Increasing dose of esketamine for supplemental analgesia after complex spinal fusion surgery: a randomized and double-blind trial

Abstract

Background

Complex spinal surgery is associated with significant trauma and pain. Herein we tested the hypothesis that increasing the dose of esketamine, as a supplement to patient-controlled analgesia, might improve pain relief after extensive spinal surgery.

Methods

In this randomized, double-blind, and controlled trial, we enrolled adults who were scheduled to undergo extensive posterior spinal fusion. Participants were randomized in a 1:1:1 ratio to receive patient-controlled sufentanil-dexmedetomidine analgesia supplemented with esketamine at 0.25 (Esk25), 0.5 (Esk50), or 0.75 (Esk75) mg/ml. Our primary endpoint was an integrated index of pain intensity and opioid consumption (ranging from − 200% to + 200%, with higher scores indicating greater pain and/or more opioid consumption) with movement within 72 h after surgery.

Results

A total of 312 patients (median 38 years; 24.4% female) were enrolled and randomized; all were included in the intention-to-treat analysis. The primary endpoint did not differ among three groups (Esk50 vs. Esk25: mean difference − 2 [95% CI -31 to 28]; Esk75 vs. Esk25: mean difference 13 [95% CI -16 to 43]. The proportions reporting nausea/vomiting and nausea were higher in the Esk75 group than in the Esk25 group (nausea/vomiting: 70.2% [73/104] with Esk75 vs. 52.9% [55/104] with Esk25; P = 0.021; nausea: 68.3% [71/104] with Esk75 vs. 49.0% [51/104] with Esk25; P = 0.005).

Conclusions

For adult patients recovering from extensive posterior spinal fusion, increasing esketamine concentration in patient-controlled analgesia from 0.25 to 0.5 and 0.75 mg/ml did not improve pain relief or reduce opioid consumption. Esketamine at a concentration of 0.75 mg/ml increased postoperative nausea and vomiting.

Trial registration

ClinicalTrials.gov, NCT06062550; first submitted on September 25, 2023.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12871-026-03962-1.

Background

Complex spinal surgery, such as thoracolumbar spinal fusion for idiopathic scoliosis or ankylosing spondylitis, is associated with significant surgical trauma due to extensive dissection, retraction, resection, and implantation of internal fixation devices. Patients recovering from complex spinal surgery usually experience moderate-to-severe pain [1, 2] which may last from 4 to 7 days [3, 4]. Persistent severe pain increases the risk of excessive opioid consumption and related adverse events, including nausea, vomiting, pruritus, urine retention, and constipation [5–8]. Additionally, the potential development of acute opioid tolerance and opioid-induced hyperalgesia represent a critical challenge in pain management [9–11].

As recommended by guidelines [12, 13], multimodal analgesia has become a cornerstone of pain management after complex spinal surgery [14–22]. Ketamine is a non-competitive antagonist of the N-methyl-D-aspartate (NMDA) receptor and plays a key role in inhibiting central sensitization and opioid-induced hyperalgesia [23–26]. Esketamine is the S-enantiomer of racemic ketamine and is approximately twice as potent as racemate in analgesia. Moreover, esketamine produces fewer psychiatric disturbances at clinically equivalent doses compared to its racemic counterpart [27, 28]. Perioperative use of both ketamine and esketamine is reported effective in improving analgesia after spinal surgery, as manifested by attenuated pain intensity, reduced moderate-to-severe pain, delayed requirement of rescue analgesia, and reduced opioid consumption [29–33].

Esketamine has been used in combination with opioids for patient-controlled intravenous analgesia (PCIA) [34, 35]. A clinical trial of Brink and colleagues investigated the effect of increasing doses of esketamine in PCIA (0.0, 0.25, 0.5, and 0.75 mg/mL; mean 0.0, 6.6, 11.8, and 14.8 µg/kg/h, respectively) after major lumber fusion surgery and found that only the highest dose esketamine reduced opioid consumption [36]. In another clinical trial, Zhang and colleagues added higher doses of esketamine to PCIA (0.0, 0.5, 1.0, and 2 mg/kg/48 h; mean 0.0, 10.4, 20.8, and 41.7 µg/kg/h, respectively) in thoracoscopic surgery patients; they also found a dose-dependent analgesia promoting effect; however, the highest dose esketamine produced sedation [37]. The optimal dosing regimen of esketamine for PCIA after complex spinal surgery remains poorly defined.

In our recent clinical trial, combined use of mini doses of esketamine (0.25 mg/mL; mean 5.5 µg/kg/h) and dexmedetomidine (1 µg/mL; mean 0.02 µg/kg/h) with sufentanil PCIA significantly improved analgesia after scoliosis correction surgery; nonetheless, the proportion with moderate-to-severe pain remained high (65.7%), indicating that analgesia should be further improved in this patient population [2]. Current trial was designed to test the hypothesis that, on the basis PCIA formula of our last trial, increasing esketamine dose might produce better analgesia in patients following complex spinal surgery.

Methods

Study design and ethics

This was a randomized, double-blind clinical trial with three parallel arms conducted at a university tertiary hospital. The trial protocol was approved by the local Biomedical Research Ethics Committee (No.2023 − 351; approved on September 7, 2023) and registered with ClinicalTrials.gov (NCT06062550; first submitted on September 25, 2023). Written informed consent was obtained from each participant prior to enrollment in the study.

Subjects

Potential participants were screened the day before surgery. We enrolled patients aged ≥ 18 years who had a body weight ≥ 40 kg, were scheduled to undergo posterior spinal fusion for idiopathic scoliosis or ankylosing spondylitis, and required PCIA after surgery. We excluded those who met any of the exclusion criteria: preoperative sick sinus syndrome, severe sinus bradycardia (heart rate < 50 beats/min), or second-degree or higher atrioventricular block without a pacemaker; comorbid with congenital heart disease, arrhythmia, or other severe cardiovascular conditions with a New York Heart Association (NYHA) functional class of III or higher; diagnosed with or judged to be at high risk of moderate-to-severe obstructive sleep apnea according to STOP-Bang score [38]; a history of hyperthyroidism or pheochromocytoma; a history of schizophrenia, epilepsy, or myasthenia gravis; severe hepatic (Child–Pugh class C) or renal dysfunction (requiring dialysis), or an American Society of Anesthesiologists (ASA) physical status classification of IV or higher; or communication barriers.

Randomization and masking

Random numbers were generated by an independent statistician using SAS software (version 9.3; SAS Institute, Cary, NC, USA) with a block size of six and an allocation ratio of 1:1:1. The generated randomization sequences were sealed in sequentially numbered, opaque envelopes and kept by a research coordinator who was otherwise not involved in the trial. Following anesthesia induction, the research coordinator opened the random envelopes in the sequence of recruitment, prepared the study drugs according to the randomization results, and provided the study drug-containing PCIA device to the responsible anesthesiologists for postoperative analgesia. The analgesic pumps had identical appearances. All participants, investigators, anesthesiologists, and other healthcare team members were masked from study group assignments. Unmasking was prohibited unless it was required for clinical reasons or patient safety.

Anesthesia, perioperative management, and intervention

Monitoring in the operating room included electrocardiogram, pulse oxygen saturation, non-invasive and invasive blood pressure, end-tidal carbon dioxide, concentration of inhalational anesthetics, nasopharyngeal temperature, bispectral index (BIS), and urine output. Low dose dexamethasone (5 mg) was administered intravenously before anesthesia induction for prevention of postoperative nausea and vomiting. Flurbiprofen axetil (50 mg) was administered before incision unless contraindicated.

General anesthesia was induced intravenously with midazolam, propofol and/or etomidate, sufentanil, and a non-depolarizing neuromuscular blocking agent. For patients with predicted difficult airway—such as limited cervical spine extension and reduced mouth opening—awake tracheal intubation was performed using a flexible bronchoscope under adequate topical anesthesia and minimal sedation. Anesthesia was maintained via intravenous infusion of propofol and remifentanil/sufentanil, with or without sevoflurane inhalation. Additional doses of opioids and muscle relaxants were administered intraoperatively as indicated to ensure adequate analgesia and optimal muscle relaxation. An appropriate depth of anesthesia was maintained, as indicated by a BIS value within the target range between 40 and 60. Nasopharyngeal temperature was maintained between 36 and 37 °C. Fluid infusion and blood transfusion were conducted according to clinical routine. As a routine practice in the study center, regional block was not performed.

At the end of surgery, a 5-hydroxytryptamine 3 receptor antagonist was administered; patients were awakened and extubated in the operating room. PCIA was connected and initiated immediately after extubation. For all patients, the PCIA pump was established with sufentanil 4 µg/kg (≤ 250 µg), dexmedetomidine 200 µg, and normal saline to a total volume of 200 mL. According to randomization results, the analgesic pumps were supplemented with esketamine at doses of 50 mg (0.25 mg/mL; Esk25 group), 100 mg (0.5 mg/mL; Esk50 group), or 150 mg (0.75 mg/mL; Esk75 group), programmed to deliver 2-mL boluses with an 8-minute lockout interval and a background infusion at 1 mL/h, and used for at least 72 h.

Patients were then transferred to the post-anesthesia care unit and monitored with electrocardiogram, pulse oxygen saturation, and non-invasive blood pressure for at least 30 min. Pain intensity was assessed at 10-minute intervals using a Numeric Rating Scale (NRS; an 11-point scale where 0 = no pain and 10 = the worst pain). In case an NRS pain score was ≥ 4, the PCIA bolus was pressed or a supplemental dose sufentanil was administered. The target was to maintain NRS pain score of ≤ 3. Patients were taught to use the PCIA pump and discharged to orthopedic wards when modified Aldrete score reached ≥ 9.

After returning to general wards, patients were monitored continuously with electrocardiogram and pulse oxygen saturation and intermittently with non-invasive blood pressure for 24 h; they were then monitored twice daily until 72 h after surgery (i.e., end of trial PCIA). Apart from PCIA, standard analgesic regimen included intravenous flurbiprofen axetil (100 mg; twice daily) and oral acetaminophen/oxycodone (oxycodone hydrochloride 5 mg and acetaminophen 325 mg; three times daily). Rescue analgesia indicated any analgesics in addition to the standard regimen described above, including oral analgesics (tramadol and/or other opioids) available on an as-needed basis and intramuscular/intravenous morphine (5–10 mg) for breakthrough pain. Non-trial esketamine and dexmedetomidine were not allowed.

Data collection and measurements

Baseline data were collected and included demographics, surgical diagnosis, comorbidity, history of smoking and drinking, current use of analgesics, and important laboratory test results. Pain intensity was evaluated with the NRS. Subjective sleep quality during the preoperative night was also evaluated with the NRS (an 11-point scale where 0 = the best sleep and 10 = the worst sleep). Subjective sleep quality over the last month was evaluated with the Pittsburgh Sleep Quality Index (PSQI; scores range from 0 to 21, with higher scores indicating worse sleep quality) [39–41]. Depressive symptoms over the last two weeks were assessed using the Patient Health Questionnaire-9 (PHQ-9; scores range from 0 to 27, with higher scores indicating more severe depressive symptoms) [42–44]. Quality of life was evaluated using the Scoliosis Research Society-22 questionnaire (SRS-22; a 22-item questionnaire; each item scores from 1 to 5, higher score better; total score represents a mean value) [45–47].

Intraoperative data were recorded and included duration of anesthesia, types and dosage of medications during anesthesia, fluid infusion and blood transfusion, estimated blood loss and urine output, and duration and extent of surgical procedures.

After surgery, pain intensity both at rest and with movement was assessed with the NRS at 2 h and then twice daily (8–10 am and 6–8 pm) until 72 h; a difference ≥ 1 point was considered clinically meaningful [48]. Analgesics administered via PCIA and as supplemental medications were documented. Subjective sleep quality was assessed daily (8–10 am) with the NRS over the first 5 days; a difference ≥ 1 point was considered clinically important [49]. Delirium was assessed twice daily (8–10 am and 6–8 pm) with the Three-Dimensional Confusion Assessment Method (3D-CAM) during the first five days [50]. Prior to delirium assessment, level of sedation/agitation was evaluated with the Richmond Agitation-Sedation Scale (RASS; a 10-point scale ranging from − 5 [unarousable] to + 4 [combative], with 0 representing a calm and alert state) [51]. Depressive symptoms were assessed again with the PHQ-9 on postoperative day 7 or before hospital discharge.

Our primary endpoint was an integrated index of Pain Intensity (with movement) and Opioid Consumption (PIOC) within 72 h after surgery [52, 53]. Pain intensity as assessed using the NRS was first summarized as the area under the curve (AUC) over time using the trapezoidal method [52], where t indicates time point and y NRS pain score:

The PIOC is then calculated as the sum of normalized rank-based deviations of AUC and total opioid consumption (OC), using the following formula, where and refer to the ranks of each patient across the entire cohort:

This yields a dimensionless value ranging from − 200% to + 200%. A positive PIOC value indicates that an individual patient has experienced greater pain intensity and/or consumed more opioids compared to the cohort average, whereas a negative value suggests better-than-average pain control with lower opioid requirements [52, 53].

Secondary endpoints included the following: proportion of patients experiencing moderate-to-severe pain (NRS ≥ 4 at any time) within 72 h, the highest NRS pain score within 72 h, AUC of NRS pain scores within 72 h, cumulative opioid consumption within 72 h, proportion requiring rescue analgesia within 72 h, PIOC at rest within 72 h, subjective sleep quality within five days, and duration of analgesia within 30 days after surgery. Cumulative opioid consumption was converted into intravenous morphine milligram equivalents [54].

Other pre-specified endpoints included RASS score at each timepoint, delirium within 5 days, depression score at 7 days, length of hospital stay, and major complications within 30 days after surgery. Major complications were generally defined as new-onset medical conditions that were deemed harmful and required clinical intervention, i.e., grade 2 or higher on the Clavien-Dindo classification [55].

Adverse events were monitored from initiation to end of trial PCIA use for up to 5 days. Potential adverse events included bradycardia (heart rate < 45 beats/min or a decrease > 30% from baseline), tachycardia (heart rate > 100 beats/min or an increase > 30% from baseline), hypotension (systolic blood pressure < 90 mmHg or a decrease > 30% from baseline), hypertension (systolic blood pressure > 180 mmHg or an increase > 30% from baseline), desaturation (pulse oxygen saturation < 90% in room air), respiratory depression (respiratory rate < 10 breaths/minute), and excessive sedation (RASS score ≤-3). We also monitored other adverse events including pruritus, nausea, and vomiting, as well as neuropsychiatric symptoms such as dizziness, hallucinations, nightmares, and myoclonus.

Statistical analysis

Sample size estimation

In our previous trial, PIOC with movement was reduced by 46.0% (95% CI 20.5% to 72.5%) following the use of esketamine-dexmedetomidine supplemented sufentanil PCIA [2]. We anticipated that PIOC with movement would exhibit a stepwise decrease of 30% across groups with increasing doses of esketamine. With a two-sided alpha level set at 0.05, a statistical power at 80%, and considering a dropout rate of about 10%, we planned to enroll 312 participants. No interim analysis was planned.

Missing data

For baseline and intraoperative variables, missing data were not replaced. Follow-ups and assessments were completed in all subjects, regardless of whether they adhered to the assigned treatment protocol. For the NRS scores of pain intensity and subjective sleep quality and the PHQ-9 score of depressive symptoms, missing data (due to stroke and coma in one patient) were replaced with median scores at that timepoints of all patients in the same group. For delirium assessment, the last assessment result was adopted as the missing data (due to stroke and coma in one patient).

Endpoint analysis

Analysis was primarily performed in the intention-to-treat (ITT) population, which included all enrolled subjects. Primary endpoint was also analyzed in the per-protocol (PP) population, after excluding patients who received PCIA for less than 72 h.

For baseline and perioperative data, categorical variables within the three arms were compared using χ² tests, χ² tests with continuity correction, or Fisher’s exact tests as appropriate; continuous variables were analyzed using either analysis of variance (ANOVA) or the Kruskal-Wallis test, depending on whether the data met the assumptions of normality and homogeneity of variances.

The primary endpoint, PIOCs with movement within 72 h, were compared with ANOVA; between-group differences were expressed as mean difference and 95% CI. Among secondary and other endpoints, categorical endpoints (proportions of patients with moderate-to-severe pain and received rescue analgesia within 72 h, and incidences of postoperative delirium, major complications, and all-cause 30-day mortality) were compared using χ² tests or Fisher’s exact tests; relative risks (RRs) and 95% CIs were provided. Numeric endpoints (highest NRS pain score, AUC of pain, morphine equivalent, and PIOC-rest within 72 h, depression score at 7 days, and scores of NRS of pain, sleep quality, and RASS) were compared using the Kruskal-Wallis test; differences between two medians and 95% CIs were calculated using the Hodges-Lehmann estimators. Time-to-event endpoints (days requiring analgesics within 30 days and length of hospital stay after surgery) were analyzed using Kaplan–Meier survival curves and log-rank tests; univariable Cox proportional hazards models were used to calculate hazard ratios (HRs) and 95% CIs.

For each hypothesis, a two-sided P value < 0.05 was considered statistically significant. Bonferroni correction was applied to all post‑hoc pairwise comparisons. All statistical analyses were performed using SPSS version 25.0 (IBM SPSS, Armonk, NY, USA) and RStudio version 4.5.1 (RStudio, Boston, MA, USA).

Results

Patients

Between October 24, 2023, and September 23, 2025, 357 patients were screened and 318 met the eligibility criteria. Among eligible patients, 312 gave consent and were randomly assigned into one of three groups (Esk25 group, n = 104; Esk50 group, n = 104; Esk75 group, n = 104); all were included in the ITT analysis. During the study period, 22 patients received PCIA for less than 72 h (5 patients in the Esk25 group, 8 in the Esk50 group, and 9 in the Esk75 group, respectively). Therefore, 290 patients were included in the PP analysis (Fig. 1). Perioperative follow-up efforts ended on October 23, 2025.

Fig. 1.

Flowchart of the study. Esk25, esketamine 0.25 mg/ml; Esk50, esketamine 0.50 mg/ml; Esk75, esketamine 0.75 mg/ml; PCIA, patient-controlled intravenous analgesia; ITT, intention-to-treat; PP, per-protocol

Of the included patients, median age was 38 (interquartile range [IQR] 34 to 45) years, 76 (24.4%) were female, and 101 (32.4%) required analgesics (all were NSAIDs) before surgery. Baseline data were well balanced among the three groups, except that the proportion with diabetes mellitus was higher in the Esk25 group (Table 1). Intraoperative variables including uses of opioids, NSAIDs, glucocorticoids, and antiemetics, as well as the number of involved vertebrae and the grade of osteotomy were comparable among the three groups (Table 2).

Table 1.

Baseline data

Age (year)	Esk25 (n = 104)	Esk50 (n = 104)	Esk75 (n = 104)	P value
	38 (33, 44)	39 (34, 45)	38 (34, 46)	0.585
Sex (female)	21 (20.2%)	26 (25.0%)	29 (27.9%)	0.426
Height (cm)	162 (153, 170)	160 (150, 168)	160 (150, 167)	0.173
Weight (kg)	65 (55, 75)	65 (55, 73)	64.5 (54, 71)	0.964
Body mass index (kg/m2)	25 (22, 29)	26 (22, 31)	25 (23, 30)	0.384
Education				0.569
Primary school	12 (11.5%)	13 (12.5%)	14 (13.5%)
Middle school	38 (36.5%)	44 (42.3%)	31 (29.8%)
High/special school	30 (28.9%)	31 (29.8%)	40 (38.5%)
College or above	24 (23.1%)	16 (15.4%)	19 (18.3%)
Comorbidity
Hypertension	12 (11.5%)	17 (16.4%)	12 (11.5%)	0.496
Diabetes	6 (5.8%)	1 (1.0%)	1 (1.0%)	0.040
Pulmonary embolism	0 (0.0%)	0 (0.0%)	1 (1.0%)	> 0.999
Asthma	2 (1.9%)	2 (1.9%)	1 (1.0%)	0.816
Gastroesophageal reflux	4 (3.9%)	6 (5.8%)	2 (1.9%)	0.353
Anemia	8 (7.7%)	2 (1.9%)	4 (3.9%)	0.123
Previous surgery	25 (24.0%)	24 (23.1%)	23 (22.1%)	0.947
Smoking index (pack⋅year) *	0 (0, 10)	1 (0, 10)	0 (0, 15)	0.201
Chronic smoking *	8 (7.7%)	9 (8.7%)	15 (14.4%)	0.254
Alcohol consumption (g/day) †	0 (0, 0)	0 (0, 0)	0 (0, 0)	0.174
Chronic drinking †	5 (4.8%)	6 (5.8%)	6 (5.8%)	0.940
ASA classification				0.892
I	35 (33.7%)	34 (32.7%)	38 (36.5%)
II	69 (66.4%)	69 (66.4%)	64 (61.5%)
III	0 (0.0%)	1 (1.0%)	2 (1.9%)
Surgical diagnosis				0.828
Idiopathic scoliosis	14 (13.5%)	13 (12.5%)	16 (15.4%)
Ankylosing spondylitis	90 (86.5%)	91(87.5%)	88 (84.6%)
Analgesics within 30 days
Opioids	0 (0.0%)	0 (0.0%)	0 (0.0%)	> 0.999
Non-opioid analgesics ‡	36 (34.6%)	31 (29.8%)	34 (32.7%)	0.757
Assessments (point)
NRS of pain-rest §	1 (0, 2)	1 (0, 2)	1 (0, 2)	0.506
NRS of pain-movement §	3 (1, 5)	3 (2, 5)	3 (1, 5)	0.653
NRS of sleep quality ΙΙ	4 (2, 6)	5 (2, 6)	4 (2, 6)	0.977
Pittsburgh sleep quality ¶	7 (4, 10)	7 (4, 9)	6 (4, 9)	0.894
PHQ-9 depressive score **	6 (3, 9)	6 (2, 12)	7 (3, 10)	0.715
SRS-22 quality of life ††	3.0 (2.7, 3.3)	3.0 (2.6, 3.3)	3.0 (2.7, 3.4)	0.928
Laboratory tests
Hemoglobin (g/dL)	14(13, 15)	14 (13, 15)	14 (13, 15)	0.610
Albumin (g/L)	43 (41, 45)	43 (41, 46)	43 (41, 45)	0.439
Creatinine (µmol/L)	73 (64, 84)	71 (62, 82) [1]	71 (62, 81)	0.570

Data are median (interquartile range) or n (%). Numbers in square brackets indicate patients with missing data. Esk25, esketamine 0.25 mg/ml; Esk50, esketamine 0.50 mg/ml; Esk75, esketamine 0.75 mg/ml; ASA, American Society of Anesthesiologists; NRS, numeric rating scale; PHQ-9, patient health questionnaire-9; SRS-22, Scoliosis Research Society-22 Questionnaire

P value in bold indicates < 0.05

^*Calculated as number of packs smoked per day (20 cigarettes/pack) × years of smoking. Chronic smoking was defined as a smoking index > 20 pack⋅years

^† Calculated as volume of daily drinking (mL/d) × alcohol by volume (%) × 0.8 (g/mL) ÷ 100. Chronic drinking was defined as daily alcohol consumption of ≥40 g/d in male or ≥20 g/d in female

^‡ Including diclofenac, loxoprofen, celecoxib, meloxicam, indomethacin, or ibuprofen

^§ An 11-point rating scale, where 0 = no pain and 10 = the worst imaginable pain

^ΙΙ An 11-point rating scale, where 0 = the best sleep and 10 = the worst sleep. Assessed on the morning of surgery

^¶ Assessed with the Pittsburgh sleep quality index; scores range from 0 to 21, with higher scores indicating worse sleep quality

^** Assessed with the PHQ-9; scores range from 0 to 27, with higher scores indicating more severe depressive symptoms

^†† Assessed with the SRS-22; scores range from 1 to 5, with higher scores indicating better quality of life

Table 2.

Perioperative data

Intraoperative data	Esk25 (n = 104)	Esk50 (n = 104)	Esk75 (n = 104)	P value
Duration of anesthesia (min)	398 (366, 438)	396 (368, 440)	382 (359, 426)	0.332
Use of sufentanil	101 (97.1%)	101 (97.1%)	99 (95.2%)	0.686
Sufentanil (µg)	45 (37, 50)	40 (35, 50)	40 (35, 50)	0.265
Use of remifentanil	101 (97.1%)	101 (97.1%)	101 (97.1%)	> 0.999
Remifentanil (mg)	2.7 (2.3, 3.1)	2.7 (2.2, 3.1)	2.6 (2.2, 2.9)	0.527
Use of other opioids *	8 (7.7%)	7 (6.7%)	5 (4.8%)	0.688
Morphine equivalent (mg) †	309 (280, 360)	305 (260, 354)	300 (266, 330)	0.362
NSAIDs
Use of flurbiprofen axetil	96 (92.3%)	97 (93.3%)	100 (96.2%)	0.483
Flurbiprofen axetil (mg)	50 (50, 50)	50 (50, 50)	50 (50, 50)	0.864
Glucocorticoids
Use of dexamethasone	100 (96.2%)	97 (93.3%)	95 (91.3%)	0.362
Dexamethasone (mg)	5 (5, 5)	5 (5, 5)	5 (5, 5)	0.563
Other glucocorticoids ‡	15 (14.4%)	8 (7.7%)	18 (17.3%)	0.109
Antiemetics
Use of ondansetron	100 (96.2%)	101 (97.1%)	101 (97.1%)	0.902
Fluid balance
Crystalloids (L)	2.6 (2.1, 3.2)	2.7 (2.1, 3.2)	2.5 (2.1, 3.1)	0.521
Artificial colloid (L)	0.5 (0.5, 1.0)	0.5 (0.5, 1.0)	0.5 (0.5, 1.0)	0.678
Allogeneic RBC (Unit) §	1 (0, 2)	2 (0, 2)	1 (0, 2)	0.665
Allogeneic plasma (Unit) §	0 (0, 2)	0 (0, 2)	0 (0, 2)	0.850
Estimated bleeding (100 mL)	8 (5, 10)	8 (5, 10)	8 (5, 10)	0.387
Urine output (L)	0.6 (0.4, 0.9)	0.7 (0.4, 0.9)	0.6 (0.4, 0.9)	0.490
Data of surgery
Duration of surgery (min)	299 (272, 331)	299 (270, 334)	288 (271, 309)	0.145
Involved vertebrae (n)	7 (7, 7)	7 (7, 7)	7 (7, 7)	0.436
Grade of osteotomy ΙΙ				0.570
No	2 (1.9%)	3 (2.9%)	1 (1.0%)
1	10 (9.6%)	11 (10.6%)	16 (15.4%)
2	2 (1.9%)	0 (0.0%)	1 (1.0%)
3	90 (86.5%)	90 (86.5%)	86 (82.7%)
Grade of osteotomy ≥ 3	90 (86.5%)	90 (86.5%)	86 (82.7%)	0.665
Postoperative data
Study drugs via PCIA
Total duration (h)	118 (90, 124)	117 (90, 124)	117 (89, 124)	0.845
Total volume (mL)	144 (109, 200)	145 (123, 200)	144 (115, 200)	0.882
Total doses via PCIA
Sufentanil (µg)	168 (123, 220)	173 (133, 225)	166 (117, 232)	0.906
Dexmedetomidine (µg)	144 (109, 200)	145 (123, 200)	144 (115, 200)	0.882
Esketamine (mg) ¶	36 (27, 50)	73 (62, 100)	108 (86, 150)	< 0.001
PCIA volume within 72 h (mL)	89 (74, 131)	89 (78, 139)	90 (74, 130)	0.873
PCIA mean rates within 72 h
Sufentanil (ng/kg/h)	20 (20, 30)	20 (20, 30)	20 (20, 30)	0.669
Dexmedetomidine (ng/kg/h)	20 (10, 30)	20 (20, 30)	20 (10, 30)	0.862
Esketamine (µg/kg/h) ¶	4.6 (3.6, 7.4)	9.8 (7.6, 13.8)	15.4 (11.0, 20.6)	< 0.001
Other opioids within 72 h	99 (95.2%)	99 (95.2%)	100 (96.2%)	0.928
Use of oxycodone	95 (91.3%)	98 (94.2%)	99 (95.2%)	0.499
Oxycodone (mg)	45 (45, 45)	45 (45, 45)	45 (45, 45)	0.175
Use of tramadol	59 (56.7%)	53 (51.0%)	54 (51.9%)	0.671
Tramadol (mg)	200 (0, 300)	100 (0, 300)	100 (0, 300)	0.735
Use of other opioids	5 (4.8%)	0 (0.0%)	2 (1.9%)	0.074
NSAIDs within 72 h	100 (100.0%)	100 (100.0%)	100 (100.0%)	> 0.999
Use of flubiprofen axetil	100 (100.0%)	100 (100.0%)	100 (100.0%)	> 0.999
Flubiprofen axetil (mg)	700 (700, 700)	700 (700, 700)	700 (700, 700)	0.445
Use of other NSAIDs	20 (19.2%)	17 (16.3%)	17 (16.3%)	0.817
Acetaminophen within 72 h	95 (91.3%)	98 (94.2%)	99 (95.2%)	0.499
Acetaminophen (g)	2.9 (2.9, 2.9)	2.9 (2.9, 2.9)	2.9 (2.9, 2.9)	0.175
Use of glucocorticoids	31 (29.8%)	24 (23.1%)	30 (28.8%)	0.499
Use of dexamethasone **	4 (3.8%)	3 (2.9%)	5 (4.8%)	0.771
Use of methylprednisolone ††	27 (26.0%)	21 (20.2%)	26 (25.0%)	0.577

Data are median (interquartile range) or n (%). Esk25, esketamine 0.25 mg/ml; Esk50, esketamine 0.50 mg/ml; Esk75, esketamine 0.75 mg/ml; NSAIDs, non-steroidal anti-inflammatory drugs; RBC, red blood cells; PCIA, patient-controlled intravenous analgesia

P value in bold indicates < 0.05

^* Including fentanyl, morphine, oxycodone, tramadol, and oliceridine

^† Converted to intravenous morphine equivalent: morphine (iv) 1 mg = fentanyl (iv) 10 µg = remifentanil (iv) 10 µg = sufentanil (iv) 1 µg = tramadol (iv) 10 mg = oxycodone (iv) 1 mg = oliceridine (iv) 3 mg

^‡ Including methylprednisolone (40 mg) and hydrocortisone (100 mg)

^§ 1 Unit RBC/plasma = 200 mL

^ΙΙ Grade 1: partial facetectomy; grade 2: pedicle removal; and grade 3: partial vertebral body removal

^¶ Esk50 vs. Esk25, P < 0.001; Esk75 vs. Esk25, P < 0.001; Esk75 vs. Esk50, P < 0.001 (P ≤ 0.017 was considered statistically after Bonferroni correction)

^** Dexamethasone 5 mg iv injection, for prophylaxis/treatment of nausea and vomiting

^†† Methylprednisolone 40 mg iv injection, for prophylaxis/treatment of nausea/vomiting or neuroedema

Intervention

During the postoperative period, PCIA was used for a median duration of 117 (IQR 90 to 124) hours. The duration and volume of study drugs, the administered doses of sufentanil and dexmedetomidine, and the mean infusion rates of sufentanil and dexmedetomidine within 72 h administered via PCIA did not differ among the three groups. For esketamine, the total dose and the mean infusion rate within 72 h increased from (median) 36 mg and 4.6 µg/kg/h in the Esk25 group to 73 mg and 9.8 µg/kg/h in the Esk50 and 108 mg and 15.4 µg/kg/h in the Esk75 groups, respectively, in accordance with group assignments (Table 2; Supplement Table S1).

Efficacy outcomes

The primary endpoint, PIOC index with movement within 72 h, were − 4±88% with Esk25, -6±88% with Esk50, and 9±91% with Esk75, respectively (P = 0.410); there were no significant differences between Esk50 and Esk25 (mean difference − 2; 95% CI -31 to 28) and between Esk75 and Esk25 groups (mean difference 13; 95% CI -16 to 43). Analyses in the PP population gave similar results.

Among secondary endpoints, the proportion of patients receiving rescue analgesics was slightly lower in the Esk50 group (relative risk 0.70; 95% CI 0.42 to 1.18) but slightly higher in the Esk75 group (relative risk 1.22; 95% CI 0.80 to 1.88) than in the Esk25 group; the differences were not statistically different (Table 3). In postoperative day 1 morning, the NRS scores of pain intensity both at rest and with movement were slightly lower in the Esk50 group than in the Esk25 group; in the fifth postoperative night, the NRS score of subjective sleep quality was slightly higher (worse) in the Esk75 group than in the Esk25 group. However, the above differences were not statistically significant (Fig. 2; Supplement Table S2). Other secondary and pre-specified endpoints did not differ among the three groups (Table 3; Supplement Table S3).

Table 3.

Efficacy endpoints

Primary endpoint	Esk25 (n = 104)	Esk50 (n = 104)	Esk75 (n = 104)	P value	Estimated effects (95% CI) *
					Esk50 vs. Esk25	Esk75 vs. Esk25
PIOC-move within 72 h (%; ITT analysis)	-4±88	-6±88	9±91	0.410	Mean D=-2 (-31, 28)	Mean D = 13 (-16, 43)
PIOC-move within 72 h (%; PP analysis)	-4±85	1±75	3±82	0.265	Mean D = 3 (-22, 27)	Mean D = 14 (-10, 39)
Secondary endpoints
Moderate-to-severe pain within 72 h †	89 (85.6%)	94 (90.4%)	92 (88.5%)	0.558	RR = 1.06 (0.96, 1.17)	RR = 1.03 (0.93, 1.15)
Highest NRS of pain within 72 h (point)	8 (6, 9)	8 (6, 9)	8 (7, 10)	0.307	MD = 0 (0, 1)	MD = 0 (0, 1)
AUC of pain-rest within 72 h (point⋅h)	236 (105, 326)	208 (118, 311)	274 (139, 354)	0.195	MD=-6 (-46, 33)	MD = 28 (-11, 70)
AUC of pain-move within 72 h (point⋅h)	321 (213, 408)	307 (192, 398)	347 (212, 443)	0.239	MD=-10 (-51, 34)	MD = 29 (-15, 72)
Morphine equivalent within 72 h (mg) ‡	435 (349, 534)	431 (373, 598)	425 (328, 595)	0.923	MD = 5 (-11, 50)	MD=-2 (-49, 48)
Received rescue analgesia §	27 (26.0%)	19 (18.3%)	33 (31.7%)	0.081	RR = 0.70 (0.42, 1.18)	RR = 1.22 (0.80, 1.88)
PIOC-rest within 72 h (%)	-5±90	-6±86	10±92	0.353	Mean D=-1 (-30, 29)	Mean D = 15 (-15, 45)
Requiring analgesics within 30 days (day) ΙΙ	30 (20, 30)	27 (19, 30)	30 (23, 30)	0.249	MD = 0 (0, 0)	MD = 0 (0, 0)
Other pre-specified endpoints
Postoperative delirium within 5 days	0 (0.0%)	0 (0.0%)	2 (1.9%)	0.134	---	---
PHQ-9 depression score at day 7 (point) ¶	8 (4, 11)	8 (4, 12)	8 (5, 10)	0.821	MD = 1 (-1, 2)	MD = 0 (-1, 2)
Hospital stay after surgery (day)	12 (10, 13)	11 (10, 12)	12 (10, 13)	0.726	HR = 1.09 (0.82, 1.45)	HR = 1.00 (0.76, 1.33)
Postoperative complications within 30 days	10 (9.6%)	14 (13.5%)	15 (14.4%)	0.540	RR = 1.40 (0.65, 3.18)	RR = 1.50 (0.71, 3.18)
All-cause 30-day mortality	0 (0.0%)	0 (0.0%)	0 (0.0%)	> 0.999	---	---

Data are mean value±standard difference, median (interquartile range), or n (%). Esk25, esketamine 0.25 mg/ml; Esk50, esketamine 0.50 mg/ml; Esk75, esketamine 0.75 mg/ml; PIOC, pain intensity and opioid consumption; CI, confidence interval; ITT, intention-to-treat; PP, per-protocol; Mean D, mean difference; RR, relative risk; NRS, numeric rating scale; MD, median difference; AUC, area under curve; HR, hazard ratio

^* Calculated as Esk50 vs. or minus Esk25, or Esk75 vs. or minus Esk25

^† Defined as NRS pain score ≥ 4 at any timepoint within 72 h after surgery

^‡ Converted to intravenous morphine equivalent: morphine (oral) 3 mg = morphine (iv) 1 mg = fentanyl (iv) 10 µg = remifentanil (iv) 10 µg = sufentanil (iv) 1 µg = tramadol (iv) 10 mg = tramadol (oral) 20 mg = oxycodone (oral) 1.5 mg = oxycodone (iv) 1 mg = oliceridine (iv) 3 mg

^§ Defined as requirement of any analgesics in addition to PCIA, intravenous flurbiprofen axetil (100 mg; twice daily), and oral acetaminophen/oxycodone (oxycodone hydrochloride 5 mg and acetaminophen 325 mg; three times daily)

^ΙΙ Cumulative number of days requiring any analgesics within 30 days postoperatively

^¶ Assessed with the PHQ-9; scores range from 0 to 27, with higher scores indicating more severe depressive symptoms

Fig. 2.

Boxplots for the NRS pain score at rest (A) and with movement (B), the NRS of subjective sleep quality (C), and the Richmond Agitation-Sedation Scale (D) among the three groups. The box and whiskers plots show medians, interquartile ranges and outer ranges; individual points indicate mild outliers (○, outside 1.5 times of interquartile range) and extreme outliers (∆, outside 3 times of interquartile range). Esk25, esketamine 0.25 mg/ml; Esk50, esketamine 0.50 mg/ml; Esk75, esketamine 0.75 mg/ml; NRS, numeric rating scale (an 11-point scale where 0 indicates no pain or the best sleep and 10 indicates the worst pain or the worst sleep); RASS, Richmond Agitation-Sedation Scale (a 10-point scale ranging from − 5 [unarousable] to + 4 [combative], with 0 representing a calm and alert state). Also see Supplement Table S2 for details

During the intervention period, the proportions of patients reporting nausea/vomiting and nausea were higher in the Esk75 group than in the Esk25 group (nausea/vomiting: 70.2% [73/104] with Esk75 vs. 52.9% [55/104] with Esk25; P = 0.021; nausea: 68.3% [71/104] with Esk75 vs. 49.0% [51/104] with Esk25; P = 0.005). The proportion of patients requiring 5-HT3 antagonists was slightly lower in the Esk50 group than in the Esk25 group, although not statistically significant. Other adverse events including neuropsychiatric symptoms did not differ among the three groups (Table 4).

Table 4.

Adverse events

	Esk25 (n = 104)	Esk50 (n = 104)	Esk75 (n = 104)	P value
	0 (0.0%)	0 (0.0%)	0 (0.0%)	> 0.999
Bradycardia *	0 (0.0%)	0 (0.0%)	0 (0.0%)	>0.999
Tachycardia †	32 (30.8%)	36 (34.6%)	38 (36.5%)	0.670
Hypotension ‡	6 (5.8%)	3 (2.9%)	7 (6.7%)	0.425
Hypertension §	0 (0.0%)	0 (0.0%)	0 (0.0%)	> 0.999
Desaturation ΙΙ	5 (4.8%)	5 (4.8%)	8 (7.7%)	0.588
Respiratory depression ¶	0 (0.0%)	0 (0.0%)	0 (0.0%)	> 0.999
Excessive sedation **	0 (0.0%)	0 (0.0%)	0 (0.0%)	> 0.999
Pruritus ††	3 (2.9%)	0 (0.0%)	1 (1.0%)	0.170
Nausea and vomiting	55 (52.9%)	60 (57.7%)	73 (70.2%) ‡‡	0.031
Nausea §§	51 (49.0%)	59 (56.7%)	71 (68.3%) ΙΙΙΙ	0.018
Vomiting ¶¶	22 (21.2%)	17 (16.3%)	27 (26.0%)	0.237
Antiemetics within 72 h	60 (57.7%)	45 (43.3%)	59 (56.7%)	0.069
Use of 5-HT3 antagonist ***	58 (55.8%)	40 (38.5%)	56 (53.8%)	0.024
Use of metoclopramide	11 (10.6%)	11 (10.6%)	17 (16.3%)	0.348
Neuropsychiatric symptoms	40 (38.5%)	37 (35.6%)	36 (34.6)	0.835
Dizziness †††	16 (15.5%)	15 (14.4%)	14 (13.5%)	0.914
Hallucination ‡‡‡	0 (0.0%)	0 (0.0%)	1 (1.0%)	> 0.999
Day-/nightmare §§§	9 (8.7%)	8 (7.7%)	10 (9.6%)	0.885
Hypnic myoclonia ΙΙΙΙΙΙ	24 (23.1%)	23 (22.1%)	20 (19.2%)	0.781

Data are presented as n (%). Esk25, esketamine 0.25 mg/ml; Esk50, esketamine 0.50 mg/ml; Esk75, esketamine 0.75 mg/ml; 5-HT3, 5-hydroxytryptamine 3 receptor

P values in bold indicates < 0.05

^* Heart rate < 45 beats/min or a reduction > 30% from baseline

^† Heart rate > 100 beats/min or an increase > 30% from baseline

^‡ Systolic blood pressure < 90 mmHg or a reduction > 30% from baseline

^§ Systolic blood pressure > 180 mmHg or an increase > 30% from baseline

^ΙΙ Pulse oxygen saturation < 90% in room air

^¶ Respiratory rate < 10 breaths/min

^** RASS score ≤-3

^†† Presence of localized or generalized pruritus without obvious rash or other allergic manifestations

^‡‡ Esk75 vs. Esk25, P = 0.021 (P < 0.025 were considered statistically significant after Bonferroni correction)

^§§ Self-reported nausea, with or without vomiting

^ΙΙΙΙ Esk75 vs. Esk25, P = 0.005 (P < 0.025 were considered statistically significant after Bonferroni correction)

^¶¶ Occurrence of ≥ 1 episode of emesis (expulsion of gastric contents via the mouth)

^*** Including ondansetron and tropisetron. Esk50 vs. Esk25, P = 0.025 (P values < 0.025 were considered statistically significant after Bonferroni correction)

^††† Self-reported sensation of mental fogginess or lightheadedness without vertigo, excluding Ménière’s disease, benign paroxysmal positional vertigo, or other vestibular pathologies

^‡‡‡ Self-reported perception of non-existent sensory stimuli (e.g., visual, auditory, tactile), in the absence of fluctuating consciousness, disorganized thinking, or impaired orientation

^§§§ Self-reported significant increase in dream frequency during daytime/nighttime sleep compared to preoperative state, accompanied by distressing emotions or impaired sleep quality

^ΙΙΙΙΙΙ Self-reported recurrent brief limb jerks occurring at the onset of sleep, with or without a sensation of falling, leading to difficulty initiating sleep or awakening

Discussion

Subanesthetic ketamine (iv bolus up to 0.35 mg/kg; iv infusion up to 1 mg/kg/h) is recommended for perioperative analgesia [56]. Substantial evidence confirmed that subanesthetic ketamine or esketamine improves analgesia following spinal surgery [29–33]. To decrease potential psychotomimetic side effects [57, 58], mini-dose esketamine is used in combination with opioids [34] and provided supplemental analgesic effect at a mean infusion rate ranging from 4 to 40 µg/kg/h [34, 35]. Nevertheless, the optimal dose of esketamine in PCIA for spinal surgery is still unclear [36, 37].

In the present study, the dosing range of esketamine administered via PCIA (at concentrations of 0.25, 0.5, and 0.75 mg/ml; corresponding to mean infusion rate of 4.6, 9.8, and 15.4 µg/kg/h, respectively) were based on our prior work (concentration 0.25 mg/ml; mean infusion rate 5.5 µg/kg/h) [2] and well within the mini-dose range reported in the literature [34–37]. Our results showed that, for patients recovering from complex spinal fusion surgery, increasing the dose of esketamine in PCIA from 0.25 to 0.5 and 0.75 mg/ml failed to improve analgesia. Furthermore, the incidence of postoperative nausea and vomiting was increased at 0.75 mg/ml of esketamine, independent from the dose-response relationship for analgesic efficacy. This dissociation underscores a critical therapeutic trade-off between potential analgesic benefits and adverse effects. Our results are different from meta-analyses but are in line with some of the previous studies in spinal surgery [29–33, 36].

Several factors may explain why our results were “neutral”. First, the absolute dose of esketamine is low in our study. Most of the previous studies in spinal surgery administered subanesthetic doses (usually including a loading dose and a maintenance dose) during anesthesia; some of these studies continued ketamine or esketamine infusion after surgery [29–33]. The analgesic effect of ketamine/esketamine is largely dose dependent, although follows a non-linear pattern in spinal surgery settings [59]. For example, in a study of 8 healthy volunteers subjected to noxious thermal stimuli, infusion of subanesthetic dose ketamine (0.71 mg/kg/h) reduced pain score, whereas infusion of subanalgesic dose ketamine (0.18 mg/kg/h) did not [60]. In a trial of patients undergoing major lumbar fusion surgery, only the highest dose esketamine (mean 14.8 µg/kg/h) produced measurable improvement [36].

Second, the dose of esketamine is insufficient relative to the severity of surgical trauma [1, 2]. In our patients, a median number of 7 vertebrae was involved during the procedure; regional anesthesia techniques which have become standard components of multimodal analgesia were not used due to concerns about surgical infections. As a result, the early postoperative median NRS pain score was 5 at rest and 7 with movement. Previous studies investigating effect of esketamine in PCIA were mainly performed in non-spinal or short-segment spinal surgery patients; some studies used regional anesthesia and/or peripheral nerve block [34, 35]. Pain intensity in those studies was less severe than in our patients.

Third, the dose of dexmedetomidine was not increased along with that of esketamine. In previous studies, esketamine was administered in combination with dexmedetomidine at a ratio of about 275–600 : 1 and produced additive or synergistic effects in improving analgesia (5.5 µg/kg/h : 0.02 µg/kg/h [2], 0.3 mg/kg : 0.5 µg/kg [61], and 0.25 mg/kg : 0.8 µg/kg [62]), producing sedation (0.5 mg/kg : 1 µg/kg [63, 64]), relieving anxiety (1 mg/kg : 2 µg/kg [63, 64]), preventing emergence delirium (0.5 mg/kg : 1 µg/kg [65]), and conducting opioid-free anesthesia (0.15 mg/kg/h : 0.3 µg/kg/h [66], 0.3 mg/kg : 0.6 µg/kg [67]). Increasing the doses of esketamine and dexmedetomidine simultaneously might produce better analgesic efficacy but requires further investigation.

We adopted PIOC as the primary endpoint of this study for the following reasons. First, AUC captures the temporal dynamics of pain, avoiding reliance on single point measurement. Second, using rank ratios instead of absolute pain scores minimizes potential bias from evaluators. Finally, the PIOC provides a standardized, quantitative measure for evaluating analgesic efficacy and opioid-sparing effect, which is consistent with clinical requirements. We did not select pain intensity, opioid consumption, or other variables alone as the primary endpoint because neither parameter alone describes the clinically relevant trade‑off between direct analgesic benefit and opioid-sparing effect. Our secondary endpoints mainly included individual parameters regarding pain intensity and opioid consumption within 72 h and thus provided more detailed evaluations of the intervention effects.

In our results, the incidence of postoperative nausea and vomiting (PONV) and, specifically, the incidence of postoperative nausea were significantly increased in the Esk75 group. This partially explains why patients in the Esk75 group experienced slightly higher pain scores and slightly worse sleep quality during the intervention period. Previous studies also found that use of ketamine or esketamine, even with a mini dose regimen via PCIA, increases the risk of nausea and vomiting [68–70]. On the other hand, ketamine or esketamine may decrease nausea and vomiting by reducing opioid consumption [71, 72]. The combined effect is that many studies did not find increase in PONV [70]. Opioid consumption was not reduced even in the highest dose group of our patients, resulting in a net increase in PONV. We did not find increase in neuropsychiatric symptoms among three groups; this can be attributed to the mini dose regimen [71]. However, increase in PONV makes it impractical to further increase esketamine dose in PCIA in this patient population.

There are some limitations. Firstly, a pilot study using the up-down sequential allocation method may help to identify the optimal dose of esketamine in PCIA and can be considered in future studies [73–77]. Secondly, the complex landscape of acute pain following extensive spinal fusion may interfere patients’ perception and self-assessment of pain intensity. For example, patients often report pain originating from regions other than the surgical site, such as lower back, shoulder, and neck [3]. We also noted radiating pain involving abdomen, hips, and lower extremities in our patients. These situations should be considered in future studies. Thirdly, our findings are applicable to patients recovering from complex spinal surgery and receiving systemic analgesics alone, but should be interpreted cautiously in other clinical settings especially when regional anesthesia is included.

Conclusions

Our results indicated that, for adult patients recovering from extensive posterior spinal fusion under multimodal analgesia, increasing esketamine dose in PCIA from 0.25 to 0.5 and 0.75 mg/ml did not significantly improve pain relief or reduce opioid consumption. Esketamine at a concentration of 0.75 mg/ml increased PONV. The optimal concentrations of components in PCIA for this patient population requires future investigation.

Abbreviations

3D-CAM

Three-Dimensional Confusion Assessment Method

ANOVA

Analysis of variance

ASA

American Society of Anesthesiologists

AUC

Area under the curve

BIS

Bispectral index

HRs

Hazard ratios

IQR

Interquartile range

ITT

Intention-to-treat

NMDA

N-methyl-D-aspartate

NRS

Numeric rating scale

NSAIDs

Non-steroidal anti-inflammatory drugs

NYHA

New York Heart Association

OC

Opioid consumption

PCIA

Patient-controlled intravenous analgesia

PHQ-9

Patient Health Questionnaire-9

PIOC

Pain Intensity with movement and Opioid Consumption

PONV

Postoperative nausea and vomiting

PP

Per-protocol

PSQI

Pittsburgh Sleep Quality Index

RASS

Richmond Agitation-Sedation Scale

RRs

Relative risks

SRS-22

Scoliosis Research Society-22 questionnaire

Authors’ contributions

All the authors contributed to the study design. Dong-Xin Wang, Fan Cui, and Jia-Hui Ma were responsible for writing proposal, study conception, data interpretation, and final manuscript revision. Zi-Meng Guo and Ying Zhang were responsible for clinical data collection, data analysis, manuscript drafting and revision of the manuscript. All authors read and approved the final version of the manuscript.

Funding

This study was support by National Natural Science Foundation of China (No. 82293644; Dong-Xin Wang). The sponsor had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, and approval of the manuscript; and decision to submit the manuscript for publication.

Data availability

For trial protocol, statistical analysis plan, and individual de-identified participant data, please contact the corresponding author ( [wangdongxin@hotmail.com](mailto: wangdongxin@hotmail.com) or [dxwang65@bjmu.edu.cn](mailto: dxwang65@bjmu.edu.cn) ).

Declarations

Ethics approval and consent to participate

The study protocol conformed to the ethical guidelines of the Declaration of Helsinki, and was approved by the Biomedical Research Ethics Committee of Peking University First Hospital (No.2023 − 351; approved on September 7, 2023). Written informed consent was obtained from each participant prior to enrollment in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.