Background

Oral tumor resection with flap repair (flap surgery) often requires delayed endotracheal tube removal. Postoperative intolerance to the endotracheal tube, pain, and immobilization increase the incidence of emergence agitation after flap surgery [1]. Therefore, patients often require moderate sedation in the early postoperative period after flap surgery to increase their tolerance to endotracheal intubation and prevent the occurrence of emergence agitation [2, 3].

Dexmedetomidine is commonly used for sedation after flap surgery in oral and maxillofacial surgeries [4]. However, the half-life of dexmedetomidine is significantly prolonged by continuous infusion, which is not conducive to the resolution of sedation after discontinuation of flap surgery, resulting in the potential risk of sedative drug accumulation when the endotracheal tube is removed after discontinuation of the drug [2]. At the same time, dexmedetomidine increases the incidence of bradycardia and hypotension during sedation, and these hemodynamic adverse events may adversely affect flap surgery because adequate tissue perfusion is an important factor in ensuring flap survival [2]. In addition to the above shortcomings, dexmedetomidine does not have a specific antagonist, so its sedative effect cannot be rapidly antagonized [5].

Remimazolam is a new type of water-soluble ultra-short-acting benzodiazepine that acts mainly on aminobutyric acid receptors, inhibits neuronal activity, reduces neuronal excitability, and thus produces a sedative effect, as well as anxiolytic and anterograde amnesic effects [6, 7]. Remimazolam has a sedative effect comparable to that of propofol; however, its respiratory and circulatory effects are minimal. In terms of pharmacokinetics, remimazolam reaches its peak blood concentration within approximately 1 min and is rapidly metabolized by tissue esterases through non-organ-dependent hydrolysis. Its metabolite zolampropionic acid has essentially no pharmacological activity and is mainly excreted by the kidneys [7]. Therefore, remimazolam has the advantages of a rapid onset of action, rapid metabolism, and no accumulation. Its pharmacological effects can also be antagonized by flumazenil. Therefore, we hypothesized that the advantages of remimazolam make it a desirable drug for postoperative sedation in patients undergoing flap surgery. However, studies on remimazolam for postoperative sedation after head and neck surgery are lacking.

Therefore, the main objective of this pilot study was to compare the efficacy and safety of remimazolam and dexmedetomidine for postoperative sedation in patients undergoing microvascular reconstruction, to provide reference data for subsequent clinical trials with a sufficiently large sample size.

Methods

Trial design

This single-center, prospective, randomized, double-blind, controlled clinical trial was conducted in the Department of Anesthesia at Peking University School of Stomatology from September 2021 to November 2022.

All procedures performed in this study conformed to the ethical guidelines of the Declaration of Helsinki, and ethical approval for this study (Number: PKUSSIRB-202060205) was provided by the Ethics Committee of the Peking University School of Stomatology, Beijing, China on February 2, 2021. The trial was registered in the Chinese Clinical Trial Registry, www.chictr.org.cn (Number: ChiCTR2100048342) on 05/07/2021. This manuscript adhered to the CONSORT guidelines.

Written informed consent was obtained before the administration of the study medication. Patients or their representatives can withdraw their consent at any stage.

Participants

Inclusion criteria

Patients who met the following criteria were eligible for the study:

Patients were eligible if they met all of the following criteria:

  1. 1)

    Aged 18–80 years;

  2. 2)

    Scheduled for elective oral and maxillofacial surgery with free flap reconstruction;

  3. 3)

    No postoperative prophylactic tracheostomy with endotracheal intubation maintained;

  4. 4)

    Voluntarily provided written informed consent.

Exclusion criteria

Patients were excluded if they:

  1. 1)

    With bradycardia before surgery (heart rate< 50 beats/min);

  2. 2)

    With atrioventricular block above II° before surgery;

  3. 3)

    With preoperative systolic blood pressure < 90mmHg (based on admission blood pressure); Long-term use of narcotic analgesics, sedatives, or nonsteroidal anti-inflammatory drugs;

  4. 4)

    Have schizophrenia, Parkinson's disease or are unable to communicate due to severe dementia or language impairment before surgery;

  5. 5)

    Severe hepatic dysfunction (Child-Pugh class C), severe renal dysfunction (preoperative dialysis), or American Society of Anesthesiology (ASA) physical status ≥ IV.

  6. 6)

    With a known history of hypersensitivity to α2 receptor agonists, remimazolam or other benzodiazepines

Randomization and blinding

Patients included in this study were randomly allocated to two parallel groups: dexmedetomidine (DEX group) and remimazolam (REM group).

A permuted block randomization scheme with undisclosed block sizes was generated, ensuring allocation concealment through sequentially numbered, opaque, sealed envelopes. Random numbers were generated in a 1:1 ratio and sealed in sequentially numbered opaque envelopes. The envelopes were opened before the end of surgery by an anesthesia nurse who prepared the study drugs but did not participate in the rest of the trial. The study drugs, remimazolam (Yichang Renfu Pharmaceutical Co., Ltd.) and/or dexmedetomidine (Yangtze River Pharmaceutical Group Co., Ltd.), were prepared with normal saline to a volume of 50 ml. Consequently, care providers, outcome assessors, and patients were blinded to the trial group assignments.

Perioperative management

Anesthesia

No premedication was administered in the patient ward. General anesthesia was induced with sufentanil 0.3 µg/kg, propofol 2 mg/kg or etomidate 0.2 mg/kg, and rocuronium 0.6 mg/kg. Nasotracheal intubation was performed. Anesthesia was maintained with a target-controlled infusion of propofol (2–6 ug/ml plasma concentration) and remifentanil (0.5-6ng/ml plasma concentration), with or without inhalational sevoflurane. Mechanical ventilation was achieved by using a mixture of oxygen and air. Depth of anesthesia was monitored using bispectral index (BIS) with a target range of 40–60. At the end of surgery, all patients received a standardized PCA regimen: sufentanil 1.5 µg/ml + tropisetron 5 mg in 100 ml normal saline, basal infusion 2 ml/h, bolus 2 ml, lockout 15 min.

After surgery, patients were transferred to the post-anesthesia care unit (PACU) with nasotracheal intubation. In the following morning, patients were extubated at 7: 30 am when they regained consciousness, fully recovered from paralysis, and had stable circulatory status. Then patients were transferred from the PACU to the general ward at 8: 00 am.

Intraoperative anesthetic management followed institutional standards, while postoperative sedation was strictly protocolized.

Administration of remimazolam and Dexmedetomidine

Sedation infusions with remimazolam or dexmedetomidine began immediately upon PACU arrival and continued until 6 am of the following morning. Sedation was initiated at 0.25 mg·kg⁻¹·h⁻¹ for remimazolam or 0.4 µg·kg⁻¹·h⁻¹ for dexmedetomidine and titrated every 10 min based on Ramsay Score assessments. Incremental adjustments of 0.05 mg·kg⁻¹·h⁻¹ (remimazolam) or 0.05 µg·kg⁻¹·h⁻¹ (dexmedetomidine) were made to maintain a Ramsay Score of 2–3. Drug concentrations were prepared to enable similar starting infusion rates (mL/h) for comparable weights, with adjustments based on Ramsay Score assessments. Maximum infusion rates were limited to 0.4 mg·kg⁻¹·h⁻¹ for remimazolam and 0.7 µg·kg⁻¹·h⁻¹ for dexmedetomidine.

Data collection and outcome assessment

Baseline data included demographic and morphometric characteristics, as well as other indicators. Intraoperative data included the duration of anesthesia, types and doses of medications, type and duration of surgery, and fluid balance.

After transferring patients to the PACU, the Richmond Agitation-Sedation Scale (RASS) score [9] was used to evaluate the occurrence of emergence agitation. Emergence agitation was defined as RASS ≥ + 2 lasting > 5 min. The RASS, which was proposed by Sessler in 2002, is a 10-point scale with four levels of anxiety or agitation (+ 1 to + 4 [combative]), one level of calm and alert state (0), five levels of sedation (− 1 to − 5), and an inability to wake up (− 5). In addition, sedation depth was monitored using BIS with target range 60–80.

Delirium was evaluated using the Confusion Assessment Methods (CAM) [10]twice daily (8:00–10:00 and 18:00–20:00) during the first five postoperative days. Pain severity was evaluated using a numeric rating scale (NRS) (0 = no pain, 10 = worst pain). Subjective sleep quality was evaluated using the NRS (0 indicated the worst sleep and 10 indicated the best sleep). To assess postoperative nausea and vomiting (PONV), nausea was diagnosed by direct questioning, with an 11-point NRS ranging from 0 (no nausea) to 10 (worst nausea). Vomiting was diagnosed when the patient retched or expelled the intragastric content.

Postoperative complications were defined as new conditions that occurred after surgery, adversely affected the patient’s recovery, and required medical intervention (Clavien-Dindo grade 2 [11]).

The primary outcome of the trial was the incidence of emergence agitation in the PACU. Secondary outcomes included the incidence of delirium, incidence and severity of postoperative pain and PONV at 0–6 h, 6–12 h, and 12–24 h postoperatively, sleep quality within 2 days after surgery, length of postoperative hospital stay, and morbidity and mortality during postoperative hospitalization.

Sample size estimation

As a pilot randomized trial, the sample size was determined according to the precision analysis method recommended for feasibility studies. Based on historical data from our center (2018–2020), the incidence of emergence agitation after oral free flap reconstruction was 38–45% with dexmedetomidine sedation [2]. Assuming a 15% absolute risk reduction with remimazolam (23–30% incidence), a sample size of 29 per group would provide 80% probability that the 95% confidence interval (CI) width for the risk difference will be ≤ 30%. Considering a 10% dropout rate, we planned to enroll 32 patients per group (total n = 64). However, due to Corona Virus Disease 2019 (COVID-19)-related recruitment constraints during the study period (September 2021-November 2022), the final sample size reached 29 vs. 28 with 60 randomized, achieving 89% of the target enrollment. Despite pandemic-related recruitment challenges, the final sample size (n = 57) achieved 89% of the pre-specified target, maintaining adequate statistical precision for preliminary analysis.

Statistical analysis

Normally distributed continuous variables were presented as mean ± SD and were compared using an independent t-test. Non-normally distributed continuous variables are expressed as medians (interquartile range (IQR)) and compared using the Mann-Whitney U test. Categorical variables were expressed as numbers (percentages) and analyzed using the chi-square test (or Fisher’s exact test, where appropriate). Risk differences with 95% confidence intervals (CIs) were calculated using the Newcombe-Wilson method to account for small sample sizes. All analyses were performed using the SPSS version 21.0. A two-tailed p < 0.05 was considered statistically significant.

Results

Participant flow and recruitment

From September 2021 to November 2022, 60 patients were randomized into two groups. However, one patient in the REM group withdrew content and two patients in the DEX group withdrew content, leaving data from 29 to 28 patients in the REM and DEX groups, respectively (Fig. 1).

Fig. 1
figure 1

Flow diagram of patients through trial

Full size image

Baseline patient demographic and perioperative characteristics

Overall, the two groups were well matched on all the variables.

The baseline demographic and preoperative characteristics of patients were similar. (Table 1). Notably, the dexmedetomidine group demonstrated higher baseline prevalence of hypertension (32.1% vs. 17.2%, p = 0.230) and coronary artery disease (10.7% vs. 3.4%, p = 0.352), though these differences did not reach statistical significance.

Table 1 Baseline patient demographic and preoperative characteristics
Full size table

The mean total sedative dosage was 0.28 ± 0.04 mg·kg⁻¹·h⁻¹ in the remimazolam group and 0.42 ± 0.07 µg·kg⁻¹·h⁻¹ in the dexmedetomidine group. Regarding perioperative characteristics (Table 2), more patients in the DEX group received sevoflurane, whereas the etomidate dosage was significantly higher in the REM group than in the DEX group (p < 0.05). In the PACU, the use of colloids was lower in the REM group than in the DEX group (p < 0.05).

Table 2 Perioperative characteristics
Full size table

Outcomes

The incidence of emergence agitation was 27.6% (8/29) in the remimazolam group versus 25.0% (7/28) in the dexmedetomidine group, with a risk difference of 2.6% (95% CI: −21.5% to + 26.7%; p > 0.999). The risk difference of 2.6% (95% CI: −21.5x% to + 26.7%) indicated no statistically significant between-group difference.

The secondary outcomes, including postoperative delirium, pain, nausea and vomiting, sleep duration, adverse events, complications, and length of hospital stay, are shown in Table 3, but without significant differences.

Table 3 Primary and secondary outcomes
Full size table

Discussion

This randomized trial establishes that remimazolam and dexmedetomidine provide statistically equivalent prevention of emergence agitation (27.6% vs. 25.0%, p > 0.999) in oral cancer patients undergoing microvascular free flap reconstruction. The most clinically significant finding was remimazolam’s 90% reduction in colloid requirements in PACU (3.4% vs. 32.1%, p = 0.005), signifying superior hemodynamic stability. This advantage carries substantial implications for microvascular surgery, where excessive fluid administration may induce interstitial edema compromising flap perfusion or increase airway compromise risk in head and neck reconstructions, as emphasized in Kovatch et al.‘s multicenter analysis of fluid management protocols for free flap patients [13].

The colloid-sparing effect likely originates from fundamental pharmacological differences between the agents. Remimazolam’s selective GABA_A receptor agonism (α1-subunit preference) confers minimal cardiovascular perturbation [15], contrasting sharply with dexmedetomidine’s α₂-adrenoceptor-mediated sympatholytic effects that frequently cause hypotension necessitating fluid boluses [2, 4]. This mechanistic distinction is supported by Wesolowski et al.‘s pharmacodynamic analysis documenting dexmedetomidine’s dose-dependent blood pressure reduction (> 35% MAP decrease at plasma concentrations > 1.2 ng/mL) [14]. Furthermore, remimazolam’s rapid hydrolysis by tissue esterases (elimination half-life: 0.75 h vs. dexmedetomidine’s 2.1 h [7]) enables metabolic stability independent of hepatic function [20], whereas dexmedetomidine’s context-sensitive half-life prolongation during prolonged infusion (> 45 min after 6-h infusion [14]) exacerbates hypotension risk. These pharmacokinetic profiles align with Gao et al.‘s bronchoscopy study demonstrating 40% lower systolic blood pressure variability with remimazolam (p < 0.01) [23], and are further corroborated by Chen et al.‘s kinetic modeling of remimazolam’s organ-independent clearance [19].

Our findings integrate with emerging comparative evidence across surgical contexts. In non-intubated orthopedic patients, Deng et al. reported comparable sedation efficacy but 68% lower hypotension incidence with remimazolam (RD: −18.2%; 95% CI: −31.4% to −5.0%) [22]. Similarly, Hong et al. observed 35% reduced vasopressor requirements during spinal anesthesia with remimazolam while maintaining equivalent sedation depth [24]. Pediatric studies by Shioji et al. further validate cardiovascular safety, showing fewer bradycardia episodes with remimazolam during MRI sedation (4% vs. 27%, p < 0.01) [17]. Tang et al.‘s meta-analysis consolidates this hemodynamic advantage, calculating a pooled OR of 0.28 (95% CI: 0.15–0.52) for hypotension with remimazolam versus dexmedetomidine [7]. Notably, despite dexmedetomidine’s established anti-delirium properties via α₂-mediated locus coeruleus suppression [8, 26], our data show remimazolam achieved equivalent agitation control (6.9% vs. 10.7% delirium, p = 0.670). This parity may reflect remimazolam’s rapid titration capability (peak effect: 1 min [7]), enabling dynamic response to agitation triggers—unlike dexmedetomidine’s 6–8 min equilibration half-time [14]. Kim et al.‘s recent trial corroborates this, showing 50% faster agitation resolution with remimazolam titrated to RASS − 1 to −2 [25].

Two methodological considerations warrant contextualization. First, higher etomidate use in the remimazolam group (13.0 mg vs. 0 mg, p = 0.030) introduces potential confounding, as etomidate’s GABAergic potentiation may synergize with benzodiazepines [27]; Sondekoppam et al. reported etomidate-induced agitation in 22% of patients receiving benzodiazepine co-administration [27]. Second, universal sevoflurane exposure in dexmedetomidine recipients (100% vs. 79.3%, p = 0.023) could mask benefits, given sevoflurane’s NMDA receptor antagonism contributing to emergence agitation [1, 26]. Future studies should validate remimazolam’s effects on endothelial biomarkers (e.g., IL-6, VEGF [23]) to elucidate microvascular protective mechanisms, compare extubation times leveraging its rapid offset (context-sensitive half-time: 7–8 min vs. dexmedetomidine’s 30–45 min [7]), and evaluate fluid-restrictive protocols integrating remimazolam into goal-directed algorithms for free flap surgery [13, 29].

Limitations of This Study​:

Several limitations warrant consideration when interpreting our findings:

  1. 1.

    ​Sample Size Constraints: As a pilot trial, the study was powered for precision analysis rather than superiority detection. The final enrollment (n = 57) reached 89% of the target sample due to COVID-19-related recruitment challenges, potentially limiting statistical power for secondary outcomes like delirium incidence (6.9% vs. 10.7%, p = 0.670).

  2. 2.

    ​Pharmacological Confounders: The significantly higher etomidate usage in the remimazolam group (13.0 mg vs. 0 mg, p = 0.030; Table 2) may have potentiated GABAergic effects [27], while universal sevoflurane exposure in dexmedetomidine recipients (100% vs. 79.3%, p = 0.023) could mask its anti-agitation properties due to known sevoflurane-agitation associations [1, 26].

  3. 3.

    ​Single-Center Design: Conducted at a tertiary stomatology center, our findings may lack generalizability to non-specialized settings or diverse surgical populations.

  4. 4.

    ​Fixed-Dosing Protocol: Sedation was titrated to Ramsay 2–3 without protocolized BIS guidance, though BIS monitoring confirmed target depth (60–80).

  5. 5.

    ​Short-Term Outcomes: Assessments were limited to the PACU and early postoperative period (≤ 5 days), precluding evaluation of long-term cognitive effects or flap survival outcomes.

  6. 6.

    ​Economic Analysis: Cost-effectiveness comparisons between agents were not performed, though remimazolam’s reduced colloid requirements suggest potential resource savings.

Conclusion

Remimazolam matches dexmedetomidine in preventing emergence agitation while significantly reducing colloid resuscitation needs. Combined with rapid metabolism and neutral hemodynamic effects, remimazolam represents a viable first-line sedative for microvascular reconstruction, warranting validation in larger trials (ChiCTR2100048342).