Efficacy of Erector Spinae Plane Block in Lumbar Fusion Surgery: A Systematic Review and Meta-Analysis on Opioid-Sparing and Postoperative Pain Outcomes

Abstract

Purpose

Lumbar fusion surgery is commonly associated with severe postoperative pain and high opioid requirements. The erector spinae plane block (ESPB) has demonstrated promising analgesic effects in spine procedures; however, its efficacy specifically in lumbar fusion surgery remains uncertain, given the heterogeneity of the existing evidence. This systematic review and meta-analysis aimed to evaluate the impact of ESPB on perioperative opioid consumption, postoperative pain, and related outcomes in patients undergoing lumbar fusion surgery.

Patients and Methods

Following the PRISMA 2020 statement, PubMed, Embase, and the Cochrane Library were searched from inception to August 2025 for randomized controlled trials (RCTs) comparing ESPB with placebo, sham, or no-block controls in adult patients undergoing lumbar fusion. The primary outcomes were intraoperative and 24-hour postoperative opioid consumption, expressed as intravenous morphine milligram equivalents (MME). Secondary outcomes included postoperative pain scores and the incidence of postoperative nausea and vomiting (PONV). Study quality was assessed using the Cochrane Risk of Bias 2.0 tool, and certainty of evidence was graded using GRADE criteria.

Results

Fifteen RCTs involving 1035 patients were included. ESPB significantly reduced intraoperative opioid use (MD −31.70 mg; 95% CI −41.69 to −21.71) and 24-hour postoperative opioid consumption (MD −11.02 mg; 95% CI −14.42 to −7.63), as well as resting pain scores (MD −0.49; 95% CI −0.77 to −0.22). ESPB was also associated with a lower risk of PONV (RR 0.71; 95% CI 0.53–0.94). Considerable heterogeneity (I² >80%) was observed, largely attributable to variations in study design, comparator rigor, and blinding quality.

Conclusion

Low- to moderate-certainty evidence suggests that ESPB probably reduces perioperative opioid consumption, early postoperative pain, and opioid-related adverse events such as PONV following lumbar fusion surgery. Although substantial heterogeneity and variability in methodological quality warrant cautious interpretation, the consistent direction and clinical relevance of observed effects support ESPB as a promising adjunct within multimodal analgesia pathways. Further well-designed, blinded RCTs with standardized protocols and extended follow-up are needed to confirm long-term effectiveness and refine clinical implementation.

Introduction

Degenerative lumbar disease represents a leading cause of chronic disability and perioperative opioid exposure worldwide. Over the past two decades, lumbar fusion procedures have increased substantially, reflecting advances in surgical techniques and expanding clinical indications.^1,2 Multilevel fusion is increasingly used for complex or multisegmental pathology. Owing to extensive soft tissue dissection, prolonged paraspinal muscle retraction, and substantial osseous manipulation, lumbar fusion is widely regarded as one of the most painful procedures in spine surgery.³ Severe postoperative pain may impede early mobilization, delay rehabilitation, compromise patient satisfaction, and increase the risk of chronic postsurgical pain (CPSP).^4,5

Perioperative analgesia for lumbar fusion has traditionally depended largely on systemic opioids, most commonly via intravenous patient-controlled analgesia (IV-PCA). Although opioids provide effective analgesia, their adverse effects, including postoperative nausea and vomiting (PONV), respiratory depression, ileus, and excessive sedation, may delay recovery and prolong hospitalization.⁶ In addition, high perioperative opioid exposure has been associated with persistent postoperative opioid use, underscoring the need for opioid-sparing strategies. Accordingly, enhanced recovery after surgery (ERAS) pathways increasingly prioritize multimodal analgesia (MMA), positioning regional anesthesia as a key component to optimize analgesia while minimizing opioid-related harm.^7,8

The erector spinae plane block (ESPB), a relatively novel fascial plane block, has gained increasing attention in perioperative pain management. Its advantages include clearly defined sonoanatomy, technical simplicity, and a favorable safety profile due to its distance from the neuraxis and major vascular structures.^9,10 Randomized trials and several meta-analyses suggest that ESPB may improve postoperative analgesia across various spinal procedures.^11,12 However, substantial clinical and methodological heterogeneity across published trials continues to complicate interpretation and limits its integration into standardized perioperative pathways.

Previous meta-analyses have frequently pooled heterogeneous spinal procedures—including decompression and fusion surgeries, as well as single- and multilevel operations—potentially obscuring the true efficacy of ESPB in surgeries associated with severe postoperative pain, such as lumbar fusion.¹³ In addition, variations in ESPB techniques—including injection level (eg, T12 vs L2), local anesthetic dosing strategies, and adjuvant use—further complicate interpretation. Moreover, most existing trials report outcomes limited to the first 24–48 postoperative hours, while data on longer-term patient-centered outcomes, such as functional recovery, quality of recovery (QoR), and chronic postsurgical pain, remain scarce and insufficient for quantitative synthesis. To address these gaps, we conducted a procedure-specific systematic review and meta-analysis focusing exclusively on lumbar fusion surgery. The primary objective was to evaluate the impact of ESPB on perioperative opioid consumption and postoperative pain intensity. Secondary aims included synthesizing available evidence on recovery-related outcomes and longer-term pain-related measures, where reported, to inform evidence-based integration of ESPB into contemporary multimodal analgesia strategies in spine surgery and anesthesiology.

Material and Methods

This systematic review and meta-analysis was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement and was prospectively registered with PROSPERO (registration ID: CRD420251140594).¹⁴

Eligibility Criteria

We included randomized controlled trials (RCTs) evaluating the analgesic effects of ESPB in adult patients undergoing lumbar fusion surgery. Eligible studies compared ESPB against a control group receiving placebo, a sham block, or standard care without any regional analgesic technique. Studies were required to report at least one of the following outcomes: postoperative opioid consumption, pain scores (VAS or NRS), time to first rescue analgesia, or anesthesia-related complications.

We excluded non-randomized studies, conference abstracts without full text, case reports, ongoing trials, animal studies, and trials with incomplete or unavailable data for pooling. Although chronic postsurgical pain (CPSP) is clinically important, it was not prespecified as an analyzable outcome because of insufficient and inconsistent long-term data across trials.

Information Sources and Search Strategy

A comprehensive search of PubMed, Embase, and the Cochrane Library was performed for all relevant studies published up to August 19, 2025, without language restrictions. The search strategy combined free-text terms and Medical Subject Headings (MeSH) related to “erector spinae plane block,” “lumbar fusion,” “postoperative analgesia,” and “opioid consumption.” Additionally, the reference lists of eligible studies and relevant reviews were manually screened to identify additional trials. The full search strategy is provided in Supplementary Table S1.

Study Selection and Data Extraction

Two reviewers (Y.J.J. and W.J.) independently screened titles, abstracts, and full texts against the eligibility criteria. Disagreements were resolved by discussion or adjudicated by a third reviewer (Z.Y).

Using a standardized form, the same two reviewers independently extracted the following data: (1) Study characteristics: first author, publication year, country, and sample size. (2) Patient demographics: age, sex, BMI, and ASA classification. (3) Procedural details: surgical approach, anesthetic technique, and intraoperative opioid use. (4) Block characteristics: timing, vertebral level, local anesthetic (type, concentration, volume), and use of adjuvants. (5) Outcomes: opioid consumption (intraoperative, 24 h, 48 h), resting and dynamic pain scores, time to first rescue analgesia, incidence of postoperative nausea and vomiting (PONV), and any ESPB-related complications.

When data were reported as medians with ranges or interquartile ranges (IQRs), means and standard deviations (SDs) were estimated using the validated methods described by Wan et al and Shi et al^15–18 Opioid doses were standardized to intravenous morphine equivalents (MME) (Supplementary Table S2). Detailed conversion tables applied to intraoperative opioids and 24-hour postoperative morphine-equivalent consumption are provided in Supplementary Tables S3 and S4.^19,20 Data presented in graphical form were digitized using WebPlotDigitizer when numerical data were unavailable.²¹

Risk of Bias Assessment

Risk of bias for each included RCT was independently evaluated by two reviewers using the Cochrane Risk of Bias 2.0 tool (RoB 2). Five domains were assessed: the randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported result.²² The overall risk of bias for each study was judged as “low risk”, “some concerns”, or “high risk”.

Certainty of Evidence

The certainty of evidence for each primary and secondary outcome was graded using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) framework. Evidence was classified as high, moderate, low, or very low based on assessments of risk of bias, inconsistency, indirectness, imprecision, and publication bias.²³

Statistical Analysis

The primary outcomes were postoperative opioid consumption and pain intensity within 48 hours postoperatively. All meta-analyses were performed using Stata (version 18.0, StataCorp) and Review Manager (version 5.4, Cochrane). For continuous outcomes, the mean difference (MD) or standardized mean difference (SMD) with 95% confidence intervals (CIs) was calculated. For dichotomous outcomes, risk ratios (RRs) with 95% CIs were calculated.

Statistical heterogeneity was quantified using the I² statistic, where values >50% were considered to indicate substantial heterogeneity.²⁴ A random-effects model was used for all analyses due to anticipated clinical and methodological variability.^25,26

Meta-regression analyses were performed using a random-effects model to explore potential sources of between-study heterogeneity and to assess dose–response relationships.²⁷ Study-level covariates included total local anesthetic dose (mg), publication year, and number of fusion levels. Because local anesthetic concentration and volume were highly correlated across studies, total dose (mg) was selected as the primary exposure variable to minimize collinearity and to better reflect overall pharmacological exposure. Only study-level aggregated variables were included.

In addition, 95% prediction intervals (PIs) were calculated for key outcomes to estimate the range in which the true effect of ESPB is expected to lie in a future clinical setting, accounting for between-study heterogeneity.²⁸

To investigate sources of heterogeneity, predefined subgroup analyses were conducted based on: (1) control group type (sham/placebo vs no intervention), (2) surgical approach (open vs minimally invasive), (3) number of fused levels (single- vs multilevel), (4) block timing (pre- vs post-anesthesia induction), and (5) local anesthetic type. The robustness of the findings was tested using leave-one-out sensitivity analysis. To assess potential measurement heterogeneity, additional sensitivity analyses were conducted stratified by pain assessment scale (VAS 0–10 vs NRS 0–10). Separate pooled analyses were performed, including only studies using the same scale type. Subgroup categories were defined according to the comparator type reported within each outcome dataset; therefore, subgroup composition varied across outcomes depending on available data.

Publication bias for outcomes reported in ≥10 studies was assessed by visual inspection of funnel plots and formally tested using Egger’s regression test. A two-tailed p < 0.05 was considered statistically significant for all analyses except for the test of heterogeneity, where p < 0.10 was used. Given the limited number of studies for most outcomes, trim-and-fill analyses were additionally conducted to evaluate the potential influence of small-study effects on pooled estimates.^29,30

Results

Literature Search and Study Characteristics

A total of 712 studies were identified through database searches (PubMed, Embase, Cochrane Library, and CNKI). After removing duplicates and screening titles and abstracts, 70 full-text articles were reviewed. Following exclusions (3 retrospective studies, 5 protocols without full data, 23 not limited to lumbar fusion, and 18 lacking relevant outcomes), 15 RCTs (27–41) involving 1,035 patients (522 ESPB, 513 control) were included (Figure 1).

Figure 1.

PRISMA flow diagram of the included studies.

Notes: n, number of records/studies at each stage; the asterisk (*) indicates records identified from the listed databases.

The included studies were published between 2019 and 2025, with sample sizes ranging from 40 to 120 participants. Most performed ultrasound-guided ESPB preoperatively, while one performed manual intraoperative injection and one postoperatively. Local anesthetics included ropivacaine (0.375–0.4%) and bupivacaine (0.25–0.5%), with three trials using dexamethasone as an adjuvant.^31–33 Control interventions included sham block (saline), local wound infiltration, or no block. Nine trials assessed both resting and dynamic pain scores, and all reported opioid consumption.^34–42 Baseline characteristics and block protocols are summarized in Table 1 and Table 2.

Table 1.

Characteristics of Included Studies

Study No	Author Year	Design	Surgical Type	Total Cases	Total Controls	Age		M/F		BMI		ASA I/II/III		Duration of Surgery		Perioperative Analgesia	Postoperative Analgesia	Rescue Analgesia
						Case	Control	Case	Control	Case	Control	Case	Control	Case	Control	Control	Case	Control
1	El Ghamry MR, 201940	RCT	Open Classic Posterior Approach	30	30	43.9±9.8	42.8±10.7	17/13	16/14	NA	NA	15/15/0	17/13/0	175.5±13.6	179.4±16.8	Fentanyl	Paracetamol+ Ketorolac	Morphine 0.1 mg/kg IV (VAS > 30)
2	Lijun Zhu, 202142	RCT	Open Posterior Lumbar Fusion	20	20	59 ± 2	60 ± 2	7/13	8/12	25.3±0.7	24.4±0.5	4/16/0	3/17/0	128±6	123±5	Remifentanil	PCIA (oxycodone)	5 μg of sufentanil IV (VAS > 4)
3	Vipin Kumar Goel, 202143	RCT	Open TLIF	51	50	52.42 ± 13.05	52.16 ± 12.05	21/30	21/29	25.69 ± 3.99	26.33 ± 2.78	24/27/0	20/30/0	131.27 ± 11.26	132.92 ± 12.10	Fentanyl	Paracetamol+ ketorolac+pregabalin capsule	Fentanyl 1 mcg/kg (NRS ≥ 5)
4	Lizhen Wang, 202136	RCT	Lumbar spine fusion surgery	102	102	53.78 ± 10.16	55.69 ± 12.01	44/48	51/51	24 ± 2.71	24.28 ± 11.26	21/81/0	11/90/1	27.29 ± 30.45 0	130.92 ± 26.91	Sufentanil+ Remifentanil	PICA (Sufentanil)	Sufentanil 5 μg/dose (NAS>5)
5	Serdar Yeşiltaş, 202138	RCT	Open Classic Posterior Approach	28	28	61.0 ± 9.4	60.1 ± 11.7	11/17	7/21	28.7 ± 3.6	28.1 ± 4.35	13/11/4	12/10/6	262.6 ± 80.9	245.0 ± 97.0	Remifentanil	PCIA (morphine)+Paracetamol thrice/day	Pethidine 25 mg (VAS≥4)
6	Qingfen Zhang, 202137	RCT	Open Posterior Lumbar Spinal Fusion	30	30	60.0 ± 9.6	59.5 ± 11.5	6/24	9/21	25.4 ± 3.2	24.7 ± 2.9	7/23/0	8/22/0	152.6 ± 38.7	143.5 ± 33.3	Remifentanil	PICA (sufentanil)	PCIA bolus dose of morphine 2 mg (NRS≥3)
7	Huifen Lin, 202234	RCT	Minimally Invasive PLIF	42	41	65 ±9.4	65 ±7.2	12/30	18/23	23.4 ±2.3	24.2 ±1.9	1/23/18	2/27/12	168 ±21.2	175 ±15.3	Remifentanil	PCIA (morphine)	PCIA (morphine)
8	Alessandro Vergari, 202244	RCT	Open Lumbar Vertebral Arthrodesis	12	12	58 ± 10	60 ± 11	5/7	4/8	25.4 ± 2.8	27.4 ± 3.9	4/8/0	1/11/0	250+51	237+65	Sufentanil	Ketoprofen +acetaminophen + sufentanil	NA
9	Gisi, G, 202333	RCT	Open Posterior Spinal Instrumentation	21	21	51.90±12.32	50.71±13.56	13/8	8/13	NA	NA	2/11/8	1/12/8	272±85.67	242±70.26	Remifentanil	PCIA (morphine)	Dexketoprofen (NAS≥4)
10	Suresh Kumar, 202432	RCT	Open Posterior Spine Fusion	28	28	48.29 ± 11.84	50.29 ± 11.55	12/16	12/16	25.04 ± 3.88	25.04 ± 3.88	9/19/0	9/19/0	NA	NA	Fentanyl	Paracetamol	Tramadol 50 mg slow IV (VAS ≥4)
11	Zhen Zhang, 202341	RCT	Open PLIF	35	30	50.06±10.42	52.13±12.33	18/17	14/16	NA	NA	22/13/0	19/11/0	151.86±37.26	159.50±37.68	Remifentanil	PICA (sufentanil)	Analgesics (opi oids and non-opioid) (VAS>3)
12	Woo-Young Jo, 202539	RCT	Minimally Invasive TLIF/OLIF	38	38	61±9	60±8	14/16	10/20	27.0±2.9	26.2±2.8	2/25/3	1/24/5	197.1836±24.5064	208.4972±22.791	Propofol+Remifentanil	PCIA (100 mL of 1.25 μg/mL sufentanil)	Flurbiprofen axetil
13	Ilse H. van de Wijgert, 202535	RCT	Dorsal Lumbar Spinal Fusion	38	36	60.7±10.2	61.6±9.1	19/19	16/20	27.6 ±3.0	28.4 ±3.2	7/30/1	2/34/0	NA	NA	Propofol+Sufentanil	PCIA (morphine)	Clonidine,+ketamine +diazepam
14	Yucel Yuce, 202445	RCT	Instrumented Lumbar Spinal Surgery	40	40	58.3±6.5	56.7±7.8	10/30	15/25	22.03—30.0	23.03—29.5	3/22/15	4/23/13	225±57	222±54	Fentanyl	Intravenous paracetamol (1 g every 6 h)	Tramadol (100 mg) (VAS≥4)
15	Daniele Bellantonio, 202331	RCT	Posterior Spinal Fusion	15	15	54.6±16.8	60.4±11.4	7/8	8/7	26.4±5.7	27.5±3.7	2/9/4	37.079	26.4±5.7	27.5±3.7	Fentanyl	PCIA (morphine)	Ketoprofen IV (100 mg)

Abbreviations: RCT, randomized controlled trial; M/F, male/female; BMI, body mass index; ASA, American Society of Anesthesiologists physical status classification; TLIF, transforaminal lumbar interbody fusion; PLIF, posterior lumbar interbody fusion; MIS, minimally invasive surgery; NA, not available.

Table 2.

ESPB Protocols of Included Studies

Author Year	Guidance	Timing	Block Position	Local Anesthetics					Control Intervention
				Laterality	Volume (mL/side)	Concentration (%)	Total Dose (mg/side)	Adjuvant
El Ghamry MR, 201940	Bilateral ultrasound-guided	Before the induction of anesthesia	At the level of L3 transverse process	Bilateral	20	0.25% bupivacaine	50mg	None	Sham/Placebo Block
Lijun Zhu, 202142	Bilateral ultrasound-guided	Before the induction of anesthesia	At the level of L2 transverse process	Bilateral	20	0.375% ropivacaine	75mg	None	Sham/Placebo Block
Vipin Kumar Goel, 202143	Bilateral ultrasound-guided	After the induction of anesthesia	At the level of L4-L5/L5-S1 transverse process	Bilateral	20	0.25% bupivacaine	50mg	None	No intervention
Lizhen Wang, 202136	Bilateral ultrasound-guided	After the induction of anesthesia	At the level of T12 transverse process	Bilateral	30	0.375% ropivacaine	112.5mg	None	No intervention
Serdar Yeşiltaş, 202138	Freehand	After the induction of anesthesia	NA	Bilateral	20	0.25% bupivacaine + 1.0% lidocaine (1:1)	50mg bupivacaine; 200mg lidocaine	Lidocaine	Sham/Placebo Block
Qingfen Zhang, 202137	Bilateral ultrasound-guided	Before the induction of anesthesia	At the level of L3 transverse process	Bilateral	20	0.4% ropivacaine	80mg	None	Sham/Placebo Block
Huifen Lin, 202234	Bilateral ultrasound-guided	After the induction of anesthesia		Bilateral	20	0.375% ropivacaine	75mg	None	No Intervention
Alessandro Vergari, 202244	Bilateral ultrasound-guided	After the induction of anesthesia	NA	Bilateral	20	0.375% ropivacaine	75mg	None	Wound Infiltration
Gisi, G, 202333	Bilateral ultrasound-guided	Before the induction of anesthesia	At the level of L3 transverse process	Bilateral	20	0.25% bupivacaine	50mg	4 mg dexamethason	No intervention
Suresh Kumar, 202432	Bilateral ultrasound-guided	After the induction of anesthesia	At the level of L3-L4 transverse process	Bilateral	20	0.25% bupivacaine	50mg	4 mg dexamethason	Sham/Placebo Block
Zhen Zhang, 202341	Bilateral ultrasound-guided	After the induction of anesthesia	NA	Bilateral	20	0.375% ropivacaine	75mg	None	Wound Infiltration
Woo-Young Jo, 202539	Bilateral ultrasound-guided	Before the induction of anesthesia	At the level of T12 transverse process	Bilateral	20	0.375% bupivacaine	75mg	None	No intervention
Ilse H. van de Wijgert, 202535	Bilateral ultrasound-guided	After the induction of anesthesia		Bilateral	30	0.375% ropivacaine	112.5mg	None	Sham/Placebo Block
Yucel Yuce, 202445	Bilateral ultrasound-guided	After the induction of anesthesia	At the level of L2 -L3 transverse process	Bilateral	20	0.5% bupivacaine	100mg	None	Wound Infiltration
Daniele Bellantonio, 202331	Bilateral ultrasound-guided	After the induction of anesthesia	NA	Bilateral	20	0.4% ropivacaine	80mg	4 mg dexamethason	No intervention

Abbreviations: US, ultrasound; L, lumbar vertebra; T, thoracic vertebra; S, sacral vertebra; NA, not available; mg, milligram; mL, milliliter; ESPB, erector spinae plane block.

Risk of Bias Assessment

Overall, methodological quality, evaluated using RoB 2, was moderate (Figure 2). Randomization procedures were adequately reported in most studies, although six trials had unclear allocation concealment.^{31,32,38,42–44} Eight studies lacked participant or operator blinding, leading to potential performance bias.^{31,33,34,36,39,41,43,45} Measurement bias was possible in two studies due to unblinded outcome assessors.^32,42 Selective reporting was suspected in one trial lacking preregistration.³¹ GRADE quality assessment results are summarized in Supplementary Table S5.

Figure 2.

Risk of bias graph. (A) Risk of bias assessment of included studies. (B) Risk of bias summary: each risk of bias item for each included study. Green, red, and yellow circles indicate low, high, and unclear risks of bias, respectively.

Primary Outcomes

Intraoperative opioid consumption: Seven studies (n = 578) reported intraoperative opioid use.^{31,34,36,37,40,42,43} ESPB significantly reduced opioid consumption compared with controls (MD = −31.70 mg, 95% CI [−41.69, −21.71], p < 0.001; I² = 99%) (Figure 3). Subgroup analysis by control type showed a significant reduction in trials using sham or placebo controls (MD = −18.35 mg, 95% CI [−24.30, −12.41]; I² = 93%). In studies using no-intervention comparators, the pooled estimate indicated a reduction in opioid use that did not reach statistical significance (MD = −45.78 mg, 95% CI [−95.28, 3.73]; I² = 99%) (Supplementary Figure S1).

Figure 3.

Forest plot of primary outcomes (milligram morphine equivalents). (A) Intraoperative Opioid Consumption. (B) Postoperative Opioid Consumption at 24 and 48 hours. (C) Pain Scores at rest at 24 hours.

Eight trials reported 24-hour and five reported 48-hour morphine-equivalent consumption. ESPB significantly reduced postoperative opioid requirements at 24 h (MD = −11.02 mg, 95% CI [−14.42, −7.63], I² = 84%) and 48 h (MD = −11.46 mg, 95% CI [−14.91, −8.00], I² = 42%) (Figure 3). Subgroup analyses were performed to explore potential sources of heterogeneity. Stratification by local anesthetic regimen (local anesthetic alone vs combined with adjuvants) demonstrated consistent opioid-sparing effects in both subgroups, without significant subgroup differences (Supplementary Figure S2). Similarly, analyses based on block timing (pre-incision vs post-incision) showed comparable reductions in opioid consumption.

Although heterogeneity was attenuated in certain subgroup comparisons, moderate-to-high between-study variability persisted. Prediction interval estimates were wide, indicating variability in the expected magnitude of treatment effect across different clinical settings.

Pain Scores

Fifteen studies reported resting and/or dynamic pain scores across multiple time points. Compared with controls, ESPB significantly reduced resting pain at all postoperative intervals, with the largest effects observed within the first 6 hours (MD = −1.88 immediately, −1.67 at 2 h, and −1.38 at 4 h). The analgesic effect persisted but gradually diminished by 24–48 h (MD = −0.49 and −0.31, respectively) (Supplementary Figure S3).

Dynamic (movement) pain followed a similar trend: significant reduction at 4–8 h (MD = −1.44 at 4 h; −1.14 at 8 h), but non-significant differences at 12–48 h (Supplementary Figure S4). These findings suggest that ESPB provides clinically meaningful early postoperative analgesia after lumbar fusion surgery. Overall, ESPB provided superior analgesia during the first 24 postoperative hours in both resting and dynamic states, though its effect diminished beyond 48 hours, consistent with the pharmacologic duration of a single-shot local anesthetic injection.

Sensitivity analyses stratified by pain scale type (VAS vs NRS) yielded consistent effect directions and comparable magnitude estimates across most postoperative time points. No material changes in statistical significance were observed, indicating that the pooled analgesic effects were not driven by differences in pain assessment scales (Supplementary Table S6 and Supplementary Figures S5–S8).

For pain scores at 24 hours, prediction intervals crossed the null value, suggesting that while the average effect favored ESPB, the true effect in individual clinical settings may be minimal or absent.

Secondary Outcomes

ESPB significantly prolonged the time to first rescue analgesia (MD = 169.46 min, 95% CI [39.89, 299.04], p = 0.01) and reduced the proportion of patients requiring rescue analgesia (RR = 0.42, 95% CI [0.30, 0.58], p < 0.001).

Patients in the ESPB group had shorter PACU stay (MD = −13.98 min, p = 0.04) and shorter hospital stay (MD = −0.35 days, p = 0.03) (Supplementary Figure S9).

Regarding adverse events, ESPB significantly reduced the incidence of PONV (RR = 0.71, 95% CI [0.53, 0.94]), drowsiness (RR = 0.27, 95% CI [0.10, 0.75]), and pruritus (RR = 0.40, 95% CI [0.16, 0.99]), with minimal heterogeneity across studies. Postoperative dizziness showed a non-significant trend toward reduction (RR = 0.75, p = 0.49). Patient satisfaction scores were higher with ESPB (MD = 1.37, 95% CI [1.12, 1.62], p < 0.0001) (Figure 4).

Figure 4.

Forest plot of satisfaction and adverse effects. (A) Patient satisfaction; (B) Postoperative nausea and vomiting (PONV); (C) Somnolence; (D) Skin pruritus; (E) Postoperative dizziness. Effect sizes are presented as standardized mean difference (SMD) for continuous outcomes and risk ratios (RR) for dichotomous outcomes with 95% confidence intervals.

Subgroup and Heterogeneity Analyses

To explore potential sources of heterogeneity, pre-specified subgroup analyses were conducted on intraoperative and postoperative opioid consumption, as well as 24-hour resting and dynamic pain scores. A summary of the subgroup analyses conducted in this study is presented in Supplementary Table S7.

(1) Number of fused segments: ESPB provided greater analgesic benefit in single-level fusion (I² = 41%) than in multilevel fusion (I² = 94%), suggesting that diffusion limitations may reduce block efficacy in extensive procedures (Figure 5) (Supplementary Figure S10).

Figure 5.

Forest plot for subgroup analysis of pain scores during movement at 24 hours. (a) According to the surgical approach. (b) According to the number of fusion segments.

(2) Surgical approach: Significant analgesic effects were observed in open lumbar fusion procedures (Rest: SMD = −0.57; 95% CI [−0.91, −0.23]; p = 0.001; I² = 82%; Movement: SMD = −0.59; 95% CI [−0.76, −0.42]; p < 0.0001; I² = 88%), whereas no statistically significant differences were identified in minimally invasive surgery (Rest: SMD = −0.13; p = 0.68) (Figure 5) (Supplementary Figure S10).

(3) Control group type: Trials using placebo/sham blocks demonstrated smaller yet significant effects (SMD = −0.48; 95% CI [−0.89, −0.08]; I² = 71%) compared with no-intervention trials (SMD = −0.66; 95% CI [−1.22, −0.10]; I² = 89%), reinforcing that methodological rigor substantially influences effect estimates (Supplementary Figure S1 and S10).

Publication Bias and Sensitivity Analysis

Publication bias and sensitivity analyses were performed for outcomes with sufficient studies. Funnel plots for postoperative morphine consumption and 24-hour pain scores were visually inspected (Supplementary Figures S11 and S12).

For 24-hour pain scores (15 studies), Egger’s test showed no evidence of small-study effects (P = 0.625) (Supplementary Figures S13). Trim-and-fill analysis under a random-effects model did not impute additional studies, and the pooled estimate remained unchanged.

For 24-hour postoperative opioid consumption (8 studies), Egger’s test suggested potential small-study effects (P = 0.009). Trim-and-fill analysis imputed one potentially missing study (Supplementary Figures S14). The pooled mean difference changed slightly from −10.427 mg (95% CI −11.784 to −9.070) to −10.207 mg (95% CI −11.552 to −8.861), and statistical significance was preserved. These findings indicate that the overall conclusions are robust to potential small-study effects.

Galbraith plots (Supplementary Figure S12) demonstrated that all studies lay within 95% confidence boundaries. GRADE certainty assessments are summarized in Supplementary Table S5, and the detailed GRADE downgrading criteria applied in this review are provided in Supplementary Table S8.

Leave-one-out sensitivity analyses were performed to assess the robustness of pooled estimates. Sequential exclusion of individual studies did not materially alter effect sizes or heterogeneity for primary outcomes (Supplementary Figures S15, S16 and Supplementary Table S9). Specifically, exclusion of the single free-hand ESPB trial did not meaningfully change pooled effect estimates, confidence intervals, or I² values. These findings indicate that overall results were not driven by the guidance modality used in that study and support the robustness of the primary conclusions.

Meta-Regression and Prediction Intervals

Meta-regression analyses were conducted to explore potential sources of heterogeneity (Supplementary Table S10).

For 24-hour postoperative opioid consumption (8 studies), total local anesthetic dose was not significantly associated with effect size (β = −0.013, P = 0.802). However, the number of fusion levels (β = 2.12, P = 0.008) and adjuvant use (β = −0.631, P = 0.041) showed statistical associations. Notably, inclusion of the number of fusion levels reduced between-study variance (τ² decreased to 0.3845), suggesting a partial explanation of heterogeneity.

For 24-hour resting pain scores (12 studies), both the number of fusion levels (P = 0.025) and adjuvant use (P = 0.024) reached statistical significance, whereas total dose and publication year were not associated with treatment effect. However, the proportion of residual heterogeneity remained substantial (τ² ranging from 0.228 to 0.269), indicating that these variables explained only a limited fraction of between-study variability.

For intraoperative opioid consumption, publication year was statistically associated with treatment effect (β = 4.08, P = 0.025), while other covariates were not significant. Despite this, heterogeneity remained considerable, suggesting multifactorial contributors.

Despite statistically significant pooled effects for several outcomes, the corresponding prediction intervals were wide. For example, intraoperative opioid consumption showed a pooled MD of −31.7 mg with a 95% prediction interval ranging from −58.02 to −5.38 mg. In contrast, prediction intervals for 24-hour postoperative opioid consumption (−3.20 to 0.48) and 24-hour resting pain scores (−1.51 to 0.50) crossed the null value, indicating variability in potential treatment effects across clinical settings (Supplementary Table S11).

Discussion

Over the past decade, the volume of lumbar fusion surgery has increased markedly, accompanied by a growing shift toward multilevel procedures, as evidenced by recent epidemiological and systematic review studies.⁴⁶ Compared with single-level surgery, multilevel fusion typically involves longer operative time, more extensive paraspinal muscle dissection, and greater surgical trauma, which are associated with intensified postoperative pain and an elevated risk of perioperative complications.³ Consequently, these factors are associated with prolonged hospital stay and reduced patient satisfaction.

Conventional postoperative analgesia continues to rely predominantly on intravenous patient-controlled opioid administration (IV-PCA); however, opioid-related adverse effects, including postoperative nausea and vomiting (PONV), respiratory depression, constipation, and pruritus, remain prevalent and may contribute to opioid tolerance and persistent use.^47,48 In response to these limitations, multimodal analgesia (MMA), a cornerstone of the Enhanced Recovery After Surgery (ERAS) strategy, has gained increasing attention for its potential to reduce perioperative opioid consumption and facilitate functional recovery.^49–51 Previous systematic reviews have demonstrated that integrating regional anesthesia techniques into multimodal regimens significantly decreases opioid requirements and improves postoperative functional outcomes.⁵² Moreover, accumulating evidence indicates that inadequately controlled acute postoperative pain is a major risk factor for chronic postsurgical pain (CPSP), underscoring the importance of effective perioperative analgesia beyond the immediate postoperative period.⁵³

In this procedure-specific systematic review and meta-analysis encompassing 15 randomized trials, we provide robust quantitative evidence supporting the integration of ESPB into multimodal analgesia for lumbar fusion surgery. ESPB significantly reduced perioperative opioid consumption and acute postoperative pain intensity while also decreasing the incidence of opioid-related adverse events, particularly PONV, and shortening hospital stay. These findings are highly consistent with the ERAS philosophy and provide comprehensive quantitative evidence supporting the clinical adoption of ESPB for lumbar fusion procedures.

One of the most notable findings of this study is the pronounced opioid-sparing effect of ESPB. Compared with controls, ESPB markedly reduced total opioid consumption both intraoperatively and within 48 hours postoperatively. This finding carries several clinically meaningful implications. First, excessive perioperative opioid exposure is a well-established risk factor for PONV, excessive sedation, ileus, and urinary retention.^54,55 Our analysis demonstrated a significantly lower incidence of PONV in the ESPB group, most likely mediated by reduced opioid consumption—consistent with previous meta-analyses in spinal and thoracic surgeries. Second, minimizing perioperative opioid exposure, particularly during the initial postoperative period, is crucial for preventing persistent opioid use and dependence, an issue amid the ongoing global opioid crisis.⁵⁵

Regarding analgesic efficacy, ESPB demonstrated superior pain control in the early postoperative period, especially within 24 hours, for both rest and movement pain. Interestingly, several studies observed that the benefit of dynamic pain was more sustained than that of resting pain. This could be attributed to blockade of the dorsal rami of spinal nerves, potentially reducing nociceptive input from deep paraspinal muscles such as the multifidus and erector spinae. These muscles are frequently traumatized or in spasm following surgical manipulation, and their activation during mobilization is a major contributor to postoperative discomfort.^56–58 However, the analgesic advantage of ESPB gradually diminished after 24–48 hours, aligning with the pharmacological duration of a single-shot local anesthetic injection and explaining why some studies found no intergroup differences beyond this time point.⁵⁹

Differences in analgesic effects across surgical approaches may have several explanations. In the present analysis, ESPB appeared to provide greater analgesic benefit in open lumbar fusion procedures compared with minimally invasive techniques, suggesting that surgical invasiveness may influence the magnitude of clinical response. Lumbar fusion surgery is commonly associated with severe postoperative pain and high opioid requirements. Because procedure-specific subgroup analyses based on individual fusion techniques (eg, PLIF, TLIF, OLIF, ALIF) were not feasible due to limited sample sizes and inconsistent reporting across randomized trials, stratification by surgical approach (open versus minimally invasive) provides a pragmatic and clinically meaningful framework to partially account for variability in surgical complexity and nociceptive burden when more granular analyses are underpowered.

From a clinical perspective, differences in tissue trauma and perioperative nociceptive input likely contribute to variability in ESPB effectiveness. Open lumbar fusion procedures typically involve more extensive paraspinal muscle dissection and fascial disruption, which may increase baseline nociceptive signaling and thereby enhance the relative analgesic impact of regional techniques.^57,58 In contrast, minimally invasive approaches preserve paraspinal musculature and utilize narrower surgical corridors, potentially reducing both nociceptive input and the relative incremental benefit achievable with a fascial plane block.⁶⁰

From a mechanistic standpoint, anatomical characteristics of the erector spinae plane may further explain these differential effects. Cadaveric and imaging studies consistently demonstrate that local anesthetics injected during lumbar ESPB spread predominantly in a craniocaudal direction, deep to the erector spinae muscle, and primarily involve the dorsal rami, with variable, often limited anterior extension toward the ventral rami or paravertebral space.^61,62 Magnetic resonance imaging investigations also reveal substantial interindividual variability in both longitudinal and anterior diffusion, highlighting the influence of fascial integrity and surrounding tissue structures on injectate distribution.⁶³

In minimally invasive lumbar fusion, preservation of paraspinal muscle architecture and the use of a limited surgical corridor may constrain expansion of the fascial plane and limit effective segmental blockade. Conversely, the broader tissue disruption characteristic of open fusion procedures may facilitate wider spread of local anesthetic along adjacent fascial planes and neural structures. Taken together, these anatomical and procedural considerations offer a biologically plausible framework through which the surgical approach may modulate the clinical effectiveness of ESPB.

While pooled estimates demonstrated statistically significant analgesic benefits, prediction interval analyses indicated that the magnitude and even direction of treatment effects may vary across future clinical settings. These findings suggest that the observed heterogeneity likely reflects genuine clinical and methodological differences rather than statistical variability alone. In practical terms, although ESPB appears beneficial on average, individual institutions or specific procedural contexts may experience smaller or non-significant effects depending on baseline analgesic protocols and operative characteristics. Therefore, pooled results should be interpreted as estimates of average treatment effect rather than universally reproducible outcomes, reinforcing the importance of individualized clinical implementation and context-specific evaluation. These observations provide a conceptual framework for understanding the sources of between-study heterogeneity observed in subsequent analyses.

Despite overall favorable findings, moderate-to-high statistical heterogeneity was observed across several outcomes, particularly opioid consumption and length of stay. Sensitivity and subgroup analyses suggest that methodological and clinical variability contributed substantially to between-study differences. One important source relates to control group design. Trials employing sham or placebo injections represent more rigorous comparators that better isolate the specific analgesic effect of ESPB. In contrast, studies using no block or local wound infiltration introduce potential performance bias or compare ESPB against alternative analgesic interventions rather than true controls. These differences may influence both the magnitude of the treatment effect and the variability of pooled estimates.

Clinical heterogeneity also likely played a role. Surgical invasiveness represents a key modifier of nociceptive burden, with open fusion procedures typically involving greater tissue trauma than minimally invasive techniques, potentially amplifying the relative analgesic benefit of fascial plane blocks.^64,65 In addition, variability in fusion extent, including single-level versus multilevel procedures, may influence baseline pain intensity and responsiveness to regional anesthesia. Collectively, these methodological and clinical differences provide plausible explanations for the observed heterogeneity and underscore the importance of procedure-specific analgesic strategies rather than uniform application across surgical contexts.

Chronic postsurgical pain (CPSP) is a key determinant of long-term recovery after lumbar fusion surgery, yet remains substantially underexplored in the context of erector spinae plane block (ESPB). From a mechanistic perspective, CPSP is closely associated with intense perioperative nociceptive input and subsequent central sensitization, providing biological plausibility that effective regional analgesia may influence long-term pain trajectories.⁶⁶ Evidence from other high-pain surgical models, including thoracotomy and mastectomy, further supports the potential role of perioperative neural blockade in reducing persistent pain.^5,67,68 Although biologically plausible, this hypothesis remains unproven in lumbar fusion surgery due to the absence of adequately powered long-term randomized data.

Long-term pain outcomes in lumbar fusion surgery have largely been neglected. Although CPSP was prespecified as an exploratory outcome in our protocol, the available data were insufficient to support quantitative synthesis. Among the 15 included randomized controlled trials, the vast majority focused exclusively on short-term analgesic outcomes within the first 48 postoperative hours, reflecting a predominant emphasis on acute pain control rather than long-term recovery.

Only one trial (Wang et al, 2021) reported follow-up beyond three months, assessing quality-of-life scores at six months rather than CPSP as a predefined endpoint.³⁶ In that study, both ESPB and transmuscular quadratus lumborum block (TLIP) groups demonstrated improved outcomes compared with patients who did not receive regional anesthesia, but no significant difference was observed between the two regional techniques. Similarly, Jo et al (2025) evaluated health-related quality of life using the EQ-5D-5L at four weeks postoperatively and reported transiently higher scores in the ESPB group, whereas van de Wijgert et al (2025) found no significant differences in pain intensity or opioid consumption at 30 days.^35,39 Importantly, none of these studies predefined CPSP as a primary outcome or employed standardized CPSP definitions, and all were insufficiently powered to evaluate long-term pain prevention.

Collectively, these heterogeneous and methodologically limited findings do not allow reliable conclusions regarding the effect of ESPB on CPSP. Rather than indicating an absence of long-term benefit, they highlight an important evidence gap in ESPB research for lumbar fusion surgery. Given the well-established association between inadequately controlled acute postoperative pain, central sensitization, and CPSP development, the lack of systematic long-term pain assessment represents a major limitation of existing trials. These findings underscore the need for future randomized studies to incorporate CPSP as a prespecified outcome. CPSP should be defined as pain persisting beyond three months and assessed with extended follow-up of at least 6–12 months, in order to better characterize the long-term clinical value of ESPB.

Technical variability across studies also warrants consideration. Although ultrasound guidance predominated, one randomized trial employed an intraoperative free-hand technique.³⁸ In response to methodological concerns regarding potential bias introduced by differing guidance techniques, leave-one-out sensitivity analyses were conducted and demonstrated that exclusion of the free-hand study did not materially alter pooled estimates or overall heterogeneity, supporting the statistical robustness of the primary findings.

From a clinical perspective, however, alternative guidance strategies remain relevant. Free-hand ESPB has been proposed as a simplified technique suitable for intraoperative or resource-limited settings but requires further validation regarding reproducibility, anatomical accuracy, and safety. Limited comparative data suggest that while postoperative pain scores may be similar, opioid consumption may differ, indicating potential variability in block performance.⁶⁹ Fluoroscopic guidance has also been explored as a workflow-adapted alternative.⁷⁰ Overall, current evidence is insufficient to determine comparative efficacy among guidance modalities, and ultrasound guidance remains the most established approach pending adequately powered head-to-head trials.

Likewise, some trials performed ESPB pre-induction, enabling immediate efficacy assessment and possible intraoperative opioid reduction, while others administered it post-induction to enhance patient comfort. The timing of block administration, local anesthetic type (ropivacaine vs bupivacaine), concentration, volume, and adjuvant use (eg, dexamethasone) all influence block onset and duration, contributing to interstudy variability. Importantly, all included RCTs employed single-shot ESPB, inherently limiting analgesic duration. Recent studies have shown that continuous ESPB via catheter infusion or programmed intermittent bolus delivery can extend analgesia and reduce opioid use after major spine, thoracic, and cardiac surgeries. For instance, a paired cohort study in multilevel lumbar fusion patients demonstrated that continuous bilateral ESPB infusion for 48 hours significantly decreased morphine consumption compared with a single-shot injection.⁷¹ Comparable benefits have been reported in pediatric thoracic and rib-fracture patients and in thoracic surgery using programmed intermittent boluses.^72,73 These advances highlight the potential of continuous ESPB for sustained postoperative analgesia in extensive lumbar fusion.

In addition to analgesic efficacy, potential cost-effectiveness is an important consideration for the broader implementation of ESPB in lumbar fusion surgery. Although ESPB entails additional procedural time and material costs, particularly when catheter-based techniques are used, including catheter kits, infusion systems, and extended ultrasound utilization, these costs should be interpreted within the broader perioperative economic framework and may be partially offset by reductions in opioid-related adverse events, postoperative nausea and vomiting (PONV), and length of hospital stay. Opioid-related complications and prolonged hospitalization represent major drivers of perioperative healthcare expenditure in spine surgery.⁷⁴ In the present analysis, ESPB was associated with a lower incidence of PONV and modest reductions in PACU and hospital length of stay, outcomes that are known to contribute substantially to perioperative resource utilization and cost burden.⁷⁵ However, the magnitude of these reductions was modest, and whether such improvements fully compensate for additional material and staffing costs likely depends on institutional reimbursement structures, baseline opioid-related complication rates, and the surgical context (eg, open multilevel fusion versus minimally invasive procedures).

Importantly, none of the included randomized trials incorporated formal cost-effectiveness analyses, and the available data do not permit quantitative economic modeling.⁷⁶ Accordingly, any inference regarding economic value remains exploratory and hypothesis-generating rather than definitive.⁷⁷ Future prospective studies incorporating predefined health-economic endpoints—including direct catheter-related material costs, resource utilization metrics, payer perspectives, and quality-adjusted life years—are warranted to determine whether ESPB provides net economic value, particularly in high-pain, resource-intensive procedures such as open multilevel lumbar fusion.

From a clinical translational perspective, ESPB should not be viewed as a standalone intervention but rather as a regional component within contemporary multimodal perioperative analgesic frameworks.⁷⁸ Current enhanced recovery after surgery (ERAS) pathways for lumbar fusion emphasize opioid stewardship, early mobilization, and the combined use of systemic non-opioid analgesics together with regional anesthesia techniques.⁷⁹ A proposed perioperative analgesic pathway illustrating how ESPB may be integrated into contemporary multimodal ERAS-based pain management is presented in Supplementary Figure S17.

Importantly, given the substantial heterogeneity observed across trials and the predominance of short-term outcome reporting, ESPB should currently be viewed as an adjunctive regional technique rather than a replacement for established multimodal strategies. Its role is likely to be individualized based on surgical complexity, expected postoperative pain burden, and institutional ERAS protocols.

Future studies should focus on protocol standardization, including optimal dosing strategies, block timing, and integration with other multimodal interventions, as well as evaluation of functional recovery and longer-term patient-centered outcomes to better define its position within standardized lumbar fusion analgesia pathways.

Limitations

Despite these promising results, several limitations should be acknowledged. First, methodological quality varied across included studies, with more than half exhibiting potential risk of bias related to inadequate blinding, which may influence subjective outcomes such as pain scores.

Second, substantial clinical and methodological heterogeneity—including differences in comparator design, surgical approach, and perioperative analgesic protocols—necessitates cautious interpretation of pooled estimates despite the use of subgroup and sensitivity analyses.

Third, long-term pain outcomes, including chronic postsurgical pain (CPSP), were not systematically assessed or prespecified in most included trials, precluding reliable evaluation of the sustained impact of ESPB beyond the early postoperative period.

Fourth, secondary functional outcomes such as time to ambulation and recovery metrics were inconsistently reported, limiting assessment of broader patient-centered benefits.

Finally, lumbar fusion encompasses a wide spectrum of surgical techniques (eg, PLIF, TLIF, OLIF, ALIF, and minimally invasive variants), yet insufficiently detailed reporting prevented procedure-specific subgroup analyses. Consequently, residual heterogeneity related to surgical complexity cannot be fully excluded.

Conclusion and Future Perspectives

In summary, low- to moderate-certainty evidence indicates that the erector spinae plane block (ESPB), when incorporated into a multimodal analgesia framework, probably reduces early postoperative pain, perioperative opioid consumption, and opioid-related adverse events following lumbar fusion surgery.

Although the direction of effect was consistent across outcomes, substantial heterogeneity and variability in methodological quality warrant cautious interpretation. The magnitude of benefit appears to be context dependent and may vary according to comparator design, surgical invasiveness, and block protocol. Current evidence primarily reflects short-term outcomes, and standardized integration strategies within multimodal pathways remain to be defined.

Future research should prioritize multicenter, double-blind, placebo-controlled trials; direct comparisons of ESPB techniques and delivery strategies; rigorous evaluation of long-term outcomes, including CPSP and functional recovery; and formal health-economic analyses. Addressing these priorities will help delineate the optimal clinical positioning of ESPB within individualized, procedure-specific multimodal analgesia strategies for lumbar fusion surgery.

Abbreviations

ALIF, Anterior lumbar interbody fusion; ASA, American Society of Anesthesiologists; BMI, Body mass index; CI, Confidence interval; CNKI, China National Knowledge Infrastructure; CPSP, Chronic postsurgical pain; ERAS, Enhanced Recovery After Surgery; ESPB, Erector spinae plane block; EQ-5D-5L, EuroQol five-dimension five-level questionnaire; GRADE, Grading of Recommendations, Assessment, Development and Evaluation; IQR, Interquartile range; IV-PCA, Intravenous patient-controlled analgesia; MD, Mean difference; MeSH, Medical Subject Headings; MME, Morphine milligram equivalents; MMA, Multimodal analgesia; MRI, Magnetic resonance imaging; NRS, Numeric rating scale; OLIF, Oblique lumbar interbody fusion; PACU, Post-anesthesia care unit; PI, Prediction interval; PLIF, Posterior lumbar interbody fusion; PONV, Postoperative nausea and vomiting; PROSPERO, International prospective register of systematic reviews; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses; QoR, Quality of recovery; RCT, Randomized controlled trial; RoB 2, Risk of Bias 2 tool; RR, Risk ratio; SD, Standard deviation; SMD, Standardized mean difference; TLIF, Transforaminal lumbar interbody fusion; TLIP, Thoracolumbar interfascial plane block; VAS, Visual analog scale.

Disclosure

The authors report no conflicts of interest in this work.