Hemodynamic safety of remimazolam vs propofol in elderly patients

Abstract

BACKGROUND

Endoscopy and colonoscopy are gastrointestinal procedures which are frequently performed in diagnosing conditions like gastrointestinal inflammatory disorders, strictures, or malignancy. Sedation in these procedures is necessary for patient comfort and successful completion of the procedure, especially in elderly people. This meta-analysis compares the efficacy and safety profiles of two commonly used sedatives, propofol and remimazolam, a new agent with a potentially safer profile.

AIM

To compare the efficacy and safety profiles of two commonly used sedatives in the elderly population, propofol and remimazolam, a new agent with a potentially safer profile.

METHODS

PubMed, Google Scholar, and Cochrane Library databases were searched from inception until 2025 to identify studies comparing remimazolam to propofol in elderly patients undergoing gastrointestinal endoscopic procedures. Statistical analyses were performed using RevMan with a random-effects model. Heterogeneity was assessed using the I² test.

RESULTS

Our study involved 10 randomized controlled trials with a total of 1994 patients, 997 (50.0%) of whom were in the remimazolam group and 997 (50.0%) were in the propofol group. Hypotension, the primary outcome, occurred significantly less often in patients receiving remimazolam (risk ratio = 0.48, 95% confidence interval: 0.38-0.61; P < 0.00001; I² = 52%). Other outcomes, including bradycardia, respiratory depression, and pain at the injection site, were also significantly lower with remimazolam. No significant difference was found between the groups in terms of nausea and vomiting and time to recovery outcome.

CONCLUSION

Remimazolam appears to be a safer alternative to propofol for elderly patients undergoing gastrointestinal endoscopy. Remimazolam was associated with a lower incidence of hypotension, bradycardia, respiratory depression, and injection-site pain. The recovery time and incidence of nausea and vomiting were similar after administration of either drug. Further high-quality trials are warranted to confirm these findings.

Key Words: Remimazolam; Propofol; Endoscopy; Colonoscopy; Procedural sedation

Core Tip: This meta-analysis of 10 randomized controlled trials involving 1994 elderly patients establishes the superior hemodynamic safety of remimazolam over propofol during gastrointestinal endoscopy. Remimazolam reduced the risk of hypotension by more than half compared to propofol, while significantly lowering incidences of bradycardia, respiratory depression, and injection pain. Crucially, these safety advantages were achieved without compromising recovery times or increasing post-operative nausea. These quantitative findings demonstrate that remimazolam serves as a safer, highly stable sedative alternative for mitigating cardiorespiratory risks in vulnerable older populations.

INTRODUCTION

Endoscopic procedures, including upper endoscopy and colonoscopy, are commonly performed to diagnose and evaluate various gastrointestinal conditions[1]. Advances in endoscopic technology and peri-procedural care have increased the emphasis on patient comfort, procedural efficiency, and the safe use of sedation. For example, the use of anesthesia assistance during outpatient colonoscopy procedures in the United States increased substantially between 2006 and 2015; the percentage of outpatient colonoscopies performed with the assistance of anesthesia increased from 16.7% to 58.1% in a nationwide analysis of more than 4.6 million procedures[2]. It is this growth in the use of anesthesia for these procedures that has led to the need to evaluate the safety of various sedative agents that are used during endoscopic procedures.

Several drug regimens can be used during gastrointestinal endoscopy. Each of these agents, however, has certain drawbacks that may preclude their use for older patients undergoing endoscopy. One of the most common agents is midazolam, a water-soluble benzodiazepine that acts rapidly and with a short half-life; its action is reversible with flumazenil[3]. However, midazolam can lead to sedation failure, hypoxemia, and prolonged recovery in older patients. Opioids such as fentanyl are often used to aid analgesia during endoscopy but can lead to nausea, vomiting, gastrointestinal motility issues, and respiratory depression, especially in the elderly[4,5]. Finally, sedation agents such as dexmedetomidine can also be problematic for older patients due to their potential to cause hemodynamic instability[6].

Propofol is a phenolic compound, rapid onset and short duration of action are its point of distinction. It performs its action by potentiating gamma-aminobutyric acid type A receptor activity and inhibiting N-methyl-D-aspartate receptors; meanwhile, it also modulates calcium influx[7]. Propofol sedation displays a quicker recovery time, which makes it perfect for short-duration procedures[8]. However, propofol usage can show some adverse effects too, such as respiratory depression, hypotension, and pain at the injection site[7]. Long-term usage of propofol can lead to propofol-related infusion syndrome, which is a rare but fatal condition characterized by metabolic acidosis, cardiac dysfunction, and renal failure[9].

In contrast, remimazolam is an ultra-short-acting benzodiazepine that can also show its action on gamma-aminobutyric acid type A receptors, but it is characterized by its rapid onset and metabolism by tissue esterases, leading to a much safer option. Through clinical studies, it has been seen that remimazolam offers a lower incidence of respiratory depression, hypotension, and injection site pain as compared to propofol[10].

Moreover, studies have proved that the geriatric population is more susceptible to adverse propofol effects during endoscopic procedures as compared to the younger population due to age-associated physiological changes, including reduced cardiovascular reserve and altered drug metabolism. For example, one study demonstrated an adverse event rate of 31% in the elderly vs 16% in non-elderly patients (P = 0.027)[11]. Moreover, the risk of hypotension also increases for each advancing decade after 60, with a marked incidence occurring within the first 10 minutes of anesthesia administration in older patients (P < 0.0001)[12]. Therefore, the selection of an appropriate sedative agent is very important for minimizing potential risks.

In light of changing trends in sedation techniques for gastrointestinal endoscopy and consideration of remimazolam as a much better and safer option as compared to propofol, this meta-analysis seeks to analyze and evaluate their clinical actions critically and to compare the efficacy and safety of remimazolam with propofol for various clinical outcomes, in addition to a pre-specified subgroup analysis performed based on the type of procedure.

MATERIALS AND METHODS

The protocol for this study was prospectively registered with PROSPERO CRD420251029678. The study was reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines[13]. The research question, eligibility criteria, outcome measures, and methods for data analyses were specified in this protocol prior to conducting the study.

Eligibility criteria

This meta-analysis examines the randomized controlled trials (RCTs) only to evaluate the hemodynamic safety of remimazolam vs propofol in elderly patients undergoing upper gastrointestinal endoscopy and colonoscopy. The inclusion criteria are: (1) Participants, elderly patients (≥ 60 years) with an American Society of Anesthesiologists (ASA) score of I-III, undergoing upper gastrointestinal endoscopy and colonoscopy. There is no restriction on race, place, sex, ethnicity, or language; (2) Studies we include should have to evaluate the effects of remimazolam as a sedative agent vs propofol in the population; (3) The outcome should be the incidence of intraoperative hypotension, adverse events, and Time to Recovery; and (4) Studies using Assessment of Alertness and Sedation scale scores to measure sedation success and time to recovery. Studies were excluded if they met any of the following criteria: (1) Observational studies, case reports, and all non-randomized studies are excluded because of high risk of confounding, and selection bias which can affect the validity of the findings of the meta-analysis; (2) Studies conducted on other than the human species and all in vitro studies are excluded; and (3) Randomized trials that do not correspond to our inclusion criteria were excluded.

Search strategy and data sources

A systematic search of the PubMed, Google Scholar, and Cochrane Central Register of Controlled Trials databases was performed to identify RCTs published in English from inception until 2025. The search strategy, which combined MeSH and free-text search terms related to the names of the drugs, types of endoscopy, and hypotension, was used to search these databases. The titles and abstracts of the studies identified through the database searches were first screened to determine whether they appeared to meet the eligibility criteria for inclusion in the review. Those articles that appeared to potentially meet the eligibility criteria were then selected for inclusion of their full-text articles for further review. Two of the authors independently performed the screenings for eligibility, with any disagreements between the two authors resolved through discussion with a third author. The search strategy used to search the databases is provided in Supplementary Table 1. The reasons for which articles were excluded from this review are provided in Supplementary Table 2.

Risk of bias assessment

Two researchers independently evaluated the potential bias in the randomized studies using the Cochrane Collaboration’s technique. This method was used to carry out an official evaluation of the quality of the chosen studies. The examination of bias included outcome blinding, the accessibility of outcome data, the exclusive reporting of outcomes in some circumstances, and the discovery of any other sources of bias. Studies were divided into three categories by the evaluation: Low risk of bias, high risk of bias, and unclear risk of bias because of incomplete data. When studies with a score higher than 1 were designated as “high risk”, a sensitivity analysis was also conducted to ensure the robustness of the findings. To address the prospect of publication bias, a visual depiction using a funnel plot was employed. Such a detailed approach ensures that the quality of studies and possible biases are thoroughly assessed, yielding reliable and unbiased results (Figure 1 and Supplementary Figure 1). A detailed risk of bias assessment is given in Supplementary Table 3.

Figure 1  Risk of bias summary: Review authors’ judgements about each risk of bias item for each included study.

Data extraction

Data extracted from all included RCTs included characteristics of the study itself, the demographics of the patients included in the study, details regarding the interventions provided to each group of patients, sample sizes, and outcome data. The primary outcome data that were extracted from these studies included data regarding the incidence of intraoperative hypotension. Data was also extracted from the studies regarding the incidence of adverse events like bradycardia, respiratory depression, nausea and vomiting, and injection-site pain. The data regarding recovery time from the medications was also extracted from these studies; recovery time was defined as the length of time between the discontinuation of either the remimazolam or propofol treatment and the attainment of an Assessment of Alertness and Sedation scale score of 5. Finally, the conclusions of the authors regarding the safety and efficacy of either remimazolam or propofol compared to the other were also extracted from each study.

Statistical analysis

We used Rev Man (Review Manager Version 5) to evaluate data in this meta-analysis. In every statistical analysis, a random effects model was used. We have used a risk ratio for the assessment of our dichotomous data, which were intraoperative hypotension (primary outcome) and adverse events, including bradycardia, respiratory depression, nausea and vomiting, and pain at the injection site. The standard mean deviation was used for continuous data (i.e., time to recovery). If the P-value was less than 0.05, we regarded the results as statistically significant. A forest plot was used to visually interpret the outcome analysis, and we ran χ² and I² tests to determine whether there was statistical heterogeneity[14]. If the measures of heterogeneity among the studies were greater than 50% (I² > 50), it was considered substantial; then, leave-one-out analyses would be performed to determine which study was contributing to the heterogeneity in the review. This step helped to ensure the reliability of the results of the review.

RESULTS

Literature search results

The initial search across the databases yielded a total of 1291 potential studies. After applying the inclusion and exclusion criteria, only 10 trials remained for further analysis. These studies compare remimazolam and propofol. The detailed information and references for these trials are available in the online databases. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow chart (Figure 2)[12,15-23] provides a visual summary of the literature search process, illustrating the process of selection, from the initial search to the final inclusion of the 10 RCTs in the analysis. The flow chart effectively outlines the number of studies identified, screened, assessed for eligibility, and ultimately included in the analysis.

Figure 2  Preferred Reporting Items for Systematic Reviews and Meta-Analyses flowchart of all the included studies.

Study characteristics

A total of 1994 patients were randomly assigned across the trials, with 997 patients in the remimazolam arm and 997 patients in the propofol arm. The mean age of the patients included in this study was 69.11 years with a standard deviation of 4.2 years. The sex distribution of the patients was almost equal, with 52.75% of the patients being male and 49.25% being female. Table 1 summarizes the baseline characteristics of the patients included in the trials. Details of the comorbidities and ASA classification of the patients are provided in Supplementary Table 4.

Table 1 Baseline characteristics of the patients.

Ref.	ASA grade	Procedure	Intervention	Control	Analgesia used	Number of PT in group (R)	Number of PT in group (P)	Age in group (R)	Age in group (P)	Sex (%) group (R)	Sex (%) group (P)
Liu et al[12], 2023	I-II	Gastrointestinal endoscopy	Remimazolam (0.10 mg/kg)	Propofol (1.5 mg/kg)	Sufentanil 5 μg	107	109	67.6 (5.7)	67.5 (4.9)	51 (47.7)	51 (46.8)
Chen et al[13], 2024	I-III	Gastrointestinal endoscopy	Remimazolam (6 mg/kg/hour)	Propofol (18 mg/kg/hour)		30	30	68.0 (67.0-72.3)	69.5 (66.0-72.0)	11 (36.7)	9 (30)
Chen et al[14], 2024	I-II	Gastroscopy	Remimazolam (0.17 mg/kg)	Propofol (1.112 mg/kg)	Sufentanil 0.1 μg/kg	122	118	71.9 (5.0)	71.7 (5.1)	66 (54)	62 (52.5)
Lin et al[15], 2024	I-III	Colonoscopy	Remimazolam (0.2 mg/kg)	Propofol (1 mg/kg)	Sufentanil 0.05 μg/kg	114	114	70 (67-75)	70 (68-75)	59 (51.8)	63 (55.3)
Ye et al[16], 2023	I-II	Gastroscopy	Remimazolam (0.2 mg/kg)	Propofol (2.0 mg/kg)	Remifentanil 0.2 μg/kg	64	65	68 (66-72)	68 (66-71)	39 (60.9)	36 (55.4)
Tan et al[17], 2022	I-II	Upper Gastrointestinal endoscopy	Remimazolam (0.2 mg/kg)	Propofol (1.0-1.5 mg/kg)	0.2 g lidocaine and butorphanol 0.01 mg/kg	33	33	65.5 (5.2)	66.2 (5.0)	22 (66.7)	21 (63.6)
Lu et al[18], 2022	I-III	Upper Gastrointestinal endoscopy	Remimazolam (300 mg/hour)	Propofol (3 g/hour)	Fentanyl citrate 50 μg	200	200	70.6 (4.7)	70.1 (4.5)	78 (39.0)	83 (46.5)
Guo et al[19], 2022	I-II	Gastrointestinal endoscopy	Remimazolam (0.15 mg/kg)	Propofol (1.5 mg/kg)	Alfentanil 5 μg/kg	39	38	70.4 (3.9)	69.1 (4.0)	25 (64)	22 (57.9)
Hu et al[20], 2022	I-III	Gastroscopy	Remimazolam (0.2 mg/kg)	Propofol (1.5mg/kg)	Sufentanil 0.1 μg/kg	173	173	70.11 (7.37)	69.92 (7.57)	69 (39.8)	72 (41.6)
Liu et al[21], 2021	I-II	Colonoscopy	Remimazolam (0.15 mg/kg)	Propofol (0.1 mL/kg)	Fentanyl citrate 0.5 μg/kg	115	117	68.87 (2.58)	69.12 (2.75)	54 (46.9)	58 (49.5)

ASA: American Society of Anesthesiologists; R: Remimazolam; P: Propofol; PT: Patient.

Outcome

Outcomes were categorized as primary and secondary outcomes. The primary outcome of interest to the authors was the occurrence of hypotension, while the secondary outcomes were bradycardia, respiratory depression, nausea and vomiting, injection-site pain, and the amount of time that it took for patients to recover from the procedure. The definitions of the terms hypotension, bradycardia, respiratory depression, and depth of sedation utilized within the included RCTs are provided in Supplementary Table 5.

Hypotension

All included studies reported the incidence of hypotension among patients receiving either remimazolam or propofol. A total of 1994 patients were analyzed, with 997 patients allocated to the remimazolam group and 997 patients allocated to the propofol group. In the remimazolam group, there were 215 instances of hypotension, compared to 431 in the propofol group. This clearly indicates that remimazolam was associated with a significantly lower incidence of hypotension [risk ratio (RR) 0.48, 95% confidence interval (CI): 0.38-0.61; P < 0.00001; I² = 52%]. The I² statistic was 52%, reflecting moderate heterogeneity among the studies (Figure 3A). A sensitivity analysis was also conducted by excluding Liu et al[23], which decreases the heterogeneity of the outcome to 36% (Supplementary Figure 2). For visual representation of the publication of bias, a funnel plot is used (Supplementary Figure 3).

Figure 3 Forest plots. A: Forest plot of incidence of hypotension in treatment groups; B: Forest plot of incidence of Bradycardia in treatment groups; C: Forest plot of incidence of respiratory depression in treatment groups.

Bradycardia

Out of all the selected studies, only seven studies reported the incidence of bradycardia in patients receiving either remimazolam or propofol. A total of 803 patients were included in the remimazolam group, with 40 events of bradycardia, while 808 patients were in the propofol group, with 90 events of bradycardia. These results suggest that remimazolam was significantly more effective than propofol in reducing the incidence of bradycardia. The RR for bradycardia in the remimazolam group compared to propofol was calculated to be 0.48 (95%CI: 0.29-0.79; P = 0.004), indicating that patients receiving remimazolam had nearly half the risk of experiencing bradycardia compared to those receiving propofol. The I² statistic was 36%, which indicates moderate heterogeneity across the studies (Figure 3B).

Respiratory depression

Among the selected studies, only six studies reported the occurrence of respiratory depression in patients receiving either remimazolam or propofol. In the remimazolam group, a total of 713 patients were included, and the number of events of respiratory depression was 29. In the propofol group, 711 patients were included, and the number of events of respiratory depression was 68. The data clearly show that remimazolam reduced the incidence of respiratory depression significantly as compared with propofol. The RR for respiratory depression was 0.45 (95%CI: 0.30-0.69; P = 0.0002). This means that patients who received remimazolam had a 45% reduced risk of respiratory depression vs those who received propofol. The I² statistic for this analysis was 0%, indicating no heterogeneity across the studies (Figure 3C).

Nausea and vomiting

Out of all the selected studies, only seven studies reported the incidence of nausea and vomiting in patients receiving either remimazolam or propofol. A total of 713 patients were included in the remimazolam group, with 29 events of nausea and vomiting, while 711 patients were in the propofol group, with 68 events of nausea and vomiting. The results suggest that there was no significant difference between the two groups regarding the occurrence of nausea and vomiting. The RR for nausea and vomiting was calculated to be 1.01 (95%CI: 0.63-1.63; P = 0.97), indicating that the risk of nausea and vomiting was nearly identical between the remimazolam and propofol groups. The I² statistic was 0%, indicating no heterogeneity between the studies (Supplementary Figure 4).

Pain at the injection site

Out of all the selected studies, eight studies reported the incidence of pain at the injection site in patients receiving either remimazolam or propofol. A total of 857 patients were included in the remimazolam group, with 47 events of pain at the injection site, while 855 patients were in the propofol group, with 196 events of pain. The results clearly show that remimazolam significantly reduced the incidence of pain at the injection site compared to propofol. The RR for pain at the injection site was calculated to be 0.26 (95%CI: 0.17-0.39; P < 0.00001). This indicates that patients receiving remimazolam had only 26% of the risk of experiencing pain at the injection site compared to those receiving propofol, demonstrating a significant risk reduction. The I² statistic was 25%, indicating low heterogeneity across the studies (Supplementary Figure 5).

Time to recovery

All the selected studies reported the time to recovery for patients receiving either remimazolam or propofol. A total of 890 patients were included in the remimazolam group, and 888 patients were in the propofol group. The results indicate that there is no significant difference in the time to recovery between remimazolam and propofol. The standard mean difference in time to recovery was calculated to be -0.39 (95%CI: -0.83 to 0.05, P = 0.08), which suggests a significant difference in recovery times between the two drugs. Additionally, the I² statistic was 95%, indicating high heterogeneity across the included studies (Supplementary Figure 6). This suggests that there was substantial variation between the studies in terms of study design, patient populations, or other factors, which may have contributed to the wide CIs and the small difference in recovery times. Despite conducting sensitivity and subgroup analyses, the high heterogeneity remained, indicating that variability between studies persisted even after adjusting for certain factors.

DISCUSSION

Sedation plays an important role in ensuring patient comfort as well as a successful therapeutic and diagnostic procedure during upper gastrointestinal endoscopy. There is an increasing demand for effective and safe strategies used in elderly patients, which requires a comparison between propofol and remimazolam. This meta-analysis evaluates efficacy, safety, pharmacokinetics, and patient recovery outcomes of propofol and remimazolam, two widely used sedative agents in gastrointestinal endoscopy procedures.

The key findings of this meta-analysis showed that remimazolam is a preferred choice for sedation in endoscopic procedures as it has lower rates of respiratory depression and cardiovascular collapse compared to propofol. It has a lower incidence of adverse clinical events, including bradycardia, hypotension, respiratory depression, and injection site pain. Our results indicate that the rates of nausea and vomiting do not significantly differ between patients receiving propofol and those receiving remimazolam.

One of the important findings of our analysis demonstrated that the patients who received remimazolam had a much lower incidence of hypotension in patients compared to those who received propofol, a finding that is consistent with the data of the previous study[15]. A recent study demonstrated that the occurrence of adverse effects is closely linked to the depth of sedation. Compared to propofol, remimazolam generally results in lighter sedation. Greater sedation depth leads to stronger suppression of circulatory and respiratory functions, accounting for the higher incidence of cardiopulmonary complications associated with propofol[12].

Bradycardia can complicate sedation in older adults who may already have compromised cardiac function. The significant reduction in bradycardia in patients administered remimazolam as compared to propofol is consistent with the findings in recent RCTs, which emphasize remimazolam’s minimal impact on autonomic cardiac regulation[20]. While pre-existing data suggest that recovery time is shorter for remimazolam compared to propofol[24], our analysis did not reveal a significant difference in recovery time between the two agents.

According to a recent study[16], propofol is metabolized via glucuronidation, a process primarily mediated by uridine diphosphate-glucuronosyltransferases, which are present in both liver and extrahepatic tissues. In contrast, remimazolam works by targeting gamma-aminobutyric acid receptors and is known for its rapid onset and recovery, as well as metabolism that is organ-independent. Unlike propofol, its sedative effects can be reversed with flumazenil[16]. Some studies have shown that elderly patients metabolize propofol at a slower rate than younger individuals, highlighting age as a key factor influencing its clearance[25,26].

Our finding that remimazolam is associated with lower incidences of injection pain in the elderly population compared to propofol is supported by a previous study[27]. This finding can be attributed to its water-soluble formulation, which does not induce the endothelial irritation leading to decreased pain events, which is often associated with the lipid emulsion of propofol[28]. Remimazolam was linked to lower rates of hypotension, bradycardia, respiratory depression, and injection site pain compared to propofol; however, these outcomes do not prove the statistical superiority of remimazolam in specific high-risk subgroups, only that the drug features a more favorable safety profile in elderly patients overall. The randomized trials that comprised these analyses did not include analyses based on significant medical conditions. Thus, while the results of this meta-analysis indicate a possible benefit of remimazolam in those with significant medical conditions, the superiority of the drug for those populations has not yet been specifically proven.

Though data regarding outcomes related to glucose regulation, end-organ complications, and complications related to diabetes were not provided in the included studies, the ability of remimazolam to stabilize hemodynamics may indicate a benefit for patients with diabetes. Similarly, while data regarding complications specific to patients with chronic obstructive pulmonary disease were not provided within the included research studies, the decreased prevalence of respiratory depression with the use of remimazolam may be especially significant for patients with chronic obstructive pulmonary disease. Similarly, the decreased incidence of bradycardia and hypotensive effects in patients who have cardiovascular illness is in line with the increasing evidence of the hemodynamic stability that remimazolam provides for patients under anesthesia. However, due to the lack of analyses regarding specific cardiovascular diagnoses within the included randomized trials, it is still not possible to prove the superiority of remimazolam in comparison to propofol for specific cardiac subgroups of patients undergoing procedures. Thus, more research is needed to prove the safety of remimazolam for these high-risk patient groups.

Our findings are consistent with a recent meta-analysis comparing remimazolam and propofol that found lower rates of hypotension, respiratory depression, injection site pain, and bradycardia. However, this study included only patients with a narrow range of ASA (I–II), body mass index limits and on an outpatient status. In contrast, our study had a broader inclusion criterion, including elderly patients aged ≥ 60 years with ASA scores of I-III, included RCTss and cluster RCTs[4], increasing the generalizability to routine clinical settings[29]. In this context, evidence from other endoscopic settings, such as endoscopic retrograde cholangiopancreatography, also suggests a consistent hemodynamic advantage of remimazolam over propofol, with preserved procedural efficacy and fewer cardiopulmonary adverse events. However, these findings are still derived from relatively small heterogeneous RCTs, and therefore should be interpreted with caution when extrapolating to broader clinical practice[30,31]. This further highlights the need for well-designed, large-scale multicenter randomized trials with standardized dosing regimens and diverse patient populations to better define the comparative safety profile of remimazolam across different endoscopic procedures.

Our analysis revealed heterogeneity across the studies, particularly in the outcomes relating to hypotension and time to recovery. Moderate heterogeneity was observed for hypotension, which improved after exclusion of the Liu et al[23]. This suggests that Liu et al’s study[23] may have some unique factors that are different from the others. Likewise, the removal of the Ye et al’s study[18] in the analysis of bradycardia resulted in a drop in variability from 36% to 0%. Sensitivity analysis of Lu et al[20] reduced heterogeneity to 0% for injection site pain. These results suggested that the inconsistency can be attributed to a few individual studies.

However, there was still high heterogeneity for time to recovery even after performing sensitivity and subgroup analyses. This may be due to differences in clinical practices, sedation protocols, or how recovery time was quantified across the studies. Yet these differences do not diminish the relevance of our findings. Some heterogeneity is expected in clinical research, particularly when comparing real-world data.

Strengths and limitations

This study has both strengths and limitations. It included both RCTs and cluster RCTs, which provided a wider understanding of the research question. The robustness of the findings is further enhanced by meta-regression and subgroup analyses. However, the study includes limited literature and has heterogeneity, which may impact the reliability. Moreover, the dose of remimazolam used in the included studies is variable, ranging from 6 mg to 300 mg, with some studies exceeding the recommended dosing ranges. Higher doses may have increased the risk of dose-dependent adverse events such as hypotension and sedation. Another potential limitation was the confounding effect of opioids such as fentanyl or sulfentanil which might have influenced the study outcomes and hemodynamic profile of patients. Future studies should focus on dose standardization of remimazolam to better understand the safety profile of the drug. Further, our study has a majority of the Asian study population, which can limit the generalizability of the findings. Most of the trials were short-term, which limited the consideration of long-term outcomes of these drugs.

CONCLUSION

In elderly patients undergoing upper gastrointestinal endoscopy or colonoscopy, remimazolam was associated with a more favorable short-term safety profile than propofol, with significantly lower risks of developing hypotension, bradycardia, respiratory depression, and injection-site pain; however, rates of nausea and vomiting and recovery time were similar between patients treated with either agent. Thus, remimazolam may be considered as an alternative to propofol for sedation of adult gastrointestinal endoscopy patients. However, due to the heterogeneity of the evidence base and the limitations of the existing studies, further research is required in order to clarify the relative benefits of each of these agents. Additional large-scale, multicenter, randomized clinical trials are needed to provide conclusive evidence of the superiority of either agent for use in endoscopic procedures.