3.1. Study Design and Setting
This single-center, two-group randomized controlled clinical trial was conducted at Alzahra Hospital in Isfahan, Iran, and included adult patients scheduled for laparoscopic cholecystectomy. The study adhered to the Declaration of Helsinki, and written informed consent was obtained from all participants.
3.2. Participants and Eligibility Criteria
Patients aged 19 – 65 years with American Society of Anesthesiologists (ASA) physical status I or II were eligible. Patients with a known allergy to the study medications, psychological disorders that could compromise pain assessment, a history of opioid or substance use, or an inability to use the visual analogue scale (VAS) reliably were excluded. Conversion from laparoscopic to open surgery or a major change in the planned anesthetic technique constituted post-enrollment exclusion criteria.
3.3. Sample Size and Randomization
The protocol specified 44 participants per group (88 total), based on a standardized effect size of 0.5, two-sided α = 0.05, 80% power, a within-participant correlation of 0.5, and one baseline plus nine repeated measurements, allowing for approximately 10% attrition. Repeated-measures ANOVA was prespecified for longitudinal postoperative VAS analysis. No clinically acceptable noninferiority or equivalence margin was prospectively defined. Therefore, any retrospective assertion that the study was powered to demonstrate dose equivalence would be inappropriate. The allocation sequence was generated using Random Allocation Software with simple randomization. Following enrollment, the admission nurse assigned participants in order of entry according to the computer-generated list. Allocation concealment was not documented in the study records.
3.4. Interventions and Anesthetic Management
Designated nursing or medication-preparation personnel prepared and coded the study medication. Patients, treating anesthesiologists, outcome assessors, and the statistician were blinded to group assignments. In the higher-dose group, the total intravenous dose was 1 µg/kg; in the lower-dose group, it was 0.1 µg/kg. Both doses were diluted with 0.9% normal saline to a final volume of 20 mL, supplied in identical coded syringes, and administered over 10 minutes beginning 15 minutes before surgery. The protocols differed by a factor of ten in cumulative drug exposure. After administration of midazolam 2 mg IV, induction consisted of fentanyl 2 µg/kg IV, propofol 2.5 mg/kg IV, and cisatracurium 0.15 mg/kg IV for endotracheal intubation, followed immediately by morphine 0.1 mg/kg IV. Anesthesia was maintained with isoflurane at approximately 1 minimum alveolar concentration (MAC). No additional intraoperative opioid was administered. Supplemental cisatracurium 0.1 mg/kg IV was permitted for inadequate muscle relaxation at the treating anesthesiologist’s discretion; cumulative dose and frequency were not systematically recorded. At the conclusion of surgery, all patients received neostigmine 0.04 mg/kg IV and atropine 0.02 mg/kg IV. Extubation was performed after confirmation of adequate spontaneous ventilation, clinically satisfactory tidal volume, command following, and the ability to raise and sustain an arm. Quantitative train-of-four monitoring was not routinely employed.
3.5. Postoperative Analgesia and Antiemetic Rescue
Postoperative pain was assessed using a 0 – 10 visual analogue scale (VAS). When the VAS score exceeded 4, rescue treatment consisted of a single rectal diclofenac suppository (100 mg); a second dose was not routinely provided within 24 hours. If diclofenac was contraindicated or insufficient, pethidine 0.5 mg/kg IV was administered. Rescue antiemetic therapy consisted of ondansetron 4 mg IV after at least three episodes of vomiting or when nausea was considered clinically severe based on reports from the patient and the treating clinician; no standardized nausea-severity instrument was used. The occurrence of nausea and vomiting, ondansetron use, and rescue analgesic administration were recorded.
3.6. Outcome Measures
The prespecified primary outcome was postoperative pain measured using the visual analogue scale (VAS) over the first 24 hours; the protocol specified repeated-measures ANOVA. Baseline postoperative VAS was recorded immediately upon admission to the post-anesthesia care unit (PACU) and before any postoperative rescue analgesia. Subsequent VAS measurements were obtained at 15, 30, 45, 60, and 75 minutes during recovery and at 2, 6, 12, and 24 hours after ward entry. Additional recorded outcomes included nausea and vomiting, rescue medication use, heart rate, mean arterial pressure (MAP), oxygen saturation, a recorded anesthesia-time variable, extubation time, PACU stay, and patient satisfaction. Extubation time was defined as the interval from discontinuation of isoflurane to removal of the endotracheal tube, and surgery duration as the time from initial incision to final skin suture. PACU stay was measured from PACU entry to actual exit. PACU discharge readiness was determined using a Modified Aldrete Score of at least 9 out of 10 and documented only at discharge. Patient satisfaction was assessed 24 hours after surgery using a five-point ordinal Likert scale (1, very dissatisfied; 5, very satisfied). Pain subtypes were not recorded separately, and individual first-ambulation times were not measured, although mobilization at approximately 8 hours was encouraged. Hemodynamic events were defined as hypotension (MAP > 20% below baseline), hypertension (MAP > 20% above baseline), bradycardia (heart rate < 60 beats/min), and tachycardia (heart rate > 100 beats/min). Clinically significant bradycardia, when symptomatic or accompanied by hypotension, was treated with intravenous atropine 0.01 mg/kg. Hypotension requiring intervention was managed with weight-adjusted intravenous 0.9% saline; the exact administered volumes were not recorded. For clinically significant hypertension with bradycardia, a 50 µg intravenous nitroglycerin bolus was administered. If the heart rate was ≥ 70 beats/min, a 5 mg intravenous labetalol bolus was administered; intermediate heart rates were managed at the clinician’s discretion. Tachycardia requiring intervention could be treated with intravenous labetalol; isolated numerical tachycardia did not necessitate medication. The number of patients receiving event-specific interventions could not be determined from the available data.
The protocol specified longitudinal VAS analysis using repeated-measures ANOVA. Covariateadjusted models accounting for baseline imbalances in age, sex, weight, and height were
performed as exploratory, non-prespecified sensitivity analyses.
The original research proposal supports the primary outcome and the prespecified repeated-measures approach. Post hoc covariate-adjusted models were conducted to evaluate robustness to baseline imbalances in age, sex, weight, and height; these analyses should not be presented as prospectively planned.
The VAS is a 10-cm horizontal scale anchored by “no pain” and “worst pain imaginable” and is widely used to assess clinical pain. Published evidence supports its validity and reliability for pain measurement (
30), and Iranian data have also demonstrated strong agreement with established pain assessment instruments and high reliability (
31).
3.7. Statistical Analysis
The initial analyses used SPSS version 27 and included repeated-measures F tests, independentsamples tests, and categorical comparisons. Additional exploratory covariate-adjusted analyses were performed on individual-patient data from 88 patients using Python statsmodels 0.14.6. For postoperative VAS scores at 15, 30, 45, 60, and 75 minutes in recovery and at 2, 6, 12, and 24 hours after transfer to the ward, a population-averaged Gaussian GEE was fitted with participant as the clustering unit, an exchangeable working correlation, categorical visit, treatment group, a treatment-by-visit interaction, and age, female sex, weight, height, and baseline VAS as covariates. The principal adjusted contrast was the equally weighted average of the between-group mean differences across the nine postoperative visits. Treatment-by-visit terms were jointly assessed using an eight-degree-of-freedom robust Wald test. Equivalent seven-visit GEE models (without a baseline outcome covariate) evaluated HR, MAP, and SpO₂. Binary outcomes with sufficient events were analyzed using binomial generalized linear models with the same four demographic covariates and HC3 robust standard errors; adjusted odds ratios were reported (and are not directly comparable with unadjusted risk ratios). Rare adverse events with zero or very few observations per arm were not submitted to unstable, fully adjusted models. Continuous recovery outcomes were fitted using OLS with HC3 robust standard errors. Satisfaction was analyzed using proportional-odds ordinal logistic regression. Conditional outcomes measured only after rescue treatment were considered exploratory and subject to selection bias. Two-sided nominal 95% confidence intervals are reported; no multiplicity adjustment was applied. These analyses represent post-review sensitivity analyses and were not documented as prespecified primary analyses.
Rare adverse events, including hypotension, hypertension, tachycardia, bradycardia, and other adverse events, occurred only 0 – 4 times in at least one study arm. Consequently, four-covariate logistic adjustment was considered unreliable because of sparse data or data separation. Therefore, unadjusted counts and exact tests are presented descriptively. In addition, the absence of event-specific source records precludes reliable adjusted analyses of adverse-event severity and intervention.