Comparison of esketamine combined with dexmedetomidine versus propofol combined with midazolam on drug-induced sleep endoscopy: a randomized controlled trial

Abstract

Background

Drug-induced sleep endoscopy (DISE) is widely used to evaluate upper airway obstruction in patients with obstructive sleep apnea syndrome (OSAS). We aimed to compare the sedative efficacy and safety of esketamine combined with dexmedetomidine versus propofol combined with midazolam for sedation in patients undergoing DISE.

Methods

This prospective, single-center, assessor-blinded, randomized controlled trial was conducted among patients scheduled for DISE between July 2025 and December 2025. Participants were enrolled and randomized into two groups: the experimental group (esketamine-dexmedetomidine) and the control group (propofol-midazolam). Polysomnography (PSG) was used to determine sleep stages, and the bispectral index (BIS) was used to assess sedation depth. Endoscopy was initiated after patients reached non-rapid eye movement sleep stage 2 (N2). The primary outcome was the time required to reach N2 sleep confirmed by PSG. Secondary outcomes included the first-pass success rate, BIS values, oxygen saturation (SpO₂), and the incidence of adverse events (body movement, hypoxemia, tachycardia, bradycardia, and other procedure-related events) during the procedure.

Results

A total of 50 patients were enrolled and randomly assigned to the experimental group (n = 25) or the control group (n = 25). All patients successfully completed the DISE procedure and achieved N2 sleep confirmed by PSG. The time required to reach N2 sleep was slightly shorter in the experimental group than in the control group, with median (interquartile range [IQR]) values of 188.0 (170.0-229.0) seconds and 214.0 (189.0-253.5) seconds, respectively; however, this difference did not reach statistical significance (p = 0.067). Meanwhile, preliminary exploratory analyses showed a higher first-pass success rate (96% vs. 36%; unadjusted p < 0.001) and a lower incidence of hypoxemia (20% vs. 48%; unadjusted p < 0.05) in the experimental group during the procedure.

Conclusions

Esketamine-dexmedetomidine and propofol-midazolam effectively induced N2 sleep and allowed completion of DISE. Exploratory analyses suggest that the esketamine-dexmedetomidine regimen may potentially improve procedural conditions and respiratory safety, though these findings are hypothesis-generating and require confirmation in larger, adequately powered studies.

Trial registration

Chinese Clinical Trial Registry (ChiCTR2500102931), first registered on 05/21/2025.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12871-026-03987-6.

Introduction

Obstructive sleep apnea syndrome (OSAS) is a prevalent chronic sleep-related breathing disorder, characterized by recurrent upper-airway luminal narrowing, leading to intermittent hypoxia and sleep fragmentation during sleep [1].The prevalence of OSAS continues to rise worldwide. According to the latest epidemiological statistics, the overall prevalence of OSAS is 43.2%, with a severe OSAS prevalence rate of 11.6% [2, 3]. Chronic OSAS is associated with daytime sleepiness, cognitive impairment, and elevated cardiovascular and cerebrovascular risks, adversely affecting daily functioning and quality of life [4].

Drug-induced sleep endoscopy (DISE) is a diagnostic technique that has been increasingly adopted in recent years to evaluate upper airway obstruction in patients with OSAS [5, 6]. DISE uses sedative-hypnotic drugs to simulate the natural sleep state and enables direct, dynamic visualization of the upper airway through nasopharyngoscopy. This approach enables objective evaluation of obstruction degree and pattern at key anatomical sites, including the soft palate, lateral pharyngeal walls, tongue base, and epiglottis, and provides more precise guidance for surgical planning and individualized treatment decisions [7–9].

There is no consensus on the sedation regimen of DISE in China. In clinical practice, propofol combined with midazolam is a commonly used sedation regimen, which has the characteristics of fast onset, reliable sedative effect, and pronounced anterograde amnesia [10]. Nevertheless, both propofol and midazolam are associated with dose-dependent respiratory and circulatory depression, which may reduce upper airway muscle tone, increase the risk of upper airway collapse, and potentially obscure the actual airway collapse, thereby affecting the DISE assessment. In recent years, dexmedetomidine and esketamine have attracted increasing attention as potentially more suitable agents for DISE sedation [11, 12]. Dexmedetomidine induces sedation with minimal respiratory depression while preserving upper airway reflexes and sleep architecture. Esketamine provides effective analgesia, also with minimal respiratory depression, and may help maintain airway patency through its sympathomimetic effects [13]. The combination of dexmedetomidine and esketamine may therefore offer complementary advantages, achieving adequate sedation while better preserving respiratory function and upper airway muscle tone, thereby enhancing the safety and procedural stability of DISE.

To date, no studies have compared the efficacy and safety of an esketamine-dexmedetomidine combination with the conventional propofol-midazolam regimen during DISE examinations. The relative advantages of these two sedation strategies in terms of procedural success, respiratory and hemodynamic stability, and sedation quality remain unclear. We conducted a prospective randomized controlled trial comparing an esketamine-dexmedetomidine regimen with a propofol-midazolam regimen for DISE.

Materials and methods

Ethical considerations

This single-center, prospective, randomized, assessor-blinded controlled trial was conducted at the Shanghai Ninth People’s Hospital between July 2025 and December 2025. The study was conducted in accordance with the Declaration of Helsinki (2013 amendment), was reviewed and approved by the Ethics Committee of Shanghai Ninth People’s Hospital (Study ID SH9H-2025-T144-2), and was registered in the Chinese Clinical Trial Registry (Registration no: ChiCTR2500102931, 05/21/2025). Written informed consent was obtained from all participants before enrollment.

Inclusion and exclusion criteria

A total of 50 patients undergoing DISE evaluation were enrolled. Inclusion criteria were: (1) aged between 18 and 65 and (2) the American Society of Anesthesiologists (ASA) physical status class I-II. The exclusion criteria were: history of upper airway surgery, allergy to study medications in the examination, psychiatric illness, pregnant or lactating women.

Randomization and sample size estimation

Patients were randomly divided into the experimental group (esketamine combined with dexmedetomidine) and the control group (propofol combined with midazolam) in a 1:1 ratio using computer-generated random numbers. Allocation concealment was achieved with sequentially numbered, sealed opaque envelopes, which were opened only immediately before the procedure. Outcome assessors were blinded to group allocation throughout the study, whereas patients and anesthesiologists were not blinded due to the nature of the interventions.

The sample size was calculated for a two-sided comparison of means between two independent groups using PASS 15.0. Based on our pilot data (n = 6 per group), the mean time to reach N2 sleep was 214.0 ± 16.7 s in the control group and 196.0 ± 19.1 s in the experimental group. Assuming a two-sided α of 0.05 and 90% power, the required sample size was estimated as 22 patients per group. To account for potential dropouts or protocol deviations, the final sample size was increased to 25 patients per group.

Sedation procedure

All examinations were performed in the operating room. All patients were fasting for 6 h without drinking water, and did not use any drugs before surgery. After establishing intravenous access, Ringer’s acetate solution was infused. Baseline patient characteristics, including age, sex, height, weight, ASA grade, body mass index (BMI), were recorded. Electrocardiogram (ECG), heart rate (HR), systolic and diastolic blood pressure (SBP and DBP) and peripheral capillary oxygen saturation (SpO₂) were routinely detected. Sleep stages were simultaneously monitored using PSG, and sedation depth was continuously assessed with the BIS. The control group received an intravenous bolus of midazolam (0.05 mg/kg), followed 2 min later by target-controlled infusion of propofol using the Schneider model, with an initial target concentration of 2.0 µg/mL. The experimental group received an intravenous bolus of dexmedetomidine (0.2 µg/kg), followed 2 min later by an intravenous bolus of esketamine (0.5 mg/kg). Sleep staging was performed using PSG, and DISE was initiated once N2 sleep was confirmed. If body movement occurred during endoscopy, the propofol target concentration was increased by 0.2-0.5 µg/mL in the control group, whereas an additional dose of esketamine (0.2 mg/kg) was administered in the experimental group. If SpO₂ is below 90% and lasted for more than 30 s during the examination, the oxygen inhalation or airway intervention would be performed. If bradycardia (HR < 50 beats/min) occurred, intravenous atropine (0.3-0.5 mg) was administered, with isoproterenol infusion (0.05-2 µg/kg/min) used if necessary. If tachycardia (HR > 100 beats/min) occurred, esmolol 0.5 mg/kg was injected intravenously. To avoid masking hypoxic events, supplemental oxygen was not routinely used during DISE. All endoscopies were performed by the same experienced surgeon with expertise in sleep medicine.

Outcomes

The primary outcome of this study was the time required to reach N2 sleep (sleep latency), defined as the interval in seconds from the administration of sedative agents to the onset of N2 sleep, as confirmed by PSG. N2 sleep onset was defined as the first appearance of either K-complexes or sleep spindles on PSG. Secondary outcomes included first-pass success rate, BIS values, hemodynamic stability, oxygenation and adverse events. Hemodynamic parameters included HR, SBP, and DBP. The lowest SpO₂ during the procedure was also recorded. All BIS values and hemodynamic parameters were recorded at four predefined time points: T0-T3. T0 was defined as the patient’s awake state upon entering the operating room. T1 corresponded to the onset of N2 sleep, confirmed by PSG. T2 represented the start of the DISE procedure, and T3 corresponded to the end of the procedure. First-pass success was defined as completion of DISE without rescue sedatives or airway intervention. Adverse events were recorded throughout the procedure and included body movement, hypoxemia (SpO₂ < 90% lasting ≥ 30 s), bradycardia (HR < 50 beats/min), and tachycardia (HR > 100 beats/min).

Statistical analysis

SPSS 26.0 statistical software was used for data analysis. Normally distributed continuous variables were presented as mean ± standard deviation (SD), and comparisons between groups were performed using independent-samples t tests. Non-normally distributed data were presented as median (interquartile range, IQR) and were compared using the Mann-Whitney U test. Nonordinal categorical variables were analyzed using chi-square tests. All statistical tests were two-tailed, and a P value < 0.05 was considered statistically significant.

This study adhered to the CONSORT 2025 guidelines for reporting randomized controlled trials.

Result

Fifty patients were enrolled and randomized into 2 groups, the control group (n = 25) and the experimental group (n = 25). All patients successfully completed endoscopy and were included in the final analysis (Fig. 1). Descriptive characteristics of patients are presented in Table 1. There are no significant differences in gender, age, height, weight and BMI between the control and experimental groups.

Fig. 1.

CONSORT flow diagram of patient enrollment and randomization

Table 1.

Characteristics of patients by group

Characteristic	Control Group (n = 25)	Experimental Group (n = 25)	p-Value
Gender (M/F)	10/15	8/17	0.556
Age (y)	26.00(22.50,31.00)	26.00(22.00,29.00)	0.800
Height (cm)	167.00(161.50,177.50)	168.00(162.00,174.00)	0.756
Weight (kg)	56.00(48.50, 71.75)	55.00(50.50,67.00)	0.992
BMI (kg/m2)	20.80(18.15,24.20)	19.60(18.10,24.40)	0.691

Data are expressed as medians with (IQRs)

IQR interquartile range, BMI body mass index

All patients completed the DISE examination and achieved N2 sleep. The time required to reach N2 sleep was slightly shorter in the experimental group than in the control group, with median (interquartile range [IQR]) values of 188.0 (170.0-229.0) seconds and 214.0 (189.0-253.5) seconds, respectively. This difference was not statistically significant (p = 0.067).

In contrast, exploratory analyses showed that the first-pass success rate was higher in the experimental group compared with the control group (96% vs. 36%; unadjusted p < 0.001) (Table 2). Representative PSG waveforms of N2 sleep are presented in Fig. 2. Vital signs recorded during the DISE procedure are summarized in Fig. 3. Baseline BIS values, HR, SBP, DBP, and SpO₂ showed no differences between the two groups. Compared with T0, BIS values in the experimental group were higher at all post-sedation time points. At T1, BIS values were 89.00 (88.00, 90.50) in the experimental group versus 76.00 (68.50, 78.00) in the control group (p < 0.001). This difference persisted at T2 (89.00 [85.50, 90.00] vs. 66.00 [60.00, 69.50], p < 0.001) and T3 (91.00 [89.00, 92.00] vs. 60.00 [56.00, 65.00], p < 0.001). HR did not differ between groups at T1 or T2. However, at T3, HR was higher in the control group compared with the experimental group (90.00 [80.00, 96.00] vs. 80.00 [69.00, 88.00], p = 0.002). At T2, SBP was higher in the experimental group than in the control group (126.00 [120.00, 135.00] vs. 112.00 [104.00, 119.50], p < 0.001), and this difference further increased at T3 (136.00 [128.50, 146.00] vs. 105.00 [100.00, 118.50], p < 0.001). In contrast, DBP was higher in the experimental group at all post-sedation time points. At T1, DBP was 72.00 (67.50, 78.50) in the experimental group compared with 65.00 (61.00, 70.00) in the control group (p = 0.003). This difference was more pronounced at T2 (75.00 [72.00, 79.00] vs. 61.00 [57.50, 68.00], p < 0.001) and T3 (81.00 [78.50, 85.00] vs. 62.00 [56.00, 67.00], p < 0.001). During DISE, SpO₂ values were higher in the experimental group than in the control group at T1 (97.28 ± 1.95% vs. 95.96 ± 2.57%, p = 0.026), T2 (95.64 ± 2.72% vs. 92.44 ± 3.57%, p = 0.005), and T3 (95.80 ± 2.20% vs. 94.00 ± 2.92%, p = 0.023). Moreover, the lowest SpO₂ recorded during the examination was higher in the experimental group than in the control group (95.00 [92.00, 96.00] vs. 90.00 [88.00, 94.50], p = 0.037).

Table 2.

Sleep latency during DISE

Characteristic during DISE	Control Group (n = 25)	Experimental Group (n = 25)	p-Value
Sleep latency (s)	214.00(189.00,253.50)	188.00(170.00,229.00)	0.067
First-Pass Success Rate (%)	9(36%)	24(96%)	<0.001

Data are expressed as medians with (IQRs). IQR, interquartile range. Sleep latency: the time (seconds) from the start of sedation to the onset of stage N2 sleep, confirmed by PSG. First-Pass Success Rate: completion of DISE without rescue sedatives or airway intervention

Fig. 2.

Representative PSG waveforms during N2 sleep. A Representative schematic of N2 PSG features (sleep spindles and K-complexes) during DISE. B PSG waveform recorded during the awake state. C PSG waveform demonstrating the typical K-complexes and sleep spindles activity

Fig. 3.

Vital signs during DISE. Bispectral index (BIS) and vital signs records. A Bispectral index (BIS). B Heart rate (HR). C Systolic blood pressure (SBP). D Diastolic blood pressure (DBP). E Peripheral capillary oxygen saturation (SpO₂). Data are expressed as median (interquartile range, IQR), compare between two groups *p < 0.05

It presents the adverse events observed during the DISE examination (Table 3). The incidence of body movement in the experimental group was lower than that in the control group (p < 0.001). Similarly, the incidence of hypoxemia was also lower in the experimental group (p = 0.037). No cases of bradycardia were observed in either group.

Table 3.

Adverse Events During DISE

Adverse Events	Control Group (n = 25)	Experimental Group (n = 25)	p-Value
Body Movement (%)	17(68.0)	1(4.0)	<0.001
Hypoxemia (%)	12(48.0)	5(20.0)	0.037
Tachycardia (%)	2(8.0)	4(16.0)	0.384
Bradycardia (%)	0	0

Data are expressed as number of patients

Discussion

Previous studies have reported that upper airway obstruction in patients with OSAS does not occur uniformly across sleep stages, but is more likely to develop during N2 sleep [14, 15]. Therefore, we selected the time required to achieve N2 sleep as the primary outcome. PSG analysis confirmed that there was no statistically significant difference in the time required to reach this sleep stage between the two groups. These findings indicate that both sedation protocols can achieve the target sleep stage required for DISE assessment, with no difference in the time taken to reach the target level of sedation.

In the present study, the first-pass success rate of DISE was higher in the experimental group. According to the 2017 European position paper, propofol combined with midazolam is one of the commonly used sedation regimens for DISE [16]. This may be related to improved procedural tolerance and reduced need for rescue sedation, given the combined sedative and analgesic properties of this regimen. However, as a secondary outcome, this finding should be interpreted as exploratory.

As an exploratory observation, analysis revealed that the experimental group consistently showed higher BIS values than the control group. This result may be related to the different mechanisms of action among various sedative drugs and their distinct effects on brain electrical activity. BIS monitoring primarily relies on brain electrical spectrum characteristics and is particularly sensitive to gamma-aminobutyric acid (GABA)-mediated sedatives [17, 18]. Esketamine primarily exerts its effects by antagonizing N-methyl-D-aspartate (NMDA) receptors [18–20], which can produce atypical EEG patterns that may affect BIS interpretation.

Respiratory safety is a critical consideration in DISE, particularly in OSAS patients. To avoid masking hypoxic events, supplemental oxygen was not routinely administered in this study. Under these conditions, the experimental group demonstrated a lower incidence of hypoxemia, as well as a higher minimum SpO₂ during the DISE. These findings suggest that the esketamine-dexmedetomidine regimen may show better preservation of respiratory function compared with propofol-midazolam. This difference may be related to the pharmacological properties of esketamine and dexmedetomidine. Esketamine, an NMDA receptor antagonist, has been shown to reduce respiratory depression while preserving airway reflexes and respiratory drive [21, 22]. Dexmedetomidine provides hypnotic, sedative, and analgesic effects with minimal impact on respiratory function [23, 24]. Consistent with previous reports, studies have demonstrated that esketamine or dexmedetomidine, when used alone, is associated with a lower risk of hypoxia than conventional sedative agents [21, 23]. The combination of these two agents may therefore contribute to improved respiratory stability during DISE. These findings suggest a potentially more favorable respiratory profile for the esketamine-dexmedetomidine regimen in patients with OSAS, although confirmation in larger studies is required.

In the present study, transient increases in blood pressure (BP) and HR were observed at certain time points in the experimental group; however, all values remained within clinically normal ranges. This phenomenon may be related to the sympathomimetic and excitatory effects of esketamine.

Nevertheless, this study has several limitations. First, most enrolled patients had mild OSAS, which may have limited the ability to detect differences related to more severe airway obstruction. Second, the sample size was relatively small and was estimated based on pilot data. Although the study was designed to achieve adequate statistical power, the effect size assumption derived from pilot observations (d = 1.0) may be optimistic, and the true effect size may be smaller, potentially limiting statistical power. Larger, well-powered studies are needed to further validate these findings. Third, some secondary outcomes were exploratory in nature and were not specifically powered for formal hypothesis testing; therefore, all secondary outcomes should be considered exploratory and not suitable for confirmatory inference. Finally, postoperative recovery after DISE was not systematically assessed because all patients proceeded directly to elective surgery under general anesthesia, which limited evaluation of recovery profile and neuropsychiatric adverse events. This limits a comprehensive assessment of the overall safety and clinical applicability of the esketamine-dexmedetomidine protocol. Future studies should include standardized postoperative recovery scales, such as the Quality of Recovery-15, and systematic monitoring of neuropsychiatric adverse events.

Conclusion

Both esketamine combined with dexmedetomidine and propofol combined with midazolam successfully induced N2 sleep and allowed completion of DISE. Exploratory analyses suggest that the esketamine-dexmedetomidine regimen may improve procedural conditions. However, because these outcomes were secondary endpoints and the study was not powered for these comparisons, these findings should be interpreted cautiously and require confirmation in future studies with adequately calculated sample sizes.

Abbreviations

DISE

Drug-induced sleep endoscopy

OSAS

Obstructive sleep apnea syndrome

PSG

Polysomnography

BIS

Bispectral index

N2 sleep

Non-rapid eye movement sleep stage 2

ASA

American Society of Anesthesiologists

BMI

Body mass index

ECG

Electrocardiogram

EEG

Electroencephalography

HR

Heart rate

SBP

Systolic blood pressure

DBP

Diastolic blood pressure

SpO₂

Peripheral capillary oxygen saturation

GABA

Gamma-aminobutyric acid

NMDA

N-methyl-D-aspartate

BP

Blood pressure

Authors’ contributions

Yanan Jiang: study design, supervision, manuscript drafting; Ziyu Wang: Patient recruitment, sedation procedures; Hongbo Wen: figure preparation; Zhirou Wen: data acquisition, clinical monitoring; Longying Chen: Polysomnography and BIS data collection and analysis; Chen Wang: Visualization; Yanyong Cheng: Literature review, manuscript editing; Ren Zhou: Statistical analysis, data interpretation; Wenwen Yu: performed DISE and interpreted the endoscopic findings; Jia Yan: Writing-review, editing Supervision, Funding acquisition; All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by the Clinical Research Program of Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine (#JYLJ202308), the Cross Disciplinary Research Fund of Shanghai Ninth People’s Hospital (#JYJC202402) and the Collaborative Research Fund of Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine(#JYHX2025039).

Data availability

The datasets used and analyzed in the current study are available from the corresponding author in response to reasonable requests.

Declarations

Ethics approval and consent to participate

This study was approved by the Medical Ethics Committee of Shanghai Ninth People’s Hospital (Study ID SH9H-2025-T144-2), and was registered in the Chinese Clinical Trial Registry (ChiCTR2500102931). Written informed consent was obtained from all participants before enrollment.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.