Abstract
Objective
Older adults undergoing ERCP are particularly susceptible to sedation-related respiratory and circulatory adverse events. This trial compared the safety and efficacy of remimazolam-esketamine with propofol-fentanyl, with identical standardized nasopharyngeal airway-assisted respiratory management in both groups.
Patients and Methods
This single-center, prospective, randomized, parallel-group superiority trial enrolled 160 patients aged 60–80 years with American Society of Anesthesiologists physical status I–III who were scheduled for elective ERCP. Participants were randomly assigned 1:1 to remimazolam-esketamine (RE group; n=80) or propofol-fentanyl (PF group; n=80). Both groups received identical nasopharyngeal airway placement, oxygen delivery, capnography, and rescue algorithms; the randomized intervention was the anesthetic drug regimen. The primary outcome was composite sedation-related adverse events (SRAEs), comprising prespecified hypotension, bradycardia, hypoxemia, or respiratory depression from anesthesia induction to endoscope withdrawal. Analyses followed the intention-to-treat principle.
Results
Composite SRAEs occurred in 14 of 80 patients (17.5%) in the RE group and 28 of 80 patients (35.0%) in the PF group (RR, 0.50; 95% CI, 0.29–0.88; risk difference, −17.5 percentage points; 95% CI, −30.9 to −4.1; P=0.016). The adjusted RR was 0.49 (95% CI, 0.31–0.78; P=0.003). Anesthetic regimen success was similar between groups (97.5% vs 98.8%). Induction was modestly longer with RE, whereas eye-opening and modified Aldrete recovery time were slightly shorter.
Conclusion
In this selected population of older patients undergoing ERCP, the remimazolam-esketamine regimen was associated with fewer composite SRAEs than propofol-fentanyl while achieving a similar anesthetic regimen success rate. Induction was modestly longer. Tachycardia, hypertension, and psychotomimetic symptoms were numerically more frequent with RE, but event counts were small and estimates were imprecise. Multicenter studies are required to confirm generalizability.
Keywords: remimazolam, esketamine, endoscopic retrograde cholangiopancreatography, elderly patients, nasopharyngeal airway, sedation-related adverse events
Plain Language Summary
ERCP is an endoscopic procedure used to diagnose and treat disorders of the bile duct and pancreas. During deep sedation, older adults may experience low blood pressure, a slow heart rate, low oxygen levels, or respiratory depression.
In this randomized trial, all participants received the same nasopharyngeal airway support, oxygen delivery, capnography monitoring, and rescue sequence. The study therefore compared two anesthetic drug regimens rather than different airway-management strategies.
Remimazolam combined with esketamine was associated with fewer sedation-related respiratory or circulatory adverse events than propofol combined with fentanyl.
The assigned anesthetic regimen was successful in a similar proportion of participants in both groups. Remimazolam-esketamine took slightly longer to induce anesthesia. Faster heart rate, higher blood pressure, abnormal dreams, and related symptoms occurred in only a small number of participants, so larger studies are needed to assess these findings.
Introduction
Endoscopic retrograde cholangiopancreatography (ERCP) is commonly performed in the prone or semi-prone position, where airway access is shared with the endoscopist and procedural stimulation and duration may be unpredictable.1,2 Anesthetic management must therefore provide adequate hypnosis and analgesia while preserving ventilation, hemodynamic stability, and timely recovery.
Propofol-based deep sedation is widely used because of its rapid onset and titratability.3 However, propofol lacks analgesic activity and is commonly combined with an opioid; propofol-fentanyl may therefore produce additive respiratory and cardiovascular depression.4 Sedation-related airway and hemodynamic events are particularly relevant in older patients and during complex ERCP.5
Remimazolam is an ultra-short-acting benzodiazepine metabolized by tissue esterases and reversible with flumazenil.6 Compared with propofol, it has generally been associated with less hypotension, respiratory depression, bradycardia, and injection pain during gastrointestinal endoscopy, although induction may be slower.7 ERCP trials and meta-analyses have reported comparable procedural efficacy with favorable cardiorespiratory safety signals.8–10
Esketamine provides analgesia with relatively limited respiratory suppression and may partly offset sedative-related hypotension through sympathetic activation.11 Low-dose esketamine can reduce propofol requirements during ERCP,12 and remimazolam-esketamine has provided effective endoscopic sedation with stable hemodynamics, including in older patients.13,14 Direct evidence against propofol-fentanyl in older patients undergoing ERCP remains limited.
Airway adjuncts can independently influence respiratory outcomes.15 We therefore standardized nasopharyngeal airway placement, oxygen delivery, capnography, and rescue algorithms in both groups. The trial tested whether remimazolam-esketamine would reduce composite SRAEs while maintaining anesthetic regimen success compared with propofol-fentanyl.
Materials and Methods
Study Design and Ethics
This single-center, prospective, randomized, parallel-group superiority trial was conducted at The Fourth Affiliated Hospital of Anhui Medical University. The original protocol was approved by the institutional ethics committee (approval number KYXM-202511-018; December 10, 2025). The trial was registered in the Chinese Clinical Trial Registry before enrollment (ChiCTR2600116715; January 14, 2026), conducted in accordance with the 2024 revision of the Declaration of Helsinki,16 and reported in accordance with the CONSORT 2025 statement.17 All participants provided written informed consent before any study-related procedure.
Before participant enrollment began, we submitted a protocol amendment on January 15, 2026. The amended study protocol, Protocol Version 2.0, together with the corresponding informed consent form Version 2.0, was reviewed as a protocol amendment and formally approved by the same institutional ethics committee on January 20, 2026. The first participant signed the written informed consent form on January 21, 2026, after approval of Protocol Version 2.0. All participants were subsequently enrolled and treated according to the ethics-approved Protocol Version 2.0.
Participants
Eligible patients were aged 60–80 years, scheduled for elective diagnostic or therapeutic ERCP, classified as American Society of Anesthesiologists physical status I–III, had a BMI of 18–28 kg/m2, and had no anticipated difficult airway. Exclusion criteria were severe cardiac, cerebral, renal, or pulmonary dysfunction; regular benzodiazepine or opioid use during the preceding 3 months; allergy to a study drug; ASA physical status IV or V; anticipated difficult airway or severe respiratory disease; diagnosed psychiatric illness, a relevant family history, or cognitive impairment (Chinese Mini Mental Status [CMMS] score <24);18 resting heart rate <55 beats/min or second-degree or higher atrioventricular block; delayed gastric emptying or a high risk of reflux and aspiration; refusal to participate; or investigator judgment that enrollment was inappropriate.
Randomization, Allocation Concealment, and Blinding
Within each stratum defined by the planned ERCP procedure type (stone extraction/lithotripsy, stricture dilation or stenting, and diagnostic or other ERCP), an independent statistician used a random-number table to generate a 1:1 allocation sequence, with equal allocation prespecified within each stratum. Assignments were placed in sequentially numbered, opaque, sealed envelopes. A circulating nurse opened the envelope immediately before induction and communicated allocation to the treating anesthesiologist. Because the drugs differed in appearance, preparation, and administration, treating anesthesiologists were not blinded. Postoperative outcomes at 24 hours were assessed by an independent investigator who was not involved in intraoperative care, did not enter the procedure room, and had no access to the anesthesia records or treatment allocation before follow-up was completed. Intraoperative physiological data, clinical responses, and rescue interventions were necessarily recorded by unblinded clinical personnel because they were directly involved in anesthesia management. To reduce ascertainment variability, these events were recorded using prespecified physiological thresholds and treatment algorithms. Statistical analyses were performed using coded groups (A and B), and the allocation code was not released until completion of the primary analysis.
Anesthesia and Airway Management
Both regimens were administered by anesthesiologists to provide deep sedation while maintaining spontaneous ventilation. For consistency with the registered trial protocol, this strategy is referred to as anesthesia throughout the manuscript. The airway-management protocol was identical in both groups and was not a randomized intervention. All participants received the same fasting and fluid regimen, pharyngeal topical anesthesia, antiemetic prophylaxis, perioperative adjuncts, preoxygenation, monitoring, nasopharyngeal airway placement, oxygen delivery, capnography, and stepwise rescue algorithm. The protocol-specified drug doses were selected primarily on the basis of prior clinical experience at our center. Published studies of remimazolam-esketamine in gastrointestinal endoscopy13,14 and an age-stratified DDDT dose-finding study showing progressively lower remimazolam induction requirements with advancing age during ERCP19 supported the clinical plausibility of conservative, age-sensitive dosing, but no single published dosing protocol or pharmacokinetic analysis determined the exact regimen used in this trial. The selected dose range also reflected the greater pharmacodynamic sensitivity and generally reduced anesthetic requirements of older adults, concomitant use of hypnotic and analgesic agents, and the goal of preserving spontaneous ventilation.
The analgesic components of the two regimens were administered differently by design. In the RE group, anesthesia was induced with intravenous remimazolam 0.2 mg/kg and esketamine 0.25 mg/kg, followed by continuous infusions of remimazolam 0.4–0.6 mg/kg/h and esketamine 0.2 mg/kg/h. The remimazolam infusion was titrated within this prespecified range according to BIS, patient movement, hemodynamic variables, and respiratory status, with a target BIS of 40–60. Processed electroencephalographic monitoring has also been prospectively evaluated for sedation-depth assessment in older patients undergoing ERCP.20 An additional remimazolam bolus of 0.05 mg/kg was administered for inadequate anesthetic depth. Because ongoing analgesic coverage was provided by the continuous esketamine infusion, supplemental esketamine boluses were not prespecified. In the PF group, anesthesia was induced with propofol 0.6 mg/kg and fentanyl 0.5 μg/kg, followed by a propofol infusion of 2–3 mg/kg/h. Fentanyl was not continuously infused; a supplemental fentanyl bolus of 0.25 μg/kg was permitted when analgesia was considered inadequate or procedural stimulation was marked. Supplemental propofol 0.25 mg/kg was administered for inadequate anesthetic depth. Inadequate anesthetic depth was defined as purposeful movement, coughing or retching that interfered with the procedure, a clinically apparent response to procedural stimulation, or a sustained BIS value >60 for at least 30 seconds after exclusion of signal artifacts. Inadequate analgesia or marked procedural stimulation was determined by the treating anesthesiologist on the basis of a sustained increase in MAP or HR of ≥20% from the preinduction baseline during a noxious procedural step, after inadequate anesthetic depth, upper-airway obstruction, hypoxemia, and other reversible causes had been excluded.
After the Ramsay Sedation Scale score reached 5, a lubricated nasopharyngeal airway was inserted in both groups and capnography was initiated. Ventilatory abnormalities were managed using a prespecified escalation sequence: nasopharyngeal airway repositioning, increased oxygen flow, jaw thrust, mask positive-pressure ventilation, temporary interruption of ERCP, and, if required, unplanned tracheal intubation. Circulatory abnormalities were managed according to a standardized protocol, and administration of vasoactive medication or atropine was recorded. Respiratory and circulatory events occurring between the prespecified observation time points were also documented.
Outcome Measures
T0 was baseline before anesthesia induction; T1 was 2 minutes after induction; T2 was the time of sphincterotomy or pancreaticobiliary duct cannulation, or 30 minutes after endoscope insertion if neither occurred; and T3 was the time after discontinuation of the study drugs when the patient first opened their eyes in response to verbal commands and recovered to a Ramsay Sedation Scale score of 2. MAP, HR, SpO2, and respiratory rate were recorded at each time point.
The primary outcome was the incidence of composite SRAEs from anesthesia induction to endoscope withdrawal. The composite and all component thresholds were prespecified in Protocol Version 2.0 before enrollment. The endpoint integrated clinically relevant respiratory and circulatory events commonly evaluated during ERCP sedation,5 while individual components were reported separately to preserve clinical interpretability, consistent with recommendations for composite outcomes.21,22 An SRAE was defined as at least one of the following: (1) MAP <65 mmHg, a decrease of ≥20% from baseline lasting ≥1 minute, or a requirement for vasoactive medication; (2) HR <50 beats/min lasting ≥1 minute or requiring atropine; (3) SpO2 <90% lasting ≥10 seconds; or (4) absent visible chest movement or loss of the end-tidal carbon dioxide waveform for ≥20 seconds, or a requirement for mask positive-pressure ventilation. Each patient was counted once for the composite outcome regardless of the number or type of events. Hypotension burden was the cumulative duration, in minutes, during which a hypotension criterion was met; hypoxemia burden was the cumulative duration, in seconds, of SpO2 <90%. Clinically significant airway intervention was defined as jaw thrust, mask positive-pressure ventilation, temporary interruption of ERCP for respiratory rescue, or unplanned tracheal intubation. Technical nasopharyngeal airway repositioning alone was recorded separately.
Anesthetic regimen success was defined as completion of the planned ERCP under the assigned regimen, including protocol-specified supplemental doses, without conversion to an alternative anesthetic regimen, administration of sedative or analgesic medication outside the prespecified protocol, unplanned tracheal intubation, or premature termination because of inadequate anesthesia or an anesthesia-related adverse event. Anesthetic regimen failure was defined as conversion to an alternative regimen because adequate anesthetic effect or physiological stability could not be maintained despite protocol-specified rescue measures. Additional sedation was defined as administration of at least one protocol-specified supplemental bolus of remimazolam or propofol after the initial induction dose. Tachycardia was defined as HR >100 beats/min or an increase of ≥20% from baseline lasting at least 1 minute, or requiring pharmacological treatment. Hypertension was defined as an increase in MAP of ≥20% from baseline, SBP ≥180 mmHg, or DBP ≥110 mmHg lasting at least 1 minute, or requiring antihypertensive treatment. Psychotomimetic symptoms were defined as any new hallucination, perceptual disturbance, dissociative experience, dysphoria, disorientation, or vivid, unusual, or distressing dream occurring from recovery of consciousness to the 24-hour follow-up; symptoms were assessed after the patient was fully alert and reassessed at 24 hours. Emergence agitation was defined as a Richmond Agitation-Sedation Scale score ≥+1 at any time between first eye opening and PACU discharge; a score ≥+3 was considered severe agitation. Other secondary outcomes included the individual SRAE components, lowest MAP, HR, and SpO2, vasoactive drug use, induction and recovery times, body movement, coughing, pain and satisfaction scores, injection pain, and other adverse reactions within 24 hours.
Sample Size
The sample size was estimated for comparison of two independent proportions. Center-specific clinical experience accumulated from June to December 2025, before randomized enrollment, informed the planning assumptions. Among 100 patients managed with the routinely used propofol-fentanyl regimen, 37 experienced cardiorespiratory events broadly corresponding to the four planned composite SRAE categories (37.0%). During more limited routine clinical use of the remimazolam-esketamine regimen, 4 of 25 patients experienced such events (16.0%). None of these 125 patients was included in the subsequent randomized trial. The unequal source-series sizes reflected the longstanding routine use of the PF regimen and the more limited prior clinical experience with the RE regimen. These historical events broadly matched the four planned categories but were not necessarily recorded using every duration threshold subsequently specified in the trial protocol. The aggregated counts were used only for internal sample-size planning; no patient-level comparative analysis of these routine-care cases was performed. The anticipated absolute risk reduction of 21 percentage points corresponded to a relative risk of approximately 0.43, a relative risk reduction of approximately 57%, and a number needed to treat of approximately five. This magnitude was considered clinically meaningful because the composite included events that could require pharmacological treatment, airway rescue, or interruption of ERCP. With a two-sided alpha of 0.05, 80% power, and 1:1 allocation, G*Power version 3.1.9.4 indicated that at least 67 participants were required per group. Allowing for approximately 15% attrition or incomplete follow-up, 80 participants per group and 160 participants in total were planned.
Statistical Analysis
The primary analysis followed the intention-to-treat principle. Continuous variables are presented as mean ± standard deviation or median (interquartile range), as appropriate, and categorical variables as number and percentage. Baseline balance was assessed descriptively using standardized differences rather than P values. The primary outcome and key binary safety outcomes were analyzed using modified Poisson regression with robust standard errors to estimate RRs, corresponding 95% CIs, and two-sided P values.23 Absolute risk differences and Wald 95% CIs are also reported. A prespecified adjusted model included age, sex, BMI, ASA III status, baseline MAP, and baseline SpO2. These covariates were selected a priori because age, sex, BMI, and ASA status represent baseline susceptibility to sedative-related cardiorespiratory events, whereas baseline MAP and SpO2 directly influence the probability of meeting the hypotension and hypoxemia criteria. Covariate selection was protocol-based rather than data-driven.
Holm correction was applied to hypotension, bradycardia, hypoxemia, respiratory depression, and clinically significant airway intervention to control the family-wise error rate for these prespecified component comparisons.24 Secondary binary outcomes were analyzed using modified Poisson regression with robust standard errors when appropriate. Sparse binary outcomes, including anesthetic regimen success, were compared using Fisher’s exact test, with RRs and 95% CIs presented descriptively. MAP, HR, SpO2, and respiratory rate were analyzed using linear mixed-effects models with fixed effects for group, time, and the group-by-time interaction and a random intercept for patient. Post hoc group contrasts at each time point were derived from these models and Holm-adjusted within each physiological outcome. Hypotension and hypoxemia burdens were highly skewed with excess zeros and were compared using Mann–Whitney U-tests; medians, interquartile ranges, ranges, and the numbers of participants with non-zero burden are reported. Other continuous outcomes were compared using independent-samples t tests or Mann–Whitney U-tests, as appropriate. Postoperative 24-hour outcomes were analyzed among participants who completed follow-up.
Prespecified exploratory analyses included circulatory and respiratory SRAE composites, treatment-needed SRAEs, no-event anesthetic regimen success, subgroup analyses, and patient-level SRAE severity. Subgroup analyses used treatment-by-subgroup interaction tests and were considered underpowered for definitive subgroup conclusions. For patient-level severity, grade 0 denoted no composite SRAE; grade 1, an SRAE without prespecified rescue; grade 2, an SRAE requiring vasoactive medication, atropine, nasopharyngeal airway adjustment, or jaw thrust; grade 3, an SRAE requiring mask positive-pressure ventilation or temporary ERCP interruption; and grade 4, unplanned tracheal intubation or a serious adverse event. Technical airway adjustments without an SRAE-defining criterion remained grade 0. The severity distribution was compared using a Mann–Whitney U-test. All exploratory analyses were interpreted as supportive and hypothesis-generating rather than confirmatory.
Results
Participant Flow and Baseline Characteristics
Routine clinical pre-screening began after amendment approval on January 20, 2026; formal study screening, written informed consent, and enrollment began on January 21, 2026. During the overall pre-screening and enrollment period from January 20 to June 13, 2026, 169 potentially eligible patients were identified. Nine were excluded before randomization: 5 did not meet the inclusion criteria, 2 met exclusion criteria, and 2 declined participation. Overall, 160 participants were randomized, with 80 assigned to each group. Primary outcome data were available for all randomized participants. The 24-hour follow-up was completed by 77 participants in the PF group and 78 in the RE group. Participant flow is shown in Figure 1, and baseline characteristics are presented in Table 1.
Figure 1.
Participant flow diagram. Primary outcome data were available for all randomized participants. The 24-hour follow-up outcomes were analyzed among participants who completed follow-up.
Abbreviations: PF, propofol–fentanyl; RE, remimazolam–esketamine.
Table 1.
Baseline Characteristics
| Variable | PF Group (n=80) | RE Group (n=80) | Standardized Difference |
|---|---|---|---|
| Age, years | 69.64 ± 4.88 | 69.10 ± 4.93 | −0.11 |
| Male sex, n (%) | 40 (50.0) | 35 (43.8) | −0.13 |
| BMI, kg/m2 | 24.11 ± 2.04 | 23.73 ± 1.83 | −0.20 |
| ASA physical status I/II/III, n | 13/49/18 | 12/51/17 | — |
| ASA physical status III, n (%) | 18 (22.5) | 17 (21.3) | −0.03 |
| Planned stone extraction/lithotripsy, n (%) | 40 (50.0) | 40 (50.0) | 0.00 |
| Planned stricture dilation or stenting, n (%) | 24 (30.0) | 24 (30.0) | 0.00 |
| Planned diagnostic or other ERCP, n (%) | 16 (20.0) | 16 (20.0) | 0.00 |
| Baseline MAP, mmHg | 93.23 ± 7.27 | 93.62 ± 7.66 | 0.05 |
| Baseline HR, beats/min | 75.64 ± 7.75 | 74.92 ± 7.24 | −0.10 |
| Baseline SpO2, % | 98.86 ± 0.54 | 98.79 ± 0.61 | −0.11 |
| Baseline respiratory rate, breaths/min | 16.30 ± 2.33 | 16.15 ± 2.08 | −0.07 |
Notes: Data are mean ± standard deviation or n (%), unless otherwise indicated. Baseline variables are described using standardized differences rather than baseline P values. Planned ERCP procedure type was determined before randomization and was used as the stratification factor.
Abbreviations: BMI, body mass index; ERCP, endoscopic retrograde cholangiopancreatography; HR, heart rate; MAP, mean arterial pressure; SpO2, peripheral oxygen saturation.
Primary Outcome
Composite SRAEs occurred in 28 of 80 participants (35.0%) in the PF group and 14 of 80 (17.5%) in the RE group. The unadjusted RR was 0.50 (95% CI, 0.29–0.88), with an absolute risk difference of −17.5 percentage points (95% CI, −30.9 to −4.1; P=0.016). The corresponding number needed to treat was approximately six, although the confidence interval around the absolute risk difference was wide. After adjustment for age, sex, BMI, ASA III status, baseline MAP, and baseline SpO2, the adjusted RR was 0.49 (95% CI, 0.31–0.78; P=0.003). Primary and key safety outcomes are summarized in Table 2 and Figure 2.
Table 2.
Primary and Key Safety Outcomes
| Outcome | PF Group | RE Group | Effect Estimate (95% CI) or Method | Risk Difference, Percentage Points (95% CI) | P value |
|---|---|---|---|---|---|
| Composite SRAE | 28 (35.0) | 14 (17.5) | 0.50 (0.29–0.88) | −17.5 (−30.9 to −4.1) | 0.016 |
| Composite SRAE, adjusted | — | — | Adjusted RR 0.49 (0.31–0.78) | — | 0.003 |
| Hypotension | 18 (22.5) | 8 (10.0) | 0.44 (0.21–0.96) | −12.5 (−23.8 to −1.2) | 0.040 |
| Bradycardia | 9 (11.2) | 4 (5.0) | 0.44 (0.14–1.38) | −6.2 (−14.7 to 2.2) | 0.162 |
| Hypoxemia | 12 (15.0) | 6 (7.5) | 0.50 (0.20–1.27) | −7.5 (−17.2 to 2.2) | 0.144 |
| Respiratory depression | 8 (10.0) | 3 (3.8) | 0.38 (0.10–1.36) | −6.2 (−14.0 to 1.5) | 0.136 |
| Vasoactive drug use | 17 (21.2) | 7 (8.8) | 0.41 (0.18–0.94) | −12.5 (−23.4 to −1.6) | 0.035 |
| Clinically significant airway intervention | 14 (17.5) | 6 (7.5) | 0.43 (0.17–1.06) | −10.0 (−20.1 to 0.1) | 0.066 |
| Mask positive-pressure ventilation | 5 (6.2) | 2 (2.5) | 0.40 (0.08–2.00) | −3.8 (−10.1 to 2.6) | 0.265 |
| Unplanned tracheal intubation | 0 (0.0) | 0 (0.0) | — | 0.0 (0.0 to 0.0) | — |
| Hypotension burden, min | 0.0 (0.0–0.0); non-zero 18/80 | 0.0 (0.0–0.0); non-zero 8/80 | Mann–Whitney U-test | — | 0.031 |
| Hypoxemia burden, s | 0.0 (0.0–0.0); non-zero 12/80 | 0.0 (0.0–0.0); non-zero 6/80 | Mann–Whitney U-test | — | 0.122 |
Notes: RRs compare the RE group with the PF group. RRs, 95% CIs, and nominal P values were derived from modified Poisson regression with robust standard errors. The adjusted primary-outcome model included age, sex, BMI, ASA III status, baseline MAP, and baseline SpO2. Holm-adjusted P values for hypotension, bradycardia, hypoxemia, respiratory depression, and clinically significant airway intervention were 0.199, 0.409, 0.409, 0.409, and 0.266, respectively. Burden outcomes are median (interquartile range) with the number of participants with non-zero burden and were compared using Mann–Whitney U-tests; ranges were 0–11.2 versus 0–6.8 minutes for hypotension burden and 0–81.5 versus 0–96.0 seconds for hypoxemia burden (PF versus RE). Individual components and burden measures should be interpreted as supportive rather than confirmatory.
Abbreviations: CI, confidence interval; MAP, mean arterial pressure; RR, risk ratio; SRAE, sedation-related adverse event; SpO2, peripheral oxygen saturation.
Figure 2.
Forest plot of primary and key safety outcomes. Risk ratios compare the remimazolam–esketamine (RE) group with the propofol–fentanyl (PF) group. Circles represent point estimates, and horizontal lines represent 95% confidence intervals. The vertical dashed line indicates a risk ratio of 1. Values <1 favor RE. Displayed P values are nominal. Holm-adjusted P values for the prespecified component comparisons are reported in Table 2. Treatment-needed SRAE was a prespecified exploratory outcome.
Abbreviation: SRAE, sedation-related adverse event.
Respiratory and Hemodynamic Variables
Linear mixed-effects models showed significant group-by-time interactions for MAP, HR, SpO2, and respiratory rate (all P<0.001). Post hoc model contrasts showed higher values in the RE group at T1 and T2 for all four outcomes (Holm-adjusted P<0.001), with no clear group differences at T0 or T3 (Figure 3 and Supplementary Table S1). The lowest intraoperative MAP, HR, and SpO2 values were also higher in the RE group (Supplementary Table S2). Hypotension burden was 0.0 (0.0–0.0) minutes in both groups, but a non-zero burden occurred in 18/80 PF participants and 8/80 RE participants (ranges, 0–11.2 and 0–6.8 minutes; P=0.031). Hypoxemia burden was 0.0 (0.0–0.0) seconds in both groups, with a non-zero burden in 12/80 and 6/80 participants, respectively (ranges, 0–81.5 and 0–96.0 seconds; P=0.122).
Figure 3.
Changes in hemodynamic and respiratory variables over time. Panels show (A) mean arterial pressure, (B) heart rate, (C) peripheral oxygen saturation, and (D) respiratory rate. Points represent observed means and error bars represent standard deviations. T0, baseline before anesthesia induction; T1, 2 minutes after induction; T2, the time of sphincterotomy or pancreaticobiliary duct cannulation, or 30 minutes after endoscope insertion if neither occurred; T3, the time after discontinuation of the study drugs when the participant first opened their eyes in response to verbal commands and recovered to a Ramsay Sedation Scale score of 2. Group-by-time interactions were assessed using linear mixed-effects models (all P<0.001). *Holm-adjusted P<0.001 for RE versus PF at the indicated time point.
Abbreviations: PF, propofol-fentanyl; RE, remimazolam-esketamine; SD, standard deviation; SpO2, peripheral oxygen saturation.
Anesthetic Regimen Efficacy and Recovery Quality
Anesthetic regimen success was achieved in 79 of 80 patients (98.8%) in the PF group and 78 of 80 patients (97.5%) in the RE group, with no clear between-group difference (descriptive RR, 0.99; 95% CI, 0.95–1.03; Fisher’s exact P = 1.000). One patient in the PF group required conversion to an alternative regimen because of hemodynamic instability, whereas two patients in the RE group required conversion because the assigned regimen did not provide an adequate anesthetic effect. None required unplanned tracheal intubation. ERCP procedure duration was similar between groups. Induction time was longer in the RE group, whereas eye-opening time and time to a modified Aldrete score ≥9 were shorter. PACU stay did not differ clearly between groups. Body movement, coughing, and additional sedation also showed no clear between-group differences (Table 3 and Supplementary Figure S1).
Table 3.
Anesthetic Regimen Efficacy and Recovery Quality
| Outcome | PF Group | RE Group | Effect Estimate (95% CI) or Method | P value |
|---|---|---|---|---|
| Anesthetic regimen success, n (%) | 79 (98.8) | 78 (97.5) | Descriptive RR 0.99 (0.95–1.03) | 1.000 |
| Induction time, min | 2.08 ± 0.46 | 2.45 ± 0.52 | 0.38 (0.22 to 0.53) | <0.001 |
| ERCP procedure time, min | 36.01 ± 7.91 | 36.16 ± 7.30 | 0.15 (−2.23 to 2.53) | 0.902 |
| Eye-opening time, min | 9.41 ± 3.16 | 8.44 ± 2.60 | −0.97 (−1.87 to −0.07) | 0.035 |
| Time to modified Aldrete score ≥9, min | 13.94 ± 3.34 | 12.94 ± 2.56 | −1.00 (−1.92 to −0.07) | 0.036 |
| PACU stay, min | 24.97 ± 3.67 | 24.17 ± 2.83 | −0.80 (−1.83 to 0.22) | 0.124 |
| Body movement, n (%) | 5 (6.2) | 7 (8.8) | RR 1.40 (0.46–4.23) | 0.551 |
| Coughing, n (%) | 6 (7.5) | 5 (6.2) | RR 0.83 (0.27–2.62) | 0.755 |
| Additional sedation, n (%) | 7 (8.8) | 11 (13.8) | RR 1.57 (0.64–3.85) | 0.323 |
| NRS pain score | 1.0 (0.0, 1.0) | 1.0 (0.0, 1.0) | Mann–Whitney U-test | 0.134 |
| Patient satisfaction | 8.9 (8.3, 9.4) | 8.9 (8.6, 9.4) | Mann–Whitney U-test | 0.651 |
| Endoscopist satisfaction | 9.0 (8.5, 9.3) | 8.9 (8.4, 9.3) | Mann–Whitney U-test | 0.324 |
Notes: Continuous mean differences are RE group minus PF group and were compared using independent-samples t tests. Mean differences were calculated from unrounded data and may therefore differ slightly from differences between the displayed rounded means. The P value for anesthetic regimen success was calculated using Fisher’s exact test; the RR and 95% CI are presented descriptively. RRs, 95% CIs, and P values for body movement, coughing, and additional sedation were derived from modified Poisson regression with robust standard errors. NRS and satisfaction scores are presented as median (interquartile range) and were compared using Mann–Whitney U-tests.
Abbreviations: CI, confidence interval; NRS, numerical rating scale; PACU, post-anesthesia care unit; RR, risk ratio.
Postoperative and Drug-Specific Adverse Reactions
Among patients who completed the 24-hour follow-up, no clear between-group differences were observed in postoperative nausea and vomiting, dizziness, nasal discomfort, or emergence agitation. Injection pain was markedly less frequent in the RE group. Tachycardia, hypertension, and psychotomimetic symptoms were numerically more frequent in the RE group, but event counts were small and confidence intervals were wide (Table 4).
Table 4.
Postoperative and Drug-Specific Adverse Reactions
| Event | PF Group | RE Group | Effect Estimate (95% CI) | P value |
|---|---|---|---|---|
| PONV | 9/77 (11.7) | 6/78 (7.7) | RR 0.66 (0.25–1.76) | 0.405 |
| Dizziness | 4/77 (5.2) | 7/78 (9.0) | RR 1.73 (0.53–5.66) | 0.367 |
| Psychotomimetic symptoms/abnormal dreams | 0/77 (0.0) | 4/78 (5.1) | RD 5.1 percentage points (95% CI, −0.6 to 12.5) | 0.120 |
| Nasal discomfort | 8/77 (10.4) | 7/78 (9.0) | RR 0.86 (0.33–2.27) | 0.766 |
| Emergence agitation | 1/77 (1.3) | 3/78 (3.8) | RR 2.96 (0.31–27.85) | 0.342 |
| Injection pain | 18/80 (22.5) | 2/80 (2.5) | RR 0.11 (0.03–0.46) | 0.003 |
| Tachycardia | 2/80 (2.5) | 8/80 (10.0) | RR 4.00 (0.88–18.26) | 0.074 |
| Hypertension | 2/80 (2.5) | 7/80 (8.8) | RR 3.50 (0.75–16.34) | 0.111 |
Notes: PONV, dizziness, psychotomimetic symptoms, nasal discomfort, and emergence agitation were analyzed among patients who completed 24-hour follow-up. Injection pain, tachycardia, and hypertension were analyzed in the full analysis set. RRs, 95% CIs, and P values were derived from modified Poisson regression with robust standard errors, except for psychotomimetic symptoms, for which the risk difference with a score-based 95% CI and Fisher exact P value is reported because no events occurred in the PF group.
Abbreviations: PONV, postoperative nausea and vomiting; RD, risk difference; RR, risk ratio.
Sensitivity and Exploratory Analyses
In prespecified exploratory analyses, circulatory SRAEs occurred in 11 of 80 participants (13.8%) in the RE group and 21 of 80 (26.3%) in the PF group; respiratory SRAEs occurred in 8 of 80 (10.0%) and 17 of 80 (21.2%), respectively. Treatment-needed SRAEs occurred in 10/80 (12.5%) and 21/80 (26.3%), and no-event anesthetic regimen success occurred in 65/80 (81.3%) and 52/80 (65.0%), respectively. These prespecified exploratory findings are summarized in Table 5 and should be interpreted as supportive rather than confirmatory.
Table 5.
Prespecified Exploratory Decomposition and Composite Clinical Success Analyses
| Outcome | PF Group | RE Group | RR (95% CI) | P value |
|---|---|---|---|---|
| Circulatory SRAE | 21/80 (26.3) | 11/80 (13.8) | 0.52 (0.27–1.01) | 0.055 |
| Respiratory SRAE | 17/80 (21.2) | 8/80 (10.0) | 0.47 (0.22–1.03) | 0.059 |
| Treatment-needed SRAE | 21/80 (26.3) | 10/80 (12.5) | 0.48 (0.24–0.95) | 0.034 |
| No-event anesthetic regimen success | 52/80 (65.0) | 65/80 (81.3) | 1.25 (1.03–1.51) | 0.023 |
Notes: Circulatory SRAE was defined as hypotension or bradycardia, and respiratory SRAE as hypoxemia or respiratory depression. Treatment-needed SRAEs were events requiring vasoactive medication, atropine, mask positive-pressure ventilation, ERCP interruption, or unplanned tracheal intubation. No-event anesthetic regimen success was defined as anesthetic regimen success without composite SRAE. These analyses were prespecified as exploratory in Protocol Version 2.0 and should be interpreted as supportive rather than confirmatory. RRs, 95% CIs, and P values were derived from modified Poisson regression with robust standard errors.
Exploratory subgroup analyses by age, ASA physical status, BMI, and planned ERCP procedure type showed generally consistent directions of effect. Interaction tests provided no statistical evidence of treatment-effect heterogeneity, although the trial was not powered for definitive subgroup conclusions (Figure 4).
Figure 4.
Exploratory subgroup analyses of the primary composite sedation-related adverse event (SRAE) outcome. Risk ratios compare the remimazolam–esketamine (RE) group with the propofol–fentanyl (PF) group. Circles represent point estimates, and horizontal lines represent 95% confidence intervals. The vertical dashed line indicates a risk ratio of 1. Interaction P values were obtained from treatment-by-subgroup interaction tests. The subgroup domains were prespecified in Protocol Version 2.0; all subgroup analyses were exploratory and should be interpreted cautiously.
Abbreviations: ASA, American Society of Anesthesiologists; BMI, body mass index; ERCP, endoscopic retrograde cholangiopancreatography; PF, propofol–fentanyl; RE, remimazolam–esketamine; SRAE, sedation-related adverse event.
Patient-Level Severity of Composite SRAEs
After severity was coded conditionally on composite SRAE, grade-0 counts matched the primary composite outcome (52/80 in the PF group and 66/80 in the RE group). The patient-level severity distribution was lower in the RE group (Mann–Whitney U-test, P=0.010). This approach prevented technical nasopharyngeal airway adjustments without an SRAE-defining criterion from being misclassified as clinically meaningful severity increases; these adjustments are reported separately in Supplementary Table S3. Because the severity analysis was prespecified as exploratory rather than confirmatory, it should be interpreted as supportive evidence (Figure 5).
Figure 5.
Patient-level severity distribution of composite sedation-related adverse events (SRAEs). Each participant was classified according to the highest severity grade reached during the procedure. Grade 0 indicates no composite SRAE; Grade 1, composite SRAE without a prespecified rescue intervention; Grade 2, composite SRAE requiring vasoactive medication, atropine, nasopharyngeal airway (NPA) adjustment, or jaw thrust; Grade 3, composite SRAE requiring mask positive-pressure ventilation or temporary ERCP interruption; and Grade 4, unplanned tracheal intubation or a serious adverse event. Bar segments represent the proportions of randomized participants in each severity grade. Group distributions were compared using the Mann–Whitney U-test.
Abbreviations: ERCP, endoscopic retrograde cholangiopancreatography; NPA, nasopharyngeal airway; PF, propofol–fentanyl; RE, remimazolam–esketamine; SRAE, sedation-related adverse event.
Discussion
In this randomized trial of older patients undergoing ERCP with standardized nasopharyngeal airway-assisted respiratory management, the remimazolam-esketamine regimen was associated with fewer composite SRAEs than propofol-fentanyl, while anesthetic regimen success was similar. The absolute risk reduction of 17.5 percentage points corresponded to a number needed to treat of approximately six, although the confidence interval was wide. In a high-volume endoscopy unit, preventing events that require vasoactive medication, airway rescue, or procedural interruption could potentially reduce workflow disruption and resource use; however, operational and economic outcomes were not measured in this trial. Because airway support and rescue algorithms were identical, unequal respiratory management is unlikely to explain the primary result.
Our findings extend the growing ERCP-specific evidence base. Dong et al reported comparable procedural efficacy and more favorable respiratory and hemodynamic profiles with remimazolam than with propofol when both were combined with alfentanil.8 A 396-patient multicenter trial similarly found comparable sedation efficacy, less injection pain, and fewer moderate-to-severe cardiopulmonary events with remimazolam.9 Meta-analytic evidence supports fewer cardiorespiratory adverse events without loss of procedural success,10 and a randomized trial in therapeutic ERCP showed more stable blood pressure with remimazolam than with propofol.25 The present trial extends these observations to a remimazolam-esketamine regimen compared directly with propofol-fentanyl in an elderly population.
The physiological pattern observed in the RE group is consistent with the complementary properties of the two-drug regimen. Remimazolam generally produces less dose-dependent cardiovascular depression than propofol, whereas esketamine provides analgesia with relatively limited respiratory suppression and sympathomimetic effects.6,11 Low-dose esketamine has reduced propofol requirements during ERCP,12 and DDDT trials of remimazolam-esketamine during colonoscopy have reported stable hemodynamics, less respiratory depression or injection pain, and effective procedural conditions.13,14 However, both the hypnotic and analgesic components differed between groups in the present study. The findings therefore apply to the regimens as complete clinical strategies and do not identify the independent contribution of any individual drug.
The individual SRAE components warrant cautious interpretation. Although nominal estimates for hypotension, bradycardia, hypoxemia, respiratory depression, and clinically significant airway intervention generally favored the RE group, none remained statistically significant after Holm correction. The trial was powered for the composite endpoint rather than for its individual components. Composite outcomes can improve statistical efficiency, but their interpretation depends on the clinical importance, frequency, and directional consistency of the component events.21,22 The generally concordant component estimates, fewer treatment-needed SRAEs, and favorable physiological trends support the overall safety signal without establishing an independent reduction in every component.
The lower SRAE rate was accompanied by similar anesthetic regimen success and ERCP duration. Induction was modestly slower with remimazolam-esketamine, whereas eye-opening and modified Aldrete recovery time were slightly shorter and PACU stay was similar. Propofol typically produces rapid loss of consciousness, whereas the conservative, titrated remimazolam dose used in older participants may require more time to reach the target depth. After drug discontinuation, rapid tissue-esterase metabolism of remimazolam may limit accumulation and contribute to prompt recovery.6 This pharmacological explanation remains speculative because pharmacokinetic sampling and cumulative drug exposure were not collected.
The safety profile was not uniformly favorable. Injection pain was less frequent in the RE group, whereas tachycardia, hypertension, and psychotomimetic symptoms were numerically more frequent. These observations are pharmacologically plausible for esketamine, but the event counts were small and confidence intervals were wide; the study was therefore underpowered to establish regimen-specific differences in these uncommon outcomes. They should be viewed as signals for further evaluation rather than as definitive clinical recommendations.
Standardized nasopharyngeal airway use was an important design feature. Airway access is limited during prone or semi-prone ERCP, and respiratory support strategies can materially influence hypoxemia rates.1,15 Applying the same airway adjunct and rescue sequence in both groups reduced confounding by airway management. It may also have lowered the overall frequency of severe hypoxemia, making between-group respiratory differences more conservative.
The prespecified exploratory patient-level severity analysis was directionally consistent with the primary endpoint. Defining severity conditionally on composite SRAE ensured that grade 0 corresponded exactly to the absence of SRAE and prevented isolated technical nasopharyngeal airway repositioning from being misclassified as a clinically meaningful severity increase. Because the ordinal analysis was exploratory, it should be considered supportive and hypothesis-generating.
Limitations
Several limitations should be acknowledged. First, this single-center trial was conducted by relatively fixed anesthesia and endoscopy teams, which may limit generalizability. Second, treating anesthesiologists and personnel recording intraoperative data could not be blinded. Although BIS and hemodynamic thresholds were prespecified, the interpretation of purposeful movement and clinically apparent responses required bedside judgment and may have introduced performance or ascertainment bias. Third, the protocol-permitted rescue strategies differed: supplemental fentanyl was available in the PF group, whereas the RE group received continuous esketamine without additional esketamine boluses. This asymmetry may have influenced supplemental-medication use and the small number of regimen failures. Fourth, the composite endpoint included events of differing severity, and the trial was not powered for confirmatory conclusions about individual components, uncommon adverse reactions, or subgroups. Fifth, both the hypnotic and analgesic components differed between groups, so their independent contributions cannot be disentangled. Sixth, the eligibility criteria selected a comparatively lower-risk older population; patients with BMI outside 18–28 kg/m2, anticipated difficult airway, severe cardiopulmonary dysfunction, or ASA physical status IV or higher were excluded. Seventh, formal postoperative delirium was not assessed with a validated instrument; recorded psychotomimetic symptoms and emergence agitation should not be considered equivalent to delirium assessment. Finally, raw end-tidal carbon dioxide waveforms, cumulative drug exposure, secretion-related outcomes, validated quality-of-recovery measures, and resource use were not collected. Future multicenter trials should evaluate higher-risk populations, neurocognitive recovery, alternative dosing, total drug exposure, and cost-effectiveness.
Conclusion
In this selected population of older patients undergoing ERCP with standardized nasopharyngeal airway-assisted respiratory management, the remimazolam-esketamine regimen was associated with fewer composite SRAEs than propofol-fentanyl while achieving a similar anesthetic regimen success rate. Multicenter trials are needed to confirm these findings, particularly in frail and otherwise higher-risk older patients. Induction was modestly longer with remimazolam-esketamine; numerically higher rates of tachycardia, hypertension, and psychotomimetic symptoms remain uncertain because event counts were small and estimates were imprecise.
Acknowledgments
We thank all study participants and the clinical and research staff who contributed to data collection and study conduct.
Funding Statement
This work was supported by the 2023 Anhui Medical University Research Fund (Grant No. 2023xkj065), awarded to Yongbo Yu.
Abbreviations
ASA, American Society of Anesthesiologists; BIS, bispectral index; BMI, body mass index; CMMS, Chinese Mini Mental Status; CI, confidence interval; ERCP, endoscopic retrograde cholangiopancreatography; HR, heart rate; MAP, mean arterial pressure; NPA, nasopharyngeal airway; NRS, numerical rating scale; PACU, post-anesthesia care unit; PF, propofol-fentanyl; PONV, postoperative nausea and vomiting; RE, remimazolam-esketamine; RR, risk ratio; SpO2, peripheral oxygen saturation; SRAE, sedation-related adverse event.
Data Sharing Statement
Deidentified individual participant data supporting the findings of this study and the study protocol will be available from the corresponding author, Xiang Gao, upon reasonable request after publication, subject to applicable institutional and ethical requirements. Requests should include a methodologically sound proposal.
Approval and Informed Consent
The original study protocol was approved by the Ethics Committee of The Fourth Affiliated Hospital of Anhui Medical University (approval number KYXM-202511-018; December 10, 2025). Protocol Version 2.0 and informed consent form Version 2.0 were approved as a protocol amendment under the same approval number on January 20, 2026, before the first participant signed written informed consent on January 21, 2026. All participants provided written informed consent before enrollment and were treated according to the ethics-approved Protocol Version 2.0. The trial was registered in the Chinese Clinical Trial Registry before enrollment (ChiCTR2600116715; January 14, 2026).
Author Contributions
All authors made substantial contributions to the conception and design of the study, acquisition of data, or analysis and interpretation of data; took part in drafting the article or revising it critically for important intellectual content; agreed to submit to the current journal; gave final approval of the version to be published; and agree to be accountable for all aspects of the work.
Disclosure
The authors report no conflicts of interest in this work.
References
- 1.Azimaraghi O, Bilal M, Amornyotin S, et al. Consensus guidelines for the perioperative management of patients undergoing endoscopic retrograde cholangiopancreatography. Br J Anaesth. 2023;130(6):763–16. doi: 10.1016/j.bja.2023.03.012 [DOI] [PubMed] [Google Scholar]
- 2.Althoff FC, Agnihotri A, Grabitz SD, et al. Outcomes after endoscopic retrograde cholangiopancreatography with general anaesthesia versus sedation. Br J Anaesth. 2021;126(1):191–200. doi: 10.1016/j.bja.2020.08.057 [DOI] [PubMed] [Google Scholar]
- 3.Sidhu R, Turnbull D, Haboubi H, et al. British Society of Gastroenterology guidelines on sedation in gastrointestinal endoscopy. Gut. 2024;73(2):219–245. doi: 10.1136/gutjnl-2023-330396 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Wang P, Chen Y, Guo Y, et al. Comparison of propofol-nalbuphine and propofol-fentanyl sedation for patients undergoing endoscopic retrograde cholangiopancreatography: a double-blind, randomized controlled trial. BMC Anesthesiol. 2022;22(1):47. doi: 10.1186/s12871-022-01578-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Smith ZL, Mullady DK, Lang GD, et al. A randomized controlled trial evaluating general endotracheal anesthesia versus monitored anesthesia care and the incidence of sedation-related adverse events during ERCP in high-risk patients. Gastrointest Endosc. 2019;89(4):855–862. doi: 10.1016/j.gie.2018.09.001 [DOI] [PubMed] [Google Scholar]
- 6.Hu Q, Liu X, Wen C, Li D, Lei X. Remimazolam: an updated review of a new sedative and anaesthetic. Drug Des Devel Ther. 2022;16:3957–3974. doi: 10.2147/DDDT.S384155 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Barbosa EC, Espirito Santo PA, Baraldo S, Meine GC. Remimazolam versus propofol for sedation in gastrointestinal endoscopic procedures: a systematic review and meta-analysis. Br J Anaesth. 2024;132(6):1219–1229. doi: 10.1016/j.bja.2024.02.005 [DOI] [PubMed] [Google Scholar]
- 8.Dong SA, Guo Y, Liu SS, et al. A randomized, controlled clinical trial comparing remimazolam to propofol when combined with alfentanil for sedation during ERCP procedures. J Clin Anesth. 2023;86:111077. doi: 10.1016/j.jclinane.2023.111077 [DOI] [PubMed] [Google Scholar]
- 9.Han SY, Oh D, Choe JW, et al. Comparative safety and efficacy of remimazolam versus propofol for ERCP with the patient under deep sedation: a multicenter randomized controlled trial. Gastrointest Endosc. 2026;103(6):1192–1200. doi: 10.1016/j.gie.2025.11.002 [DOI] [PubMed] [Google Scholar]
- 10.Kim IJ, Choi GJ, Oh HC, Kang H. Comparison of remimazolam and propofol for sedation in endoscopic retrograde cholangiopancreatography: a systematic review and meta-analysis with trial sequential analysis. Dig Endosc. 2025;37(10):1054–1067. doi: 10.1111/den.15078 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Mion G, Himmelseher S. Esketamine: less drowsiness, more analgesia. Anesth Analg. 2024;139(1):78–91. doi: 10.1213/ANE.0000000000006851 [DOI] [PubMed] [Google Scholar]
- 12.Eberl S, Koers L, van Hooft J, et al. The effectiveness of a low-dose esketamine versus an alfentanil adjunct to propofol sedation during endoscopic retrograde cholangiopancreatography: a randomized controlled multicentre trial. Eur J Anaesthesiol. 2020;37(5):394–401. doi: 10.1097/EJA.0000000000001134 [DOI] [PubMed] [Google Scholar]
- 13.Li W, Zhao J, Hao R, et al. The efficacy and safety of remimazolam besylate combined with esketamine for outpatient colonoscopy: a prospective, randomized, controlled clinical trial. Drug Des Devel Ther. 2023;17:2875–2887. doi: 10.2147/DDDT.S425860 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Deng BR, Zhang Y, Xie ZF, et al. Comparative analysis of hemodynamic effects of remimazolam and propofol combined with esketamine in colonoscopic procedures in the elderly. Drug Des Devel Ther. 2024;18:5269–5280. doi: 10.2147/DDDT.S490179 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Guo M, Lu J, Li B, Hu L, Bo L. Comparative efficacy of modified nasopharyngeal airway and nasal cannula for oxygenation in deeply sedated patients undergoing endoscopic retrograde cholangiopancreatography: a randomized controlled trial. BMC Anesthesiol. 2026;26:102. doi: 10.1186/s12871-025-03515-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.World Medical Association. World Medical Association Declaration of Helsinki: ethical principles for medical research involving human participants. JAMA. 2025;333(1):71–74. doi: 10.1001/jama.2024.21972 [DOI] [PubMed] [Google Scholar]
- 17.Hopewell S, Chan AW, Collins GS, et al. CONSORT 2025 statement: updated guideline for reporting randomised trials. BMJ. 2025;389:e081123. doi: 10.1136/bmj-2024-081123 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Katzman R, Zhang MY, Wang Z, et al. A Chinese version of the Mini-Mental State Examination; impact of illiteracy in a Shanghai dementia survey. J Clin Epidemiol. 1988;41(10):971–978 doi: 10.1016/0895-4356(88)90034-0. [DOI] [PubMed] [Google Scholar]
- 19.Meng Y, Song Y, Gao C, Wang S, He K, Xu M. Determining the median effective dose of remimazolam for anesthesia induction in patients undergoing endoscopic retrograde cholangiopancreatography: an age-stratified study. Drug Des Devel Ther. 2025;19:9797–9808. doi: 10.2147/DDDT.S537897 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Huang L, Liu LL, Lu YD, et al. Assessing sedation depth with PSI in elderly ERCP patients: a prospective cohort study. Clin Interv Aging. 2025;20:137–145. doi: 10.2147/CIA.S504212 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Freemantle N, Calvert M, Wood J, Eastaugh J, Griffin C. Composite outcomes in randomized trials: greater precision but with greater uncertainty? JAMA. 2003;289(19):2554–2559. doi: 10.1001/jama.289.19.2554 [DOI] [PubMed] [Google Scholar]
- 22.Montori VM, Permanyer-Miralda G, Ferreira-Gonzalez I, et al. Validity of composite end points in clinical trials. BMJ. 2005;330(7491):594–596. doi: 10.1136/bmj.330.7491.594 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Zou G. A modified Poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159(7):702–706. doi: 10.1093/aje/kwh090 [DOI] [PubMed] [Google Scholar]
- 24.Holm S. A simple sequentially rejective multiple test procedure. Scand J Stat. 1979;6(2):65–70. [Google Scholar]
- 25.Xiao YY, Zou HD, Qin XN, Zhu R, Dai RP. A comparison of remimazolam versus propofol on blood pressure changes during therapeutic endoscopic retrograde cholangiopancreatography: a randomized controlled trial. Anesth Analg. 2025;141(1):90–99. doi: 10.1213/ANE.0000000000007309 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Deidentified individual participant data supporting the findings of this study and the study protocol will be available from the corresponding author, Xiang Gao, upon reasonable request after publication, subject to applicable institutional and ethical requirements. Requests should include a methodologically sound proposal.