Spinal Anesthesia With Use of an Erector Spinae Plane Block Is Associated With Reduced Opioid Use and Improved Clinical Outcomes in Single-Level Transforaminal Lumbar Interbody Fusions: Evolution of Techniques and a Single-Surgeon Retrospective Analysis

Abstract

Background

General anesthesia (GA) has long been the preferred modality of anesthesia for spine surgery, including transforaminal lumbar interbody fusions (TLIFs). However, recent advances in regional anesthesia, including spinal (awake) anesthesia and regional blocks, have demonstrated potential improvements in both surgical outputs and recovery times compared with GA. This study aims to describe the associations between regional anesthesia, clinical outcomes, and opioid utilization among patients undergoing TLIF.

Methods

The electronic medical records of a single fellowship-trained spine surgeon were analyzed retrospectively, and patients undergoing a single-level TLIF between 2018 and 2024 were included. Pre- and postoperative clinical measures (eg, length of stay, operating time, and readmission rates) and both intra- and postoperative opioid use were collected. All data were analyzed using a multivariable regression model.

Results

Compared with GA, awake anesthesia was associated with reduced intraoperative opioid utilization (P = 0.007) across all TLIF groups, while the use of an erector spinae plane (ESP) block was associated with a reduced length of stay (P = 0.044) and lower postoperative opioid utilization (P < 0.001). The use of spinal anesthesia combined with the use of an ESP block was associated with significantly lower length of stay and lower postoperative morphine equivalent daily dose (P = 0.030 and P = 0.002, respectively).

Conclusions

Awake TLIF surgery under spinal anesthesia and the use of an ESP block were independently associated with reduced opioid usage, while the combined use of these techniques was associated with decreased length of stay and reduced postoperative opioid usage. Additional investigation and prospective studies are needed to further elucidate these complex relationships.

Clinical Relevance

The relationship between regional anesthesia and patient outcomes has not been comprehensively characterized in the TLIF patient population despite the growing utilization of these techniques in spine surgery. Once established, these findings have potential implications for the integration of regional anesthesia into TLIF procedures as a new standard of care.

Level of Evidence

Preliminary evidence suggests that regional anesthesia offers multiple benefits over GA, including health and mortality outcomes, time to ambulation and recovery, and incidence of postoperative complications. However, little is known about the applicability of these findings to patients undergoing TLIF.

Introduction

Spine surgery has become increasingly prevalent in recent years, driven by the rising incidence of spinal disorders, degenerative conditions, and injuries.¹ Common indications for spine surgery include herniated discs, spinal stenosis, scoliosis, and degenerative disc disease, all of which can significantly impact quality of life and capacity. Given the increasing prevalence of conditions such as chronic back pain and mobility issues, there has been a growing need for surgical intervention. At the same time, advances in surgical techniques, such as minimally invasive procedures, have further accelerated the growth of spine surgery while offering improved outcomes and reduced recovery times.² Transforaminal lumbar interbody fusion (TLIF) represents a widely used surgical approach designed to treat various spine pathologies.²

General anesthesia (GA) has been the longstanding preferred mode of anesthesia in spine surgery, providing complete unconsciousness and immobility to ensure patient safety and optimal surgical conditions.³ This approach typically involves the administration of intravenous and/or inhaled agents to induce and maintain anesthesia, combined with endotracheal intubation for airway management.⁴ GA offers the advantage of complete control over the patient’s airway, ventilation, and hemodynamic parameters, which is particularly critical during complex and lengthy spinal procedures.⁴ Additionally, it minimizes intraoperative movement and allows for optimal surgical precision in a controlled environment. However, GA has been shown to be associated with significant complications, including respiratory complications, hemodynamic instability, and prolonged recovery times, particularly in high-risk patients or those with significant comorbidities.⁵

Recent advancements have given rise to new anesthetic approaches that offer potential advantages. In particular, the use of spinal anesthesia (SA) allows for patients to be kept awake or at least partially conscious during surgery rather than inducing full sedation, with flexibility to adjust the approach based on the procedure and individual patient. In this way, it offers patients an option with a lower anesthetic load, faster recovery, and fewer complications.⁶ This method of regional anesthesia has shown significant benefits for lumbar fusion surgery, including reduced postoperative complications, shorter recovery times, and improved outcomes, especially among high-risk patients.⁷ In a similar vein, regional nerve block techniques such as the erector spinae plane (ESP) block have demonstrated improved postoperative outcomes for spine surgery patients, including reduced length of stay (LOS) and time to ambulation.^8,9 However, despite the rising popularity of SA, few studies have examined the impact of anesthesia type on surgical outcomes among patients undergoing TLIF.¹⁰ Particularly when accounting for surgical and demographic factors, data for this population remain limited, and few studies have looked at the impact of anesthesia on opioid usage and postoperative pain management. Given the evolution of surgical and anesthesia techniques for spine surgery, this single-institution, retrospective analysis aims to explore the impact of different subtypes of regional anesthesia on peri- and postoperative outcomes.

Methods

Study Design and Participants

Adult patients undergoing a single-level TLIF at an academic hospital by a single surgeon between 2018 and 2024 were included, and the electronic medical records were reviewed. Inclusion criteria included (1) significant neural compression requiring extensive decompression; (2) Meyerding grade I to II spondylolisthesis with segmental instability; and (3) disc degeneration disease with segmental instability. Exclusion criteria included spinal deformities, tumors, co-infections, and trauma.

Of the 94 patients included in this study, all reported opioid usage data and clinical outcome measures. Opioid usage and clinical outcomes were compared between patients receiving neuraxial SA (awake) and those receiving general endotracheal anesthesia (GA, asleep). Patients from either anesthesia group were further stratified based on whether or not they received a local ESP block, and their outcomes were compared. Furthermore, patients who received both SA and ESP blocks were compared with all other patients.

All ESP patients in our cohort received the ESP block with liposomal bupivacaine (Exparel, Pacira Pharmaceuticals, Parsippany, NJ) except for 1 patient, who received the ESP block with ropivacaine. This study was conducted in accordance with the Declaration of Helsinki (as revised in 2013) and approved by the Institutional Review Board with waiver of the requirement for patient consent (IRB approval number: Pro00071025).

Clinical Outcome Measures

Recorded operative variables included sociodemographic data, level of fusion, anesthesia used, and utilization of an ESP block. Clinical outcome data collected included estimated blood loss (EBL), operating time (OR time), LOS, 30-day readmission rates, and both total intraoperative and 24-hour postoperative morphine equivalent daily dose (MEDD). Due to limited follow-up and postoperative availability, pain scores were assessed but not included in the final analysis.

Statistical Methods

Statistical analyses were carried out using R software (Jamovi for MacOS, version 2.6.2). The nonparametric Wilcoxon Rank Sum Test for difference in means was used to assess significant differences across the SA and GA groups, and the Pearson χ ² test for independence was employed for categorical variables. A post-hoc Dunn test for significance was used to compare intergroup differences in significant features. A multivariable logistic regression model was then created to evaluate the risk factors associated with clinical outcomes. The covariates used in this analysis were age, race, sex, American Society of Anesthesiologists classification, body mass index, history of hypertension, diabetes, or smoking, type of anesthesia, use of ESP block, TLIF surgical technique, and level of surgery. A P value of less than 0.05 was considered significant. Results are reported in accordance with the STROBE guidelines.

Results

Patient Demographics

There were 94 total patients included in the study cohort (SA: 34; GA: 60). Average age was 68 ± 10.8 years in the cohort receiving SA and 62.2 ± 10.9 in the cohort receiving GA (P = 0.032). A total of 49 patients received an ESP block (52.1%), and 28 patients received SA with an ESP block (29.8%). Over time, the number of patients undergoing TLIF under SA increased, while the number of patients undergoing TLIF under GA decreased (Figure 1). A full breakdown of the cohort included can be found in the Table 1.

Figure 1.

Evolution of transforaminal lumbar interbody fusion cases over time, divided into spinal anesthesia (SA) and general anesthesia (GA) cases with or without erector spinae plane (ESP) block. Data are included through 2023, which was the last full year included in this study.

Table 1.

Demographic characteristics of patients undergoing single-level transforaminal lumbar interbody fusion.

Characteristic	Spinal Anesthesia (n = 34)	General Anesthesia (n = 60)	P a
Age, y	68.0 ± 10.8	62.2 ± 10.9	0.032 b
Sex, n (%)			0.092
Men	18 (52.9)	21 (35.0)
Women	16 (47.1)	39 (65.0)
Race, n (%)			0.078
White	29 (85.3)	41 (68.3)
Black	4 (11.8)	16 (26.7)
Asian	0 (0.0)	2 (3.3)
American Indian or Alaskan Native	1 (2.9)	1 (1.7)
BMI	29.7 ± 4.4	30.3 ± 4.9	0.496 b
History of smoking, n (%)	12 (35.3)	21 (35.0)	0.521
History of hypertension, n (%)	22 (64.7)	41 (68.3)	0.721
History of diabetes, n (%)	3 (8.9)	21 (35.0)	0.007
Surgical technique, n (%)			<0.001
Transfacet	14 (41.2)	3 (5.0)
Percutaneous	15 (44.1)	31 (51.7)
MIS	5 (14.7)	26 (43.3)
ESP block used, n (%)			<0.001
Yes	28 (82.4)	21 (35.0)
No	6 (17.6)	39 (65.0)
Level of surgery, n (%)			0.925
L2/L3	0 (0.0)	2 (3.3)
L3/L4	3 (8.8)	8 (13.3)
L4/L5	29 (85.3)	42 (70.0)
L5/S1	2 (5.9)	8 (13.3)
ASA classification, n (%)			0.617
1	0 (0.0)	1 (1.7)
2	12 (35.3)	18 (30.0)
3	22 (64.7)	39 (65.0)
4	0 (0.0)	2 (3.3)

Abbreviations: ASA, American Society of Anesthesiologists; BMI, body mass index; ESP, erector spinae plane; MIS, Minimally Invasive Surgery.

Note: Data presented as mean ± SD unless otherwise noted. Boldface indicates statistical significance at P < 0.05.

^a χ ² test for independence unless noted otherwise.

^bWilcoxon rank sum test for difference in means.

Differences in Clinical Outcomes

LOS was found to be significantly different between the SA and GA groups (2.4 ± 2.4 vs 3.4 ± 2.6 days, P = 0.007; Figure 2 and Table 2). Comparing those receiving SA and an ESP block to those receiving GA without an ESP block, LOS was also significantly lower (1.8 ± 1.2 vs 3.6 ± 2.2, P < 0.001; Figure 3). Thirty-day readmission rates were lower in the SA group compared with the GA group (10.0% vs 2.9%); however, this difference was not significant (P = 0.213). Furthermore, there were no cases of reoperation, wound infection, new neurological deficit, or thromboembolic events in either group.

Figure 2.

Intergroup comparisons for significant differences across anesthesia groups. ** and *** represent P < 0.05. LOS, length of stay; MEDD, morphine equivalent daily dose.

Table 2.

Clinical outcomes of patients undergoing single-level transforaminal lumbar interbody fusion.

Outcome Measure	Spinal Anesthesia (n = 34)	General Anesthesia (n = 60) a	P a
Estimated Blood Loss, mL	68.2 ± 51.8	98.0 ± 112.0	0.604
Operating (OR) time, min	211.0 ± 58.4	221.0 ± 114.0	0.593
Length of stay, d	2.38 ± 2.36	3.42 ± 2.57	0.007
30-day readmission, n (%)			0.213 b
No Readmission	33 (97.1%)	54 (90.0%)
Readmission	1 (2.9%)	6 (10.0%)
Opioid usage
Intraoperative MEDD	75.7 ± 139.0	123.0 ± 137.0	<0.001
24 h post-op MEDD	68.7 ± 132.0	41.9 ± 36.7	0.061

Abbreviation: MEDD, morphine milligram equivalent daily dose.

Note: Data presented as mean ± SD unless otherwise noted.

^aWilcoxon rank sum test for difference in means unless otherwise noted.

^bPearson χ² test for independence.

Figure 3.

Intergroup comparisons among patients receiving general anesthesia (GA) and no erector spinae plane (ESP) block compared with those receiving spinal anesthesia (SA) and ESP block, GA with ESP block, and SA with no ESP block, respectively, for (A) length of stay (LOS) and (B) 24-h postoperative morphine equivalent daily dose (MEDD). Data are represented as mean ± SD. All significance values are compared with the GA with no ESP block group as reference and represent a 2-way t test for the difference in means.

There were no statistically significant differences in EBL (68.2 ± 51.8 vs 98.0 ± 112.0 mL, P = 0.604) or OR time (211.0 ± 58 vs 221.0 ± 114 minutes, P = 0.593) between the SA and GA groups (Table 2). Although mean values were numerically lower in the SA group, these differences did not reach statistical significance and may reflect limited sample size rather than true clinical difference.

Differences in Intra- and Postoperative Opioid Utilization

As shown in Table 2, those receiving SA had significantly lower intraoperative opioid usage compared with those receiving GA (75.7 ± 139.0 vs 123.0 ± 137.0, P < 0.001; Figure 2).

Twenty-four-hour usage did not differ significantly between the SA and GA groups (68.7 ± 132.0 vs 41.9 ± 36.7 MEDD, P = 0.061). While the mean postoperative MEDD was numerically higher in the SA group, this difference was not statistically significant.

However, comparing the outcome measures of those receiving SA and an ESP block to those receiving GA without an ESP block, postoperative opioid usage was significantly lower in the SA group (26.4 ± 72.2 vs 58.2 ± 35.3, P = 0.02; Figure 3). When comparing those receiving GA with an ESP block to those receiving GA without an ESP block, postoperative opioid usage was also significantly lower in the GA with ESP block group (11.6 ± 11.2 vs 58.2 ± 35.3, P < 0.001; Figure 3).

Multivariable Regression for Clinical Outcomes

A multiple regression analysis was further performed to identify significant risk factors for clinical outcomes (Table S1). Both anesthesia type (GA vs SA) and additional utilization of an ESP block were significantly associated with several clinical outcomes in the regression analyses, accounting for all other included variables. Compared with GA, SA was associated with lower total intraoperative MEDD (P = 0.007) but higher 24-hour postoperative MEDD (P = 0.009; Figure 4A–4B). Comparatively, the use of an ESP block in the regression analyses was associated with a decreased LOS and lower 24-hour postoperative MEDD (P = 0.044 and P < 0.001, respectively; Figure 4C–4D). Of note, when comparing patients receiving both SA and an ESP block to all other patients, this group was associated with significantly lower LOS and postoperative MEDD (P = 0.030 and P = 0.002, respectively; Table S1).

Figure 4.

Estimated marginal means for outcomes significantly associated with anesthesia type and utilization of an erector spinae plane (ESP) block. (A) Anesthesia type and intraoperative morphine equivalent daily dose (MEDD). (B) Anesthesia type and postoperative MEDD. (C) ESP block and length of stay. (D) ESP block and postoperative MEDD. LOS, length of stay.

Discussion

Comparison of SA and GA

To explore the evolution of surgical and anesthetic techniques in spine surgery, this single-institution analysis compared the use of SA to GA in patients undergoing TLIF. We found potential benefits with the use of SA in spine surgery compared with traditional GA, particularly in reducing total intraoperative MEDD. In addition, we also examined the impact of regional anesthesia in the form of an ESP block and found it to be associated with a shorter LOS and reduced 24-hour postoperative MEDD, among other clinical outcomes. Of note, those receiving a combination of SA and an ESP block demonstrated significantly lower postoperative MEDD and total LOS compared with all other patients.

Previously, Amoroso et al analyzed a propensity score-matched population and found that across more than 150,000 spine surgery cases, regional anesthesia was associated with fewer complications, less blood loss, and shorter LOS than GA.¹¹ Others have examined the use of regional anesthesia in orthopedic surgery and found conflicting results, with some studies similarly indicating significant benefits of SA, such as superior perioperative analgesia, fewer systemic drug effects, and faster mobilization, while others have refuted these results and found no apparent advantages.^12,13 Our findings underscore the impact of anesthesia on TLIF surgery outcomes and the importance of tailoring anesthesia to patient-specific factors. SA may be particularly beneficial for patients at a higher risk of complications from prolonged GA, such as those with cardiopulmonary comorbidities or advanced age. Conversely, GA may remain the preferred option for cases involving prolonged or highly complex procedures, where patient immobility and surgical precision are critical.

The Evolution of Spine Surgery: Comparing GA With SA

Compared with GA, the use of SA was associated with lower intraoperative MEDD but higher 24-hour postoperative MEDD in our multivariate regression model. While there is a limited body of literature investigating intraoperative MEDD usage, the observed reduction in intraoperative MEDD with SA is unsurprising, given that prior research has linked awake surgery using SA to less total time under anesthesia with a reduced need for intraoperative pain management.^8,14–16 Still, by controlling for previously unexplored factors such as OR time and EBL, these findings confirm the value of SA in reducing intraoperative opioid utilization.

The association between SA and higher postoperative MEDD in our regression model, on the other hand, represents a departure from the existing literature. Previous studies have indicated an association between awake surgery under SA and reductions in both postoperative MEDD utilization and lower patient-reported pain scores.^17–19 These trends could reflect institution-specific norms for awake surgery or a reduced gradual taper given improved clinical outcomes in the SA group compared with the GA group. This also suggests that institutions adopting SA techniques should carefully consider postoperative opioid management protocols to fully realize the benefits of awake surgery.

Further Evolutions: Impact of ESP Blocks and Combined Anesthetic Modalities

The utilization of an ESP block was independently found to be significantly associated with a lower LOS and reduced 24-hour postoperative opioid utilization. When comparing patients receiving both SA and an ESP block to those receiving neither SA nor an ESP block, LOS was significantly reduced, and postoperative opioid usage was significantly lower (Figure 3).

These findings indicate that there is likely an exogenous factor driving high postoperative MEDD in the awake group compared with the asleep group, while ESP usage is associated with lower postoperative MEDD. Additional investigation will need to be done to elucidate the precise mechanisms driving these phenomena; however, our data suggest potential benefits to combining SA with ESP blocks for optimizing clinical outcomes. This is confirmed in our regression model, with patients receiving both an ESP block and SA also demonstrating reduced LOS and reduced postoperative opioid usage, compared with those receiving neither intervention. One physiological explanation for this trend may be the fact that SA has been associated with increased hemodynamic stability, which is particularly useful in TLIF procedures.^20,21 These results are largely supported by the existing literature, with prior studies finding a consistent association between ESP block usage and reduced postoperative pain, higher quality of life, lower postoperative opioid utilization, and a reduced need for patient-controlled analgesia.^22–27 Similarly, ESP block utilization has been linked to a reduced hospitalization duration, earlier ambulation, and improved clinical outcomes, supporting our findings relating to LOS.^28,29

Indications for Various Anesthesia Modalities in TLIF

Our findings highlight the ESP block as a potentially valuable tool in enhancing recovery protocols, especially when combined with SA. ESP blocks are particularly well suited for patients at high risk of postoperative opioid dependence, those aiming to minimize hospital stay duration, and those requiring enhanced postoperative mobility for rehabilitation. These benefits stem from the block’s ability to provide effective regional analgesia, reduce opioid requirements, and facilitate earlier ambulation.³⁰ When combined with SA, ESP blocks offer a synergistic effect, further reducing intraoperative opioid use and blood loss while enhancing recovery outcomes.³¹ Thus, our results suggest that ESP blocks are a valuable adjunct to spine surgery, particularly for patients seeking opioid-sparing techniques and faster recovery protocols.

SA, on the other hand, is particularly advantageous for patients at higher risk of complications from prolonged GA, such as those with cardiopulmonary comorbidities or advanced age.³² It is also beneficial in cases eligible for minimally invasive surgery, such as single- or double-level spondylolisthesis, where awake surgery has been associated with greater radiographic improvements.³³ GA, however, remains the preferred modality for prolonged or highly complex procedures requiring complete patient immobility or extensive surgical correction, or when regional techniques are contraindicated, such as with coagulopathies or severe spinal deformities.⁷ Tailoring anesthesia modality based on key patient characteristics ensures optimal outcomes, with SA providing a safer option for select cases and GA remaining essential for more complex surgeries.

Limitations

There are several limitations of this study to note. First and most significantly, this study represents a retrospective analysis of single surgeon TLIF cases, limiting its generalizability. A larger sample size, ideally one that spans multiple institutions, will be a valuable next step. However, the regression results from this study provide particularly novel insights into the impact of different anesthetic modalities on clinical outcomes such as opioid utilization. Although our sample size and available data did not allow for this, a matched cohort study design may also be a useful follow-up study to further elucidate the drivers of specific trends observed in our analysis. Second, the exclusion of reported pain scores may limit the clinical utility of these findings. Due to limited follow-up and high variability, pain scores and disability indices (eg, visual analog scale and Oswestry Disability Index) were not included in this analysis, limiting patient-reported insights. Still, our findings provide a comprehensive overview of how anesthesia, including various subtypes of regional anesthesia, is associated with several notable clinical outcomes.

Conclusion

The use of SA for patients undergoing TLIF is associated with lower intraoperative opioid usage, albeit with potentially higher postoperative opioid usage. When combined with an ESP block and compared against all other patients, however, this combination demonstrated reduced LOS and lower postoperative opioid usage. Although larger studies are needed, this study demonstrates the value of recent surgical and anesthetic evolutions in TLIF surgery.