		20 (48.8)	20 (48.8)
Missing data	6	3	3
Age (years)	82
mean ± SD		58.90 ± 10.77	59.56 ± 11.77
Median (interquartile range)		60.00 (51.00-66.00)	59.00 (52.00-67.00)
Missing data	6	3	3
BMI (kg/m2)	82
mean ± SD		24.44 ± 3.13	25.43 ± 3.38
Median (interquartile range)		24.61 (22.53-26.56)	25.10 (23.03-27.22)
Missing data	6	3	3
ASA score	82
1		22 (53.7)	12 (29.3)
2		15 (36.6)	22 (53.7)
3		4 (9.8)	7 (17.1)
Missing data	6	3	3
APAIS score at arrival in the ambulatory unit	82
mean ± SD		9.68 ± 4.25	9.44 ± 4.75
Median (interquartile range)		8.00 (6.00-12.00)	7.00 (6.00-11.00)
Missing data	6	3	3
APAIS score before intervention (at arrival in the endoscopy room)	80
mean ± SD		9.20 ± 4.33	9.58 ± 4.59
Median (interquartile range)		7.00 (6.00-10.50)	8.00 (6.00-12.00)
Missing data	8	4	4
Before intervention SBP	79
mean ± SD		125.01 ± 21.55	130.76 ± 21.03
Median (interquartile range)		124.00 (109.75-137.25)	129.00 (116.50-142.50)
Missing data	9	4	5
Before intervention DBP	79
mean ± SD		79.53 ± 16.22	76.64 ± 13.26
Median (interquartile range)		75.25 (69.25-87.00)	75.00 (67.50-85.00)
Missing data	9	4	5
Before intervention PP	80
mean ± SD		72.89 ± 14.85	72.32 ± 12.08
Median (interquartile range)		72.00 (62.50-80.00)	73.50 (63.00-78.00)
Missing data	8	3	5

APAIS: Amsterdam Preoperative Anxiety and Information Scale; ASA: American Society of Anesthesiologists; BMI: Body mass index; DBP: Diastolic blood pressure; PP: Pulse pressure; SBP: Systolic blood pressure.

Primary endpoint

In the modified-intention-to-treat population (n = 88; 44 per group), the mean propofol consumption rate was significantly lower in the music group than in the standard-care group (0.18 ± 0.05 mg/kg/minute vs 0.26 ± 0.07 mg/kg/minute, respectively; mean difference: -0.08 mg/kg/minute; 95%CI: -0.098 to -0.045; P < 0.0001), representing a 30.8% relative reduction in propofol consumption. The standardized effect size (Hedges’ g) was -3.55. The per-protocol sensitivity analysis produced consistent findings (n = 82; 41 per group), with mean propofol consumption rates of 0.19 ± 0.05 mg/kg/minute in the music group and 0.26 ± 0.07 mg/kg/minute in the standard-care group (unadjusted P < 0.0001; P = 0.0009 after Hochberg correction).

Secondary endpoints

Total propofol dose was significantly lower in the music group than in the standard-care group (258.24 ± 57.52 mg vs 344.88 ± 86.86 mg; mean difference -86.64 mg; 95%CI: -119.09 to -54.18; P < 0.0001; P = 0.0009 after Hochberg correction). Neither examination duration nor total ambulatory stay, measured from anesthesia induction to actual discharge, differed significantly between groups (Table 2). Postprocedural hemodynamic parameters and changes from baseline were also comparable (Table 2), as were APAIS anxiety scores upon arrival and immediately before the procedure (Table 1).

Table 2 Primary and secondary outcomes, n (%).

Characteristic	n	Music protocol (n = 44)	Standard care (n = 44)	Absolute difference (95%CI)	P value	P adjusted value
Primary outcome-modified intention-to-treat population
Propofol administered (mg/kg/minute)	88			-0.08 (-0.10 to -0.05)	< 0.0001	NA
mean ± SD		0.18 ± 0.05	0.26 ± 0.07
Median (interquartile range)		0.18 (0.16-0.20)	0.26 (0.21-0.32)
Secondary outcomes-modified intention-to-treat population
Serious adverse event				NA	NA	NA
Bronchospasm		0 (0.0)	2 (4.5)
Laryngospasm and intubation		1 (2.3)	0 (0.0)
Primary outcome – sensitivity analysis – per-protocol population
Propofol administered (mg/kg/minute)	82			-0.07 (-0.10 to -0.05)	< 0.0001	0.0009
mean ± SD		0.19 ± 0.05	0.26 ± 0.07
Median (interquartile range)		0.18 (0.17-0.20)	0.26 (0.21-0.32)
Secondary outcomes-per-protocol population
Examination duration (minutes)	82			1.24 (-1.09 to 3.58)	0.29	0.89
mean ± SD		20.44 ± 4.76	19.20 ± 5.82
Median (interquartile range)		19.00 (17.00-22.00)	19.00 (15.00-21.00)
Total ambulatory stay (minutes)	69			-0.44 (-7.00 to 6.12)	0.89	0.89
mean ± SD		82.22 ± 14.71	82.66 ± 12.59
Median (interquartile range)		81.00 (73.00-94.00)	80.50 (74.50-90.00)
Propofol administered (mg)	82			-86.64 (- 119.09 to -54.18)	< 0.0001	0.0009
mean ± SD		258.24 ± 57.52	344.88 ± 86.86
Median (interquartile range)		250.00 (216.00-283.00)	327.00 (297.00-375.00)
Propofol administered as vials	82			-0.17 (-0.24 to -0.11)	< 0.0001	0.0009
mean ± SD		0.52 ± 0.12	0.69 ± 0.17
Median (interquartile range)		0.50 (0.43-0.57)	0.65 (0.59-0.75)
PSi	81			3.50 (1.98 to 5.00)	< 0.0001	0.0009
mean ± SD		35.23 ± 3.43	31.73 ± 3.39
Median (interquartile range)		35.00 (34.00-37.50)	32.00 (30.00-34.00)
Difference between after and before intervention in SBP	78			-4.05 (-12.77 to 4.67)	0.36	0.89
mean ± SD		11.86 ± 19.18	15.91 ± 19.48
Median (interquartile range)		12.00 (-3.75-23.50)	13.25 (4.00-34.00)
Difference between after and before intervention in DBP	78			4.08 (-2.94 to 11.09)	0.25	0.89
mean ± SD		8.90 ± 15.51	4.82 ± 15.59
Median (interquartile range)		6.50 (-1.50-15.75)	5.00 (-4.50-14.00)
Difference between after and before intervention in PP	76			2.05 (-4.15 to 8.26)	0.51	0.89
mean ± SD		7.21 ± 14.32	5.16 ± 12.81
Median (interquartile range)		5.75 (1.84-15.25)	5.00 (-2.84-12.25)

DBP: Diastolic blood pressure; PP: Pulse pressure; PSi: Patient State index; SBP: Systolic blood pressure.

Processed electroencephalographic (commonly referred to as EEG) monitoring data were available for 81 participants. Mean PSi values were significantly higher in the music group than in the standard-care group (35.23 ± 3.43 vs 31.73 ± 3.39; mean difference: 3.5, 95%CI: 1.98 to 5.00; P < 0.0001; P = 0.0009 after Hochberg correction). However, mean values in both groups remained within the prespecified target range for deep sedation.

Three SAEs were recorded. Two episodes of bronchospasm occurred in the control group and were successfully managed using standard anesthetic measures, including deepening of anesthesia and supplemental oxygen. Both episodes resolved promptly without requiring tracheal intubation, allowing gastroscopy and subsequent colonoscopy to be completed. Both events were classified as grade 3 according to CTCAE version 5.0. One episode of laryngospasm occurred in the music group and was associated with oxygen desaturation. Tracheal intubation was required after initial attempts to manage the event by deepening anesthesia. The endoscopic procedure was subsequently completed, and the event did not prolong the patient’s ambulatory stay. The laryngospasm was classified as CTCAE grade 4. Given the small number of events, no formal statistical comparison between groups was undertaken. No unplanned hospital admissions occurred.

The number of propofol vials used per procedure was also significantly lower in the music group (0.52 ± 0.12 vs 0.69 ± 0.17 vials; mean difference: -0.17; 95%CI: -0.24 to -0.11; P < 0.0001; P = 0.0009 after Hochberg correction). No temporal trend in propofol consumption was observed over the recruitment period (Figure 2).

Figure 2 Propofol consumption rate by recruitment order, stratified by group (music vs standard care). No temporal trend suggestive of a learning effect was observed. PP: Pulse pressure.

Linear model

In the prespecified multivariable linear regression model, assignment to standard care was independently associated with a higher propofol consumption rate (P < 0.0001; P = 0.0009 after Hochberg correction). Higher BMI was associated with a lower propofol consumption rate (P = 0.0008; P = 0.0064 after Hochberg correction). In the post hoc model, shorter intervention duration was also associated with a higher propofol consumption rate (P < 0.0001; P = 0.0009 after Hochberg correction). Although age was associated with propofol consumption rate before adjustment for multiple comparisons (P = 0.02), this association was no longer statistically significant after Hochberg correction (P = 0.14) (Table 3).

Table 3 Multivariable linear regression model for propofol consumption rate (mg/kg/minute) in per-protocol population.

Characteristic	Propofol dose/mg/kg/min administered, n = 82			t value	P value	P adjusted (Hochberg)
	Estimate	95%CI	Standard error
Music protocol	0
Standard care	0.07	0.05 to 0.1	0.01	7.29	< 0.0001	0.0009
Female	0
Male	-0.02	-0.04 to 0.00007	0.01	-1.99	0.05	0.3
Age in years	-0.001	-0.002 to -0.0002	0.0005	-2.44	0.02	0.14
BMI	-0.006	-0.009 to -0.002	0.002	-3.48	0.0008	0.0064
Duration of intervention in minutes	-0.006	-0.008 to -0.003	0.001	-5.51	< 0.0001	0.0009

BMI: Body mass index.

DISCUSSION

In this prospective single-center study of adults undergoing combined diagnostic gastroscopy and colonoscopy under propofol TCI, patient-selected music initiated before anesthesia induction and continued throughout the procedure was associated with an approximately 31% reduction in propofol consumption rate. This reduction was not accompanied by longer procedure duration, delayed actual discharge, or detectable between-group differences in hemodynamic parameters or SAEs.

These findings build on previous randomized and observational studies of music during GI endoscopy, in which music has been associated with improved patient experience and, in some reports, reduced sedative requirements[10,11,16,17,26-29]. Spagnuolo et al[30] reported that music during digestive endoscopy was associated with lower pain and anxiety levels, particularly among patients undergoing conscious sedation, supporting its potential role as a simple non-pharmacological adjunct to improve procedural comfort. More recently, Bozkurt and Kocasli[31] found that pain and anxiety decreased after endoscopy in both the music and control groups, without a significant between-group difference, and observed no significant effect of music on vital signs. Together, these studies highlight the heterogeneity of the available evidence for subjective outcomes such as pain and anxiety, whereas the absence of a consistent effect on vital signs is consistent with our findings. A recent scoping review also reported potential beneficial effects of music interventions on pain, anxiety, and patient satisfaction during GI endoscopy[32]. The present study extends this literature by focusing specifically on patients undergoing combined upper GI endoscopy and colonoscopy under a standardized propofol TCI protocol. By selecting propofol consumption as the primary endpoint, the study provides an objective pharmacological measure that complements prior investigations focused predominantly on anxiety, pain, and patient satisfaction. Our findings indicate that a simple patient-selected music intervention may exert a clinically measurable propofol-sparing effect even when anesthetic administration is standardized and protocol driven.

In parallel, the assessment of perioperative outcomes is evolving beyond narrowly defined clinical metrics toward multidimensional frameworks of patient experience that incorporate affective and cognitive appraisal, domains that music may plausibly influence, including anxiety, attention, and perceived meaning. Indeed, the benefits of music may extend beyond reduced anesthetic consumption to encompass patient-centered outcomes and procedural conditions. In our routine clinical practice, we have observed that patients who appear more relaxed often experience a smoother peri-procedural course. Moreover, our group has conducted dedicated patient-experience surveys assessing perceived anxiety, comfort, and the perceived benefits of music during endoscopic procedures. Although these findings remain unpublished and fall outside the scope of the present study, the patient feedback obtained was positive. This broader perspective supports incorporating more comprehensive patientreported experience measures alongside anesthetic consumption endpoints in future trials[33].

Interestingly, preprocedural APAIS anxiety scores did not differ between groups despite the observed reduction in propofol consumption. This finding may be explained by the timing of assessment and the psychometric scope of APAIS, which was validated to measure preoperative anxiety; potential floor or ceiling effects in an ambulatory endoscopy population; or mechanisms unrelated to consciously perceived anxiety, such as attentional distraction and reduced sensory salience. The open-label design may also have introduced performance-related effects during anesthetic titration. Nevertheless, propofol was administered via effect-site TCI and titrated according to clinical endpoints, and no temporal trend was observed that would suggest a learning effect. Previous studies evaluating music interventions have used alternative instruments to assess anxiety, including the Spielberger State-Trait Anxiety Inventory, the Faces Anxiety Scale, and a 100-mm visual analog scale[29].

The strengths of this trial include its randomized design, a standard hypnotic regiment consisting exclusively of propofol administered via TCI, the use of objective drug consumption as the primary endpoint, and the pragmatic delivery of patient-selected music through a speaker, reflecting the practical constraints of routine endoscopy workflows.

Several limitations should be acknowledged. First, this was a single-center study, and although sedation targets and propofol administration were standardized, the open-label nature of the intervention may have introduced performance bias. Because music was audible during the procedure, the anesthesiologist could not be blinded to treatment allocation. Despite the use of an identical TCI protocol and predefined clinical sedation criteria in both groups, knowledge of group allocation may have influenced propofol titration. Given that propofol consumption was the primary outcome, this potential source of bias cannot be excluded. Second, anxiety was assessed only before the procedure, and post-procedural patient-reported outcomes, including satisfaction, recall, and perceived control, were not collected. Third, although the predefined clinical sedation targets were achieved in both groups, the higher mean PSi observed in the music group suggests a slightly lighter electrophysiological depth of sedation. The clinical significance of this difference remains uncertain and warrants consideration when interpreting the observed reduction in propofol consumption. Fourth, examination duration was comparable between groups; however, this finding does not rule out potential differences in the frequency or type of additional endoscopic interventions, which were not prospectively documented as separate outcomes.

From an implementation perspective, patient-selected music is a low-cost, low-risk intervention that can be readily integrated into ambulatory endoscopy workflows. Our findings support its potential use as an adjunct to standard care and are consistent with previous reports of its benefits. However, further studies are needed to confirm these results across different clinical settings. Beyond patient outcomes, music has been reported to be compatible with clinical workflows and may help reduce stress among procedural staff[34,35]. As a non-pharmacological intervention, music may also be particularly suitable for resource-limited settings[36]. Reducing anesthetic drug consumption may offer additional environmental benefits, given the contribution of pharmaceutical production and waste to the environmental footprint of anesthesia. However, these potential sustainability benefits require formal life-cycle assessment and cannot be inferred from reduced vial use alone[37]. To establish vial consumption as a robust economic endpoint, future studies should account for drug concentration, vial size, medication waste and disposal practices, and direct costs through a formal economic analysis. These considerations are consistent with evidence from randomized multicenter studies showing that anesthesia delivery strategies can influence patient experience in ambulatory care. Together, they highlight the potential importance of simple, workflowcompatible interventions for both objectively measured clinical outcomes and patients’ experiences of care[38].

CONCLUSION

In this single-center randomized trial, patient-selected music initiated before and continued during combined gastroscopy and colonoscopy was associated with a reduction in propofol consumption. Although both groups underwent the same predefined sedation protocol, the potential effects of unblinded propofol titration should be taken into account when interpreting the results. Neither examination duration nor total ambulatory stay differed between groups. Patient-selected music may therefore offer a simple, non-pharmacological adjunct that can be readily integrated into routine ambulatory endoscopy care. Larger, multicenter randomized trials are warranted to confirm these findings and assess their generalizability.