Analgesic effect of intraoperative sufentanil on postoperative pain in patients undergoing robot-assisted urologic surgery: a retrospective propensity score-matched analysis

Abstract

Background

Robot-assisted urologic surgery is associated with reduced intraoperative bleeding and shortened hospital stays; however, postoperative pain remains concerning. Thus, the authors investigated whether intraoperative sufentanil is more effective than remifentanil in reducing postoperative pain and analgesic consumption.

Methods

Patients undergoing robot-assisted urologic surgery (March 2023 to March 2024) were divided into the sufentanil or remifentanil groups and matched by propensity score. The primary outcome was the distribution of pain severity after surgery. Secondary outcomes included pain scores and rescue analgesic consumption.

Results

Among 832 eligible patients, 194 were matched from each group. The distribution of pain severity after surgery was significantly different between groups (rank-biserial correlation −0.34, 95% confidence interval [CI] −0.43 to −0.24; P < 0.001). The sufentanil group had more patients with no/mild pain (27.8% vs. 4.6%, risk difference [RD] 23.2%, 95% CI 16.2% to 30.2%; P < 0.001) and fewer patients with severe pain (32.5% vs. 57.2%, RD −24.7%, 95% CI −34.3% to −15.2%; P < 0.001). Immediate postoperative pain score was lower with sufentanil (6 [3, 7] vs. 7 [6, 8], Hodges–Lehmann estimator [HL estimator] −1, 95% CI −2 to −1; P < 0.001). During the 48-hour postoperative period, the sufentanil group required significantly lower opioid rescue analgesic consumption (13.3 [5, 22.5] vs. 20 [10, 29.6] mg of morphine-equivalent dose, HL estimator −5 mg, 95% CI −6.7 to −2.5 mg; P < 0.001).

Conclusions

Intraoperative sufentanil was associated with lower postoperative pain and analgesic consumption than was remifentanil in patients undergoing robot-assisted urologic surgery.

INTRODUCTION

Robot-assisted surgery is currently the standard of care for major procedures, including radical prostatectomy, nephrectomy, nephroureterectomy, and radical cystectomy [1]. This approach provides superior three-dimensional visualization and precise tissue manipulation, leading to less intraoperative blood loss, shorter hospital stays, and lower postoperative pain levels than does open surgery [2]. Consequently, robot-assisted urologic surgery is increasingly being adopted worldwide, particularly in the United States, Europe, Japan, and Republic of Korea [3,4]. Despite these benefits, patients undergoing robot-assisted urologic surgery experience substantial postoperative pain owing to factors such as port incisions, pneumoperitoneum, and organ manipulation. This emphasizes the need to optimize intraoperative pain management approaches to enhance recovery and patient satisfaction [5–7].

Given the concerns about opioid dependence and adverse effects such as respiratory depression and postoperative nausea and vomiting (PONV) [8,9], finding an optimal intraoperative opioid remains critical. Sufentanil, a potent μ-opioid receptor agonist, has demonstrated postoperative analgesic efficacy that is superior to that of remifentanil (an ultra-short-acting μ-opioid receptor agonist). Sufentanil reduces pain scores and analgesic requirements across colorectal [10], gynecological [11], and neurosurgical procedures [12].

However, evidence supporting its use in robot-assisted urologic surgery remains limited. Given the absence of robust data in this surgical field, the authors aimed to determine whether intraoperative sufentanil administration is associated with improved postoperative pain scores and reduced postoperative rescue analgesic consumption compared with intraoperative remifentanil administration in patients undergoing robot-assisted urologic surgery.

MATERIALS AND METHODS

1. Ethics statement

This retrospective study was approved by the Seoul National University Bundang Hospital’s Institutional Review Board on April 29, 2024 (IRB no. B-2405-901-101); the IRB waived the need for informed consent because of the retrospective design of the study.

2. Inclusion and exclusion criteria

The authors included adult patients (aged ≥ 19 years) who received general anesthesia for robot-assisted urologic surgery between March 2023 and March 2024. The exclusion criteria comprised: (1) a history of transfer to the intensive care unit after surgery, (2) a history of undergoing single-port surgery, (3) a history of being subjected to intraoperative administration of both sufentanil and remifentanil, (4) a history of being subjected to intraoperative administration of neither sufentanil nor remifentanil, (5) a history of refusing patient-controlled analgesia (PCA), and (6) a history of use of fentanyl-based PCA.

3. Anesthesia protocol

Anesthetic and postoperative analgesic management performed in this study was based on standardized institutional protocol, which has been described previously [13]. Patients were monitored intraoperatively using electrocardiography, pulse oximetry, noninvasive blood pressure measurement, processed electroencephalography (SedLine Patient State Index, PSi; Masimo Inc.), and neuromuscular monitoring (Philips Intellivue NMT monitor; Royal Philips Electronics). Anesthesia was induced using propofol with the assigned opioid (sufentanil or remifentanil). Once patients lost consciousness, they were administered 0.6 mg/kg of rocuronium for neuromuscular blockade. Subsequently, they underwent endotracheal intubation with a video laryngoscope (KoMAC®, KoMAC Co., Ltd.) using a 7.0–7.5 mm tube. Each opioid was administered intravenously via target-controlled infusion. In the sufentanil group, the effect-site concentration was 0.3 ng/mL for induction and 0.1–0.5 ng/mL for maintenance, based on the Gepts model. In the remifentanil group, the effect-site concentration was 3 ng/mL for induction and 1–5 ng/mL for maintenance, based on the Minto model. In both groups, the effect-site concentration was adjusted to maintain mean arterial blood pressure within 20% of baseline values. Sufentanil infusion was stopped at robot de-docking, whereas remifentanil infusion was discontinued after skin closure.

On completion of the surgery, 0.3 mg of ramosetron was administered, and a PCA device (Accufuser®, Woo Young Medical) was connected intravenously. The PCA comprised 180 mg (for patients aged < 65 years) or 120 mg (for patients aged ≥ 65 years) ketorolac in a total volume of 50 mL. Sugammadex 2 or 4 mg/kg was administered to reverse the neuromuscular blockade, and patients were transferred to the post-anesthetic care unit (PACU) after extubation.

During the PACU stay, a nurse assessed patients using the modified Aldrete score and an 11-point numerical rating scale (NRS; 0 = no pain, 10 = worst pain) at fixed 15-minute intervals, including at PACU admission, at discharge, and additionally whenever patients reported pain and required rescue analgesics. When the NRS score was ≥ 4, rescue analgesics (fentanyl at 50, 50, 25 µg increments) were administered. If patients continued to complain of pain or requested additional analgesics after receiving the full fentanyl regimen, 1 g of acetaminophen or 20 mg of nefopam was administered at the discretion of the attending anesthesiologist. Patients were discharged from the PACU when the modified Aldrete score was ≥ 9. In the ward, when patients had an NRS score ≥ 4 or requested additional analgesics, 100 mg of tramadol or 25 mg of meperidine was administered according to the surgeon’s preference.

4. Clinical data collection

The authors reviewed the electronic medical records and collected patients’ demographic and clinical data, including sex, age, height, body weight, body mass index, American Society of Anesthesiologists (ASA) physical status, and medical history (diabetes mellitus, hypertension, dyslipidemia, or hyperlipidemia). Intraoperative variables included surgeon, surgical and anesthetic type, surgical and anesthetic duration, crystalloid administration, estimated blood loss, and intraoperative opioid consumption. Postoperative data included pain scores, rescue analgesic consumption, PONV, incidence of oxygen desaturation (SpO2 < 95%), PACU duration, time to first ambulation, time to canalization (return of bowel function), and length of hospital stay (LOS).

The primary outcome was the distribution of pain severity upon arrival at the PACU. Pain scores were classified as no/mild pain (0–3), moderate pain (4–6), and severe pain (7–10) [14,15]. Secondary outcomes included recovery profiles during PACU stay (pain scores, time to first rescue analgesic administration, PONV, SpO₂ < 95%, and PACU duration), cumulative rescue opioid consumption (at PACU admission and 6, 12, 24, and 48 hours after PACU admission), rescue non-opioid (acetaminophen and nefopam) requirement (during PACU stay), and postoperative recovery indicators (time to first ambulation, time to canalization, and LOS). Opioid analgesics were converted to morphine-equivalent doses (MEDs) for standardized comparison (Supplementary Table 1) [16–18].

5. Statistical analysis

Normality of continuous variables was examined with Shapiro–Wilk test; such variables are reported as mean ± standard deviation when normally distributed and as median with interquartile range [Q1, Q3] when not normally distributed. Comparisons between groups were performed using Student’s t-test (for normally distributed data), with mean differences and 95% confidence interval (CI) reported, or Mann–Whitney U-test (for non-normally distributed data), with Hodges–Lehmann estimator (HL estimator) and its 95% CI reported. Categorical variables are reported as numbers (percentages) and were analyzed using chi-square or Fisher’s exact test; rank-biserial correlations with 95% CI were reported as effect size for ordered categorical comparisons, and risk difference (RD) with 95% CI for binary comparisons. The ordinal pain-severity distribution was the single confirmatory analysis for the primary outcome, assessed using Mann–Whitney U-test and the rank-biserial correlation. The category-specific comparisons (no/mild and severe pain) were prespecified descriptive analyses to characterize where the distributional shift occurred; they were reported as RD with 95% CIs and interpreted as exploratory rather than as confirmatory; therefore, they were not adjusted for multiplicity.

Propensity score matching (PSM) was conducted using the MatchIt package in R (R Project for Statistical Computing) with a 1:1 nearest-neighbor method and a caliper of 0.2; the matched dataset was then analyzed using Jamovi v.2.4.14 (Jamovi Project) for all subsequent comparisons. Covariates included age, sex, weight, surgeon, surgical type, anesthetic type (volatile vs. intravenous), specific volatile anesthetic used, ASA class, duration of surgery and anesthesia, crystalloid administration, and PCA dosage. Covariate balance was assessed using absolute standardized differences; values < 0.2 indicated an adequate match.

To assess the robustness of the primary outcome against potential temporal or practice-pattern effects, the authors performed three prespecified sensitivity analyses: (1) restriction of the cohort to the period after sufentanil was introduced into routine practice (June 2023 to March 2024); (2) a Cochran–Mantel–Haenszel (CMH) analysis of the primary outcome stratified by successive intervals from the start of the study period (March to May 2023, June to August 2023, September to November 2023, and December 2023 to March 2024), with a Tarone test of homogeneity; and (3) logistic regression for the no/mild-pain and severe-pain categories including the surgical month as a covariate to adjust for temporal trends. P values < 0.05 were considered indicative of statistical significance.

RESULTS

1. Patient selection and group matching

Between March 2023 and March 2024, 1,110 patients underwent robot-assisted urologic surgery at the authors’ hospital. Among these, after excluding 278 patients who met the exclusion criteria (Fig. 1), the remaining 832 patients were assigned to the sufentanil group (n = 241) or remifentanil group (n = 591). After PSM, 194 patients were included in each group (Fig. 1). Tables 1, 2, and 3, respectively, present the preoperative, intraoperative surgical, and anesthetic characteristics before and after PSM. An absolute standardized difference of < 0.2 confirmed an adequate balance between the groups.

Fig. 1.

Flow diagram of patient selection and propensity score matching. PCA: patient-controlled analgesia.

Table 1.

Comparison of patient preoperative data

Variable	Entire cohort				After PSM
	Sufentanil (n = 241)	Remifentanil (n = 591)	ASD		Sufentanil (n = 194)	Remifentanil (n = 194)	ASD
Age (yr)	65.0 [55.0, 71.0]	66.0 [59.0, 71.0]	0.247		67.0 [58.0, 72.0]	65.0 [54.3, 70.0]	0.119
Sex (male/female)	195 (80.9)/46 (19.1)	521 (88.2)/70 (11.8)	0.201		167 (86.1)/27 (13.9)	162 (83.5)/32 (16.5)	0.072
Height (cm)	167.0 [161.7, 172.0]	167.0 [162.8, 171.6]	0.017		167.1 ± 7.8	167.4 ± 7.0	0.037
Weight (kg)	67.5 [60.0, 75.5]	68.2 [61.5, 75.8]	0.026		67.8 [59.9, 75.3]	69.3 [61.1, 76.5]	0.087
BMI (kg/m2)	24.3 [22.4, 26.6]	24.7 [22.5, 26.6]	0.036		24.1 [22.3, 26.4]	25.1 [22.3, 26.7]	0.094
ASA class			0.067				0.061
I	40 (16.6)	84 (14.2)			24 (12.4)	28 (14.4)
II	175 (72.6)	439 (74.3)			148 (76.3)	145 (74.7)
III	26 (10.8)	68 (11.5)			22 (11.3)	21 (10.8)
Diabetes mellitus	18 (7.5)	36 (6.1)	0.055		16 (8.2)	11 (5.7)	0.101
Hypertension	47 (19.5)	117 (19.8)	0.007		40 (20.6)	34 (17.5)	0.079
Dyslipidemia/hyperlipidemia	70 (29.0)	175 (29.6)	0.012		61 (31.4)	59 (30.4)	0.022

Values are presented as median [interquartile range Q1, Q3], number (%), or mean ± standard deviation.

PSM: propensity score matching, ASD: absolute standardized difference, BMI: body mass index, ASA: American Society of Anesthesiologists.

Table 2.

Comparison of surgical data

Variable	Entire cohort				After PSM
	Sufentanil (n = 241)	Remifentanil (n = 591)	ASD		Sufentanil (n = 194)	Remifentanil (n = 194)	ASD
Surgical type			0.488				0.045
Nephrectomy, partial	116 (48.1)	155 (26.2)			73 (37.6)	77 (39.7)
Nephrectomy, radical	11 (4.6)	21 (3.6)			9 (4.6)	10 (5.2)
Nephroureterectomy	12 (5.0)	34 (5.8)			12 (6.2)	10 (5.2)
Prostatectomy	102 (42.3)	381 (64.5)			100 (51.5)	97 (50.0)
Surgeon			0.572				0.137
Surgeon 1	72 (29.9)	52 (8.8)			32 (16.5)	31 (16.0)
Surgeon 2	24 (10.0)	164 (27.7)			24 (12.4)	27 (13.9)
Surgeon 3	102 (42.3)	185 (31.3)			96 (49.5)	95 (49.0)
Surgeon 4	4 (1.7)	113 (19.1)			4 (2.1)	4 (2.1)
Surgeon 5	0 (0)	5 (0.8)			0 (0)	0 (0)
Surgeon 6	38 (15.8)	65 (11.0)			37 (19.1)	33 (17.0)
Surgeon 7	1 (0.4)	7 (1.2)			1 (0.5)	4 (2.1)
Operation time (min)	125 [100, 150]	120 [100, 145]	0.043		130 [105, 150]	130 [110, 150]	0.025

Values are presented as number (%) or median [interquartile range Q1, Q3].

PSM: propensity score matching, ASD: absolute standardized difference.

Table 3.

Comparison of anesthetic data

Variable	Entire cohort				After PSM
	Sufentanil (n = 241)	Remifentanil (n = 591)	ASD		Sufentanil (n = 194)	Remifentanil (n = 194)	ASD
Main anesthetics			0.489				N/A
Volatile anesthetics	240 (99.6)	528 (89.3)			194 (100)	194 (100)
Intravenous anesthetics	1 (0.4)	63 (10.7)			0 (0)	0 (0)
Type of volatile anesthetics			0.795				0.064
Sevoflurane	165 (68.5)	208 (35.2)			118 (60.8)	124 (63.9)
Desflurane	76 (31.5)	320 (54.1)			76 (39.2)	70 (36.1)
Anesthesia time (min)	150 [130, 175]	145 [125, 170]	0.139		155 [131.25, 175]	155 [135, 175]	0.001
Crystalloid (mL)	700 [550, 800]	650 [500, 800]	0.163		700 [550, 800]	700 [500, 800]	0.065
Estimated blood loss (mL)	50 [50, 50]	50 [50, 100]	0.015		50 [50, 50]	50 [50, 50]	0.027
Postoperative variable
PCA (mg)	180 [120, 180]	180 [120, 180]	0.043		120 [120, 180]	180 [120, 180]	0.097

Values are presented as number (%) or median [interquartile range Q1, Q3].

PSM: propensity score matching, ASD: absolute standardized difference, PCA: patient-controlled analgesia, N/A: not applicable.

2. Sufentanil improved postoperative pain control without delaying recovery

The distributions of pain severity at PACU admission were significantly different between the two groups (rank-biserial correlation −0.34, 95% CI −0.43 to −0.24; P < 0.001) (Fig. 2). Consistent with this shift, more patients in the sufentanil group experienced no/mild pain (27.8% vs. 4.6%, RD 23.2%, 95% CI 16.2% to 30.2%; P < 0.001) and fewer experienced severe pain (32.5% vs. 57.2%, RD −24.7%, 95% CI −34.3% to −15.2%; P < 0.001) than those in the remifentanil group. A consistent pattern was observed in the unmatched entire cohort (rank-biserial correlation −0.29, 95% CI −0.35 to −0.22; P < 0.001). Notably, more patients experienced no/mild pain (27.0% vs. 7.4%, RD 19.5%, 95% CI 13.5% to 25.5%; P < 0.001) and fewer patients experienced severe pain (31.5% vs. 52.5%, RD −20.9%, 95% CI −28.0% to −13.8%; P < 0.001) (Fig. 2) in the sufentanil group than in the remifentanil group. The patients in the sufentanil group experienced significantly better analgesic effects than did those in the remifentanil group, with lower pain scores, longer time to first rescue analgesic administration, and lower rescue analgesic consumption; the unmatched cohort revealed concordant findings (Table 4). No significant differences were noted between the groups in terms of PONV, desaturation, and duration of PACU stay. Postoperative recovery outcomes including time to first ambulation, time to canalization, and LOS showed no significant differences between the groups (Supplementary Table 2). The superior analgesic effect of sufentanil on the primary outcome remained robust across all sensitivity analyses (Table 5). The association was unchanged when the cohort was restricted to the post-adoption period, in the interval-stratified CMH analysis (no/mild pain: odds ratio [OR] 9.15, 95% CI 3.84 to 21.83; severe pain: OR 0.39, 95% CI 0.25 to 0.60; both P < 0.001), and after adjustment for surgical month in logistic regression (no/mild pain: OR 7.83, 95% CI 3.72 to 16.50; severe pain: OR 0.36, 95% CI 0.23 to 0.54; both P < 0.001). The surgical month itself was not associated with either outcome (P = 0.808 and P = 0.747 for no/mild and severe pain, respectively), and the Tarone test indicated homogeneity of the effect across intervals (P = 0.337 and P = 0.233 for no/mild and severe pain, respectively). Remifentanil was used exclusively from March to April 2023; sufentanil was introduced in May 2023, after which both opioids were used concurrently throughout the remainder of the study period (Supplementary Table 3).

Fig. 2.

Distribution of pain severity at PACU admission in the entire (unmatched) cohort and the propensity-matched cohort. Each stacked bar represents the percentage of patients with no/mild, moderate, and severe pain in the sufentanil and remifentanil groups. In both cohorts, the sufentanil group had a significantly higher proportion of patients with no/mild pain and a lower proportion of patients with severe pain than did the remifentanil group (P < 0.001). PACU: post-anesthetic care unit.

Table 4.

Recovery profiles during PACU stay

Variable	Entire cohort					After PSM
	Sufentanil (n = 241)	Remifentanil (n = 591)	Effect size (95% CI)	P value		Sufentanil (n = 194)	Remifentanil (n = 194)	Effect size (95% CI)	P value
Pain score
Admission	6 [3, 7]	7 [6, 8]	−1 (−1 to −1)	< 0.001a		6 [3, 7]	7 [6, 8]	−1 (−2 to −1)	< 0.001a
15 minutes	6 [4, 7]	6 [6, 8]	−1 (−1 to 0)	< 0.001a		6 [3.25, 7]	6.5 [6, 8]	−1 (−1 to 0)	< 0.001a
30 minutes	4 [2, 6]	6 [3, 7]	0 (−1 to 0)	0.003a		4.5 [3, 6]	6 [3, 6.75]	0 (−1 to 0)	0.020a
Maximal score	6 [5, 8]	7 [6, 8]	−1 (−1 to −1)	< 0.001		6 [5.25, 8]	7 [6, 8]	−1 (−1 to 0)	< 0.001
Time to first rescue analgesics (min)	15 [5, 24.5]	5 [3, 10]	5 (5 to 8)	< 0.001		13 [5, 22.5]	5 [3, 10]	5 (3 to 7)	< 0.001
Rescue analgesic consumption
Opioid (MED, mg)	5.8 [5, 10]	10 [10, 12.5]	−3.3 (−5 to −2.5)	< 0.001		7.5 [5, 10]	10 [10, 12.5]	−3.3 (−5 to −2.5)	< 0.001
Non-opioid
Acetaminophen	53 (22.0)	233 (39.4)	−17.4 (−24 to −10.9)	< 0.001		45 (23.2)	76 (39.2)	−16 (−25.1 to −6.9)	0.001
Nefopam	6 (2.5)	29 (4.9)	−2.4 (−5.0 to 0.2)	0.166		6 (3.1)	8 (4.1)	−1.0 (−4.7 to 2.7)	0.785
PONV	18 (7.5)	28 (4.7)	2.7 (−1.0 to 6.5)	0.163		14 (7.2)	8 (4.1)	3.1 (−1.5 to 7.7)	0.188
SpO2 < 95%	36 (14.9)	82 (13.9)	1.1 (−4.2 to 6.4)	0.772		33 (17.0)	24 (12.4)	4.6 (−2.4 to 11.7)	0.197
Modified Aldrete score
Admission	8 [7, 8]	8 [7, 8]	0 (0 to 0)	0.023		8 [7, 8]	8 [7, 8]	0 (0 to 0)	0.896
Discharge	10 [10, 10]	10 [10, 10]	0 (0 to 0)	0.087		10 [10, 10]	10 [10, 10]	0 (0 to 0)	0.754
Duration of PACU stay (min)	40 [32, 50]	40 [30, 50.5]	0 (−2 to 3)	0.810		39 [31, 49]	42 [30, 53]	−2 (−5 to 1)	0.201

Values are presented as median [interquartile range Q1, Q3] or number (%). Effect sizes for variables presented as median [Q1, Q3] are reported as the Hodges–Lehmann estimator and its 95% CI.

PACU: post-anesthetic care unit, PSM: propensity score matching, CI: confidence interval, MED: morphine-equivalent doses, PONV: postoperative nausea and vomiting.

^aBonferroni-adjusted significance threshold: P < 0.017.

Table 5.

Sensitivity analyses for the primary outcome (pain severity at PACU admission)

Analysis	No/mild pain (sufentanil vs. remifentanil)	Severe pain (sufentanil vs. remifentanil)
Primary analysis (unadjusted)	27.8% vs. 4.6%; P < 0.001	32.5% vs. 57.2%; P < 0.001
Restricted to post-adoption period (June 2023 to March 2024)	27.7% vs. 4.3%; P < 0.001	32.5% vs. 56.0%; P < 0.001
Interval-stratified CMH analysis	OR 9.15 (3.84–21.83); P < 0.001	OR 0.39 (0.25–0.60); P < 0.001
Logistic regression adjusted for surgical month	OR 7.83 (3.72–16.50); P < 0.001	OR 0.36 (0.23–0.54); P < 0.001

Values are presented as percentages or OR with 95% CI. The OR for the stratified analysis is the pooled Mantel–Haenszel estimate, with strata defined as successive intervals from the start of the study period (March to May 2023, June to August 2023, September to November 2023, and December 2023 to March 2024). A Tarone test confirmed homogeneity of the effect across intervals (P = 0.337 and P = 0.233 for no/mild and severe pain, respectively).

PACU: post-anesthetic care unit, CMH: Cochran–Mantel–Haenszel, OR: odds ratio, CI: confidence interval.

3. Sufentanil reduced perioperative opioid consumption

Postoperative rescue opioid consumption was significantly lower in the sufentanil group than in the remifentanil group during the PACU stay (7.5 [5, 10] vs. 10 [10, 12.5] mg morphine-equivalent, HL estimator −3.3 mg, 95% CI −5 to −2.5 mg; P < 0.001). Perioperative opioid consumption was also significantly lower in the sufentanil group than in the remifentanil group, both intraoperatively (37 [30, 50] vs. 50 [31.25, 70] mg morphine-equivalent, HL estimator −10 mg, 95% CI −15 to −5 mg; P < 0.001) and cumulatively at 24 hours (10 [5, 15] vs. 12.5 [10, 20] mg morphine-equivalent, HL estimator −5 mg, 95% CI −5 to −2.5 mg; P < 0.001) and 48 hours (13.3 [5, 22.5] vs. 20 [10, 29.6] mg morphine-equivalent, HL estimator −5 mg, 95% CI −6.7 to −2.5 mg; P < 0.001) postoperatively (Table 6).

Table 6.

Perioperative opioid consumption

Variable	Entire cohort					After PSM
	Sufentanil (n = 241)	Remifentanil (n = 591)	Effect size (95% CI)	P value		Sufentanil (n = 194)	Remifentanil (n = 194)	Effect size (95% CI)	P value
Intraoperative	38 [30, 50]	50 [35, 75]	−12 (−15 to −10)	< 0.001a		37 [30, 50]	50 [31.25, 70]	−10 (−15 to −5)	< 0.001a
Postoperative
Cumulative 6 hours	10 [5, 10]	10 [10, 15]	−5 (−5 to −2.5)	< 0.001a		10 [5, 12.5]	10 [10, 17.3]	−5 (−5 to −2.5)	< 0.001a
Cumulative 12 hours	10 [5, 12.5]	10 [10, 20]	−5 (−5 to −2.5)	< 0.001a		10 [5, 13.3]	10 [10, 20]	−5 (−5 to −2.5)	< 0.001a
Cumulative 24 hours	10 [5, 15]	12.5 [10, 20]	−5 (−5 to −2.5)	< 0.001a		10 [5, 15]	12.5 [10, 20]	−5 (−5 to −2.5)	< 0.001a
Cumulative 48 hours	13.3 [5, 22.5]	16.6 [10, 30]	−5 (−5 to −2.5)	< 0.001a		13.3 [5, 22.5]	20 [10, 29.6]	−5 (−6.7 to −2.5)	< 0.001a

Values are presented as median [interquartile range Q1, Q3]. Opioid consumption is expressed in mg of morphine-equivalent dose.

PSM: propensity score matching, CI: confidence interval.

^aBonferroni-adjusted significance threshold: P < 0.01.

DISCUSSION

In this study, intraoperative sufentanil administration was associated with significantly lower postoperative pain and rescue analgesic consumption than was remifentanil administration in patients undergoing robot-assisted urologic surgery. These analgesic benefits were achieved without evidence of delayed recovery or respiratory depression.

Sufentanil administration was associated with reduced incidence of patients reporting severe pain and increased incidence of patients reporting no/mild pain after surgery. Moreover, postoperative rescue opioid consumption over 48 hr was lower with sufentanil (HL estimator −5 mg of MED, 95% CI −6.7 to −2.5 mg). A secondary analysis of perioperative opioid consumption confirmed that this was significantly lower in the sufentanil than in the remifentanil group. Furthermore, pain scores were 1 point lower in the sufentanil group than in the remifentanil group. Previous studies have established the minimal clinically important difference for postoperative pain score and analgesic consumption as 1 point [19] and 2.5–10 mg MED [20–22], respectively. Therefore, the observed reductions in both pain score and analgesic consumption suggest a clinically meaningful advantage associated with sufentanil.

The present findings align with those of previous studies comparing the postoperative analgesic effects of sufentanil and remifentanil in various minimally invasive surgeries. Sung et al. [11] reported that, among patients undergoing robot-assisted gynecologic surgery, those receiving sufentanil had lower postoperative pain scores and required fewer rescue analgesics during their PACU stay than those receiving remifentanil. Similarly, findings from two randomized controlled trials involving patients with morbid obesity undergoing laparoscopic gastric surgery demonstrated significantly lower cumulative analgesic consumption at 2 or 4 hours postoperatively in the sufentanil group than in the remifentanil group [23,24]. To the authors’ knowledge, no previous study has evaluated the postoperative analgesic efficacy of sufentanil in patients undergoing robot-assisted urologic surgery. Given that recent review articles emphasize the necessity of optimized pain management strategies for this surgical cohort [2,5], the present findings provide meaningful evidence supporting the role of sufentanil in improving postoperative pain control. Furthermore, sufentanil may serve as a viable option for opioid minimization strategies, which have been increasingly emphasized in recent literature on postoperative pain management [25].

The greater analgesic efficacy of sufentanil observed in this study can be attributed to its pharmacological properties. Its high affinity for μ-opioid receptors and prolonged context-sensitive half-life likely contribute to the sustained analgesic effects. Sufentanil is approximately 10 times more potent than remifentanil [26], and its context-sensitive half-time is approximately 30 minutes, whereas remifentanil has a significantly shorter half-life of 3–6 minutes [27].

Despite these pharmacological advantages, Casati et al. [28] observed no significant difference in postoperative pain scores between the sufentanil and remifentanil groups in patients undergoing upper abdominal surgery. However, their study protocol included epidural analgesia for all patients, which may have mitigated the differences in opioid effects between the groups.

Notably, the between-group difference in rescue analgesic consumption was not confined to the immediate postoperative period but persisted through the 48-hour period (Table 6). Because this interval extends well beyond the pharmacokinetic window of sufentanil, the sustained late-phase difference is unlikely to be attributable to residual sufentanil analgesia alone. A complementary explanation could be acute opioid tolerance and remifentanil-induced hyperalgesia following the rapid offset of remifentanil, a phenomenon linked to N-methyl-D-aspartate receptor-mediated central sensitization and enhanced descending facilitation, which can heighten nociceptive sensitivity and analgesic requirements for hours to days postoperatively [29–33]. Accordingly, the early advantage of sufentanil may reflect its sustained analgesic effect, whereas the persistence of the difference into the late postoperative phase may additionally reflect avoidance of remifentanil-associated hyperalgesia. As the present study did not directly assess hyperalgesia, this mechanism remains hypothetical and warrants further investigation.

Herein, the recovery parameters were comparable between the sufentanil and remifentanil groups. Although concerns have been raised regarding the potential of sufentanil to delay recovery because of its longer duration of action [34,35], no significant differences in recovery parameters were observed. This finding may be attributed to the earlier cessation of opioid infusion in the sufentanil group than in the remifentanil group. The timing of sufentanil discontinuation was anchored to robot de-docking, a precise and reproducible intraoperative landmark. This timepoint was chosen on the basis of the drug's pharmacokinetic profile. The discontinuation at de-docking allows the effect-site concentration to decline over approximately one context-sensitive half-time (~30 minutes) by the time of emergence and PACU transfer [27], allowing restoration of spontaneous ventilation and patient readiness for transfer to the recovery room, while preserving a clinically meaningful residual analgesic effect. Consistent with this design, the median interval from opioid discontinuation to PACU admission was 35 [30, 41] minutes in the sufentanil group versus 11 [8, 12] minutes in the remifentanil group (discontinued at skin closure). Therefore, despite being discontinued earlier, relative to PACU arrival, sufentanil provided superior analgesia without delayed recovery. These results suggest that sufentanil can provide effective postoperative pain control without compromising recovery in patients undergoing robot-assisted urologic surgery.

The incidences of PONV (7.2% vs. 4.1%) and oxygen desaturation (17.0% vs. 12.4%) were numerically higher in the sufentanil group than in the remifentanil group (Table 4), although neither reached statistical significance and the CI crossed zero. These trends may relate to the longer-acting profile of sufentanil, whose residual effect-site concentration could contribute to mild respiratory and emetic effects during the early recovery phase. Notably, this same residual concentration also underlies the preserved early analgesic effect shown by sufentanil; therefore, earlier discontinuation would be expected to attenuate both effects to some degree. Because the analgesic advantage was substantial and statistically significant whereas the adverse-event differences were small and non-significant, a modestly earlier cessation may plausibly reduce these residual effects while retaining a clinically meaningful, if slightly attenuated, analgesic benefit. This trade-off warrants prospective evaluation through future dedicated timing- and dose-finding studies.

This study represents the foundational investigation of sequential research, which aims to progressively refine postoperative pain control after robot-assisted urologic surgery. As the first step, this study isolates the effects of intraoperative opioids, with subsequent studies evaluating the addition of preventive [13] or preemptive strategies to a standardized sufentanil-based anesthesia. Because the present study preceded the introduction of those strategies, the comparison reported here reflects independent effects of opioid selection. Consistent with this, the analgesic advantage of sufentanil was robust to adjustment for the surgical period, indicating that it was not attributable to temporal practice-pattern changes.

This study has some limitations. First, this was a retrospective analysis and potential confounding variables may have influenced the findings. Although PSM may itself have introduced selection bias, the concordance between the unmatched full-cohort and analyses of matched groups argues against substantial matching-induced bias. Second, the absence of a standardized postoperative pain management protocol may have contributed to variability in analgesic consumption. Third, the study’s single-center design and inclusion of only patients undergoing robot-assisted urologic surgery may limit the generalizability of these findings to other surgical populations. Fourth, because the cohort was managed without postoperative opioid PCAs, the findings may not be generalizable to settings where opioid-based PCAs are routinely used, wherein a postoperative opioid background could attenuate or modify the observed between-group differences in rescue analgesic requirements. Fifth, the intraoperative opioid was selected by the attending anesthesiologist, and each anesthesiologist predominantly used a single opioid. Nonetheless, first-line postoperative analgesia followed routine institutional practice, including factors such as the PCA composition, the PACU rescue sequence, and discharge criteria, all of which were applied uniformly to both groups, limiting anesthesiologist-level variation beyond the choice of intraoperative opioid, although some late analgesic decisions were left to clinical discretion. To address these limitations, future prospective multicenter studies involving various surgical types and following standardized pain management protocols are warranted. Investigating the long-term outcomes and potential side effects of sufentanil use could also provide a more comprehensive understanding of its efficacy and safety.

In conclusion, in patients undergoing robot-assisted urologic surgery, intraoperative sufentanil was associated with improved postoperative analgesia compared with intraoperative remifentanil, without evidence of delayed recovery. These findings have significant clinical implications, particularly for optimizing pain management strategies and reducing opioid-related complications. Further prospective multicenter studies are warranted to validate these results and assess the broader applicability of sufentanil in other surgical settings.

SUPPLEMENTARY MATERIALS

Supplementary materials can be found via https://doi.org/10.3344/kjp.26196.