Abstract
Objective
To evaluate whether erector spinae plane block (ESPB) combined with local infiltration analgesia (LIA) provides better postoperative analgesia than ESPB alone, LIA alone, or placebo in patients undergoing posterior lumbar interbody fusion (PLIF).
Methods
This prospective, double-blind, randomized controlled trial enrolled adults undergoing open PLIF for degenerative lumbar disease. Patients were randomly assigned in a 1:1:1:1 ratio to placebo, LIA, ESPB, or combined ESPB with LIA. Fluoroscopy-guided ESPB or placebo injection was performed before induction of general anesthesia, while LIA or placebo infiltration was given before wound closure. All patients received standardized general anesthesia and postoperative patient-controlled intravenous analgesia. The primary outcomes were postoperative sufentanil consumption and visual analogue scale (VAS) pain scores at rest and during movement at 4, 8, 12, 24, and 48 hours. Analyses were based on the modified intention-to-treat population. Longitudinal VAS trajectories were assessed using linear mixed-effects models, together with multiplicity-adjusted pairwise comparisons and factorial interaction analyses. A right-censoring sensitivity analysis of time to first patient-controlled intravenous analgesia (PCIA) bolus was performed using Kaplan-Meier, log-rank, and adjusted Cox proportional hazards analyses.
Results
Of the 188 randomized patients, 182 were included in the modified intention-to-treat analysis. Baseline characteristics were similar across the 4 groups. Body-weight- and operative-time-normalized remifentanil and propofol consumption rates were lower in the ESPB-containing groups, whereas the rocuronium consumption rate did not differ significantly among groups. Linear mixed-effects models showed significant group×time interactions for both resting and movement VAS scores, indicating different postoperative pain trajectories among the groups. The combined ESPB with LIA group showed the most consistent short-term analgesic benefit, with lower pain scores at several early postoperative time points, lower postoperative sufentanil consumption, fewer analgesia pump activations, and a longer time to first PCIA bolus. After multiplicity adjustment, the postoperative rescue analgesia count was also lower in the combined group than in the control group. Factorial analyses identified a greater-than-additive effect only for time to first PCIA bolus, whereas the other analgesic outcomes showed no consistent evidence of statistical synergy. Right-censoring sensitivity analyses supported the longer time to first PCIA bolus in the combined group. Adverse-event rates were similar among the 4 groups, and no significant between-group differences were observed in VAS or Oswestry Disability Index scores at 3 months.
Conclusion
Fluoroscopy-guided ESPB combined with LIA improved short-term postoperative analgesia after open PLIF. The combined regimen reduced postoperative opioid requirements and analgesia pump activations and prolonged the time to first PCIA bolus without increasing adverse events. Its benefit is better interpreted as an incremental, layered multimodal analgesic effect rather than a consistently synergistic effect across outcomes. No between-group differences in pain or functional outcomes were observed at 3 months.
Keywords: Posterior lumbar interbody fusion, Erector spinae plane block, Local infiltration analgesia, Postoperative pain, Multimodal analgesia
INTRODUCTION
Lumbar degenerative diseases are common indications for spine surgery, and posterior lumbar interbody fusion (PLIF) is a commonly used posterior fusion technique for degenerative lumbar conditions requiring decompression and stabilization [1]. Lumbar fusion surgery is associated with a high risk of acute postoperative pain and may also be followed by persistent pain, making adequate postoperative analgesia an important component of perioperative care [2]. In spine surgery, postoperative tissue injury and irritation or compression of the nerve roots and dorsal root ganglia may contribute to inflammatory responses, peripheral sensitization, and central sensitization, mechanisms implicated in chronic pain after spine surgery [3]. Effective postoperative analgesia is therefore an important component of Enhanced Recovery After Surgery (ERAS) pathways. Systematic review evidence suggests that ERAS protocols in spine surgery are associated with shorter hospital stays, fewer complications, and lower postoperative pain scores, although the strength of evidence varies among procedures [4].
Postoperative pain after lumbar spine surgery remains a major clinical concern, and conventional systemic analgesic approaches may be limited by incomplete analgesia and treatment-related adverse effects [5]. Opioids remain important for acute postoperative pain but are associated with adverse effects including nausea, vomiting, pruritus, sedation, urinary retention, and respiratory depression [6]. Persistent postoperative opioid use is an additional concern after major spine surgery [7]. Epidural and other neuraxial techniques can provide effective analgesia in spine surgery, but their use requires consideration of rare but serious neuraxial complications and the timing of perioperative antithrombotic therapy [8-10]. These considerations support a multimodal analgesic approach that combines pharmacological and regional techniques with different mechanisms to improve pain control and reduce opioid requirements [11]. Systematic reviews and network meta-analyses support multimodal and regional analgesic strategies in adult spine surgery [11,12].
Regional nerve blocks have gained attention in postoperative pain management. The erector spinae plane block (ESPB) was first described by Forero et al. [13] in 2016 for thoracic neuropathic pain. It has since been evaluated across various surgical procedures [14]. During ESPB, local anesthetic is injected into the fascial plane between the erector spinae muscle and the transverse process. Mechanistic evidence supports consistent involvement of the dorsal rami, whereas spread to the paravertebral or epidural spaces and ventral rami appears variable [15]. This dorsalramus involvement provides a mechanistic rationale for ESPB in posterior lumbar surgery [15]. Randomized clinical evidence has demonstrated postoperative analgesic efficacy of ESPB in lumbar spinal decompression surgery [16]. Recent meta-analyses have shown that ESPB can reduce postoperative pain and opioid consumption after lumbar spine surgery, supporting its role as a regional component of multimodal analgesia [17,18].
Local infiltration analgesia (LIA) has been used as a component of multimodal analgesia in lumbar spine surgery. In this technique, local anesthetic is infiltrated into the wound-adjacent paraspinal muscles and subcutaneous tissues, typically around the time of wound closure [19,20]. Systematic review and meta-analysis evidence suggests that local anesthetic wound infiltration can reduce early postoperative opioid requirements and may provide modest early pain relief, although the magnitude and duration of benefit are variable [19,20]. Wound infiltration has also been used as an analgesic technique in patients undergoing open PLIF, supporting its practical applicability in this surgical setting [21].
Both ESPB and LIA have shown value for postoperative pain control after PLIF. Still, routine analgesic regimens may be insufficient when pain is severe, as often occurs after extensive tissue injury and nerve root manipulation. Lennon et al. [22] studied combined ESPB and LIA in total hip arthroplasty (THA) and found that adding ESPB did not provide a clear extra analgesic benefit in patients who were already receiving LIA and oral multimodal analgesia. This raises a practical question for PLIF. Because the pain pattern and extent of tissue injury in PLIF differ from those in THA, it remains unclear whether combining ESPB with LIA provides stronger analgesia than either technique alone. Recent randomized controlled trials have shown that fluoroscopy-guided ESPB can provide effective early postoperative analgesia in lumbar spine surgery [23,24].
ESPB and wound infiltration have been directly compared in spine surgery. Beltrame et al. [24] reported that radioscopically guided ESPB provided better early postoperative analgesia than standard wound infiltration in open lumbar spine surgery, especially in multilevel decompression. In a separate randomized study involving a lumbotomy surgical incision, Hakimoğlu et al. [25] found that LIA reduced intraoperative remifentanil consumption, whereas ESPB provided longer postoperative analgesia. These studies mainly used two-arm designs and did not evaluate the individual and combined contributions of ESPB and LIA in the same PLIF population. The present trial adds value through its four-arm comparative design, including placebo, LIA alone, ESPB alone, and combined ESPB with LIA. Therefore, this prospective, double-blind, randomized, placebo-controlled, four-arm trial aimed to determine whether fluoroscopy-guided ESPB combined with LIA provides superior short-term postoperative analgesia compared with ESPB alone, LIA alone, or placebo in patients undergoing PLIF.
MATERIALS AND METHODS
1. Study Design
This prospective, double-blind, randomized controlled trial was conducted at the Second Hospital of Shanxi Medical University. The Ethics Committee of the Second Hospital of Shanxi Medical University approved the protocol under approval number 2024YX Number 354. The trial was registered in the Chinese Clinical Trial Registry under registration number ChiCTR2400088118. The trial is reported in accordance with the Consolidated Standards of Reporting Trials (CONSORT) 2025 statement [26]. All patients provided written informed consent before surgery.
2. Inclusion and Exclusion Criteria for Patients
Patients were eligible if they met all of the following criteria: (1) age older than 18 years; (2) a diagnosis of lumbar disc herniation, lumbar spondylolisthesis, or lumbar spinal stenosis; (3) scheduled for posterior lumbar open decompression and fusion with internal fixation surgery; (4) surgical fusion spanning 1 to 3 levels; (5) American Society of Anesthesiologists physical status classification grade of I to III; (6) New York Heart Association heart function classification of I to III; (7) a signed informed consent form. Patients were excluded if any of the following conditions were present: (1) coagulation dysfunction; (2) severe cardiac, pulmonary, hepatic, renal, or other major organ functional abnormalities; (3) a history of previous lumbar spine surgery via a posterior approach; (4) coexisting chronic pain conditions or long-term pain management, defined as chronic pain unrelated to the index lumbar degenerative disease lasting more than 3 months, regular preoperative opioid use, or long-term use of neuropathic pain medications; (5) infection at the puncture site or anatomical structural variations; (6) known allergy to local anesthetics; (7) communication barriers due to psychiatric or intellectual problems; (8) spinal deformity; (9) refusal to participate in the study. Patients were withdrawn from the study if any of the following occurred: (1) development of severe perioperative complications or anesthetic/surgical accidents; (2) loss to follow-up; (3) voluntary withdrawal requested by the patient or a family member; (4) failure of ESPB in the ESPB-containing groups.
3. Randomization and Blinding
Eligible patients were randomly assigned in a 1:1:1:1 ratio to 4 groups using a computer-generated random number table. Group A was the placebo group and received ESPB with 0.9% saline and LIA with 0.9% saline. Group B was the LIA group and received ESPB with 0.9% saline and LIA with 0.375% ropivacaine. Group C was the ESPB group and received ESPB with 0.375% ropivacaine and LIA with 0.9% saline. Group D was the combined ESPB plus LIA group and received 0.375% ropivacaine for both ESPB and LIA. Just before the intervention, a research nurse who was not involved in outcome assessment opened sequentially numbered opaque sealed envelopes and prepared either 0.9% saline or 0.375% ropivacaine in identical opaque syringes labeled only with the patient number. Fluoroscopy-guided ESPB or placebo injection was performed before induction of anesthesia, and LIA or placebo infiltration was performed before wound closure according to group assignment. Patients, anesthesiologists responsible for postoperative follow-up, ward staff, all physicians involved in the surgical procedure, outcome assessors, and statisticians remained blinded to group allocation. Sensory testing after ESPB was performed by an anesthesiologist who was not involved in postoperative outcome assessment, and the results were withheld from all blinded personnel.
4. Study Interventions and Perioperative Workflow
After randomization and allocation concealment, the assigned interventions were performed according to the actual perioperative workflow: ESPB or placebo injection was completed before induction of general anesthesia, standardized general anesthesia and PLIF surgery were then performed, LIA or placebo infiltration was administered before wound closure, and postoperative PCIA was initiated after skin closure.
1) ESPB or placebo injection before anesthesia induction
After arrival in the operating room, standard monitoring was established and intravenous access was secured. Before induction of general anesthesia, patients were placed in the prone position while awake and were closely monitored. The skin was disinfected and draped in a sterile manner. The target block level was selected according to the center of the planned surgical incision. Under C-arm fluoroscopy, a standard anteroposterior view of the target segment was obtained to identify the tips of both transverse processes. The needle entry point was selected lateral to the skin projection of the transverse process tip. After local infiltration, the block needle was advanced toward the lateral edge of the transverse process until bone contact was reached. The needle was then slightly withdrawn and positioned in the subfascial plane. After negative aspiration for blood or air, 20 mL of either 0.375% ropivacaine or 0.9% saline was injected slowly. The same procedure was repeated on the opposite side.
Sensory testing was performed 30 minutes after the bilateral injection and before induction of general anesthesia using a standardized pinprick test. Pinprick sensation was assessed bilaterally over the posterior lumbar region corresponding to the planned incision and adjacent dermatomal areas. The result was compared with the preblock baseline sensation and with an area outside the expected block distribution. Decreased sharp sensation or conversion from sharp pain to dull pressure was considered a positive sensory change. Sensory block success was defined as decreased pinprick sensation covering the planned incision-related posterior lumbar region and involving at least 2 adjacent dermatomal levels. The pinprick test was performed and recorded by an anesthesiologist who was not involved in later postoperative outcome assessment. The results were not disclosed to patients, ward staff, attending surgeons involved in the operation, outcome assessors, or statisticians. After sensory assessment, patients were returned to the supine position for induction of general anesthesia and endotracheal intubation. Once the airway was secured, patients were repositioned prone for PLIF surgery. Patients in the ESPB-containing groups who did not meet the predefined criteria for successful sensory blockade were classified as block failures.
2) Anesthesia and surgical procedures
After ESPB or placebo injection and pinprick sensory testing, patients were returned to the supine position for induction of general anesthesia. General anesthesia was induced intravenously with sufentanil 0.4 μg/kg, medium- and long-chain propofol 2 mg/kg, and rocuronium 0.6 mg/kg. Tracheal intubation was performed with a video laryngoscope, and mechanical ventilation was then started. After airway security was confirmed, patients were repositioned prone for surgery. Anesthesia was maintained with 1% sevoflurane, propofol initially at 4–6 mg/kg/hr, and remifentanil initially at 0.1–0.2 μg/kg/min. The infusion rates of propofol and remifentanil were adjusted according to bispectral index (BIS) values and hemodynamic responses during surgery. Blood pressure and heart rate were maintained within 20% of baseline values by adjusting the remifentanil infusion rate when necessary. If additional intraoperative rescue analgesia was required, sufentanil 5 μg was administered intravenously according to the anesthesiologist’s judgment. The BIS was maintained at 40–60 by adjusting the propofol dose as needed. Rocuronium was given intermittently to maintain muscle relaxation. All anesthetic drugs were stopped 15 minutes before the end of surgery. At that time, flurbiprofen axetil 50 mg was given intravenously for postoperative analgesia, and ondansetron 8 mg was given intravenously to prevent nausea and vomiting. The tracheal tube was removed after the patient was fully awake and spontaneous breathing had recovered. The patient was then transferred to the post-anesthesia care unit for observation.
3) LIA or placebo infiltration before wound closure
Before wound closure, the incision was thoroughly irrigated and careful hemostasis was achieved. A drainage tube was inserted as usual. Using a 20-mL syringe, 20 mL of either 0.375% ropivacaine or 0.9% saline was evenly infiltrated into the bilateral paraspinal muscles, fascia, and subcutaneous tissues on both sides of the incision. The wound was then closed layer by layer.
4) Postoperative analgesia methods
A patient-controlled intravenous analgesia (PCIA) pump was prepared before surgery with sufentanil 100 μg diluted in 0.9% sodium chloride to a total volume of 100 mL. The pump was started after skin closure. The background infusion rate was 1 mL/h, the bolus dose was 5 mL, and the lockout interval was 15 minutes. Analgesia continued for 48 hours after surgery. All patients also received flurbiprofen axetil 100 mg by intravenous infusion twice daily after surgery. When the visual analogue scale (VAS) pain score was greater than 4, or whenever additional analgesia was required, patients pressed the analgesia pump for extra pain relief. If pain control remained inadequate after 2 consecutive bolus doses, parecoxib sodium 40 mg could be given intramuscularly as rescue analgesia. PCIA could be discontinued if severe and intolerable nausea or vomiting occurred, if symptoms did not improve after antiemetic treatment, or if nausea and vomiting recurred after initial improvement. The time to first PCIA bolus, total number of pump activations, postoperative cumulative sufentanil consumption, and number of rescue analgesia administrations within 48 hours were recorded.
5. Observation Indicators
The primary outcomes were total postoperative sufentanil consumption within the first 48 hours after surgery and postoperative VAS scores at rest and during movement at 4, 8, 12, 24, and 48 hours. Key secondary outcomes included analgesia pump activations, time to first PCIA bolus, postoperative rescue analgesia count, early recovery indicators, adverse events, and 3-month VAS and Oswestry Disability Index (ODI) scores. Postoperative VAS pain scores at rest and during movement were assessed at 4, 8, 12, 24, and 48 hours after surgery in all 4 groups. Analgesia-related outcomes included total perioperative opioid consumption, cumulative postoperative sufentanil consumption within 48 hours, time to first PCIA bolus, number of PCIA bolus activations, and number of rescue analgesia administrations. For patients who did not receive a PCIA bolus within the first 48 postoperative hours, the time to first PCIA bolus was recorded as 48 hours for the primary restricted-time analysis. Postoperative sufentanil consumption was calculated as the fixed background PCIA infusion dose plus the bolus dose. The PCIA regimen provided continuous background sufentanil infusion at 1 μg/hr for 48 hours, corresponding to 48 μg, and each PCIA bolus delivered an additional 5 μg of sufentanil. Thus, cumulative postoperative sufentanil consumption within 48 hours was calculated as 48 μg+5 μg×number of PCIA bolus activations. Recovery-related outcomes included time to first ambulation, length of postoperative hospital stay, and patient satisfaction. Adverse events within 48 hours after surgery were monitored and recorded, including nausea, vomiting, pruritus, drowsiness, respiratory depression, bradycardia, hypotension, surgical site infection, urinary retention, and motor block. At 3 months after surgery, all patients were followed up using ODI and VAS scores to assess postoperative recovery. Satisfaction with pain management was rated on a 5-point scale. Scores of 0, 1, 2, 3, and 4 represented very dissatisfied, dissatisfied, neutral, satisfied, and very satisfied, respectively. Patients with a score of 3 or higher were considered satisfied and were included in the satisfaction rate analysis. The ODI evaluates 10 aspects of thoracolumbar function, including self-care, pain intensity, social life, sexual activity, and travel. Each item is scored from 0 to 5 points. The final score is calculated as total points scored/(5×number of questions answered)×100. The total ODI score ranges from 0 to 100, with lower scores indicating better function and quality of life [27]. The VAS score ranges from 0 to 10, with higher scores indicating more severe pain [28].
6. Sample Size Estimation
Sample size was calculated with PASS 15 software from NCSS, LLC (USA). Total postoperative sufentanil consumption was used as the primary outcome for the calculation. In the pilot study, mean total postoperative sufentanil consumption was 65.50±3.55 μg in 10 patients from group A, 59.50±4.74 μg in 10 patients from group B, 59.00±5.68 μg in 10 patients from group C, and 53.00±4.08 μg in 10 patients from group D. With a significance level of 0.05 and a statistical power of 0.90, the estimated sample size was 28 patients per group. To allow for a 20% dropout rate, at least 35 patients were required in each group.
7. Statistical Analysis
Statistical analyses were performed using IBM SPSS Statistics ver. 22.0 (IBM Co., USA) and RStudio, and figures were generated using GraphPad Prism 9.5 (GraphPad Software Inc., USA) and RStudio. Reporting followed the CONSORT 2025 recommendations and the CONSORT extension for multiarm parallel-group randomized trials [26,29]. Continuous variables were assessed for normality using the Shapiro-Wilk test together with histogram inspection. Normally distributed variables were presented as mean±standard deviation and compared using one-way analysis of variance. Nonnormally distributed or ordinal variables were presented as median (interquartile range) and compared using the Kruskal-Wallis test. Categorical variables were presented as n (%) and compared using Pearson χ2 test or the Fisher-Freeman-Halton exact test, as appropriate. Pairwise comparisons used Mann-Whitney U-tests for continuous or count outcomes and Pearson χ2 test for categorical outcomes. Primary efficacy analyses were based on the modified intention-to-treat (mITT) population, which included randomized participants with available postoperative outcome data and analyzed them according to their assigned groups [30,31]. Missing outcome data were not imputed.
Repeated resting and movement VAS scores measured at 4, 8, 12, 24, and 48 hours were analyzed using linear mixed-effects models. Group, time, and the group×time interaction were included as fixed effects, with a patient-level random intercept to account for repeated measurements [32]. Adjusted models also included surgical segments, operative time, and preoperative VAS score. Models were fitted by maximum likelihood, and fixed effects were assessed using Wald χ2 tests. Group A, the control group, and the 4-hour assessment were used as the reference levels. The group×time interaction was the main longitudinal test for differences in postoperative pain trajectories. Time-specific group comparisons were assessed using Kruskal-Wallis tests and pairwise Mann-Whitney U-tests. For the 2×2 factorial analyses, the 4 treatment groups were recoded according to the presence or absence of ESPB and LIA. Models included ESPB, LIA, and the ESPB×LIA interaction to assess departure from additivity [33,34]. Longitudinal factorial analyses used mixed-effects models with a patient-level random intercept. The ESPB× LIA and ESPB×LIA×time interactions were assessed using likelihood-ratio tests. Time-specific factorial interaction estimates were obtained using linear regression with HC3 heteroskedasticity-robust standard errors. For VAS outcomes, the interaction estimate was calculated as group D−group C−group B+group A. Negative values indicated a greater-than-additive reduction in pain, whereas positive values indicated a less-than-additive effect. Postoperative analgesia requirement outcomes were compared overall using the Kruskal-Wallis test and pairwise using Mann-Whitney U-tests. Multivariable sensitivity analyses were performed to assess the robustness of the findings [35]. These analyses and the factorial ESPB×LIA interaction analyses for postoperative analgesia requirement outcomes used linear regression with HC3 heteroskedasticity-robust standard errors. Adjusted models included surgical segments, operative time, and preoperative VAS score, and model effects were assessed using Wald χ2 tests. In the primary restricted-time analysis of time to first PCIA bolus, patients who did not receive a PCIA bolus within 48 hours were assigned a value of 48 hours. A further sensitivity analysis accounted for right censoring by treating these patients as administratively censored at 48 hours. Kaplan-Meier methods and log-rank tests were used to compare time to first PCIA bolus among groups [36]. Pairwise log-rank comparisons were adjusted across the 6 group comparisons using the Holm-Bonferroni method. An adjusted Cox proportional hazards model was also fitted and included surgical segments, operative time, and preoperative VAS score [37]. Hazard ratios below 1 indicated a lower hazard of receiving the first PCIA bolus and therefore a longer time to first PCIA bolus.
Holm-Bonferroni correction was used to control for multiple testing [38]. Holm adjustment was applied across the 10 time-specific VAS omnibus tests, including 5 resting and 5 movement pain assessments. Pairwise VAS comparisons were adjusted globally across 30 comparisons within each pain modality, based on 5 time points and 6 group pairs. Time-specific factorial VAS interaction analyses were adjusted across the 5 time points within each modality. For postoperative analgesia requirement outcomes, Holm adjustment was applied across 4 overall comparisons, globally across 24 pairwise comparisons based on 4 outcomes and 6 group pairs, and across 4 factorial interaction tests when applicable. For intraoperative outcomes, Holm adjustment was applied across the 4 overall comparisons and across 6 pairwise comparisons within each variable. For early postoperative recovery outcomes, Holm adjustment was applied across the 3 overall comparisons, while pairwise comparisons of satisfaction rate were adjusted across the 6 group comparisons. The clinical relevance of between-group VAS differences was assessed using a minimal clinically important difference (MCID) of 1 point on the 0–10 VAS scale [39]. All statistical tests were 2-sided. A p-value below 0.05 was considered statistically significant, and Holm-adjusted p-values were used when correction for multiple testing was required.
RESULTS
1. Demographic Data
From September 2024 to September 2025, 209 patients with lumbar degenerative diseases scheduled for PLIF were screened according to the predefined criteria. Among them, 13 patients did not meet the eligibility criteria and 8 declined to participate. A total of 188 patients were enrolled and randomly assigned to groups A, B, C, and D, with 47 patients in each group. During the study period, 6 patients were excluded after randomization and were not included in the mITT population because complete analyzable postoperative outcome data were unavailable or considered inapplicable due to major protocol deviations or postoperative complications. In group A, 1 patient was excluded because of postoperative delirium. In group C, 3 patients were excluded, including 1 patient with intraoperative cerebrospinal fluid leakage and 2 patients with prolonged hospitalization caused by changes in medical condition. In group D, 2 patients were excluded for similar reasons, including 1 case of intraoperative cerebrospinal fluid leakage and 1 case of prolonged hospitalization related to changes in medical condition. The mITT population included 182 patients who completed the full 3-month postoperative follow-up, with 46 patients in group A, 47 in group B, 44 in group C, and 45 in group D, as shown in Fig. 1. Pinprick sensory testing 30 minutes after injection showed successful ESPB sensory blockade in all analyzed patients assigned to ropivacaine ESPB: 44 of 44 patients (100%) in group C and 45 of 45 patients (100%) in group D. No block failure was recorded among the analyzed ESPB-containing groups. Sensory testing results were recorded only by the anesthesiologist responsible for block assessment and were not disclosed to patients, attending surgeons, postoperative follow-up anesthesiologists, ward staff, outcome assessors, or statisticians. Outcome assessment blinding was therefore considered maintainable, although possible sensory-related unblinding cannot be fully excluded. Baseline demographic characteristics and preoperative clinical data were comparable among the 4 groups. Preoperative pain intensity and functional status were also comparable, with no significant differences in preoperative VAS or ODI scores. Intraoperative variables, including operative time, number of surgical segments, and intraoperative blood loss, were similar across groups, as shown in Table 1.
Fig. 1.

CONSORT (Consolidated Standards of Reporting Trials) flow diagram of patient enrollment, allocation, follow-up, and analysis. mITT, modified intention-to-treat; ESPB, erector spinae plane block; LIA, local infiltration analgesia; PCIA, patientcontrolled intravenous analgesia; PONV, postoperative nausea and vomiting.
Table 1.
Baseline demographic and clinical characteristics
| Variable | Group A (n = 46) | Group B (n = 47) | Group C (n = 44) | Group D (n = 45) | F/H/χ2/Fisher | p-value |
|---|---|---|---|---|---|---|
| Age (yr) | 65.1 ± 8.3 | 64.7 ± 7.1 | 62.5 ± 9.1 | 64.1 ± 9.2 | 0.807† | 0.491 |
| Sex, male:female (n) | 22:24 | 23:24 | 24:20 | 21:24 | 0.650§ | 0.885 |
| Weight (kg) | 69.00 (61.75–78.50) | 65.00 (60.00–75.00) | 70.50 (61.25–79.75) | 67.00 (60.00–75.00) | 2.291‡ | 0.514 |
| Height (m) | 1.66 ± 0.08 | 1.65 ± 0.10 | 1.66 ± 0.08 | 1.65 ± 0.09 | 0.228† | 0.877 |
| BMI (kg/m2) | 24.22 (22.72–29.33) | 24.84 (22.49–27.13) | 25.73 (23.77–27.30) | 25.20 (22.94–27.54) | 0.817‡ | 0.845 |
| Disease type (n) | 0.416§ | 0.999 | ||||
| Lumbar disc herniation | 20 | 19 | 20 | 18 | ||
| Lumbar spondylolisthesis | 10 | 10 | 9 | 10 | ||
| Lumbar spinal stenosis | 16 | 18 | 15 | 17 | ||
| No. of surgical segments | ||||||
| 1:2:3 | 31:13:2 | 30:15:2 | 27:14:3 | 28:15:2 | 1.005∥ | 0.992 |
| Comorbidities | ||||||
| Hypertension | 15/46 (32.6) | 18/47 (38.3) | 16/44 (36.4) | 14/45 (31.1) | 0.667§ | 0.881 |
| Diabetes | 18/46 (39.1) | 16/47 (34.0) | 18/44 (40.9) | 20/45 (44.4) | 1.084§ | 0.781 |
| ASA PS classification grade | ||||||
| I:II:III | 0:37:9 | 1:33:13 | 1:33:10 | 1:35:9 | 2.699∥ | 0.910 |
| NYHA grade | ||||||
| I:II:III | 8:37:1 | 12:33:2 | 9:34:1 | 13:30:2 | 3.156∥ | 0.826 |
| Operative time (min) | 139.00 (114.25–162.25) | 130.00 (110.00–151.00) | 124.00 (110.00–161.25) | 124.00 (112.00–156.00) | 0.739‡ | 0.864 |
| Intraoperative blood loss (mL) | 200.00 (200.00–300.00) | 250.00 (200.00–350.00) | 275.00 (200.00–300.00) | 200.00 (200.00–400.00) | 0.686‡ | 0.877 |
| Preoperative VAS score | 7.00 (7.00–8.00) | 7.00 (6.00–7.00) | 7.00 (6.00–8.00) | 7.00 (6.00–8.00) | 2.914‡ | 0.405 |
| Preoperative ODI score | 64.57 ± 6.54 | 62.89 ± 7.93 | 64.20 ± 7.07 | 63.11 ± 8.26 | 0.546† | 0.652 |
Values are presented as mean±standard deviation, median (interquartile range), or number (%) otherwise indicated.
Group A, control; Group B, LIA; Group C, ESPB; Group D, ESPB+LIA; LIA, local infiltration analgesia; ESPB, erector spinae plane block; BMI, body mass index; ASA PS, American Society of Anesthesiologists physical status; NYHA, New York Heart Association; VAS, visual analogue scale; ODI, Oswestry Disability Index.
F statistic.
H statistic.
χ2 statistic.
Fisher-Freeman-Halton exact test.
2. Comparative Analysis of Normalized Anesthetic Drug Consumption Rates and Intraoperative Rescue Analgesia Rate
As shown in Fig. 2 and Table 2, body-weight- and operative-time-normalized intraoperative anesthetic drug consumption rates and the intraoperative rescue analgesia rate were compared among the 4 groups. Remifentanil consumption rates differed significantly among groups (group A: 0.28 [0.22–0.34] μg/kg/min; group B: 0.30 [0.23–0.36] μg/kg/min; group C: 0.23 [0.16–0.27] μg/kg/min; group D: 0.23 [0.17–0.30] μg/kg/min; H=24.029, Holm-adjusted p<0.001). Pairwise comparisons with Holm-Bonferroni correction showed significantly lower remifentanil rates in groups C and D than in groups A and B. Propofol consumption rates also differed significantly among groups (group A: 6.86 [6.04–7.77] mg/kg/hr; group B: 7.15 [6.68–8.74] mg/kg/hr; group C: 5.82 [4.96–6.74] mg/kg/hr; group D: 6.43 [5.36–6.83] mg/kg/hr; H=37.594, Holm-adjusted p<0.001), with lower rates in groups C and D than in groups A and B. In contrast, the rocuronium consumption rate did not differ significantly among the 4 groups (group A: 0.50 [0.36–0.74] mg/kg/hr; group B: 0.53 [0.37–0.74] mg/kg/hr; group C: 0.41 [0.33–0.59] mg/kg/hr; group D: 0.49 [0.33–0.62] mg/kg/hr; H=7.324, Holm-adjusted p=0.062). The intraoperative rescue analgesia rate also differed among groups (group A: 21 of 46 [45.7%]; group B: 18 of 47 [38.3%]; group C: 8 of 44 [18.2%]; group D: 8 of 45 [17.8%]; χ2=12.977, Holm-adjusted p=0.009). After Holm-Bonferroni correction, rescue analgesia rates were significantly lower in groups C and D than in group A. Detailed pairwise comparisons with raw and Holm-Bonferroni-adjusted p-values are presented in Supplementary Table 1.
Fig. 2.

Comparison of body-weight- and operative-time-normalized intraoperative anesthetic drug consumption rates among the 4 groups. (A) Remifentanil consumption rate (μg/kg/min). (B) Propofol consumption rate (mg/kg/hr). (C) Rocuronium consumption rate (mg/kg/hr). Pairwise comparisons were performed using Mann-Whitney U-tests, with Holm-Bonferroni correction across the 6 group comparisons within each variable. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. Group A, control; Group B, LIA; Group C, ESPB; Group D, ESPB+LIA; LIA, local infiltration analgesia; ESPB, erector spinae plane block; ns, not significant.
Table 2.
Intraoperative anesthetic drug consumption rates and rescue analgesia among the 4 groups
| Variable | Group A (n = 46) | Group B (n = 47) | Group C (n = 44) | Group D (n = 45) | H/χ2 | Raw p-value | Holm-adjusted p-value |
|---|---|---|---|---|---|---|---|
| Remifentanil rate (μg/kg/min) | 0.28 (0.22–0.34) | 0.30 (0.23–0.36) | 0.23 (0.16–0.27)a,b) | 0.23 (0.17–0.30)a,b) | 24.029† | < 0.001 | < 0.001 |
| Propofol rate (mg/kg/hr) | 6.86 (6.04–7.77) | 7.15 (6.68–8.74) | 5.82 (4.96–6.74)a,b) | 6.43 (5.36–6.83)a,b) | 37.594† | < 0.001 | < 0.001 |
| Rocuronium rate (mg/kg/hr) | 0.50 (0.36–0.74) | 0.53 (0.37–0.74) | 0.41 (0.33–0.59) | 0.49 (0.33–0.62) | 7.324† | 0.062 | 0.062 |
| Rescue analgesia rate | 21/46 (45.7) | 18/47 (38.3) | 8/44 (18.2)a) | 8/45 (17.8)a) | 12.977‡ | 0.005 | 0.009 |
Values are presented as median (interquartile range) or number (%).
Group A, control; Group B, LIA; Group C, ESPB; Group D, ESPB+LIA; LIA, local infiltration analgesia; ESPB, erector spinae plane block.
Remifentanil, propofol, and rocuronium consumption rates were normalized to body weight and operative time.
Overall comparisons were performed using the Kruskal-Wallis test for continuous outcomes and Pearson χ2 test for intraoperative rescue analgesia rate, with Holm adjustment across the 4 omnibus tests. Pairwise comparisons were performed using Mann-Whitney U-tests for continuous outcomes and Pearson χ2 tests for intraoperative rescue analgesia rate, with Holm-Bonferroni correction across the 6 group comparisons within each variable. Superscripts indicate significant favorable pairwise differences after correction:
Adjusted p<0.05 versus group A.
Adjusted p<0.05 versus group B.
H statistic.
χ2 statistic.
3. Comparison of Resting Pain VAS Scores at Various Time Points After Surgery
As shown in Fig. 3 and Table 3, resting pain VAS scores differed significantly among the 4 groups at 4, 8, 12, and 24 hours after surgery after Holm correction across the 10 VAS omnibus tests (all Holm-adjusted p<0.001), whereas no significant between-group difference was observed at 48 hours (H=3.550, Holm-adjusted p=0.314). At 4 and 8 hours, all 3 treatment groups showed significantly lower resting pain VAS scores than group A. At 12 hours, groups B and C had lower resting pain scores than group A, and group D showed lower resting pain scores than both groups A and B. At 24 hours, group D had significantly lower resting pain scores than groups A, B, and C. No statistically significant pairwise difference was found at 48 hours after global Holm-Bonferroni correction (Supplementary Table 2).
Fig. 3.

Comparison of postoperative resting pain visual analogue scale (VAS) scores among the 4 groups at different time points. Pairwise comparisons were performed using Mann- Whitney U-tests, with Holm-Bonferroni correction applied globally across all 30 resting-pain comparisons (5 time points×6 group pairs). Asterisks indicate statistically significant adjusted comparisons. Group A, control; Group B, LIA; Group C, ESPB; Group D, ESPB+LIA; LIA, local infiltration analgesia; ESPB, erector spinae plane block. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.
Table 3.
Comparison of resting pain VAS scores of patients at various time points after surgery
| Variable | 4 Hours | 8 Hours | 12 Hours | 24 Hours | 48 Hours |
|---|---|---|---|---|---|
| Group A (n = 46) | 2.00 (1.00–3.00) | 3.00 (3.00–4.00) | 4.00 (3.00–5.00) | 3.00 (2.75–4.00) | 2.00 (2.00–3.00) |
| Group B (n = 47) | 1.00 (0.00–2.00)a) | 2.00 (2.00–3.00)a) | 3.00 (2.00–4.00)a) | 3.00 (2.00–3.00) | 2.00 (1.00–3.00) |
| Group C (n = 44) | 1.00 (0.00–2.00)a) | 2.00 (1.00–3.00)a) | 3.00 (2.00–4.00)a) | 3.00 (2.00–4.00) | 2.00 (2.00–3.00) |
| Group D (n = 45) | 1.00 (0.00–1.00)a) | 2.00 (1.00–2.00)a) | 2.00 (2.00–3.00)a,b) | 2.00 (1.00–3.00)a,b,c) | 2.00 (2.00–2.50) |
| H | 23.407 | 49.200 | 48.207 | 29.508 | 3.550 |
| Raw p-value | < 0.001 | < 0.001 | < 0.001 | < 0.001 | 0.314 |
| Holm-adjusted p-value | < 0.001 | < 0.001 | < 0.001 | < 0.001 | 0.314 |
Values are presented as median (interquartile range).
VAS, visual analogue scale; Group A, control; Group B, LIA; Group C, ESPB; Group D, ESPB+LIA; LIA, local infiltration analgesia; ESPB, erector spinae plane block.
Overall comparisons at each time point were performed using the Kruskal-Wallis test, with Holm adjustment across the 10 VAS omnibus tests, including 5 resting-pain and 5 movement-pain tests. Pairwise comparisons were performed using Mann-Whitney U-tests, with Holm-Bonferroni correction applied globally across all 30 resting-pain comparisons (5 time points×6 group pairs).
Superscripts indicate significant favorable pairwise differences after correction:
Adjusted p<0.05 versus group A.
Adjusted p<0.05 versus group B.
Adjusted p<0.05 versus group C.
Longitudinal mixed-effects analysis further confirmed that resting pain trajectories differed significantly among the 4 groups. In the unadjusted model, the group×time interaction was significant (Wald χ2=68.271, df=12, p<0.001). After adjustment for surgical segments, operative time, and preoperative VAS score, the group×time interaction remained significant (Wald χ2=68.271, df=12, p<0.001). With treatment coding using group A (control) and 4 h as the reference levels, the conditional main effects of time (Wald χ2=199.255, df=4, p<0.001) and group (Wald χ2=62.110, df=3, p<0.001) were also significant. Surgical segments were significantly associated with resting pain VAS scores (Wald χ2=139.062, df=1, p<0.001), whereas operative time (Wald χ2=1.503, df=1, p=0.220) and preoperative VAS (Wald χ2=3.699, df=1, p=0.054) were not statistically significant covariates (Supplementary Table 3).
In the longitudinal 2×2 factorial mixed-effects analysis, the overall ESPB×LIA interaction across time was significant (likelihood-ratio χ2=30.224, df=5, p<0.001), and the ESPB×LIA×time interaction was also significant (likelihood-ratio χ2=26.654, df=4, p<0.001), indicating that the interaction between ESPB and LIA varied across postoperative time points (Supplementary Table 4). However, time-specific factorial estimates showed that the interaction estimates at 4 and 8 hours were positive and statistically significant after Holm correction, suggesting a less-than-additive rather than synergistic pattern at these early time points. At 12, 24, and 48 hours, the ESPB×LIA interaction was not statistically significant after correction. Therefore, the combined intervention should be interpreted as providing a combined or incremental short-term analgesic benefit rather than formal statistical synergy for resting pain (Supplementary Table 5).
To evaluate the clinical relevance of the observed differences, an MCID-based analysis was performed using a threshold of 1 point on the 0–10 VAS scale. Compared with group A, groups B, C, and D all achieved statistically significant and clinically meaningful reductions in resting pain at 4, 8, and 12 hours. At 12 hours, group D showed a 2-point median reduction versus group A. At 24 hours, only group D maintained a statistically significant and clinically meaningful reduction versus group A. At 48 hours, no treatment group met the MCID threshold compared with group A (Supplementary Table 6). In the detailed pairwise MCID analysis, group D also met both statistical significance and MCID criteria compared with group B at 12 hours and compared with groups B and C at 24 hours, supporting a clinically meaningful short-term advantage of the combined ESPB with LIA regimen mainly within the first 24 hours after surgery (Supplementary Table 7).
Overall, these findings indicate that ESPB, LIA, and their combination reduced early postoperative resting pain compared with control, while the combined ESPB with LIA regimen provided the most sustained short-term benefit up to 24 hours. However, the benefit did not persist at 48 hours, and the factorial interaction analysis did not support a formal synergistic effect for resting pain.
4. Comparison of Movement Pain VAS Scores at Various Time Points After Surgery
As shown in Fig. 4 and Table 4, movement pain VAS scores differed significantly among the 4 groups at all postoperative time points after Holm correction across the 10 VAS omnibus tests (4 hours: H=52.327, adjusted p<0.001; 8 hours: H=61.770, adjusted p<0.001; 12 hours: H=29.071, adjusted p<0.001; 24 hours: H=28.671, adjusted p<0.001; 48 hours: H=15.932, adjusted p=0.002). At 4 and 8 hours, all 3 treatment groups showed significantly lower movement pain VAS scores than group A. At 4 hours, group D also had lower movement pain scores than group C. At 8 hours, group D had lower movement pain scores than both groups B and C. At 12 hours, group D had lower movement pain scores than groups A and C, whereas the differences between group A and groups B or C did not remain statistically significant after correction. At 24 hours, group D had significantly lower movement pain scores than groups A, B, and C. At 48 hours, statistically significant pairwise differences were still observed for group A versus group D and group C versus group D after global Holm-Bonferroni correction (Supplementary Table 8).
Fig. 4.

Comparison of postoperative movement pain visual analogue scale (VAS) scores among the 4 groups at different time points. Pairwise comparisons were performed using Mann-Whitney U-tests, with Holm-Bonferroni correction applied globally across all 30 movement-pain comparisons (5 time points×6 group pairs). Asterisks indicate statistically significant adjusted comparisons. Group A, control; Group B, LIA; Group C, ESPB; Group D, ESPB+LIA; LIA, local infiltration analgesia; ESPB, erector spinae plane block. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.
Table 4.
Comparison of movement pain VAS scores of patients at various time points after surgery
| Variable | 4 Hours | 8 Hours | 12 Hours | 24 Hours | 48 Hours |
|---|---|---|---|---|---|
| Group A (n = 46) | 4.00 (3.00–4.25) | 5.00 (4.00–5.00) | 5.00 (3.75–6.00) | 4.00 (3.00–5.00) | 3.00 (3.00–4.00) |
| Group B (n = 47) | 3.00 (2.00–4.00)a) | 3.00 (3.00–4.00)a) | 4.00 (3.00–5.00) | 4.00 (3.00–4.00) | 3.00 (2.00–4.00) |
| Group C (n = 44) | 3.00 (2.00–3.00)a) | 3.00 (3.00–4.00)a) | 4.00 (3.00–5.00) | 4.00 (3.00–4.00) | 3.00 (2.25–4.00) |
| Group D (n = 45) | 2.00 (1.00–3.00)a,c) | 3.00 (2.00–3.00)a,b,c) | 3.00 (2.00–4.00)a,c) | 3.00 (2.00–4.00)a,b,c) | 3.00 (2.00–3.00)a,c) |
| H | 52.327 | 61.770 | 29.071 | 28.671 | 15.932 |
| Raw p-value | < 0.001 | < 0.001 | < 0.001 | < 0.001 | 0.001 |
| Holm-adjusted p-value | < 0.001 | < 0.001 | < 0.001 | < 0.001 | 0.002 |
Values are presented as median (interquartile range).
VAS, visual analogue scale; Group A, control; Group B, LIA; Group C, ESPB; Group D, ESPB+LIA; LIA, local infiltration analgesia; ESPB, erector spinae plane block.
Overall comparisons at each time point were performed using the Kruskal-Wallis test, with Holm adjustment across the 10 VAS omnibus tests, including 5 resting-pain and 5 movement-pain tests. Pairwise comparisons were performed using Mann-Whitney U-tests, with Holm-Bonferroni correction applied globally across all 30 movement-pain comparisons (5 time points×6 group pairs).
Superscripts indicate significant favorable pairwise differences after correction:
Adjusted p<0.05 versus group A.
Adjusted p<0.05 versus group B.
Adjusted p<0.05 versus group C.
Longitudinal mixed-effects analysis further confirmed that postoperative movement pain trajectories differed significantly among the 4 groups. In the unadjusted model, the group×time interaction was significant (Wald χ2=48.011, df=12, p<0.001). After adjustment for surgical segments, operative time, and preoperative VAS score, the group×time interaction remained significant (Wald χ2=48.011, df=12, p<0.001). With treatment coding using group A (control) and 4 h as the reference levels, the conditional main effects of time (Wald χ2=96.722, df=4, p<0.001) and group (Wald χ2=88.552, df=3, p<0.001) were also significant. Surgical segments were significantly associated with movement pain VAS scores (Wald χ2=52.601, df=1, p<0.001), whereas operative time (Wald χ2=1.146, df=1, p=0.284) and preoperative VAS (Wald χ2=1.355, df=1, p=0.244) were not statistically significant covariates (Supplementary Table 9).
In the longitudinal 2×2 factorial mixed-effects analysis, the overall ESPB×LIA interaction across time was significant for movement pain VAS scores (likelihood-ratio χ2=12.649, df=5, p=0.027), and the ESPB×LIA×time interaction was also significant (likelihood-ratio χ2=12.460, df=4, p=0.014), indicating that the interaction between ESPB and LIA varied across postoperative time points (Supplementary Table 10). However, time-specific factorial interaction estimates did not remain statistically significant after Holm correction at any individual time point. Therefore, although the longitudinal model suggested departure from simple additivity over time, the time-specific analyses did not provide sufficient evidence to claim a formal synergistic effect for movement pain reduction. The combined intervention should therefore be interpreted as providing a combined or incremental short-term analgesic benefit rather than formal statistical synergy for movement pain (Supplementary Table 11).
MCID analysis provided further context for the clinical importance of the observed differences. Using an MCID threshold of 1 point on the 0–10 VAS scale, clinically meaningful reductions in movement pain were seen mainly in the early postoperative period. Compared with group A, all active treatment groups reached the MCID threshold at 4, 8, and 12 hours. At 12 hours, group D still showed both a statistically significant difference after multiplicity adjustment and a clinically meaningful reduction, while groups B and C reached the MCID threshold but were no longer statistically significant after correction. By 24 hours, only group D continued to show both statistical significance and a clinically meaningful reduction compared with group A. At 48 hours, some pairwise differences remained statistically significant, although none of the median between-group differences reached the MCID threshold. Taken together, these findings suggest that the combined ESPB with LIA regimen provided the most consistent clinically meaningful improvement in movement pain during the first 24 hours after surgery, with no clinically meaningful between-group difference evident at 48 hours (Supplementary Tables 12 and 13).
Overall, these findings indicate that ESPB, LIA, and their combination reduced early postoperative movement pain compared with control, while the combined ESPB with LIA regimen provided the most consistent short-term benefit for movement pain within the first 24 hours. However, the clinical benefit was not sustained at 48 hours, and the factorial interaction analysis did not support a formal synergistic effect for movement pain.
5. Comparison of Postoperative Analgesia Requirements
Postoperative analgesia requirements within 48 hours after surgery are shown in Fig. 5 and Table 5. Significant overall between-group differences were observed in total postoperative sufentanil consumption (H=63.942, Holm-adjusted p<0.001), analgesia pump activations (H=63.942, Holm-adjusted p<0.001), time to first PCIA bolus (H=111.807, Holm-adjusted p<0.001), and postoperative rescue analgesia count (H=25.649, Holm-adjusted p<0.001). In Holm-Bonferroni-adjusted pairwise comparisons, groups B, C, and D had lower total postoperative sufentanil consumption and fewer analgesia pump activations than group A, while group D also had lower values than groups B and C. For time to first PCIA bolus, groups B, C, and D had longer times than group A, and group D had longer times than groups B and C. No significant difference was observed between groups B and C. For postoperative rescue analgesia count, only group D remained significantly lower than group A after correction (Supplementary Table 14).
Fig. 5.

Comparison of postoperative analgesia requirements among the 4 groups within 48 hours after surgery. (A) Total postoperative sufentanil consumption. (B) Analgesia pump activations. (C) Time to first PCIA bolus. (D) Postoperative rescue analgesia count. Pairwise comparisons were performed using Mann-Whitney U-tests, with Holm-Bonferroni correction applied globally across all 24 comparisons (4 outcomes×6 group pairs). Asterisks indicate statistically significant adjusted comparisons. Group A, control; Group B, LIA; Group C, ESPB; Group D, ESPB+LIA; LIA, local infiltration analgesia; ESPB, erector spinae plane block; PCIA, patient-controlled intravenous analgesia; ns, not significant. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.
Table 5.
Comparison of postoperative analgesia requirements among the 4 groups within 48 hours
| Variable | Total postoperative sufentanil consumption (μg) | Analgesia pump activations (count) | Time to first PCIA bolus (hr) | Postoperative rescue analgesia (count) |
|---|---|---|---|---|
| Group A (n = 46) | 63.00 (61.75–68.00) | 3.00 (2.75–4.00) | 4.72 (2.48–6.08) | 1.00 (1.00–2.00) |
| Group B (n = 47) | 58.00 (58.00–63.00)a) | 2.00 (2.00–3.00)a) | 10.15 (7.03–16.20)a) | 1.00 (0.00–2.00) |
| Group C (n = 44) | 58.00 (54.25–63.00)a) | 2.00 (1.25–3.00)a) | 11.18 (6.70–15.08)a) | 1.00 (0.00–1.00) |
| Group D (n = 45) | 53.00 (53.00–55.50)a,b,c) | 1.00 (1.00–1.50)a,b,c) | 25.75 (18.71–31.59)a,b,c) | 0.00 (0.00–1.00)a) |
| H | 63.942 | 63.942 | 111.807 | 25.649 |
| Raw p-value | < 0.001 | < 0.001 | < 0.001 | < 0.001 |
| Holm-adjusted p-value | < 0.001 | < 0.001 | < 0.001 | < 0.001 |
Values are presented as median (interquartile range).
Group A, control; Group B, LIA; Group C, ESPB; Group D, ESPB+LIA; LIA, local infiltration analgesia; ESPB, erector spinae plane block; PCIA, patient-controlled intravenous analgesia.
Overall comparisons were performed using the Kruskal-Wallis test, with Holm adjustment across the 4 postoperative analgesia outcomes.
Pairwise comparisons were performed using Mann-Whitney U-tests, with Holm-Bonferroni correction applied globally across all 24 comparisons (4 outcomes×6 group pairs). For time to first PCIA bolus, patients who did not receive a PCIA bolus within 48 hours were assigned a value of 48 hours in the primary restricted-time analysis. A right-censoring sensitivity analysis is reported in Supplementary Table 17.
Superscripts indicate significant favorable pairwise differences after correction:
Adjusted p<0.05 versus group A.
Adjusted p<0.05 versus group B.
Adjusted p<0.05 versus group C.
Multivariable sensitivity analyses adjusting for surgical segments, operative time, and preoperative VAS score showed that the group effect remained significant for all 4 postoperative analgesia requirement outcomes: time to first PCIA bolus (Wald χ2=223.969, df=3, p<0.001), postoperative rescue analgesia count (Wald χ2=37.774, df=3, p<0.001), total postoperative sufentanil consumption (Wald χ2=94.909, df=3, p<0.001), and analgesia pump activations (Wald χ2=94.909, df=3, p<0.001). Surgical segments were significantly associated with postoperative rescue analgesia count (Wald χ2=28.392, df=1, p<0.001), whereas operative time was not significant in any model and preoperative VAS showed only borderline associations with sufentanil consumption and pump activations (Supplementary Table 15).
In the 2×2 factorial analysis, the ESPB×LIA interaction was significant only for time to first PCIA bolus. The unadjusted interaction estimate was 9.710 hours (95% confidence interval [CI], 5.700–13.719; Holm-adjusted p<0.001), and the adjusted estimate remained significant after controlling for surgical segments, operative time, and preoperative VAS score (9.631 hours; 95% CI, 5.575–13.687; Holm-adjusted p<0.001), indicating a greater-than-additive favorable prolongation in the combined ESPB with LIA group. No significant ESPB×LIA interaction remained after Holm correction for postoperative rescue analgesia count, total postoperative sufentanil consumption, or analgesia pump activations (Supplementary Table 16).
Because patients who did not receive a PCIA bolus within 48 h were assigned a value of 48 hours in the primary restricted-time analysis, an additional right-censoring sensitivity analysis was performed. Seven of 45 patients (15.6%) in group D did not receive a PCIA bolus within 48 hours and were administratively censored at 48 hours. No censoring occurred in groups A–C. Kaplan-Meier analysis showed a significant overall difference in time to first PCIA bolus among the 4 groups (log-rank χ2=196.043, df=3, p<0.001). After Holm-Bonferroni correction, groups B and C had significantly longer times to first PCIA bolus than group A. Group D also had a significantly longer time to first PCIA bolus than groups A–C (all adjusted p<0.001). Groups B and C did not differ significantly (adjusted p=0.891). In the adjusted Cox model, the overall group effect remained significant (likelihood-ratio χ2=178.809, df=3, p<0.001). Compared with group A, the hazard of receiving the first PCIA bolus was lower in group B (hazard ratio [HR], 0.153; 95% CI, 0.093–0.252), group C (HR, 0.150; 95% CI, 0.090–0.248), and group D (HR, 0.016; 95% CI, 0.008–0.032). Group D also had a lower hazard of receiving the first PCIA bolus than group B (HR, 0.107; 95% CI, 0.060–0.188) and group C (HR, 0.109; 95% CI, 0.062–0.190). Groups B and C did not differ significantly. These findings were consistent with the primary analysis (Supplementary Table 17).
Overall, the combined ESPB with LIA regimen was associated with lower postoperative analgesic requirements and a longer time to first PCIA bolus than the control and single-intervention regimens. The factorial analysis demonstrated a greater-than-additive effect only for time to first PCIA bolus, while the right-censoring sensitivity analysis confirmed the robustness of this finding. For the remaining analgesia requirement outcomes, the results support an incremental multimodal analgesic benefit rather than consistent statistical synergy.
6. Comparison of Early Postoperative Recovery Indicators
As shown in Table 6 and, for ambulation and hospital stay, Fig. 6, early postoperative recovery indicators were generally similar among the 4 groups. Time to first ambulation did not differ significantly, with a median of 2 days in all groups (H=3.612, Holm-adjusted p=0.613). Postoperative hospital stay was also comparable, with a median of 4 days across the groups (H=2.836, Holm-adjusted p=0.613). Patient satisfaction differed significantly among the 4 groups (χ2=12.558, Holm-adjusted p=0.017), with the highest satisfaction rate observed in group D. In pairwise comparisons, only the difference between groups D and A remained statistically significant after Holm-Bonferroni correction (adjusted p=0.009). No other pairwise differences in satisfaction were statistically significant after correction (Supplementary Table 18).
Table 6.
Comparison of early postoperative recovery indicators
| Time to first ambulation (days) | Postoperative hospital stay (days) | Satisfaction rate | |
|---|---|---|---|
| Group A (n = 46) | 2.00 (2.00–3.00) | 4.00 (3.00–5.00) | 29/46 (63.04) |
| Group B (n = 47) | 2.00 (2.00–3.00) | 4.00 (3.00–5.00) | 39/47 (82.98) |
| Group C (n = 44) | 2.00 (2.00–3.00) | 4.00 (3.00–6.00) | 37/44 (84.09) |
| Group D (n = 45) | 2.00 (2.00–3.00) | 4.00 (3.50–5.00) | 41/45 (91.11)a) |
| H/χ2 | 3.612† | 2.836† | 12.558‡ |
| Raw p-value | 0.306 | 0.418 | 0.006 |
| Holm-adjusted p-value | 0.613 | 0.613 | 0.017 |
Values are presented as median (interquartile range) or number (%).
Group A, control; Group B, LIA; Group C, ESPB; Group D, ESPB+LIA; LIA, local infiltration analgesia; ESPB, erector spinae plane block.
Overall comparisons were performed using the Kruskal-Wallis test for continuous outcomes and Pearson χ2 test for satisfaction rate, with Holm adjustment across the 3 outcomes. Pairwise comparisons for satisfaction rate were performed using Pearson χ2 tests, with Holm-Bonferroni correction across the 6 group comparisons.
Holm-adjusted p<0.05 versus group A.
H statistic.
χ2 statistic.
Fig. 6.

Comparison of early postoperative recovery indicators among the 4 groups. (A) Time to first ambulation. (B) Postoperative hospital stay. Overall comparisons were performed using the Kruskal-Wallis test, with Holm adjustment across the 3 early postoperative recovery outcomes. Holm-adjusted p-values are shown in the figure. Group A, control; Group B, LIA; Group C, ESPB; Group D, ESPB+LIA; LIA, local infiltration analgesia; ESPB, erector spinae plane block.
7. Comparison of Postoperative Adverse Events
Postoperative adverse-event rates were similar among the 4 groups, with no significant between-group differences (all p>0.05) (Table 7). Although postoperative nausea and vomiting occurred in some patients, all cases were mild and improved after antiemetic treatment. No patient discontinued PCIA early because of nausea, vomiting, or other opioid-related adverse events. Therefore, all PCIA pumps were continued for 48 hours postoperatively, with no protocol deviations or missing PCIA data related to PCIA intolerance. No surgical site infections or cases of local anesthetic toxicity were observed.
Table 7.
Comparison of postoperative adverse-event rates among 4 patient groups
| Variable | Group A (n = 46) | Group B (n = 47) | Group C (n = 44) | Group D (n = 45) | χ2/Fisher | p-value |
|---|---|---|---|---|---|---|
| Nausea | 12 (26.09) | 10 (21.28) | 8 (18.18) | 5 (11.11) | 3.460† | 0.326 |
| Vomiting | 4 (8.70) | 5 (10.64) | 6 (13.64) | 3 (6.67) | 1.337‡ | 0.720 |
| Pruritus | 1 (2.17) | 0 (0) | 0 (0) | 1 (2.22) | 2.135‡ | 0.740 |
| Drowsiness | 0 (0) | 1 (2.13) | 0 (0) | 0 (0) | 2.757‡ | 1.000 |
| Respiratory depression | 0 (0) | 0 (0) | 0 (0) | 0 (0) | - | - |
| Bradycardia | 0 (0) | 0 (0) | 0 (0) | 0 (0) | - | - |
| Hypotension | 0 (0) | 0 (0) | 0 (0) | 0 (0) | - | - |
| Surgical site infection | 0 (0) | 0 (0) | 0 (0) | 0 (0) | - | - |
| Urinary retention | 0 (0) | 0 (0) | 0 (0) | 0 (0) | - | - |
| Motor block | 0 (0) | 0 (0) | 0 (0) | 0 (0) | - | - |
Values are presented as number (%).
Group A, control; Group B, LIA; Group C, ESPB; Group D, ESPB+LIA; LIA, local infiltration analgesia; ESPB, erector spinae plane block.
Overall comparisons were performed using Pearson χ2 test or the Fisher-Freeman-Halton exact test, as appropriate.
Comparisons were not performed for adverse events with no cases in any group, as indicated by “-”.
Pearson χ2 test.
Fisher-Freeman-Halton exact test.
8. Comparison of VAS and ODI Scores at 3 Months After Surgery
As shown in Table 8 and Fig. 7, resting pain VAS scores, movement pain VAS scores, and ODI scores at 3 months were similar among the 4 groups. Resting pain VAS scores did not differ significantly, with median values of 2 (1–2) in groups A, B, and C and 1 (1–2) in group D (H=2.166, p=0.539). Movement pain VAS scores were also comparable, with median values of 2 (2–3) in groups A and B and 2 (1–3) in groups C and D (H=1.661, p=0.646). ODI scores showed no significant between-group difference, with mean values of 19.57±7.55, 18.68±7.07, 17.77±6.22, and 17.20±6.72 in groups A, B, C, and D, respectively (F=1.024, p=0.383).
Table 8.
Comparison of VAS and ODI scores among the 4 groups at 3 months after surgery
| Variable | VAS |
ODI | |
|---|---|---|---|
| Resting | Movement | ||
| Group A (n = 46) | 2.00 (1.00–2.00) | 2.00 (2.00–3.00) | 19.57 ± 7.55 |
| Group B (n = 47) | 2.00 (1.00–2.00) | 2.00 (2.00–3.00) | 18.68 ± 7.07 |
| Group C (n = 44) | 2.00 (1.00–2.00) | 2.00 (1.00–3.00) | 17.77 ± 6.22 |
| Group D (n = 45) | 1.00 (1.00–2.00) | 2.00 (1.00–3.00) | 17.20 ± 6.72 |
| H/F | 2.166† | 1.661† | 1.024‡ |
| p-value | 0.539 | 0.646 | 0.383 |
Values are presented as median (interquartile range) or mean±standard deviation.
VAS, visual analogue scale; ODI, Oswestry Disability Index; Group A, control; Group B, LIA; Group C, ESPB; Group D, ESPB+LIA; LIA, local infiltration analgesia; ESPB, erector spinae plane block.
Overall comparisons were performed using the Kruskal-Wallis test for VAS scores and 1-way analysis of variance for ODI scores. Pairwise comparisons were not performed because no overall between-group differences were statistically significant.
H statistic.
F statistic.
Fig. 7.

Comparison of VAS and ODI scores among the 4 groups at 3 months after surgery. (A) Resting pain VAS score. (B) Movement pain VAS score. (C) ODI score. Boxes indicate the median and interquartile range, whiskers indicate the minimum and maximum values, and dots represent individual observations. Overall comparisons were performed using the Kruskal-Wallis test for VAS scores and 1-way analysis of variance for ODI scores. VAS, visual analogue scale; ODI, Oswestry Disability Index; Group A, control; Group B, LIA; Group C, ESPB; Group D, ESPB+LIA; LIA, local infiltration analgesia; ESPB, erector spinae plane block.
DISCUSSION
Traditional PLIF requires a posterior surgical approach that may injure the paraspinal muscles and require retraction of the thecal sac and nerve roots [40]. Pain after major spine surgery arises not only from the surgical incision but also from deeper structures, including muscles, ligaments, bone, intervertebral discs, facet joints, and injured nerve roots [41]. Patients undergoing lumbar fusion are therefore at substantial risk of acute postoperative pain, particularly during mobilization, which makes effective perioperative analgesia important for recovery [2]. Opioids remain effective for postoperative pain control, but their use is limited by adverse effects such as nausea, vomiting, pruritus, sedation, urinary retention, and respiratory depression [42]. The ERAS framework supports multimodal and opioid-sparing analgesic strategies to reduce these concerns [11,43]. Recent meta-analyses have also supported the analgesic efficacy of ESPB in lumbar spine surgery [17,18]. Against this background, the main contribution of the present study is not simply to confirm the efficacy of ESPB, but to provide a prospective, double-blind, randomized, placebo-controlled, four-arm comparison of placebo, LIA alone, ESPB alone, and combined ESPB with LIA in patients undergoing PLIF.
Combined ESPB with LIA provided favorable short-term postoperative analgesia in this study. Compared with the control and single-intervention groups, the combined group had lower resting and movement pain VAS scores at several early postoperative time points, lower postoperative sufentanil consumption, fewer analgesia pump activations, and a longer time to first PCIA bolus. Repeated-measures mixed-effects models supported these findings, with significant group-by-time interactions for both resting and movement pain VAS scores, indicating that pain changed differently over time across the 4 groups. The results remained robust after adjustment for surgical segments, operative time, and preoperative VAS score. Holm-Bonferroni correction was applied to pairwise comparisons to reduce the risk of type I error from multiple testing, and several between-group differences remained significant after correction. Although some time-specific differences in VAS scores were statistically significant, their magnitude was modest, and not all reached the predefined MCID threshold. The findings therefore support a short-term analgesic benefit but do not indicate consistent clinical superiority across all postoperative time points.
The 4-arm design also allowed the ESPB×LIA interaction to be examined directly. In the 2×2 factorial analyses, ESPB and LIA were treated as binary factors, and the interaction term was tested explicitly. The results did not show consistent evidence of statistical synergy across VAS outcomes after adjustment for multiple testing. For resting pain VAS, the significant early interaction estimates were positive, indicating a less-than-additive effect rather than synergy. For movement pain VAS, the interaction estimates at 24 and 48 hours were negative, but they did not remain significant after Holm-Bonferroni correction. A significant ESPB×LIA interaction was found for time to first PCIA bolus, suggesting a greater-than-additive prolongation of time to first PCIA bolus. The right-censoring sensitivity analysis produced consistent findings and further supported the longer time to first PCIA bolus in the combined group. Overall, combined ESPB with LIA appears to provide an incremental and layered multimodal analgesic effect. Evidence of statistical synergy, however, was limited to time to first PCIA bolus and was not consistent across the other outcomes.
One possible explanation for this incremental benefit is that ESPB and LIA act at different anatomical levels. ESPB delivers local anesthetic into the fascial plane deep to the erector spinae muscle near the transverse process. Mechanistic and anatomical studies have shown consistent involvement of the dorsal rami, while spread to other neural structures appears more variable [15,44]. LIA acts more locally by delivering anesthetic directly to tissues around the surgical wound at the time of closure, and previous evidence suggests that it can reduce early postoperative analgesic requirements after lumbar spine surgery [20]. Combining ESPB with LIA may therefore provide complementary coverage of different nociceptive sources after open PLIF. This explanation is mechanistic and should not be taken as direct evidence of pharmacological or statistical synergy.
The present findings should also be viewed in the context of mixed evidence on ESPB in spine surgery. Some randomized trials have reported lower postoperative opioid consumption or less early pain after lumbar spine surgery, while others have found limited or no additional benefit for their primary analgesic outcomes [45-47]. These differences may be related to variation in surgical procedures, block timing and technique, background multimodal analgesia, comparators, and outcome selection [8,12,45-47]. In a randomized trial of non-instrumented lumbar surgery conducted within an ERAS pathway, Lonjon et al. [45] found no additional analgesic benefit of ESPB compared with placebo. Another randomized trial in patients undergoing instrumented lumbar spinal surgery found that ESPB reduced opioid consumption compared with wound infiltration [46]. More recently, Zamudio et al. [47] reported that postoperative ESPB improved some early pain and recovery measures after lumbar fusion but did not significantly reduce 24- or 48-hour morphine consumption. Reviews of regional analgesia in spine surgery have also highlighted substantial variation in techniques, comparators, and reported outcomes [8,12]. In open PLIF, where surgical injury involves several tissue layers, a combined regional strategy may offer additional value, although this remains a procedure-specific hypothesis rather than an established effect modifier. Our postoperative analgesia requirement findings are consistent with this view. The combined regimen showed its clearest advantage in prolonging time to first PCIA bolus, while the reductions in sufentanil consumption and pump activation are better viewed as incremental multimodal benefits rather than evidence of statistical synergy.
Patient satisfaction was highest in the combined ESPB with LIA group, and the overall comparison among the 4 groups was significant. After adjustment for multiple comparisons, only the difference between the combined group and the control group remained significant. Early postoperative recovery outcomes should still be interpreted with caution. Time to first ambulation and postoperative hospital stay did not differ significantly among the 4 groups. These outcomes depend on more than pain intensity and may also be influenced by ward routines, rehabilitation protocols, drain management, surgeon preference, patient confidence, and institutional discharge criteria. The present data support better acute postoperative analgesia and greater patient-reported satisfaction, but they do not show a clear reduction in time to ambulation or hospital discharge. The rates of nausea, vomiting, pruritus, drowsiness, respiratory depression, bradycardia, hypotension, surgical site infection, urinary retention, and motor block were also similar among the 4 groups.
The early analgesic advantage of combined ESPB with LIA did not lead to detectable differences in pain intensity or functional recovery at the 3-month follow-up. Resting VAS scores, movement VAS scores, and ODI scores were similar among the 4 groups at 3 months. This finding suggests that the benefit of the combined regimen was mainly limited to the acute postoperative period and was not associated with measurable differences in pain or disability at later follow-up. Its clinical value should therefore be described as an improvement in short-term postoperative analgesia rather than long-term functional superiority. Any possible effect on chronic postsurgical pain should also be interpreted cautiously. Poor control of acute postoperative pain has been considered a possible risk factor for persistent postsurgical pain, but the present study was not designed to evaluate chronic pain trajectories. Follow-up was limited to 3 months, and no serial outpatient assessments were performed during the 1- to 4-week postoperative period or beyond 3 months. Because VAS and ODI scores at 3 months did not differ significantly among groups, the present data do not show that combined ESPB with LIA prevents chronic postsurgical pain or improves long-term functional recovery after PLIF. Longer followup with repeated assessments will be needed to determine whether early analgesic benefits have any lasting effect on chronic pain, functional recovery, or quality of life.
Several limitations should be considered. This was a single-center trial conducted within a standardized perioperative pathway, so multicenter studies are needed to determine whether the findings can be applied more broadly. The study also included patients undergoing 1- to 3-level PLIF. Surgical segments were comparable among groups and were included in the adjusted analyses, but some residual variation in surgical extent and nociceptive burden may remain. Pain was assessed during the first 48 hours and again at 3 months, leaving the intermediate recovery period unexamined. Time to first PCIA bolus was also observed only during the 48-hour postoperative period, so events occurring after this point could not be assessed. The right-censoring sensitivity analysis, however, produced results consistent with the primary analysis. The primary analyses were based on an mITT population rather than a strict ITT population, and the 6 postrandomization exclusions may have introduced some attrition bias. Sensory differences after regional anesthesia may also have affected blinding despite protocolized sensory testing. Because the study focused on open PLIF, it also remains uncertain whether the combined ESPB with LIA regimen would produce similar effects in minimally invasive lumbar fusion procedures.
CONCLUSION
Fluoroscopy-guided ESPB combined with LIA improved short-term postoperative analgesia after open PLIF, reduced postoperative opioid requirements and analgesia pump activations, and prolonged the time to first PCIA bolus without increasing adverse events. The combined regimen should be interpreted as an incremental layered multimodal analgesic strategy rather than definitive statistical synergy. No significant differences were observed in 3-month VAS or ODI outcomes.
Footnotes
Conflict of Interest
The authors have nothing to disclose.
Funding/Support
This study received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Author Contribution
Conceptualization: MYF, SWW, LL; Data curation: YHS; Formal analysis: FZ, RRZ, WPR, YKL; Methodology: MYF, FZ, XCW, SWW, LL; Project administration: SWW, LL; Visualization: MYF, YHS, JX; Writing – original draft: MYF, FZ; Writing – review & editing: PCL, XCW, SWW, LL.
Supplementary Materials
Supplementary Tables 1-18 are available at https://doi.org/10.14245/ns.26520656.0328.
Pairwise Comparisons of Body-Weight- and Operative-Time-Normalized Anesthetic Drug Consumption Rates and Intraoperative Rescue Analgesia Rate With Unadjusted and Holm-Bonferroni-Adjusted P Values
All 30 Pairwise Comparisons of Postoperative Resting Pain VAS Scores With Global Holm-Bonferroni Correction
Unadjusted and Adjusted Linear Mixed-Effects Models for Repeated Postoperative Resting Pain VAS Scores
Longitudinal 2×2 Factorial Mixed-Effects Analysis of ESPB and LIA for Postoperative Resting Pain VAS Scores
Time-Specific 2×2 Factorial Interaction Estimates of ESPB and LIA for Postoperative Resting Pain VAS Scores
MCID Analysis of Resting Pain VAS Scores in Active Treatment Groups Versus the Control Group
Detailed Pairwise MCID Analysis for Resting Pain VAS Scores
All 30 Pairwise Comparisons of Postoperative Movement Pain VAS Scores With Global Holm-Bonferroni Correction
Unadjusted and Adjusted Linear Mixed-Effects Models for Repeated Postoperative Movement Pain VAS Scores
Longitudinal 2×2 Factorial Mixed-Effects Analysis of ESPB and LIA for Postoperative Movement Pain VAS Scores
Time-Specific 2×2 Factorial Interaction Estimates of ESPB and LIA for Postoperative Movement Pain VAS Scores
MCID Analysis of Movement Pain VAS Scores in Active Treatment Groups Versus the Control Group
Detailed Pairwise MCID Analysis for Movement Pain VAS Scores
Pairwise Comparisons of Postoperative Analgesia Requirements With Holm-Bonferroni Correction
Adjusted Multivariable Sensitivity Analysis of Postoperative Analgesia Requirements
2×2 Factorial ESPB×LIA Interaction Estimates for Postoperative Analgesia Requirements
Right-Censoring Sensitivity Analysis of Time to First PCIA Bolus Within 48 Hours
Pairwise Comparisons of Patient Satisfaction Rates With Holm-Bonferroni Correction
REFERENCES
- 1.Bamps S, Raymaekers V, Roosen G, et al. Lateral lumbar interbody fusion (direct lateral interbody fusion/extreme lateral interbody fusion) versus posterior lumbar interbody fusion surgery in spinal degenerative disease: a systematic review. World Neurosurg. 2023;171:10–8. doi: 10.1016/j.wneu.2022.12.033. [DOI] [PubMed] [Google Scholar]
- 2.Geisler A, Zachodnik J, Køppen K, et al. Postoperative pain treatment after spinal fusion surgery: a systematic review with meta-analyses and trial sequential analyses. Pain Rep. 2022;7:e1005. doi: 10.1097/PR9.0000000000001005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Wu Q, Cui X, Guan LC, et al. Chronic pain after spine surgery: insights into pathogenesis, new treatment, and preventive therapy. J Orthop Translat. 2023;42:147–59. doi: 10.1016/j.jot.2023.07.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Zaed I, Bossi B, Ganau M, et al. Current state of benefits of Enhanced Recovery After Surgery (ERAS) in spinal surgeries: a systematic review of the literature. Neurochirurgie. 2022;68:61–8. doi: 10.1016/j.neuchi.2021.04.007. [DOI] [PubMed] [Google Scholar]
- 5.Prabhakar NK, Chadwick AL, Nwaneshiudu C, et al. Management of postoperative pain in patients following spine surgery: a narrative review. Int J Gen Med. 2022;15:4535–49. doi: 10.2147/IJGM.S292698. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Wheeler M, Oderda GM, Ashburn MA, et al. Adverse events associated with postoperative opioid analgesia: a systematic review. J Pain. 2002;3:159–80. doi: 10.1054/jpai.2002.123652. [DOI] [PubMed] [Google Scholar]
- 7.Dunn LK, Yerra S, Fang S, et al. Incidence and risk factors for chronic postoperative opioid use after major spine surgery: a cross-sectional study with longitudinal outcome. Anesth Analg. 2018;127:247–54. doi: 10.1213/ANE.0000000000003338. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ní Eochagáin A, Singleton BN, Moorthy A, et al. Regional and neuraxial anaesthesia techniques for spinal surgery: a scoping review. Br J Anaesth. 2022;129:598–611. doi: 10.1016/j.bja.2022.05.028. [DOI] [PubMed] [Google Scholar]
- 9.Cook TM, Counsell D, Wildsmith JA, et al. Major complications of central neuraxial block: report on the Third National Audit Project of the Royal College of Anaesthetists. Br J Anaesth. 2009;102:179–90. doi: 10.1093/bja/aen360. [DOI] [PubMed] [Google Scholar]
- 10.Kopp SL, Vandermeulen E, McBane RD, et al. Regional anesthesia in the patient receiving antithrombotic or thrombolytic therapy: American Society of Regional Anesthesia and Pain Medicine Evidence-Based Guidelines (fifth edition) Reg Anesth Pain Med. doi: 10.1136/rapm-2024-105766. 2025 Oct 17:rapm-2024-105766. doi: 10.1136/rapm-2024-105766. [Epub] [DOI] [PubMed] [Google Scholar]
- 11.Licina A, Silvers A. Perioperative multimodal analgesia for adults undergoing surgery of the spine-a systematic review and meta-analysis of three or more modalities. World Neurosurg. 2022;163:11–23. doi: 10.1016/j.wneu.2022.03.098. [DOI] [PubMed] [Google Scholar]
- 12.Bae S, Alboog A, Esquivel KS, et al. Efficacy of perioperative pharmacological and regional pain interventions in adult spine surgery: a network meta-analysis and systematic review of randomised controlled trials. Br J Anaesth. 2022;128:98–117. doi: 10.1016/j.bja.2021.08.034. [DOI] [PubMed] [Google Scholar]
- 13.Forero M, Adhikary SD, Lopez H, et al. The erector spinae plane block: a novel analgesic technique in thoracic neuropathic pain. Reg Anesth Pain Med. 2016;41:621–7. doi: 10.1097/AAP.0000000000000451. [DOI] [PubMed] [Google Scholar]
- 14.Gonçalves JPF, Duran ML, Barreto ESR, et al. Efficacy of erector spinae plane block for postoperative pain management: a meta-analysis and trial sequential analysis of randomized controlled trials. J Clin Anesth. 2025;103:111831. doi: 10.1016/j.jclinane.2025.111831. [DOI] [PubMed] [Google Scholar]
- 15.Chin KJ, El-Boghdadly K. Mechanisms of action of the erector spinae plane (ESP) block: a narrative review. Can J Anaesth. 2021;68:387–408. doi: 10.1007/s12630-020-01875-2. [DOI] [PubMed] [Google Scholar]
- 16.Yayik AM, Cesur S, Ozturk F, et al. Postoperative analgesic efficacy of the ultrasound-guided erector spinae plane block in patients undergoing lumbar spinal decompression surgery: a randomized controlled study. World Neurosurg. 2019;126:e779–85. doi: 10.1016/j.wneu.2019.02.149. [DOI] [PubMed] [Google Scholar]
- 17.Muthu S, Viswanathan VK, Annamalai S, et al. Bilateral erector spinae plane block for postoperative pain relief in lumbar spine surgery: a PRISMA-compliant updated systematic review & meta-analysis. World Neurosurg X. 2024;23:100360. doi: 10.1016/j.wnsx.2024.100360. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Ye J, Wu J, Zhou Y, et al. Efficacy of erector spinae plane block in lumbar fusion surgery: a systematic review and meta-analysis on opioid-sparing and postoperative pain outcomes. J Pain Res. 2026;19:573516. doi: 10.2147/JPR.S573516. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kjærgaard M, Møiniche S, Olsen KS. Wound infiltration with local anesthetics for post-operative pain relief in lumbar spine surgery: a systematic review. Acta Anaesthesiol Scand. 2012;56:282–90. doi: 10.1111/j.1399-6576.2011.02629.x. [DOI] [PubMed] [Google Scholar]
- 20.Perera AP, Chari A, Kostusiak M, et al. Intramuscular local anesthetic infiltration at closure for postoperative analgesia in lumbar spine surgery: a systematic review and meta-analysis. Spine (Phila Pa 1976) 2017;42:1088–95. doi: 10.1097/BRS.0000000000001443. [DOI] [PubMed] [Google Scholar]
- 21.Zhang Z, Zhu RL, Yue L, et al. Bilateral ultrasound-guided erector spinae plane block versus wound infiltration for postoperative analgesia in lumbar spinal fusion surgery: a randomized controlled trial. Eur Spine J. 2023;32:301–12. doi: 10.1007/s00586-022-07453-y. [DOI] [PubMed] [Google Scholar]
- 22.Lennon MJ, Isaac S, Currigan D, et al. Erector spinae plane block combined with local infiltration analgesia for total hip arthroplasty: a randomized, placebo controlled, clinical trial. J Clin Anesth. 2021;69:110153. doi: 10.1016/j.jclinane.2020.110153. [DOI] [PubMed] [Google Scholar]
- 23.Jahangiri P, Bagherzadeh S, Roohollahi F, et al. Fluoroscopy-guided erector spinae plane block reduces early postoperative pain following lumbar fusion: a randomized controlled trial. Global Spine J. 2026;16:1232–9. doi: 10.1177/21925682251378470. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Beltrame SA, Fasano F, Jalón P. Bilateral radioscopically guided erector spinae plane block for postoperative analgesia in spine surgery: a randomized clinical trial. J Neurol Surg A Cent Eur Neurosurg. 2023;84:360–9. doi: 10.1055/a-1768-3797. [DOI] [PubMed] [Google Scholar]
- 25.Hakimoğlu S, Özdemir T, Çömez MS, et al. Comparison of the postoperative effects of the erector spinae plane block and local infiltration analgesia in patients operated with lumbotomy surgery incision: randomized clinical study. Medicine (Baltimore) 2024;103:e39054. doi: 10.1097/MD.0000000000039054. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Hopewell S, Chan AW, Collins GS, et al. CONSORT 2025 statement: updated guideline for reporting randomised trials. BMJ. 2025;389:e081123. doi: 10.1136/bmj-2024-081123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Fairbank JC, Pynsent PB. The Oswestry Disability Index. Spine (Phila Pa 1976) 2000;25:2940–52. doi: 10.1097/00007632-200011150-00017. discussion 2952. [DOI] [PubMed] [Google Scholar]
- 28.Delgado DA, Lambert BS, Boutris N, et al. Validation of digital visual analog scale pain scoring with a traditional paper-based visual analog scale in adults. J Am Acad Orthop Surg Glob Res Rev. 2018;2:e088. doi: 10.5435/JAAOSGlobal-D-17-00088. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Juszczak E, Altman DG, Hopewell S, et al. Reporting of multiarm parallel-group randomized trials: extension of the CONSORT 2010 Statement. JAMA. 2019;321:1610–20. doi: 10.1001/jama.2019.3087. [DOI] [PubMed] [Google Scholar]
- 30.Abraha I, Montedori A. Modified intention to treat reporting in randomised controlled trials: systematic review. BMJ. 2010;340:c2697. doi: 10.1136/bmj.c2697. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Montedori A, Bonacini MI, Casazza G, et al. Modified versus standard intention-to-treat reporting: are there differences in methodological quality, sponsorship, and findings in randomized trials? a cross-sectional study. Trials. 2011;12:58. doi: 10.1186/1745-6215-12-58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Gueorguieva R, Krystal JH. Move over ANOVA: progress in analyzing repeated-measures data and its reflection in papers published in the Archives of General Psychiatry. Arch Gen Psychiatry. 2004;61:310–7. doi: 10.1001/archpsyc.61.3.310. [DOI] [PubMed] [Google Scholar]
- 33.Montgomery AA, Peters TJ, Little P. Design, analysis and presentation of factorial randomised controlled trials. BMC Med Res Methodol. 2003;3:26. doi: 10.1186/1471-2288-3-26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Kahan BC, Hall SS, Beller EM, et al. Reporting of factorial randomized trials: extension of the CONSORT 2010 Statement. JAMA. 2023;330:2106–14. doi: 10.1001/jama.2023.19793. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Thabane L, Mbuagbaw L, Zhang S, et al. A tutorial on sensitivity analyses in clinical trials: the what, why, when and how. BMC Med Res Methodol. 2013;13:92. doi: 10.1186/1471-2288-13-92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Schober P, Vetter TR. Survival analysis and interpretation of time-to-event data: the Tortoise and the Hare. Anesth Analg. 2018;127:792–8. doi: 10.1213/ANE.0000000000003653. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Katz MH, Hauck WW. Proportional hazards (Cox) regression. J Gen Intern Med. 1993;8:702–11. doi: 10.1007/BF02598295. [DOI] [PubMed] [Google Scholar]
- 38.Aickin M, Gensler H. Adjusting for multiple testing when reporting research results: the Bonferroni vs Holm methods. Am J Public Health. 1996;86:726–8. doi: 10.2105/ajph.86.5.726. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Myles PS, Myles DB, Galagher W, et al. Measuring acute postoperative pain using the visual analog scale: the minimal clinically important difference and patient acceptable symptom state. Br J Anaesth. 2017;118:424–9. doi: 10.1093/bja/aew466. [DOI] [PubMed] [Google Scholar]
- 40.Mobbs RJ, Phan K, Malham G, et al. Lumbar interbody fusion: techniques, indications and comparison of interbody fusion options including PLIF, TLIF, MI-TLIF, OLIF/ATP, LLIF and ALIF. J Spine Surg. 2015;1:2–18. doi: 10.3978/j.issn.2414-469X.2015.10.05. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Sharma S, Balireddy RK, Vorenkamp KE, et al. Beyond opioid patient-controlled analgesia: a systematic review of analgesia after major spine surgery. Reg Anesth Pain Med. 2012;37:79–98. doi: 10.1097/AAP.0b013e3182340869. [DOI] [PubMed] [Google Scholar]
- 42.Vandermeulen E. Systemic analgesia and co-analgesia. Acta Anaesthesiol Belg. 2006;57:113–20. [PubMed] [Google Scholar]
- 43.Debono B, Wainwright TW, Wang MY, et al. Consensus statement for perioperative care in lumbar spinal fusion: Enhanced Recovery After Surgery (ERAS®) Society recommendations. Spine J. 2021;21:729–52. doi: 10.1016/j.spinee.2021.01.001. [DOI] [PubMed] [Google Scholar]
- 44.Kokar S, Ertaş A, Mercan Ö, et al. The lumbar erector spinae plane block: a cadaveric study. Turk J Med Sci. 2022;52:229–36. doi: 10.3906/sag-2107-83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Lonjon G, Ross N, Dhenin A, et al. Non-efficacy of ultrasound-guided erector spinae plane block in postoperative analgesia of non-instrumented spine surgery: a randomized, double-blind, controlled trial. Brain Spine. 2025;5:104379. doi: 10.1016/j.bas.2025.104379. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Yuce Y, Karakus SA, Simsek T, et al. Comparative efficacy of ultrasound-guided erector spinae plane block versus wound infiltration for postoperative analgesia in instrumented lumbar spinal surgeries. BMC Anesthesiol. 2024;24:374. doi: 10.1186/s12871-024-02754-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Zamudio D, Fernández L, Rodríguez A, et al. Postoperative erector spinae plane block does not reduce morphine consumption after lumbar spinal fusion: a randomized controlled trial. J Clin Anesth. 2025;107:112022. doi: 10.1016/j.jclinane.2025.112022. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Pairwise Comparisons of Body-Weight- and Operative-Time-Normalized Anesthetic Drug Consumption Rates and Intraoperative Rescue Analgesia Rate With Unadjusted and Holm-Bonferroni-Adjusted P Values
All 30 Pairwise Comparisons of Postoperative Resting Pain VAS Scores With Global Holm-Bonferroni Correction
Unadjusted and Adjusted Linear Mixed-Effects Models for Repeated Postoperative Resting Pain VAS Scores
Longitudinal 2×2 Factorial Mixed-Effects Analysis of ESPB and LIA for Postoperative Resting Pain VAS Scores
Time-Specific 2×2 Factorial Interaction Estimates of ESPB and LIA for Postoperative Resting Pain VAS Scores
MCID Analysis of Resting Pain VAS Scores in Active Treatment Groups Versus the Control Group
Detailed Pairwise MCID Analysis for Resting Pain VAS Scores
All 30 Pairwise Comparisons of Postoperative Movement Pain VAS Scores With Global Holm-Bonferroni Correction
Unadjusted and Adjusted Linear Mixed-Effects Models for Repeated Postoperative Movement Pain VAS Scores
Longitudinal 2×2 Factorial Mixed-Effects Analysis of ESPB and LIA for Postoperative Movement Pain VAS Scores
Time-Specific 2×2 Factorial Interaction Estimates of ESPB and LIA for Postoperative Movement Pain VAS Scores
MCID Analysis of Movement Pain VAS Scores in Active Treatment Groups Versus the Control Group
Detailed Pairwise MCID Analysis for Movement Pain VAS Scores
Pairwise Comparisons of Postoperative Analgesia Requirements With Holm-Bonferroni Correction
Adjusted Multivariable Sensitivity Analysis of Postoperative Analgesia Requirements
2×2 Factorial ESPB×LIA Interaction Estimates for Postoperative Analgesia Requirements
Right-Censoring Sensitivity Analysis of Time to First PCIA Bolus Within 48 Hours
Pairwise Comparisons of Patient Satisfaction Rates With Holm-Bonferroni Correction
