Remimazolam versus propofol for procedural sedation in hysteroscopic surgery: a meta-analysis of randomized controlled trials

Abstract

Procedural sedation is widely used to improve patient tolerability and satisfaction and minimize complications. Propofol is the most commonly used agent in procedural sedation, but its use is associated with a number of adverse effects. Remimazolam, which has a different mechanism of action compared to propofol, is a short-acting GABA-A receptor agonist. It is being increasingly used in procedural sedation. This meta-analysis aims to clarify the efficacy and safety of remimazolam versus propofol for procedural sedation in hysteroscopic surgery. We systematically searched studies published in PubMed, EMBASE, Web of Science, and the Cochrane Central Register of Controlled Trials from the inception of the databases to February 28, 2025, for randomized controlled trials (RCTs) comparing the effect of remimazolam and propofol in procedural sedation for hysteroscopic surgery. Seven RCTs (n = 664 patients) met our inclusion criteria. The pooled results showed that remimazolam significantly reduced the risk of injection pain, hypoxemia, and hypotension (all P < 0.0001) in hysteroscopic surgery, with a sedation success rate comparable to that of propofol (P = 0.05). There were no significant differences in the risks of postoperative dizziness, postoperative nausea and vomiting (PONV), or bradycardia. Findings from sensitivity analyses suggested a potentially higher risk of body movement with remimazolam. We conclude that the sedation success rate of remimazolam is non-inferior to that of propofol, and patients who receive remimazolam for hysteroscopy exhibit lower incidences of hypotension, hypoxemia, and injection pain than patients who receive propofol. Therefore, remimazolam represents a safer alternative to propofol for procedural sedation in hysteroscopic procedures, with a distinct safety profile. Nevertheless, its risk-benefit profile should be carefully evaluated in clinical practice.

Introduction

With the rapid development of medicine and the advancement of comfort-oriented care, hysteroscopic surgery has been widely adopted in modern gynecological practice for both diagnostic and therapeutic purposes, owing to advantages such as minimal invasiveness and rapid recovery (Salazar & Isaacson, 2018). Although outpatient hysteroscopic procedures are relatively simple and require less time to perform, they impose stricter requirements on anesthetic safety and efficacy. Hysteroscopy is considered the gold standard for evaluating intrauterine lesions and abnormal uterine bleeding. The use of sedative drugs plays a pivotal role in the diagnosis and treatment of gynecological diseases (Clark, 2021). Consequently, the judicious selection of anesthetic drugs is crucial to ensure procedural success, relieve pain, and ensure safety (Ma et al., 2017).

Various sedatives are administered during hysteroscopy, including benzodiazepines, opioids (meperidine and fentanyl), propofol, ketamine, and other agents (Crespo & Terán, 2018). These drugs are widely used for hysteroscopic sedation to alleviate patient anxiety and ensure successful procedure completion. In China, propofol has gradually become the most popular sedative used by anesthesiologists and endoscopists in endoscopic surgery (Zhou et al., 2021) because it has a rapid onset and short half-life, can reduce anxiety and pain in patients, and improve comfort (Bingol Tanriverdi et al., 2019). However, the use of propofol is associated with adverse effects, such as injection pain, hypoxemia, and circulatory inhibition, which may increase the incidence of intraoperative and postoperative unexpected risks, such as hypoxemia, hypotension, and even cardiac arrest, in patients undergoing endoscopic procedures (Daza et al., 2018).

As a new type of benzodiazepine, remimazolam has characteristics such as rapid onset, rapid metabolism, and a low incidence of adverse reactions (Philip, Brohan & Goudra, 2024). Remimazolam has better hemodynamic stability and has caused fewer incidences of injection pain than propofol, and flumazenil, a postoperative antagonist, can better shorten the recovery time (Brohan, Brohan & Goudra, 2024). Propofol does not have an antagonist (Sneyd & Rigby-Jones, 2020). Remimazolam has been successfully used for endoscopy and surgery as well as for the induction and maintenance of general anesthesia. The onset and recovery time of remimazolam is short and predictable, the respiratory and hemodynamic fluctuations caused by remimazolam are minor, and no serious drug-related adverse reactions have been reported (Chang et al., 2023; Wu et al., 2023; Zhang et al., 2022b). In a meta-analysis of the safety and efficacy of remimazolam and propofol in intravenous anesthesia for endoscopic surgery, the use of remimazolam was found to reduce postoperative injection pain, hypotension, and respiratory depression compared with the use of propofol. A recent meta-analysis compared the safety and efficacy of remimazolam and propofol for intravenous anesthesia in elderly patients undergoing gastrointestinal endoscopic surgery. The findings showed that remimazolam was a safer alternative to propofol, causing a lower incidence of hypotension, respiratory depression, injection pain, and bradycardia (Li et al., 2024). A recent systematic review and meta-analysis comparing the efficacy and adverse effects of remimazolam and propofol in gastrointestinal endoscopy demonstrated that while remimazolam significantly reduced the incidence of hypoxia and bradycardia compared to propofol, its use was associated with higher rates of tachycardia and body movements (Kim et al., 2025). In this meta-analysis, we systematically evaluated the efficacy and safety of remimazolam and propofol in hysteroscopic surgery, determined the risk of adverse reactions associated with remimazolam and propofol, and explored the sedative effects of both. While previous meta-analyses have compared the effects of remimazolam and propofol in various endoscopic procedures, this study is the first comprehensive evaluation focusing specifically on their effects in hysteroscopic surgery. We attempted to systematically evaluate the differences in safety profiles between remimazolam and propofol in hysteroscopic procedures, thereby providing high-quality evidence-based medical guidance for clinical decision-making.

Methods

Research methods

We presented our findings in accordance with the Cochrane Handbook for Systematic Reviews of Interventions and the Preferred Reporting Items of the Guidelines for Systematic Reviews and Meta-Analyses (PRISMA). Our review is registered in PROSPERO (CRD42024581005).

Search strategy

Two authors searched for studies on the use of remimazolam in hysteroscopic surgery published in PubMed, Embase, Cochrane, and Web of Science databases on or before February 28, 2025. The search terms included “remimazolam” and “hysteroscopy” as subject words and the associated free words. The search scope was “Title and Summary”. The detailed search strategy is comprehensively documented in the Supplementary Materials. We attempted to review all studies on the efficacy and safety of remimazolam versus propofol in hysteroscopy.

Types of studies included

The inclusion criteria were as follows: (a) subjects: patients undergoing hysteroscopy; (b) intervention: remimazolam was used as an intravenous anesthetic for complete anesthesia; (c) control measures: other sedative drugs were used as intravenous anesthetics; (d) study outcomes: the incidence of postoperative adverse events, length of stay in the post-anesthesia care unit, and length of discharge were assessed; (e) study design: the inclusion criterion was a clinical randomized controlled trial (RCT) study design.

The exclusion criteria were as follows: (a) text not written in Chinese or English; (b) non-RCT study design; (c) duplicate literature; (d) reviews, abstracts, conference documents, letters, basic research, and duplicate documents, among others. (e) unavailable full text, experiments underway, and studies with incomplete data; (f) combination with other treatments or drugs.

Literature screening

The researchers independently screened the literature according to predefined inclusion and exclusion criteria. The selection process comprised four steps: (a) all identified records were imported into EndNote X9 (Clarivate Analytics) for duplicate removal; (b) the title and abstract of the literature were read and the literature was selected according to the inclusion and exclusion criteria; (c) the full text was read, and studies with unavailable original data were excluded; (d) the included documents were finally determined, and data extraction and analysis was performed thereafter.

Data extraction

The author, year of publication, and country of publication of each study were retrieved. The number of participants in the experimental and control groups of each study was determined. The specific intervention received by the participants (including the name of the drug), the induction and maintenance doses and dosing regimen, and the number of patients in the remimazolam and control groups, were also noted. Data and outcome indicators were recorded and collated by two independent researchers (Jing Liu and Keke Lu) for the included studies. The two researchers independently extracted data from all eligible studies and resolved discrepancies through discussion with the co-author (Jiahang Chen). If no consensus could be reached, the final decision was made by the corresponding author (Junjie Song).

Outcome measures

Our primary outcome indicator was the comparison of the safety of remimazolam and propofol by analyzing the incidence of adverse events after remimazolam and propofol administration, respectively. In addition, as secondary outcomes, we assessed the efficacy of both sedatives by measuring their success rate, the time to recovery from sedation, and the time to discharge.

Quality assessment

The quality of the included literature was assessed according to the Cochrane bias risk assessment tool in the Cochrane Handbook 5.1.0 for Systematic Review. The Cochrane bias risk assessment included the following seven items: random sequence generation (selection bias), allocation concealment (selection bias), participant and person blindness (performance bias), outcome assessment blindness (detection bias), incomplete outcome data (attrition bias), selective reporting (reporting bias), and others (bias owing to acquired financial benefits and academic bias). For each item, the included RCTs were evaluated with a low risk of bias, a high risk of bias, and an unclear risk of bias.

Statistical analysis

Review Manager 5.4.1 software was used to create the bias risk map and perform the analysis. Heterogeneity was assessed using I² and P values. The values of I² < 25%, 25% ≥ I² ≥ 50%, and I²> 50% indicated low, medium, and high heterogeneity, respectively. For continuous variables, the results were represented using standardized mean difference (SMD) and 95% confidence intervals (CIs). For binary variables, the odds ratio (OR) was used to represent the summary effect. A P-value of ≤ 0.05 was considered statistically significant for all outcomes.

Handling of multi-arm trials

We followed the recommendations provided in the Cochrane Handbook for RCTs with multiple intervention arms (e.g., different doses of remimazolam) sharing a single control group (propofol) (Section 16.5.4). To avoid unit-of-analysis errors and the duplication of control data, we combined the intervention arms to create a single pairwise comparison for the primary analysis. For dichotomous outcomes, the sample sizes and number of events from all remimazolam arms were summed. For continuous outcomes, the means, standard deviations, and sample sizes of the remimazolam arms were combined using the formulae provided in the Cochrane Handbook (Section 7.7.3.8). This resulted in a single pooled experimental group that had to be compared against the shared control group. This approach addresses the review question of “remimazolam versus propofol” while preserving data integrity.

Results

Search results

A selective literature search yielded 127 potentially eligible articles, including 34 articles from Embase, 50 articles from Cochrane, 21 articles from PubMed, and 22 articles from Web of Science. Through systematic screening, we excluded 62 duplicate records, eight meta-analyses, one letter to the editor, and 30 trial registrations. After reviewing titles and abstracts, we eliminated six studies irrelevant to our meta-analysis objectives. Based on a full-text assessment of the remaining 11 articles, we excluded three non-randomized controlled trials and one study investigating E50. Following the comprehensive evaluation of all potentially eligible studies, we selected seven RCTs for this meta-analysis (Fan et al., 2023; Lin, Chen & Liang, 2024; Xie et al., 2024; Zhang et al., 2022a; Zhang, Li & Liu, 2021). The flow diagram of study selection is shown in Fig. 1.

Figure 1. Flow diagram for study identified and included into this meta-analysis.

Study characteristics

Seven studies with a total of 664 patients were included; all studies were RCTs (published between May 2021 and February 2025). Based on the type of sedative used for hysteroscopy, we divided the patients in each study into two groups: remimazolam and propofol. In the remimazolam group, 273/281 patients were successfully sedated, whereas in the propofol group, 254/255 patients were successfully sedated. In addition, the incidence of adverse events, especially injection pain, hypoxic saturation, and hypotension, among others, was documented extensively. The clinical characteristics of the included studies are shown in Table 1.

Table 1. Baseline characteristics of the included study population.

Authors, year	Country	Study period	Study design	Age	BMI (kg/m2)	ASA	Dose of medicine
							Dose of Propofol	Dose of Remimazolam
Zhang, Li & Liu (2021)	China	2020.9–2020.12	RCT	18–65	19–30	I–II	Induction dose: 1.5∼2.0 mg/kg, maintenance dose: 3.0∼6.0 mg/kg/h (n = 41)	Induction dose: 0.2 mg/kg, Maintenance dose: 1.0 mg/kg) (n = 41)
Zhang et al. (2022a)	China	2022.2–2022.7	RCT	18–60	NA	I–II	Induction dose: 1.5 mg/kg, maintenance dose: 4∼10 mg/kg/h (n = 64)	Induction dose: 0.2 mg/kg, maintenance dose: 0.4∼1.0 mg/kg) (n = 64)
Zhang et al. (2022c)	China	2020.11–2021.6	RCT	18–55	18–30	I–II	Induction dose: 2 mg/kg, maintenance dose: 5 mg/kg/h (n = 30)	Induction dose: 0.25 mg/kg, maintenance dose: 0.48 mg/kg/h) (n = 30)
								Induction dose: 0.25 mg/kg maintenance dose: 0.6 mg/kg/h) (n = 30)
Fan et al. (2023)	China	2022.3–2022.9	RCT	20–60	NA	I–II	Induction dose: 2.5 mg/kg, Each subsequent addition: 0.5 ∼1 mg/kg (n = 43)	Induction dose: 0.25 mg/kg, Each subsequent addition: 2.5 mg (n = 40)
Lin, Chen & Liang (2024)	China	NA	RCT	18–50	18.5–30	I–II	Induction dose: 2.27 mg/kg, maintenance dose: 1 mg/kg/h (n = 47)	Induction dose 0.2 mg/kg, maintenance dose: 1 mg/kg/h (n = 48)
Xie et al. (2024)	China	2021.6–2022.5	RCT	>60	NA	II–III	Induction dose: 2 mg/kg, maintenance dose:4∼8 mg/kg/h (n = 30)	Induction dose: 0.2 mg/kg, maintenance dose: 0.5∼1 mg/kg/h (n = 30)
Shan et al. (2024)	China	2023.3–2023.7	RCT	18–65	19–30	I–II	Induction dose: 2 mg/kg, maintenance dose:3∼6 mg/kg/h (n = 64)	Induction dose: 0.2 mg/kg, maintenance dose: 1 mg/kg/h (n = 62)

Risk of bias

We used the Cochrane Collaboration tool to assess the risk of bias in all included RCTs. The quality assessment of this method is shown in Fig. 2. Across all included studies, the authors’ judgments on each risk of bias item are presented as percentages in Fig. 3. The Cochrane Collaborative Bias Risk Assessment tool was used to assess the risk of bias in RCTs. The risk of bias assessment of the seven included studies yielded inconsistent results. Most studies demonstrated a low risk of bias with respect to random sequence generation (six studies, 85.7%), incomplete outcome data (seven studies, 100%), and selective reporting (seven studies, 100%). The body of evidence showed some methodological heterogeneity, particularly with respect to the reporting of safeguards against bias. The blinding of participants, personnel, and outcome assessors was not universally thorough, and the allocation concealment was not always explicitly described. To assess the potential impact of these limitations on the results, we conducted subgroup analyses. The stability of the effect estimates in these analyses reassured us of the directional robustness of our primary conclusions, particularly for objective endpoints, though precision varied across analyses.

Figure 2. Risk of bias graph.

Figure 3. Risk of bias summary.

Adverse reactions

Injection pain

Seven trials involving 664 patients reported details of injection pain associated with remimazolam and propofol use during hysteroscopic surgery (Fan et al., 2023; Lin, Chen & Liang, 2024; Shan et al., 2024; Xie et al., 2024; Zhang et al., 2022a; Zhang et al., 2022c; Zhang, Li & Liu, 2021). The findings of this meta-analysis demonstrated a significant reduction in the incidence of injection pain with remimazolam than with propofol (OR = 0.03; 95% CI [0.01–0.07]; P < 0.00001). The included studies showed moderate heterogeneity (I² = 35%) (Fig. 4).

Figure 4. Forest plot comparing adverse outcomes, including incidence of injection pain, hypoxemia, hypotension, bradycardia, PONV, dizziness, Body movement.

Hypoxemia

Four trials involving 426 patients reported hypoxemia associated with the two sedative regimens (Shan et al., 2024; Zhang et al., 2022a; Zhang et al., 2022c; Zhang, Li & Liu, 2021). Based on the limited trial data available, our analyses indicated that remimazolam anesthesia was associated with a significantly lower incidence of hypoxemia during hysteroscopic procedures than propofol anesthesia (OR = 0.14; 95% CI [0.06–0.29]; P < 0.00001). The included studies showed low heterogeneity (I² = 16%) (Fig. 4).

Hypotension

Four trials involving 431 patients reported details of hypotension associated with remimazolam and propofol use during hysteroscopic surgery (Lin, Chen & Liang, 2024; Shan et al., 2024; Zhang et al., 2022a; Zhang, Li & Liu, 2021). Our meta-analysis demonstrated a significantly lower incidence of hypotension in the remimazolam group than in the propofol group (OR = 0.16; 95% CI [0.09–0.28]; P < 0.00001). The included studies showed low heterogeneity (I² = 0%) (Fig. 4).

Bradycardia

Three trials involving 336 patients reported detailed information on bradycardia associated with remimazolam and propofol use in hysteroscopic surgery (Shan et al., 2024; Zhang et al., 2022a; Zhang, Li & Liu, 2021). Our meta-analysis suggested that the incidence of bradycardia in the remimazolam group was not statistically significant compared with that in the propofol group (OR = 0.34; 95% CI [0.07–1.77]; P = 0.20). The included studies showed low heterogeneity (I² = 0%) (Fig. 4).

Postoperative nausea and vomiting

Three trials involving 278 patients reported details of postoperative nausea and vomiting (PONV) associated with remimazolam and propofol use in hysteroscopic surgery (Xie et al., 2024; Zhang et al., 2022a; Zhang et al., 2022c). The meta-analysis revealed no statistically significant difference in PONV incidence between the groups (OR = 1.44; 95% CI [0.25–8.11]; P = 0.68). The included studies showed low heterogeneity (I² = 0%) (Fig. 4).

Postoperative dizziness

Three RCTs involving 270 patients evaluated postoperative dizziness events caused by remimazolam- and propofol-based sedation in hysteroscopic procedures (Xie et al., 2024; Zhang et al., 2022a; Zhang, Li & Liu, 2021). The meta-analysis demonstrated a non-significant trend toward the reduction of dizziness incidence with remimazolam (OR = 0.17; 95% CI [0.02–1.82]; P = 0.14). The included studies showed high levels of heterogeneity (I² = 79%). Sensitivity analysis excluding one outlier study (Xie et al., 2024) eliminated heterogeneity (I²= 0%, P for heterogeneity <0.01), with the recalculated effect size maintaining consistent direction (OR =0.06, 95% CI [0.02–0.22]). However, this finding should be interpreted with caution, as the results remained sensitive to the inclusion or exclusion of individual studies, and the precision of the effect estimate varied substantially when the outlier was removed (Fig. 4).

Body movement during surgery

Four trials involving 396 patients reported detailed information on body movements associated with remimazolam and propofol use in hysteroscopic surgery (Shan et al., 2024; Xie et al., 2024; Zhang et al., 2022a; Zhang, Li & Liu, 2021). We performed sensitivity analyses by sequentially excluding each study to identify heterogeneity sources. Despite exhaustive sensitivity analyses (such as the sequential exclusion of each study), substantial heterogeneity persisted. The initial pooled analysis indicated a directionally consistent higher incidence of body movement in the remimazolam group, although this difference did not show conventional statistical significance (OR = 2.79; 95% CI [0.76–10.17]; P = 0.12). High heterogeneity was observed among the studies (I² = 81%) (Fig. 4). We performed leave-one-out sensitivity analyses and influence diagnostics. According to these analyses, a study by Zhang, Li & Liu (2021) was a highly influential outlier. After we excluded this study, the pooled estimate shifted substantially, showing a statistically significant increase in the risk of body movement with remimazolam (OR = 4.90; 95% CI [1.45–16.52]; P = 0.01). Heterogeneity was notably reduced but remained substantial (I² = 62%), indicating that the precision of this estimate is not fully stable. The study by Zhang, Li & Liu (2021) acted as a conservative outlier, as it was the only study reporting a lower (though non-significant) point estimate for body movement with remimazolam. Its relatively large sample size and divergent effect direction exerted a strong downward pull on the initial pooled estimate, obscuring the clearer trend observed in the other three studies. Taken together, these sensitivity analyses suggest that while the direction of effect is robust, the magnitude and precision of the pooled estimate are sensitive to individual studies. Therefore, the findings indicate that remimazolam may be associated with a higher risk of intraoperative body movement compared to propofol in hysteroscopy, although this conclusion requires confirmation in future studies with standardized outcome definitions.

Anesthesia recovery time

Five trials involving 492 patients reported details about remimazolam and propofol leading to different postoperative anaesthetic recovery times after hysteroscopy (Fan et al., 2023; Lin, Chen & Liang, 2024; Shan et al., 2024; Xie et al., 2024; Zhang et al., 2022a), The meta-analysis did not reveal a statistically significant difference in recovery times (MD = 1.20; 95% CI [−0.11–2.51]; P = 0.07). Heterogeneity was extremely high in the included studies (I² = 98%). We performed sensitivity analyses by sequentially excluding each study to identify heterogeneity sources. Despite exhaustive sensitivity analyses (i.e., sequential exclusion of each study), substantial heterogeneity persisted among the studies (I² remained 86% after all exclusion attempts), indicating that the heterogeneity remains substantial and largely unexplained by available data (Fig. 5).

Figure 5. Forest plot comparing procedural outcomes, including anesthesia recovery time, discharge time.

Discharge time

Two RCTs involving 185 patients compared postprocedural discharge times between remimazolam and propofol-based sedation following hysteroscopy (Lin, Chen & Liang, 2024; Zhang et al., 2022c). The meta-analysis indicated significantly shorter discharge times in the remimazolam group (MD −2.11 min; 95% CI [−3.70 to −0.53]; P = 0.009), with no observed heterogeneity (I² = 0%). The interpretation of this finding requires caution owing to the limited number of studies (Fig. 5).

Success rate of anesthesia and sedation

Six RCTs involving 536 patients compared the efficacy of remimazolam versus propofol for achieving successful anesthetic sedation during hysteroscopic procedures (Fan et al., 2023; Lin, Chen & Liang, 2024; Shan et al., 2024; Xie et al., 2024; Zhang, Li & Liu, 2021). The meta-analysis showed no statistically significant difference in sedation success rates between groups (OR = 0.21; 95% CI [0.04–0.97]; P = 0.05). We observed no heterogeneity among the studies (I² = 0%) (Fig. 6).

Figure 6. Forest plot comparing success rate of anesthesia and sedation.

Discussion

This meta-analysis compared the efficacy and safety of remimazolam versus propofol for hysteroscopic procedures. There was no statistically significant difference in the sedation success of remimazolam and propofol. The incidence of hypotension, injection pain, and postoperative dizziness was reduced with remimazolam, and its discharge time after anesthesia was shorter than that of propofol. However, we observed no differences in the incidence of bradycardia, postoperative nausea and vomiting, and body movement during surgery. A meta-analysis investigating the safety and efficacy of endoscopic procedures showed that remimazolam is a safe and effective sedative for patients undergoing endoscopic procedures (Zhu et al., 2021). A conclusion similar to our study indicates that remimazolam may be safe and exhibit satisfactory efficacy in hysteroscopy.

Injection pain

While injection pain is not considered a major complication, it frequently induces patient anxiety, fear, and physical or emotional stress. This pain can lead to systemic adverse effects. The findings of the pooled results demonstrated that remimazolam significantly reduced the incidence of injection pain during hysteroscopic procedures compared to propofol. Propofol is an alkylphenol (2,6-diisopropylphenol). The immediate pain caused by propofol injection is owing to the stimulation of the venous endothelium and delayed pain caused by the release of mediators such as kininogen from the kinin cascade (Desousa, 2016). Remimazolam is an ester-based benzodiazepine (Kim & Fechner, 2022), which is supposed to reduce injection pain in theory. In addition, the short metabolic half-life of remimazolam may reduce local and in vivo accumulation and the release of active metabolites (Wesolowski et al., 2016). Multiple published studies support our findings that remimazolam significantly reduces injection pain compared to propofol. In a meta-analysis involving intravenous anesthesia in endoscopic surgery with remimazolam versus propofol, remimazolam caused a more significant reduction in injection pain than propofol (Zhao et al., 2023). Similarly, Li et al. (2024) reported reduced injection pain in elderly patients undergoing gastroenteroscopy with remimazolam-based sedation. These consistent results across different clinical settings confirm that remimazolam offers superior injection tolerability than propofol.

Hypotension

Intraoperative hypotension during non-cardiac surgery is common and associated with a greater likelihood of major adverse cardiac or cerebrovascular events at 30 days (Gregory et al., 2021). The smaller reduction in mean arterial blood pressure after the intravenous administration of remimazolam compared to propofol was discussed in two articles (Frölich et al., 2011; Lim et al., 2012). This may be attributed to the fact that many anesthetics, including propofol, induce an elevation of intracellular calcium, as these anesthetics have different effects on the intracellular calcium homeostasis of endothelial cells, and therefore lead to different haemodynamic effects (Urabe et al., 2022). One reason for this is that propofol acts directly on organelles, such as the endoplasmic reticulum. This increases the concentration of calcium within the cell, which in turn activates intracellular signaling pathways. It also promotes the phosphorylation of NO synthase, which induces vasodilation and leads to NO synthesis. This phenomenon may be related to excessive hypotension and propofol-induced vascular pain (Urabe et al., 2020). Urabe et al. (2022) studied the effects of remimazolam on intracellular calcium concentrations. They reported that remimazolam increased calcium concentrations in endothelial and neuronal cells. They investigated the regulatory effects of remimazolam and propofol on calcium ions in malignant hyperthermia and reported that this effect is reversible with remimazolam. The administration of propofol resulted in the fragmentation and aggregation of the endoplasmic reticulum and mitochondria, accompanied by an increase in calcium levels. These alterations persisted throughout the observation period, suggesting that the propofol-induced increase in calcium levels was attributable to calcium leakage from these organelles and was irreversible (Watanabe et al., 2021). These findings may explain the different hemodynamic effects of remimazolam and propofol and provide further evidence supporting our research. In an analytical retrospective study comparing the hemodynamic effects of remimazolam and propofol during general anesthesia, the mean arterial pressure reduction rate of remimazolam was lower than that of propofol, and the incidence of intraoperative hypotension was 32.1% for remimazolam and 67.9% for propofol (Tsukimoto et al., 2024). This is consistent with our findings that remimazolam exerts a less significant effect than propofol on circulation.

Hypoxemia

During surgery, respiratory depression and hypoxia caused by sedatives are usually temporary but can cause serious complications in elderly patients with poor cardiopulmonary reserve. In a study involving elderly patients undergoing gastroscopy, the incidence of respiratory depression was significantly lower in the remimazolam tosilate (0.2 mg/kg) group than in the propofol (1.5 mg/kg) group (9.8% vs 17.9%, P = 0.042) (Hu et al., 2022). In another study on the safety and efficacy of remimazolam versus propofol during gastrointestinal endoscopy, the incidence of cardiopulmonary adverse events was lower in the remimazolam group than in the propofol group (8.5% vs 16%, P = 0.022) (Choe et al., 2024). In the current analysis, only 12 patients in the remimazolam group experienced hypoxemia, which was significantly lower than the number of patients in the propofol group who experienced the same (n = 63) (P < 0.00001). This may be related to alterations in cardiopulmonary physiology, the primary adverse effect of propofol, including the loss of airway reflexes, hypoventilation, and even apnea (Gregory et al., 2021). The pharmacological features of remimazolam minimize the risk of severe hemodynamic instability during anesthesia and contribute to patient safety during hysteroscopy.

Bradycardia

In this meta-analysis, the data indicated no difference in the incidence of bradycardia between the remimazolam and propofol groups. A cardiac electrophysiology study showed that remimazolam did not prolong cardiac repolarization (Kleiman et al., 2020). In another analysis of 12-lead digital recorders in 20 adult men receiving a continuous intravenous infusion of remimazolam at 5 mg/min, remimazolam exerted no clinical effect on PR intervals or QRS duration (Schüttler et al., 2020). A meta-analysis of remimazolam versus propofol for procedural sedation showed a lower incidence of bradycardia in patients using remimazolam for procedural sedation than in patients using propofol (Chang et al., 2023).

Postoperative nausea, vomiting, and dizziness

Postoperative nausea, vomiting, and dizziness are uncomfortable experiences that can considerably reduce a patient’s quality of life after surgery. In a study on the prevention of postoperative vomiting in patients undergoing elective cesarean delivery, sub-hypnotic doses of propofol were shown to be effective in preventing maternal vomiting during cesarean section with the use of intrathecal morphine for spinal anesthesia (Kampo et al., 2019). In another study comparing the effects of remimazolam and propofol for procedural sedation, no difference in the incidence of PONV was observed between the remimazolam and propofol groups (Chang et al., 2023). Compared to surgeries that require general anesthesia, hysteroscopy is shorter, and fewer anesthetic drugs are injected during anesthetic sedation, which may weaken the inhibitory effect of propofol on PONV. This may explain why this meta-analysis showed that remimazolam and propofol led to similar incidences of post-hysteroscopic nausea and vomiting. Conversely, we believe that PONV development is multifactorial and may be related not only to the sedative itself but also to the type of surgery and the use of different drugs during the operation. However, further research is needed to determine whether remimazolam exerts the same effect as propofol in reducing PONV. Postoperative dizziness is a common adverse reaction following hysteroscopic surgery, and the reason may be related to the residual effects of anesthetic drugs. Both remimazolam and propofol belong to the class of sedative-hypnotic agents, and their residual effects may lead to postoperative symptoms such as dizziness and drowsiness in patients. As a novel benzodiazepine, remimazolam is metabolized more rapidly, suggesting that the incidence of postoperative dizziness after its use may be lower than that observed with propofol (Yaqiu et al., 2025). In this meta-analysis, high heterogeneity was observed in postoperative dizziness. However, after excluding the study by Xie et al. (2024), the heterogeneity decreased significantly (I² = 0%, P < 0.01). The heterogeneity may be attributed to the broad doses of remimazolam and propofol used in the study. Variations in dosing could lead to inconsistencies in anesthetic depth, thereby influencing the incidence of postoperative dizziness. In addition, it may be owing to the fact that the study conducted by Xie et al. (2024) exclusively enrolled patients aged over 65 years, which is an established contributor to high heterogeneity.

Success rate of anesthesia and sedation

A high sedation success rate implies that the patient can maintain an appropriate depth of sedation during the surgical procedure, cooperate with the surgical operation, and avoid issues such as surgical interruption and difficulties in operation (Chen et al., 2025). A meta-analysis of remimazolam versus propofol in gastrointestinal endoscopy and colonoscopy with elderly patients showed a higher sedation success rate after the first administration of propofol (Ahmer et al., 2024). In another study of upper gastrointestinal endoscopy, the success rate of sedation in the remimazolam group was no worse than that in propofol group (Chen et al., 2021). The efficacy and safety of remimazolam sedation during endoscopy were investigated in a 2021 meta-analysis. The results show that remimazolam exerts a strong sedative effect, which is significantly higher than that of midazolam and lower than that of the traditional sedative propofol (Zhu et al., 2021). However, our meta-analysis findings showed that remimazolam was not inferior to propofol in terms of its sedative effect. This may be owing to the influence of multiple factors, such as age, inclusion criteria, the combination of analgesic drugs used, and operator technique. In the present meta-analysis, we observed substantial heterogeneity in the definition of key outcomes among the included RCTs, including sedation success, recovery time, and body movement, which may compromise the homogeneity of results and the reliability of conclusions. These are not merely differences in measurement thresholds but rather conceptually distinct endpoint definitions. Pooling such outcomes may therefore reduce clinical interpretability and affect both the validity and generalizability of the conclusions. For instance, Lin, Chen & Liang (2024) defined sedation success as the absence of eyelash reflex upon cotton swab stimulation 1 min after initial sedative administration, whereas Fan et al. (2023) considered sedation failure as the requirement of an induction dose plus five additional boluses insufficient to complete hysteroscopy (for procedures not exceeding 30 min). In contrast, Zhang et al. (2022a) and Zhang et al. (2022b) defined sedation failure as a Ramsay score <5 after three supplemental doses during the induction phase. Notably, the latter definition may reflect a deeper level of sedation, potentially leading to an underestimation of the success rate of remimazolam’s owing to its pharmacological profile favoring moderate sedation rather than profound hypnotic states.

Body movement

Contrary to initial expectations, our influence-adjusted analysis suggests that remimazolam may be associated with a higher incidence of intraoperative body movement than propofol during hysteroscopy. This finding warrants careful consideration. The high and persistent heterogeneity (I² = 62–81%) reflects variability in outcome definitions, dosing regimens, and patient populations. Body movement during hysteroscopy is a critical safety endpoint, as it can interfere with delicate intrauterine instrument use and increase the risk of unintended tissue injury. This potential disadvantage of remimazolam must be balanced against its clear benefits in reducing injection pain, hypotension, and hypoxemia. The divergent result underscores the procedure-specific nature of sedative drug effects and highlights the need for anesthesiologists to consider this trade-off when selecting agents for hysteroscopic sedation.

Anesthesia recovery time

With respect to the onset and recovery time, a previous Phase III clinical trial showed a shorter sedation induction and recovery time in patients undergoing bronchoscopy sedated with remimazolam than with midazolam (Pastis et al., 2019). In a study on the pharmacokinetics of remimazolam (Kim & Fechner, 2022), it was shown that the steady-state total volume of distribution (Vdss) of remimazolam was approximately 35 L, whereas the Vdss of propofol was estimated to be 400 L, which was almost 10 times that of remimazolam. A smaller Vdss speeds up drug elimination and patient recovery because it indicates lesser drug accumulation in the body during dosing and therefore a lower requirement for drug clearance after discontinuation. That is, the recovery time after remimazolam anesthesia is often approximately 1–5 min longer than that of propofol. In this meta-analysis, we observed a high heterogeneity in the recovery time from anesthesia, which persisted after a comparison. One reason may be that these studies (Fan et al., 2023; Lin, Chen & Liang, 2024; Shan et al., 2024; Zhang et al., 2022a) defined recovery time of anesthesia differently. Additionally, patients in the remimazolam group received the antagonist flumazenil at 30 min after the operation, when the bispectral index of patients was not greater than 90. Also, the Observer’s Assessment of Alertness/Sedation did not reach a value of 5 in Zhang’s study. This may explain the high heterogeneity in the recovery time from anesthesia. Our meta-analysis included all studies on remimazolam versus other sedatives for hysteroscopic sedation. We included seven studies involving 664 patients, all of which were RCTs. The results of the different control drug analyses indicate that the results are convincing and reliable. While previous studies have compared the effects of remimazolam and propofol in procedures such as gastrointestinal endoscopy, there remains a relative lack of systematic reviews focusing specifically on hysteroscopic procedures. This study is the first to focus on the effects of the two agents in hysteroscopic surgery, providing a comprehensive comparison of the efficacy and safety profiles of both agents in this context, thereby offering direct evidence for anesthetic choice in gynecological day surgery. Nevertheless, the interpretation of these findings requires consideration of the substantial variability in outcome definitions across the included trials. For instance, sedation success was defined as the absence of eyelash reflex, achievement of a Ramsay score ≥ 5, or completion of the procedure without additional interventions. Recovery time endpoints ranged from early recovery (Aldrete score ≥9) to full consciousness or PACU discharge. Body movement definitions varied from “any movement” to “movement requiring intervention.” While such definitional heterogeneity is common in procedural sedation research, it represents a notable source of clinical heterogeneity that may limit the comparability of results across studies and affect the precision of pooled estimates. We have attempted to mitigate this by transparently documenting definitions (Supplementary File 4 and interpreting outcomes with high heterogeneity (e.g., recovery time, body movement) with appropriate caution.

Several limitations warrant consideration. First, most included studies were conducted in China, limiting generalizability to other populations. Second, leave-one-out sensitivity analyses showed that for certain outcomes (e.g., body movement, dizziness), the magnitude and precision of pooled estimates varied with the inclusion or exclusion of individual studies, indicating sensitivity to specific studies. Third, substantial and largely unexplained heterogeneity persisted for recovery time (I² = 86–98%) and body movement (I² = 62–81%), limiting interpretability. Quantitative explorations (e.g., subgroup analysis, meta-regression) were not feasible due to the limited number of studies. Fourth, remimazolam dosing regimens varied across studies, precluding specific dosage recommendations. Fifth, our approach to combining multi-dose studies had minimal impact on primary conclusions, but more precise methods could be considered in future research. Consequently, high-quality, large-sample RCTs with standardized outcome definitions are needed to confirm these findings.

Conclusion

In summary, this meta-analysis suggests moderate-certainty evidence that remimazolam may reduce injection pain, hypoxemia, and hypotension while achieving a non-inferior sedation success rate compared with propofol. For other outcomes—including body movement, bradycardia, PONV, dizziness, and recovery times—the evidence certainty is low, precluding definitive conclusions. Taken together, the findings of this meta-analysis suggest that remimazolam may represent a promising alternative for hysteroscopic sedation, particularly when hemodynamic stability is a priority. However, the strength of this recommendation is tempered by the moderate-to-low certainty of the underlying evidence, as well as by substantial unresolved heterogeneity for certain outcomes. Future high-quality trials with standardized outcome definitions are warranted to confirm these findings.

Supplemental Information

Supplemental Information 15. GRADE (Grading of Recommendations, Assessment, Development and Evaluations).

Evidence profile table summarizing the quality of evidence for each outcome, including injection pain, hypoxemia, hypotension, bradycardia, PONV, dizziness, body movement, anesthesia recovery time, and discharge time.

Funding Statement

This work was supported by the Scientific Research Program of Henan Education Institutions (Nos. 24B320001); Science and Technology Department of Henan Province (Nos. 242102310274). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.