The median effective dose (ED₅₀) of remimazolam combined with esketamine for conscious sedation in patients with dental anxiety undergoing outpatient dental surgery: a prospective, dose-finding study

Abstract

Background

Dental anxiety presents a significant barrier to treatment. Conscious sedation is widely used to alleviate anxiety. Although remimazolam has demonstrated its efficacy and advantages for outpatient dental procedures, research on the effective dose of remimazolam combined with esketamine remains limited. Thus, the purpose of our research was to estimate the effective dose of remimazolam with esketamine for conscious sedation in patients with dental anxiety undergoing outpatient dental surgery.

Methods

The study enrolled outpatients with dental anxiety aged 18–65 years who received dental operations and were classified under ASA grades I–II. Using the modified Dixon up-and-down method, each patient was first administered by infusing esketamine 0.1 mg/kg, followed by an assigned dose of remimazolam. The initial dose of remimazolam was 0.1 mg/kg, with a dose gradient of 0.01 mg/kg. Sedation success was defined as achieving a MOAA/S score of 3–4 within 5 min after the initial loading dose, without requiring additional remimazolam. The primary outcome was the ED₅₀ of remimazolam and its 95% confidence interval (CI). Secondary outcomes included security indicators, patient satisfaction and memory of the procedure, dentist satisfaction, duration of treatment, and discharge time.

Results

The study achieved 10 transition points (from sedation success to failure) with a total enrollment of 32 individuals. The ED₅₀ and ED₉₅ of remimazolam with 0.1 mg/kg esketamine, for conscious sedation during outpatient dental surgery, were 0.069 mg/kg (95% CI: 0.050–0.080 mg/kg) and 0.099 mg/kg (95% CI: 0.085-0.200 mg/kg), respectively. No circulatory and respiratory adverse events requiring intervention were observed during anesthesia induction.

Conclusions

For outpatient dental procedures in patients with dental anxiety, the ED₅₀ and ED₉₅ of remimazolam with esketamine for conscious sedation were 0.069 mg/kg and 0.099 mg/kg, respectively. The ED₅₀ provides a preliminary reference for initial dose selection in this clinical setting. However, the wide confidence interval of the ED₉₅ limits its clinical utility and requires validation in larger studies.

Trial registration

This study was registered in the Chinese Clinical Trial Registry (ChiCTR2500104775; first registered on 23/06/2025).

Background

Dental treatment commonly involves painful stimuli and invasive procedures, which may lead patients to avoid or refuse dental care, resulting in the continued deterioration of their oral health [1–3]. Data indicate that approximately 30.0% of adults experience anxiety or fear when receiving dental procedures [4, 5]. Dental anxiety (DA) is a major obstacle for patients in receiving dental treatment [6]. To alleviate patient anxiety and fear, conscious sedation techniques have been widely adopted and have demonstrated efficacy in outpatient dental surgery [7–9]. Intravenous sedation is a commonly used method of conscious sedation [10], which is preferred by anesthesiologists for its swift onset and the ease of adjusting sedation levels.

Propofol and midazolam are extensively utilized intravenous sedatives in dental surgery [11], each possessing distinct advantages and disadvantages. Although propofol has a rapid onset of action, it may induce adverse reactions, including cardiovascular and respiratory depression [12, 13]. Midazolam has amnesic, anxiolytic, and anticonvulsant effects [14]. However, its prolonged onset and recovery times present limitations [15, 16]. Consequently, further investigation is essential to develop a new anesthesia protocol with enhanced safety, comfort, and efficacy for outpatient dental procedures.

Remimazolam is an innovative ultra-short-acting benzodiazepine sedative that is rapidly metabolized by tissue esterases into an inactive metabolite, and its effects can be effectively reversed by flumazenil. Compared to midazolam, it provides a rapid onset and a short recovery time, making it particularly suitable for short-duration procedures [15–18]. Furthermore, these characteristics may increase patient throughput [19]. In contrast to propofol, remimazolam exhibits a reduced incidence of cardiovascular depression, respiratory depression, and injection pain [20]. Previous studies have demonstrated that remimazolam provides safe and effective sedation for outpatient dental procedures [18, 21, 22].

Esketamine, the dextrorotatory enantiomer of ketamine, possesses a greater affinity for N-methyl-D-aspartic acid (NMDA) receptors than ketamine and exhibits analgesic effects approximately twice to three times higher than (R)-ketamine [23]. Studies have shown that esketamine exhibits antidepressant effects [24] and has received FDA approval for the treatment of treatment-resistant depression [25]. Moreover, in dental procedures, esketamine provides safe and effective sedation while concurrently mitigating postoperative dental anxiety [26, 27].

Currently, the combination of remimazolam and esketamine has been preliminarily explored in procedures such as painless gastrointestinal endoscopy [28], ambulatory surgery [29], and painless abortion [30], demonstrating superior recovery quality and safety compared with conventional regimens. However, research on the concomitant use of remimazolam and esketamine in outpatient dental surgery remains limited, and the effective dose of remimazolam for this combination has not been determined. This study aimed to determine the effective dose of remimazolam combined with esketamine for conscious sedation in DA patients undergoing outpatient dental procedures, thereby providing preliminary data for subsequent controlled trials.

Methods

Design and patients

The trial was conducted from July 2025 to August 2025. Under conscious sedation with remimazolam and esketamine, the patients received dental surgery, comprising periodontal therapy and the extraction of impacted teeth. The research was conducted in strict compliance with the principles of the Declaration of Helsinki and received approval from the Institutional Review Board of The Eighth Affiliated Hospital of Southern Medical University (The First People’s Hospital of Shunde, Foshan ) on April 22, 2025 [No.KYLS20250405]. All patients provided informed consent. The trial was registered at www.chictrorg.cn before patient enrollment (ChiCTR2500104775, registration date: 23 June 2025).

Participants

Thirty-two outpatients scheduled for dental surgery were recruited. The inclusion criteria were the Modified Dental Anxiety Scale (MDAS) score ≥ 12, an age range of 18 to 65 years, an American Society of Anesthesiologists (ASA) classification of grade I-II, and a body mass index (BMI) between 18.5 and 30.0 kg/m². The exclusion criteria encompassed patients with a history of excessive alcohol consumption, alcohol or drug dependence, regular benzodiazepine use, contraindications to benzodiazepines and esketamine, severe uncontrolled preoperative hypertension (baseline blood pressure SBP > 180 mmHg or DBP > 110 mmHg), preoperative intracranial hypertension, schizophrenia or other psychiatric disorders, severe cardiovascular, pulmonary, hepatic, or renal dysfunction, adrenocortical impairment, structural brain disorders, pregnant patients, and those with obstructive sleep apnea-hypopnea syndrome or airway difficulties.

Study procedures and interventions

On the same day, patients presenting to the clinic received evaluations by an anesthesiologist, and informed permission was secured from the patient. After signing the informed consent form, each participant’s baseline information was recorded, including age, gender, BMI, Electrocardiogram data, physical state, medical history, ASA PS classification, and MDAS score. Before the procedure, patients were fasted for 6 h and restricted to clear fluids for 2 h.

Upon the patient’s entry into the operating room, the electrocardiogram, blood pressure (BP), heart rate (HR), respiratory rate (RR), and pulse oxygen saturation (SpO₂) were routinely monitored. The Narcotrend-Compact M (MT Monitor Technik, Germany) was utilized for the continuous monitoring of the NI value to objectively evaluate the depth of consciousness [31]. A peripheral venous infusion was established for drug administration. The patients were administered 3 L/min of oxygen using a nasal catheter. HR, BP, RR, and SpO₂ of the patients were taken as baseline values five minutes after entering the operation room, with continuous monitoring and documenting at six time points: (T0, 5 min after entering the operating room; T1, immediately after starting intravenous medication; T2, achieving an adequate level of sedation; T3, 15 min after starting the procedure; T4, immediately after stopping the infusion of remimazolam; T5, exiting the operating room).

Anesthesia was induced by infusing esketamine (50 mg/2 ml, Jiangsu Hengrui Pharmaceuticals Co, Ltd, Jiangsu, PR, China, diluted as 1 mg/ml into 10 ml syringe) 0.1 mg/kg, followed by the initiation of remimazolam (36 mg/powder, Jiangsu Hengrui Pharmaceuticals Co, Ltd, Jiangsu, PR, China, diluted as 1 mg/ml into 50 ml syringe) loading at the indicated dose via a micropump, with the injection duration of 5 min. Following the administration of the loading dose, a second anesthesiologist blinded to the remimazolam dose assessed the level of sedation/anesthesia utilizing the Modified Observer’s Assessment of Alertness/Sedation (MOAA/S) scores. The procedure commenced when the patient attained an MOAA/S score 3–4. If the target level of sedation was not attained, 1.0 mg of remimazolam was delivered intravenously until the MOAA/S score was 3–4. However, the maximum of additional remimazolam was 3 times within 15 min. If the desired sedation level remained unattainable, propofol was administered as a rescue medicine. During the anesthesia maintenance phase, remimazolam was continuously infused at a rate of 0.1-0.3 mg/kg/h by a micropump. The infusion rate was modified to maintain the MOAA/S score 3–4).

After successful sedation, the dentist administered infiltration anesthesia to the surgical area using a local anesthetic. For all patients, 4% articaine hydrochloride and epinephrine tartrate injection was selected as the local anesthetic, with a maximum dose of 7 mg/kg. Surgery commenced 5 min after local anesthetic administration. Five minutes before the completion of the procedure, the administration of remimazolam was discontinued. After completion of surgery, the patient was transferred to the post-anesthesia care unit (PACU). The modified Aldrete scale was employed to assess recovery quality based on five criteria: RR, BP, SpO₂, activity, and consciousness [32]. Patients were eligible for release when their Aldrete score reached ≥ 9, or they had recovered to their preoperative baseline status. Before discharge, patients needed to complete a survey regarding their satisfaction and memory of the dental surgery. Furthermore, satisfaction scores and patients’ recollection of the dental operation were rated on a five-point scale. Patients were contacted by telephone on postoperative days 1 and 7 to assess their recovery status and any adverse events occurring after discharge from the treatment room.

Adverse event management: SpO₂ < 93% was considered hypoxic. In such case, the anesthesiologist immediately performed a jaw thrust maneuver. If hypoxia did not ameliorate, the procedure was stopped, and pure oxygen was administered via a mask. Atropine 0.25 mg was delivered intravenously when HR < 50 beats/min. When HR > 100 beats/min, esmolol 0.5-1.0 mg/kg was administered intravenously. When MAP was 20% lower than the baseline value, ephedrine 6–12 mg was injected intravenously. And if MAP was 20% higher than the baseline value, urapidil 10 mg was administered intravenously. When the patient experienced nausea and vomiting, tropisetron 2 mg was administered intravenously.

Outcome measure

The primary outcome was the ED₅₀ and ED₉₅ of remimazolam for successful conscious sedation in patients with DA.

Secondary outcomes included: (1) SBP/DBP, HR, RR, and SpO₂ were recorded at six time points; (2) Security indicators(e.g., incidence of adverse events such as hypotension, hypertension, bradycardia, tachycardia, respiratory depression (RR < 8 breaths/min), hypoxemia (SpO₂ < 93%), nausea and vomiting); (3) Patient satisfaction and memory of the procedure; (4) Dentist satisfaction; (5) Time to treatment; (6) Discharge time(time from patients were moved to the PACU to they were discharged).

Modified dixon’s up-and-down method

The modified Dixon’s up-and-down method was used to explore the effective dose of remimazolam with esketamine 0.1 mg/kg for conscious sedation in patients with DA. The dose of remimazolam administered in the next patient was determined by the reaction of the previous patient. According to the preliminary experiments and the recommended loading dose specified in the Expert consensus on risk prevention and control of comfort care in dental treatment issued by the Chinese Stomatological Association, we selected the initial remimazolam dose of 0.1 mg/kg. A fixed incremental or decremental dose of 0.01 mg/kg was used, corresponding to 10% of the initial dose. The dosage was sequentially decreased by 0.01 mg/kg if the previous patient had a positive reaction. Conversely, in the case of a negative reaction, the subsequent patient received a 0.01 mg/kg increment of remimazolam. A positive response (successful sedation) was defined as achieving an MOAA/S score 3–4 within 5 min of the initial loading dose, without the need for rescue remimazolam, while a negative response (sedation failure) was indicated by an MOAA/S score>4. Drug administration and patient assessments were performed by two independent anesthesiologists to maintain blinding.

Sample size

Pace et al. [33] suggest enrolling 20–40 patients in sequential dose-finding studies using modified Dixon’s up-and-down method, and the study must enroll them in sequence. This method requires at least six crossovers (positive to negative) for statistical analysis. Based on the stopping rules and previous relevant studies [34, 35], the pre-defined stopping criterion of this study was set at seven crossover points. To enhance the stability and reliability of the dose-response estimation, the sample size was appropriately increased. Ultimately, ten crossover points were obtained.

Statistical analysis

SPSS (version 27.0) was employed to conduct statistical analysis and data processing. Normally distributed continuous data were reported as mean and standard deviation, while those not conforming to normality are summarized as medians and quartiles. The Shapiro–Wilk test was employed to assess the normality of continuous data. Comparisons between groups were conducted with the independent samples t-test for normally distributed variables and the Mann–Whitney U test for non-normally distributed ones. For categorical variables, data were presented as frequencies (percentages) and analyzed using Pearson’s chi-square test, Fisher’s exact test, or McNemar chi-square test. Repeatedly measured data were analyzed using repeated-measures ANOVA or the Friedman test, based on distributional assumptions. For multiple comparisons, P-values were adjusted with the Bonferroni method to control the Type I error rate. Probit analysis was applied to generate dose-response curves and to calculate the ED₅₀ and ED₉₅ values (with corresponding 95% CI). P value < 0.05 was deemed statistically significant.

Results

This study included 32 patients (Fig. 1). Table 1 demonstrates the demographic and surgical characteristics of the patients. There were no statistically significant differences regarding age, sex ratio, BMI, ASA status, MDAS score baseline, surgery length, and discharge time between the success and failure groups. Of the 32 patients, 18 achieved successful sedation and 14 had unsuccessful sedation. 14 patients received rescue remimazolam, among whom 9 patients responded to a single 1 mg dose, 4 required a second 1 mg dose, and 1 patient needed a third dose. No patient required rescue propofol (Table 2). Additionally, 22 patients underwent extraction of impacted teeth, and 10 patients received periodontal treatment. There was no statistically significant difference in preoperative dental anxiety levels between the two operation types (Table 3).

Fig. 1.

Flow diagram for the Dixon’s up-and-down method

Table 1.

Demographics data, clinical characteristics and operation type

Characteristic	Success (n = 18)	Fail (n = 14)	Total (n = 32)	P-value
Age (years)	34.3 ± 10.8	31.8 ± 6.6	33.2 ± 9.2	0.437
Sex (M/F)	8/10	8/6	16/16	0.722
BMI (kg/m2)	22.0 ± 1.6	23.4 ± 2.9	22.5 ± 2.6	0.136
ASA PS (I/II)	11/7	11/3	22/10	0.446
MDAS score baseline	12.0 (12.0–16.0)	14.0 (12.0–16.0)	12.5 (12.0–16.0)	0.555
Duration of procedure (min)	37.0 ± 8.7	38.5 ± 10.9	37.6 ± 9.6	0.686
Discharge time (min)	23.9 ± 7.0	25.2 ± 5.8	24.5 ± 6.5	0.597
Operation type, n (%)				1.000
impacted teeth extraction	12 (66.7)	10 (71.4)	22 (68.8)
periodontal treatment	6 (33.3)	4 (28.6)	10 (31.2)

Note: Data are provided as mean ± standard deviation, median (interquartile range), or number (%). There was no significant difference between the success and failure groups

BMI Body mass index, ASA PS American Society of Anesthesiologists Physical Status, MDAS the Modified Dental Anxiety Scale

Table 2.

Summary of the patient’s required remedial medications

Characteristic		Success	Fail
Requiring remimazolam rescue, n (%)		14 (43.8)	0 (0.0)
	Requiring a first rescue remimazolam (1 mg), n (%)	9 (28.1)	5 (15.6)
	Requiring a second rescue remimazolam (1 mg), n (%)	4 (12.5)	1 (3.1)
	Requiring a third rescue remimazolam (1 mg), n (%)	1 (3.1)	0 (0.0)
Requiring propofol rescue, n (%)		0 (0.0)	0 (0.0)

Note: Data are provided as number (%)

Table 3.

Comparison of baseline MDAS scores between the two operation types

Characteristic	Impacted teeth extraction (n = 22)	Periodontal treatment (n = 10)	P-value
MDAS score baseline	12.0 (12.0-13.5)	13.0 (12.0–17.0)	0.760

Note: Data are provided as median (interquartile range) or rate (%)

MDAS the Modified Dental Anxiety Scale

Figure 2 shows the dose–response sequence data obtained from 32 patients with dental anxiety. Probit analysis revealed that, for patients with dental anxiety undergoing dental procedures under conscious sedation, the ED₅₀ and ED₉₅ of remimazolam after receiving 0.1 mg/kg esketamine were 0.069 mg/kg (95% CI: 0.050–0.080 mg/kg) and 0.099 mg/kg (95% CI: 0.085-0.200 mg/kg), respectively (Fig. 3). However, the wide CI for the ED₉₅ reflects the inherent limitations of the Dixon up-and-down method when estimating high quantiles. Hence, the estimated ED₉₅ should be considered exploratory and requires validation in larger-scale studies.

Fig. 2.

Process for determining the 50% and 95% effective dose of remimazolam with a fixed dose of esketamine (0.1 mg/kg) for conscious sedation in patients with dental anxiety receiving dental surgery

Fig. 3.

Dose-response curve of remimazolam with a fixed dose of esketamine (0.1 mg/kg) for conscious sedation in patients with dental anxiety receiving dental surgery, plotted using probit analysis

The satisfaction and patient memory of the dental procedure were assessed using five-point scales. The results are presented in Table 4. The median total patient satisfaction score was 2.0 (1.0–2.0). Among patients, 43.8% reported being very satisfied and 50.0% reported being satisfied. Regarding procedural memory, the median score was 1.0 (1.0–2.0), 59.4% of patients reported no memory, and 31.3% reported little memory. The median total dentist satisfaction score was 1.0 (1.0-1.8). Among dentists, 75.0% reported being very satisfied and 21.9% reported being satisfied.

Table 4.

Summary of the satisfaction and patient memory of procedure

Characteristic	1	2	3	4	5	Median
Patient satisfaction, n (%)	14 (43.8)	16 (50.0)	1 (3.1)	1 (3.1)	0 (0.0)	2.0 (1.0–2.0)
Patient memory of procedure, n (%)	19 (59.4)	10 (31,3)	3 (9.4)	0 (0.0)	0 (0.0)	1.0 (1.0–2.0)
Dentist satisfaction, n (%)	24 (75.0)	7 (21.9)	0 (0.0)	1 (3.1)	0 (0.0)	1.0 (1.0-1.8)

Note: Data are provided as median (interquartile range) or number (%)

Patient and dentist satisfaction: 1 = very satisfied, 2 = satisfied, 3 = neutral, 4 = unsatisfied, 5 = very unsatisfied

Memory of procedure: 1 = none, 2 = remember little, 3 = remember some, 4 = remember most, 5 = remember all

Table 5 shows the distribution of each MOAA/S score at different time points. The mean MOAA/S score was 3.5 (95% CI: 3.3–3.7) at T2 (time point of achieving adequate sedation) and remained at 3.5 (95% CI: 3.4–3.6) throughout the surgical period (T2 - T4). Figure 4 depicts the changes in Narcotrend values across different time points. At T2, the Narcotrend value was 85.9 ± 1.5; during surgery (T2 - T4), values averaged 86.4 ± 1.7.

Table 5.

Distribution of MOAA/S scores at different time points

MOAA/S scores	T0	T1	T2	T3	T4	T5
0: Does not respond to deep stimulus, n (%)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)
1: Does not respond to mild prodding or shaking, n (%)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)	0 (0.0)
2: Responds only after mild prodding or shaking, n (%)	0 (0.0)	0 (0.0)	2 (6.3)	0 (0.0)	0 (0.0)	0 (0.0)
3: Responds only after name is called loudly and/or repeatedly, n (%)	0 (0.0)	0 (0.0)	13 (40.6)	19 (59.4)	12 (37.5)	0 (0.0)
4: Lethargic response to name spoken in normal tone, n (%)	0 (0.0)	0 (0.0)	17 (53.1)	13 (40.6)	20 (62.5)	0 (0.0)
5: Responds readily to name spoken in normal tone (alert), n (%)	32 (100.0)	32 (100.0)	0 (0.0)	0 (0.0)	0 (0.0)	32 (100.0)

Note: Scores 3–4 will be classified as satisfactory sedation score

Fig. 4.

Changes in the Narcotrend value during study period. Data are provided as the mean ± standard deviation and were analyzed by repeated-measures ANOVA, followed by post-hoc tests with Bonferroni correction. T1: immediately after starting intravenous medication; T2: achieving an adequate level of sedation; T3: 15 min after starting the procedure; T4: immediately after stopping the infusion of remimazolam; T5: exiting the operating room. ^aP < 0.05, vs. T0; ^bP < 0.05, vs. T1; ^cP < 0.05, vs. T2; ^dP < 0.05, vs. T3; ^eP < 0.05, vs. T4

Figure 5 demonstrates the trends in HR, MAP, RR, and SpO₂ among patients at different time points. There were no statistically significant differences in HR, MAP, and SpO₂ at different time points (Fig. 5a and 5b, and 5d). Compared with baseline values, RR increased significantly at T4 (P = 0.006) (Fig. 5c).

Fig. 5.

Changes in heart rate (a), mean arterial pressure (b), respiratory rate (c), and SpO₂ (d) during study period. Data are provided as the mean ± standard deviation and were analyzed by repeated-measures ANOVA, followed by post-hoc tests with Bonferroni correction. HR: heart rate; MAP: mean arterial pressure; RR: respiratory rate; SpO₂: percutaneous oxygen saturation. T0: 5 min after entering the operating room; T1: immediately after starting intravenous medication; T2: achieving an adequate level of sedation; T3: 15 min after starting the procedure; T4: immediately after stopping the infusion of remimazolam; T5: exiting the operating room^aP < 0.05, vs. T0

Table 6 details the occurrence of adverse events. Nausea and vomiting were observed in two patients (6.3%). Mild dizziness was reported in seven patients (21.9%). Hypertension developed in two patients (6.3%). Tachycardia occurred in one patient (3.1%). No adverse events, including hypotension, bradycardia, respiratory depression, or hypoxemia, were observed.

Table 6.

Summary of the adverse events

Characteristic	Success (n = 18)	Fail (n = 14)	Total (n = 32)	P-value
Nausea, vomiting, n (%)	2 (11.1)	0 (0.0)	2 (6.3)	0.492
Dizziness, n (%)	4 (22.2)	3 (21.4)	7 (21.9)	1.000
Hypertension, n (%)	1 (5.6)	1 (7.1)	2 (6.3)	1.000
Hypotension, n (%)	0 (0.0)	0 (0.0)	0 (0.0)	1.000
Respiratory depression, n (%)	0 (0.0)	0 (0.0)	0 (0.0)	1.000
Hypoxemia, n (%)	0 (0.0)	0 (0.0)	0 (0.0)	1.000
Bradycardia, n (%)	0 (0.0)	0 (0.0)	0 (0.0)	1.000
tachycardia, n (%)	1 (5.6)	0 (0.0)	1 (3.1)	1.000

Note: Data are provided as the number (%). There was no significant difference between the success and failure groups

Discussion

This dose-response study found that, during outpatient dental procedures, the ED₅₀ and ED₉₅ of remimazolam in combination with 0.1 mg/kg of esketamine for achieving conscious sedation were 0.069 mg/kg (95% CI: 0.050–0.080 mg/kg) and 0.099 mg/kg (95% CI: 0.085-0.200 mg/kg), respectively. However, the ED₉₅ value is an exploratory estimate derived from the Dixon up-and-down method and requires validation in larger studies.

Dental anxiety (DA) is a primary reason for patients delaying or even avoiding treatment, leading to a vicious cycle of worsening oral health status [1–3, 6]. The MDAS is a typical psychometric tool for evaluating dental anxiety levels in patients. Kassem et al. [36] identified an optimal cut-off value of ≥ 12 for dental anxiety. Therefore, in this study, patients with an MDAS score of ≥ 12 were included as the study population to represent individuals with DA. To alleviate patients’ fear and anxiety during dental procedures, conscious sedation has been widely employed in outpatient dental procedures [7–10]. Although traditional intravenous anesthetic agents (e.g., midazolam, propofol) have been widely used, each has its own limitations.

Remimazolam is a novel ultra-short-acting benzodiazepine drug characterized by rapid onset, quick recovery, and no accumulation of metabolites [19]. Consistent with our research findings, the studies by Oue et al. [21], Guo et al. [22] and Wang et al. [18] have shown that remimazolam can achieve adequate depth of sedation during dental procedures, without significantly compromising respiratory or hemodynamic stability. Dhuvad et al. [37] demonstrated that propofol offers advantages over midazolam and local anesthesia in terms of hemodynamic stability, operator comfort, and patient satisfaction. In contrast, Zhao et al. [38] reported that remimazolam is a safer and more effective alternative for outpatient sedation compared with propofol, reducing recovery and discharge times and decreasing the incidence of perioperative adverse events. Although remimazolam offers the advantage of having minimal respiratory effects, it still presents a risk of dose-dependent respiratory depression. Considering the requirement for precise control of sedation depth during dental surgery and that procedure preparation time typically exceed 5 min, the remimazolam loading dose was administered using an infusion pump. The infusion time is set to 5 min.

Esketamine, a non-competitive antagonist of the N-methyl-D-aspartate (NMDA) receptor, provides both sedative and analgesic effects. It exhibits mild inhibition of the respiratory and circulatory systems, allows rapid awakening, and has no adverse impact on postoperative recovery [39]. Previous studies have shown that esketamine has anxiolytic and antidepressant effects, and that when used as an adjunct for induction, low-dose (< 0.2 mg/kg) esketamine exhibits mild and transient adverse reactions [40]. The results of studies by Zhan et al. [41] and Wan et al. [42] indicated that the addition of esketamine at 0.1 mg/kg and 0.2 mg/kg reduced the propofol requirement and lowered patients’ HADS-A and HADS-D scores, without significant differences in the awakening and recovery status. However, in our clinical practice of outpatient dental sedation, we observed that administering 0.2 mg/kg of esketamine commonly leads to postoperative dizziness, which significantly impairs clinical turnover. Given that dental outpatient procedures require lighter sedation compared to other surgeries, that analgesia is primarily achieved through local anesthesia, and that the adverse effects of esketamine are dose-dependent, this study selected 0.1 mg/kg of esketamine as the induction dose.

Dixon’s up-and-down method is a classic and efficient design for estimating the ED₅₀. Compared with traditional designs, this method typically provides a reliable estimate of the ED₅₀ with a smaller sample size [43]. However, this method is not designed to estimate extreme percentiles (e.g., ED₉₅) or to directly compute CI. Therefore, we performed probit analysis on the data. Probit analysis is a widely used method for determining ED₅₀, ED₉₅, and dose-response relationships [44, 45]. It assumes that dose logarithms in the target population follow a normal distribution and the response is dichotomous. In this study, the logarithms of doses in patients with DA followed a normal distribution, and sedation outcomes (success vs. failure) were dichotomous. The dose-response relationship is fundamental to the action of any anesthetic agent [33]. The ED₅₀ is a commonly used measure of drug potency and often serves as the initial dose in clinical trials [45]. In contrast, the ED₉₅ is derived by extrapolation from the dose-response curve and relies on limited data from the extreme response region [46]. Consequently, the 95% CI for the ED₅₀ is generally narrower, enabling more reliable comparisons of drug potency. In this trial, the combination of remimazolam and 0.1 mg/kg esketamine successfully achieved conscious sedation in patients with DA undergoing outpatient dental surgery. The ED₅₀ and ED₉₅ of remimazolam for conscious sedation were 0.069 mg/kg (95% CI: 0.050–0.080 mg/kg) and 0.099 mg/kg (95% CI: 0.085-0.200 mg/kg), respectively. Consistent with previous findings, the 95% CI for the ED₅₀ of remimazolam in this study was narrower than that for its ED₉₅. This reflects an inherent limitation of the up-and-down method when estimating extreme percentiles, highlighting the necessity for validation in larger sample sizes. The biased coin design up-down method effectively addresses this limitation by significantly reducing the variability in the estimation of extreme percentiles, thereby resulting in narrower and more reliable CI. Although the combination of Dixon’s up-and-down method with probit analysis is widely used in anesthesiology, a previous study found that the 95% CI for ED₅₀ estimated by probit analysis can be impractically narrow [47]. Future studies should consider using isotonic regression and leveraging the biased-coin-design up-and-down sequential method to explore the ED₉₅ of the drug [33, 46].

Few studies have investigated the appropriate dose of remimazolam for conscious sedation in outpatient dental procedures. A prospective, exploratory clinical trial recommended an induction dose of 0.08 mg/kg (0.07–0.09 mg/kg) as suitable for outpatient dental surgery [21]. Our results showed that the ED₅₀ was 0.069 mg/kg (95% CI: 0.050–0.080 mg/kg), which is slightly lower than the dose of the single agent used in the study by Oue et al. [21]. This difference may be attributed to a potential pharmacodynamic interaction between esketamine and remimazolam, which enhances sedative efficacy. In clinical practice, for patients with DA undergoing outpatient dental surgery, when co-administered with esketamine 0.1 mg/kg, we recommend an initial induction dose of remimazolam of 0.07 mg/kg, followed by 1 mg increments based on patient response until the desired sedation depth is achieved. This strategy ensures sedative efficacy while avoiding the risk of overdose.

Previous studies show that conscious sedation (MOAA/S 3–4) suffices for outpatient dental procedures, including impacted tooth extraction, periodontal treatment, and implant surgery [48]. Both periodontal treatment and impacted tooth extraction require sedation depth maintained at MOAA/S 3–4, which enables subsequent invasive procedures like local anesthesia [21, 49]. Different surgical types vary in invasiveness and psychological stress, which may influence dental anxiety outcomes. Accordingly, we performed an analysis of preoperative dental anxiety levels across procedure types and found no statistically significant difference between the two surgical types in this study. Furthermore, our primary endpoint was successful achievement of target sedation depth (MOAA/S 3–4) within 5 min after induction, without requiring rescue remimazolam. All subsequent procedures began only after this endpoint was reached. Thus, the primary outcome—ED₅₀ (with 95% CI) of remimazolam—was observed within the first 5 min, before any operator-driven procedures. Hence, surgery type did not confound the primary outcome assessment. Although this heterogeneity did not affect the assessment of the primary endpoint, it may still introduce variability into the measurement of secondary endpoints, such as satisfaction. Dental anxiety levels and BMI may affect sedation requirements and ED₅₀. We enrolled patients with DA (MDAS > 12) and a BMI of 18.5–30.0 kg/m², including those with dental phobia (MDAS > 19) and overweight patients (BMI 25.0–29.9 kg/m²). However, we did not perform subgroup analysis by anxiety level or BMI. Future studies should restrict to a single surgical type and stratify patients by dental anxiety level and BMI to further explore their impact on ED₅₀.

Post-treatment questionnaires indicated high levels of satisfaction among both dentists and patients. Masuda et al. demonstrated that the amnestic effect reduces patients’ fear and anxiety associated with dental treatment, and that the absence of intraoperative memory is the strongest independent predictor of both patient and dentist satisfaction [50]. In our study, one case was reported as unsatisfactory by both the patient and the dentist, which may be associated with the occurrence of nausea and vomiting that required medical intervention, resulting in extended observation time and delayed discharge.

In terms of vital signs and adverse reactions, hypertension was observed in 2 patients and tachycardia in 1 patient. These events were generally transient, asymptomatic, and did not lead to serious cardiovascular sequelae. This is likely attributable to esketamine’s indirect cardiovascular stimulatory effects, which may be mediated by catecholamine release, inhibition of norepinephrine reuptake, and subsequent sympathetic nervous system stimulation [51]. It is noteworthy that the local anesthetic used in our dental practice was articaine with epinephrine (1:100,000). Therefore, the intraoperative hypertension and tachycardia could also be attributed to the administration of epinephrine. In this study, the incidence of postoperative dizziness was 21.9%, which was higher than the incidence reported for monotherapy by Oue et al. [21]. This difference may be attributable to a potential synergistic effect between remimazolam and esketamine. Furthermore, adverse events including dizziness were assessed immediately after surgery completion, which may have contributed to the higher observed incidence. Nevertheless, all affected patients experienced complete resolution after a brief rest (within 15 min), without requiring medical intervention. Given the relatively high incidence of dizziness, albeit mostly mild, its clinical significance should not be overlooked. This finding highlights the need for close monitoring during the early recovery period, and careful assistance is recommended when transferring patients from the dental chair to avoid a secondary risk of falls. Before the procedure, patients must be clearly informed of possible transient effects, including dizziness or gait instability, and advised to refrain from driving, operating machinery, or participating in activities that require full motor coordination for the remainder of the day. Additionally, it is recommended that patients undergo objective balance assessments before discharge, such as the Romberg test. In our study, patients were not discharged until dizziness had resolved and they could ambulate independently; no fall-related events occurred. Two cases of postoperative nausea and vomiting (PONV) were observed, which could be potentially attributed to drug-induced effects or local stimuli such as blood ingestion and inflammatory responses following oral dental surgery [52]. Moreover, in this study, although the sedative regimen combining remimazolam with low-dose esketamine did not observe any cases of circulatory and respiratory depression needing intervention, or significant adverse events, the small sample size may have been insufficient to detect rare or uncommon adverse events. Therefore, no definitive safety conclusions can be drawn, and the safety of this regimen still requires confirmation in larger-scale studies.

This study has several limitations. First, the sample size of this study was small. Although it was adequate for the modified Dixon’s up-and-down method, it was insufficient for precise estimation of higher effective doses such as ED₉₅. Larger studies are needed to accurately determine ED₉₅, potentially using alternative designs such as the biased-coin design. Second, this study included two distinct dental procedures. While this heterogeneity does not affect the assessment of our primary endpoint, it could introduce variability into the measurement of secondary outcomes. Third, similar to other studies utilizing the modified Dixon’s up-and-down methodology, our study lacked a control group; therefore, it is unclear whether the observed effects were attributable to esketamine, remimazolam, or their combination. Furthermore, as a single-arm dose-finding study, this investigation did not employ a stratified randomized design to explore the potential impact of dental anxiety levels or BMI on the effective dose of remimazolam.

Conclusions

For outpatient dental procedures in patients with dental anxiety, the ED₅₀ and ED₉₅ of remimazolam with esketamine for conscious sedation were 0.069 mg/kg and 0.099 mg/kg, respectively. In this clinical setting, the ED₅₀ provides a preliminary reference for initial dose selection. However, the wide confidence interval of the ED₉₅ limits its clinical utility, necessitating validation in larger studies. Future randomized controlled trials are required to confirm the efficacy and safety of this regimen for outpatient dental procedures.

Abbreviations

CI

Confidence interval

DA

Dental anxiety

ASA

American Society of Anesthesiologists

ED₅₀

The 50% effective dose

ED₉₅

The 95% effective dose

MOAA/S

Modified observer’s assessment of alert

MDAS

The Modified Dental Anxiety Scale

NMDA

N-methyl-D-aspartic

BMI

Body mass index

HR

Heart rate

BP

Blood pressure

RR

Respiratory rate

SpO₂

Pulse oxygen saturation

EEG

Electroencephalogram

PACU

Post-Anesthesia Care Unit

POD 1

Postoperative days 1

POD 7

Postoperative days 7

Authors’ contributions

Mengjiao Che and Cuiyao Kang conceived the study, obtained approval, reviewed the literature, performed chart reviews, and wrote and revised the manuscript. Mengjiao Che, Cuiyun Zhou, Yaoheng Yang, Yuxuan Zhou, Yifeng Yang, Feifei Li, Danru Xie, and Yanjing Zhang conceived data acquisition, statistical analysis, and data interpretation. Yiwen Zhang assisted with study conception, created the study design, and revised the manuscript. All the authors approved the final version of the manuscript and agreed to be accountable for all aspects of the work.

Funding

This study was supported by the 2023 Key Scientific Research Platforms and Projects of Ordinary Universities in Guangdong Province (No. 2023ZDZX2008), the 2023 Basic and Applied Basic Research Joint Fund in Guangdong Province (Guangdong-Foshan) (No. 2023A1515140142), and the 2024 Scientific Research Program for Innovative Drug Clinical Research of Development Center for Medical Science and Technology National Health Commission of the People’s Republic of China (No. WKZX2024CX301206).

Data availability

The corresponding author can provide all datasets generated and/or analysed during this study, if they are requested reasonably.

Declarations

Ethics approval and consent to participate

The study was approved by the Medical Ethics Committee of the Eighth Affiliated Hospital of Southern Medical University (The First People’s Hospital of Shunde, Foshan) (Approval No. KYLS20250405) and had been registered at the China Clinical Trials Registry (ChiCTR2500104775). All procedures were performed following the ethical guidelines and regulations established by the World Medical Association Declaration of Helsinki. Before enrollment, written informed consent was obtained from all participants or their legal guardians. Specifically, for adult participants who possessed the capacity to give informed consent, written consent was obtained directly from them. For a subset of adult participants with (e.g., a low level of education or a hearing or visual impairment), informed consent was provided by their LAR.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.