Enhanced Recovery after Surgery (ERAS) Reduces Hospital Stay and Postoperative Nausea in ACDF: A Meta-Analysis of 2330 Patients

Abstract

Study Design

Retrospective Study.

Objective

Enhanced Recovery After Surgery (ERAS) is gaining traction in spine surgery, yet high-level evidence specifically for Anterior Cervical Discectomy and Fusion (ACDF) remains sparse. This meta-analysis systematically evaluates the clinical efficacy and safety of ERAS protocols in patients undergoing ACDF.

Methods

Following PRISMA 2020 and AMSTAR-2 guidelines, we searched Embase, PubMed, Cochrane Library, and Web of Science databases from their inception to December 2025. The Newcastle-Ottawa Scale (NOS) was employed to appraise quality. Pooled estimates of mean differences (MDs) or odds ratios (ORs) were derived using a random-effects model. Sensitivity and subgroup analyses were additionally executed. Primary outcomes encompassed postoperative complications and length of stay (LOS) following ACDF with ERAS. The certainty of evidence was assessed using the GRADE (Grading of Recommendations, Assessment, Development, and Evaluation) approach. Statistical analysis was performed using Review Manager 5.4.

Results

Five studies involving 2330 patients (ERAS: 1108; conventional: 1222) were included. All studies demonstrated high methodological quality (NOS 7-9). Implementation of ERAS was associated with a significant reduction in length of stay (MD=-1.56; 95% CI: -1.83 to -1.28; P<0.001; Moderate-certainty evidence) and a markedly lower incidence of postoperative nausea and vomiting(OR: 0.32; 95% CI: 0.15 to 0.68; P=0.03;High-certainty evidence). No significant differences were observed in overall complications(OR: 0.78; 95% CI: 0.37 to 1.63; Low-certainty evidence), hardware failure, surgical site infections, or dysphagia/dysphonia (all P> 0.05). Patient satisfaction showed a non-significant positive trend(OR: 1.31; 95% CI: 0.99 to 1.73; Low-certainty evidence).

Conclusion

Based on moderate-to-high certainty evidence, ERAS protocols significantly optimize perioperative recovery for ACDF by reducing hospital stay and gastrointestinal morbidity without compromising safety or patient satisfaction. These findings support the standardization and broader implementation of ERAS pathways in cervical spine surgery.

Introduction

ACDF is the gold-standard surgical intervention for symptomatic disc disease, aimed at alleviating neurological deficits and restoring spinal stability.^1-3 Despite its high success rate, the perioperative recovery process significantly dictates long-term patient-reported outcomes and quality of life.^4,5 Traditionally, ACDF recovery involves variable management protocols-such as prolonged fasting, strict bed rest, and reliance on opioids-which may exacerbate postoperative discomfort, consequently prolonging hospital stays and imposing a substantial economic burden on healthcare systems. ⁶

Enhanced Recovery After Surgery (ERAS), a multidisciplinary, evidence-based perioperative care model, was introduced to optimize recovery by minimizing surgical stress and stabilizing organ function.^7-10 While ERAS was initially standardized in colorectal surgery, its principles have increasingly been integrated into various orthopedic and neurosurgical procedures.^11-13 However, the implementation of ERAS in ACDF remains a subject of ongoing debate. While several clinical cohorts suggest that ERAS can significantly shorten the length of stay (LOS) without compromising safety, others argue that its benefits regarding patient satisfaction and complication rates-such as surgical site infections or ACDF-specific issues like dysphagia-are not as pronounced as in other surgical fields.^14-19

This inconsistency in the literature creates a significant evidence gap for clinicians and hospital administrators. Although contemporary studies suggest that ERAS may reduce LOS while maintaining safety, a rigorous, systematic evaluation focused exclusively on ACDF is notably absent. To date, no systematic review or meta-analysis has synthesized the comparative efficacy of ERAS versus conventional care specifically for this procedure.

Therefore, this meta-analysis was designed to evaluate the clinical outcomes of ERAS in ACDF cohorts, focusing on LOS, postoperative complications, and patient satisfaction. By addressing the methodological constraints of individual studies, this work constitutes the first comprehensive synthesis of ERAS efficacy specifically for ACDF surgery, providing a higher level of evidence for perioperative optimization.

Methods

This study followed PRISMA 2020 guidelines ²⁰ and AMSTAT-2 (A Measurement Tool to Assess Systematic Reviews) guidelines. ²¹ The meta-analysis was prospectively registered in the PROSPERO website (ID: CRD420251048519).

Search Strategy and Selection Criteria

Two independent reviewers (P.L. and W.Z.) systematically searched PubMed, Embase, the Cochrane Library, and Web of Science from database inception through May 2024. A combination of MeSH terms and keywords was utilized, including “cervical spondylosis,” “Enhanced Recovery After Surgery,” “ERAS,” and “Anterior Cervical Discectomy and Fusion” (or “ACDF”). No language restrictions were applied. References of retrieved articles and relevant reviews were manually screened to identify additional eligible studies. Figure 1 shows the PRISMA flow diagram.

Figure 1.

Study selection process. The flowchart showed the selection of included studies for meta-analysis

Eligibility Criteria

Inclusion criteria were defined as: (1) Study Design: Randomized controlled trials (RCTs) or non-randomized comparative cohort studies; (2) Population: Adult patients (≥18 years) undergoing ACDF for degenerative cervical diseases; (3) Intervention: Implementation of a multidisciplinary ERAS protocol; (4) Comparator: Conventional perioperative care; (5) Outcomes: Reporting of at least one primary outcome (Length of Stay or postoperative complications). Exclusion criteria included: (1) Case reports, case series (n < 10), reviews, editorials, and conference abstracts; (2) Studies involving pediatric populations, revision surgeries, or emergency procedures; (3) Studies with insufficient data for effect size calculation or inaccessible full texts.

Data Extraction and Quality Assessment

Data were extracted independently by two researchers using a standardized form, encompassing study characteristics (author, year, design), patient demographics (age, ASA grade), specific ERAS components, and clinical outcomes. Discrepancies were resolved through consensus or consultation with a senior investigator (L.Q.C.).

Methodological quality of observational studies was appraised via the Newcastle-Ottawa Scale(NOS), ²² while RCTs were assessed for bias risk using the Cochrane Collaboration tool. ²³

The certainty of evidence for each outcome was assessed using the Grading of GRADE approach. Additionally, we qualitatively synthesized the components of ERAS protocols across the included studies to evaluate the intensity of interventions.

Statistical Analysis

Statistical analyses were performed using Review Manager software (version 5.4). ²⁴ For dichotomous variables, ORs with 95% Confidence Intervals (CIs) were calculated using the Mantel-Haenszel method. For continuous variables, MDs with 95% CIs were derived. A random-effects model was preferentially utilized to provide conservative estimates regardless of statistical heterogeneity. ²⁵

Heterogeneity was quantified using I ² statistics, ²⁶ with I ² >50% indicating significant heterogeneity. Sensitivity analysis was performed by sequentially omitting individual studies to evaluate the stability of the pooled results. Potential publication bias was assessed via Egger’s test and visual inspection of funnel plots if the number of included studies exceeded ten. ²⁷ Statistical significance was defined as two-tailedP<0.05.

Results

Literature Search

The PRISMA ²⁸ flow diagram depicts the literature screening process(Figure 1). Initial database searches retrieved 59 records, with duplicates removed and eligibility screening (title/abstract review) excluding 47 non-relevant studies. Full-text review of 12 articles led to exclusion of 7 for unmet criteria, resulting in 5 studies^29-33 eligible for meta-analysis. The reasons for exclusion after full-text review are listed in eTable 1 in Supplement 1. A total of 2330 patients were included (ERAS group: n=1108; conventional care group: n=1222). The studies’ procedures of ERAS are listed in eTable 2 in Supplement 1, respectively.

Characteristics of Included Studies

As summarized in Table 1, five retrospective cohort studies published between 2019 and 2025 were included in this meta-analysis.^29-33 Geographically, three studies were conducted in China^31-33 and two in France.^29,30 The total population comprised 2330 patients, with 1108 in the ERAS group and 1222 in the conventional care (non-ERAS) group.

Table 1.

Characteristics of the Studies Included in the Meta-analysis(Non-ERAS/ERAS)

Author-year	Journal	Country	Study type	Age(year),mean ± SD	Sample size(n)	Sex(male/female)	Body mass index	Smoker(n)	ASA 1
Xue Leng,2022	BMC Musculoskeletal Disorders	China	Cohort study	52.07±10.6/53.2±9.3	73/70	36/37 vs 39/31	24.65 ± 3.2/23.94 ± 3.0	-	0/2
Bertrand Debono, 2019	Neurosurgical focus	France	Cohort study	47.6±9.9/48.7±8.7	749/612	405/342 vs 312/300	25.3 ± 5.1/27.3 ± 5.0	382/287	427/320
Bertrand Debono, 2021	European Spine Journal	France	Cohort study	48.7±9.2/48.5±10.6	202/202	106/96 vs 103/99	25.4 ± 4.7/25.4 ± 4.6	106/89	118/118
Peng Wang, 2023	Clinical Interventions in Aging	China	Cohort study	66.03±5.71/67.28±5.84	122/135	43/79 vs 51/84	24.66±3.52/25.13±3.08	20/33	5/5
Guilin Liu, 2025	Frontiers in Medicine	China	Cohort study	55.81 ± 11.98/59.15 ± 10.52	76/80	46/30 vs 54/26	Not Reported	Not Reported	68/75

The mean age of the participants ranged from 47.6 to 67.28 years, providing a broad representation across middle-aged and elderly populations. Notably, Wang et al (2023) ²⁰ focused specifically on an older cohort (mean age > 66 years). Gender distribution was relatively balanced across the studies, with a total of 1210 males and 1120 females. Baseline characteristics, including Body Mass Index (BMI) and American Society of Anesthesiologists (ASA) physical status, were generally comparable between the ERAS and control groups within each study. Follow-up durations across the studies varied from 30 days to 12 months (Debono 2021: 12 months; Liu 2025: 4.38 months; Wang 2023: 30 days), while two studies (Debono 2019, Leng 2022) did not explicitly specify the long-term follow-up period. Despite this variation, all studies provided sufficient perioperative data to assess the primary short-term outcomes (length of stay and PONV) and specific complication measures (e.g., hematoma, dysphagia, and infection).

Risk of Bias Assessment

Methodological quality of the four included observational studies was assessed using the NOS, with all scores ranging from 7 to 9 stars, indicating high quality (Table 2).

Table 2.

Risk of Bias of Cohort or Case–control Studies (NOS Evaluation)

study, publication year	Representativeness of cohort	Ascertainment of exposure	Selection of nonexposed cohort	Outcome lacking at the beginning	Comparability of cohorts	Outcome assessment	Sufficient follow-uptime	Follow-upadequacy	Score
Xue Leng,2022	*	*	*	*	**	*	*	*	9
Bertrand Debono,2019	*	*	*	*	**	*	*	*	9
Bertrand Debono,2021	*	*	*	*	**	*	*	*	9
Peng Wang, 2023	*	*	*	*	**	*	*	*	9
Guilin Liu, 2025	*	*	*	*	**	*	*	*	9

Note. Risk of bias was assessed using the Newcastle-Ottawa Scale (NOS). A study can be awarded a maximum of one star ( ) for each numbered item within the Selection and Outcome categories. A maximum of two stars (**) can be given for Comparability, indicating that the study controlled for both the most important confounding factor (e.g., age or sex) and any additional factor (e.g., BMI or comorbidities). A total score of ≥7 indicates high methodological quality.*.

Primary Outcome: LOS

All five studies reported LOS. Random-effects meta-analysis revealed that ERAS protocols significantly reduced LOS by 1.56 days (95% CI: -1.83 to -1.28; P< 0.00001), despite significant heterogeneity (P< 0.00001, I ² = 78%) (Figure 2). Subgroup analysis by ethnicity confirmed the consistency of this effect, with significant reductions in both Caucasian (MD: -1.69; 95% CI: -1.90 to -1.48) and Asian populations (MD: -1.44; 95% CI: -2.01 to -0.86). No significant interaction was found between subgroups (P= 0.42).

Figure 2.

Comparison of LOS between ERAS group and non-ERAS group. LOS=length of stay; CI = confidence interval, and df = degrees of freedom

Complications

The incidence of overall complications was reported in all five studies(Table 3). In our analysis, we classified surgical complications into major and minor categories based on their clinical severity and impact on recovery. Major complications included prolonged dysphagia, hardware failure, dyspnea (or bronchial aspiration), neck hematoma, new neurological deficits, and surgical site infections. Minor complications primarily included transient dysphagia/dysphonia and postoperative nausea and vomiting (PONV). No statistically significant difference was observed between the ERAS and conventional groups (OR:0.78; 95% CI: 0.37 to 1.63; P=0.51; heterogeneity: Chi²=21.23, df = 4,P _{heterogeneity} = 0.003, I ² = 81%). Similarly, ERAS did not significantly impact the rates of major complications (OR: 0.94, 95% CI: 0.42 to 2.11, P= 0.88; I ² = 59%) or minor complications (OR: 0.81, 95% CI: 0.42 to 1.58, P= 0.54; I ² = 65%), with general heterogeneity across all analyses (Figure 3).

Table 3.

Complications and the Number of Cases in the Included Studies

Complication	Study	No. of cases (ERAS VS non-ERAS)
Bronchial aspiration or Pneumonia	Xue Leng, 2022; Bertrand Debono,2021; Peng Wang, 2023	0/3
Heart disease	Peng Wang, 2023	3/0
Neck hematoma	Bertrand Debono,2021	2/0
Deep vein thrombosis	Peng Wang, 2023	1/0
Spinal fluid leakage	Peng Wang, 2023	1/1
New neurological deficit	Peng Wang, 2023; Bertrand Debono,2021	7/5
Nausea; Emesis	Xue Leng, 2022; Peng Wang, 2023;Guilin Liu, 2025	10/27
Surgical site infection	Bertrand Debono,2021; Peng Wang, 2023; Bertrand Debono,2019;Guilin Liu, 2025	6/7
Dysphagia/dysphonia	Xue Leng, 2022; Bertrand Debono,2021;Guilin Liu, 2025	13/20
Hardware failure	Xue Leng, 2022; Bertrand Debono,2021	3/4
Prolonged dysphagia	Xue Leng, 2022	1/9

Figure 3.

Comparison of complications between ERAS group and non-ERAS group. CI = confidence interval, and df = degrees of freedom

Surgical Site Infection

Three studies reported surgical site infection outcomes, with a total of 10 cases observed (6/958 patients in the ERAS group vs. 4/1073 patients in the non-ERAS group). The meta-analysis demonstrated no statistically significant difference in the rate of surgical site infection between the two groups (OR: 1.52; 95% CI: 0.42 to 5.44, P= 0.52; heterogeneity: Chi²= 0.05, df = 2,P_{heterogeneity}= 0.98, I ² = 0%). Subgroup analyses confirmed the consistency in the results, with no evidence of increased risk attributable to either intervention(eFigures 1 in Supplement 1).

Hardware Failure

Two studies documented hardware failure, with 3 cases (3/272) reported in the ERAS group and 4 cases (4/275) observed in the non-ERAS group. No significant difference was observed in the occurrence of hardware failure between the groups (OR: 0.77, 95% CI: 0.17 to 3.52, P= 0.73; heterogeneity: Chi²= 0.05, df = 1,P_{heterogeneity}= 0.82, I ² = 0%) (eFigures 2 in Supplement 1).

Bronchial Aspiration

Two studies documented bronchial aspiration, with 0 cases (3/272) reported in the ERAS group and 2 cases (4/275) observed in the non-ERAS group. No significant difference was observed in the occurrence of bronchial aspiration between the groups (OR: 0.34; 95% CI:0.03 to 3.27, P= 0.73; heterogeneity: Chi ² = 0.00, df = 1,P_{heterogeneity}= 0.99, I ² = 0%) (eFigures 3 in Supplement 1).

New Neurological Deficit

Two studies documented new neurological deficit, with five cases (5/337) reported in the ERAS group and seven cases (7/324) in the non-ERAS group. No significant difference was observed in the occurrence of new neurological deficit between groups (OR: 1.37, 95% CI: 0.43 to 4.39; P= 0.60; heterogeneity: Chi ² = 0.07, df = 1,P_{heterogeneity}= 0.79, I ² = 0%) (eFigures 4 in Supplement 1).

Dysphagia/Dysphonia

No significant difference was observed in the occurrence of dysphonia between groups (OR: 0.64; 95%CI: 0.29 to 1.39, P= 0.26; heterogeneity: Chi ² = 3.67, df = 1, P_{heterogeneity}= 0.06, I ² = 73%). Two studies reported dysphagia/dysphonia (ERAS group, 11/691; non-ERA group, 18/822). These findings suggest comparable rates of dysphagia/dysphonia, with no evidence of intervention-related effects (eFigures 5 in Supplement 1).

Postoperative Nausea and vomiting(PONV)

The ERAS group demonstrated a significantly lower incidence of PONV (ERAS: 10/285 vs control: 27/271) compared to conventional care (OR: 0.32; 95% CI: 0.15 to 0.68; P= 0.003), with no observed heterogeneity (I ² =0%, Figure 4).

Figure 4.

Comparison of PONV between ERAS group and non-ERAS group. PONV=postoperative nausea and vomiting; CI = confidence interval, and df = degrees of freedom

Satisfaction

Patient satisfaction was assessed in three studies using varying instruments. The pooled results indicated a trend toward higher satisfaction in the ERAS group, although this did not reach statistical significance (OR: 1.31, 95% CI: 0.99 to 1.73, P= 0.06; heterogeneity: Chi ² = 4.37, df = 2,P_{heterogeneity}= 0.24, I ² = 29%, Figure 5).

Figure 5.

Comparison of satisfaction between ERAS group and non-ERAS group. CI = confidence interval, and df = degrees of freedom

Sensitivity Analysis and Publication Bias

To evaluate the robustness of the pooled estimates, leave-one-out sensitivity analyses were performed for LOS, overall complications, PONV, and patient satisfaction.

For the primary outcome of LOS, the significant reduction in the ERAS group remained highly stable across all iterations (MD range: -1.46 to -1.71 days; all P< 0.001). Notably, the exclusion of Xue Leng (2022) ²⁰ resulted in a substantial decrease in statistical heterogeneity, with I2 dropping from its original level 78% to 25.6% (eTable 3). This identifies that specific study as the primary source of heterogeneity, while the overall clinical benefit of ERAS on LOS remains robust.

For overall complications, the meta-analysis results were non-significant and stable across all iterations (P-value range: 0.30 to 0.83). Similar to the LOS analysis, the omission of Xue Leng (2022) ²⁰ markedly reduced the heterogeneity from 81% to 55.5% (eTable 4), without altering the non-significant direction of the effect.

Regarding PONV, the results demonstrated remarkable consistency. The significant advantage of ERAS in reducing PONV persisted regardless of which study was omitted, with I² remaining at 0.0% throughout the analysis (eTable 5), further validating the reliability of this multidisciplinary intervention.

Crucially, the sensitivity analysis for patient satisfaction yielded a noteworthy finding. While the original pooled result showed a non-significant trend (P = 0.06), the exclusion of Bertrand Debono (2021) shifted the result to statistical significance (OR: 1.41; 95% CI: 1.05 to 1.90; P= 0.022) and completely eliminated the heterogeneity (I ² =0.0%) (eTable 6). This suggests that the impact of ERAS on satisfaction is generally positive, but its statistical significance in the overall analysis was likely attenuated by the results of the aforementioned study. Due to the limited number of studies (<10), publication bias was assessed via visual inspection of funnel plots, which revealed generally symmetrical distributions.Visual inspection of funnel plots for LOS and overall complications indicated a generally symmetrical distribution, suggesting minimal publication bias (eFigures 6-8 in Supplement 1).

ERAS Components Analysis and Certainty of Evidence

As shown in Table 4, the adherence to ERAS components varied across studies, with scores ranging from 9 to 13. All five studies consistently implemented preoperative counseling, early oral intake, and early mobilization. However, intraoperative goal-directed fluid therapy and standardized swallowing drills remained less frequently adopted. According to the GRADE assessment, the certainty of evidence was High for PONV reduction and Moderate for LOS shortening (Table 5). The certainty for overall complications and satisfaction was Low, primarily due to significant statistical heterogeneity and imprecision (Table 5).

Table 4.

Detailed Components of ERAS Protocols Across Included Studies (NP,Not Reported)

Phase	ERAS Components (interventions)	Xue Leng, 202	Debono 2019	Peng Wang, 2023	Bertrand Debono,2021	Guilin Liu, 2025
Preoperative	Detailed Patient Counseling: Psychological preparation & expectations	√	√	√	√	√
	Optimized Fasting: 6h for solids, 2h for clear liquids	√	√	√	√	√
	Carbohydrate Loading: Pre-op carbohydrate drinks (2h before)	√	√	√	√	×
	Preemptive Analgesia: Pre-op NSAIDs or Gabapentinoids	√	NP	√	√	×
	Prehabilitation exercises((e.g., cervical range-of-motion training, swallowing drills))	NP	NP	×	NP	√
Intraoperative	Short-acting Anesthesia: Use of Propofol/Remifentanil (TIVA)	√	NP	√	NP	×
	Local Anesthetic Infiltration: Incision site (e.g., Bupivacaine)	√	NP	√	NP	×
	Hemodynamic stability(within ±20% of baseline values)	NP	NP	NP	NP	√
	Prophylactic antiemetics (ondansetron 4 mg and dexamethasone 5 mg)	√	NP	NP	NP	√
	Fluid Management: Goal-directed or restricted fluid therapy	NP	NP	NP	NP	×
	Normothermia Maintenance: Use of forced-air warming blankets	√	NP	√	NP	√
	No Routine Urinary Catheterization: Or early removal	×	√	NP	√	×
Postoperative	Multimodal Analgesia: Scheduled non-opioids (NSAIDs/Acetaminophen)	√	√	√	√	√
	Early Oral Intake: Sips/solids initiated within 2–6 hours	√	√	√	√	√
	Early Mobilization: Out-of-bed activity within 6h or same day	√	√	√	√	√
	Standardized Drain Management: No drain or removal within 24h	√	√	√	√	NP
	PONV Prophylaxis: Use of Dexamethasone or 5-HT3 antagonists	NP	NP	NP	NP	×
	Structured Discharge Criteria: Based on functional milestones	√	√	NP	√	×
Total Counts	Adherence Score (Sum of ✓)	13	9	11	10	10

Table 5.

Summary of Findings (GRADE): ERAS vs. Conventional Care for ACDF

Outcomes	Anticipated absolute effects		Relative effect(95% CI)	No. of participants (studies)	Quality of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
LOS	The mean day in the control groups was 5.91day	The mean LOS in the intervention groups was 1.56 lower (1.83 lower to 1.28 lower)	​	2330(5)	⨁⨁⨁◯ Moderate Upgraded due to large magnitude of effect.	Random-effects estimate presented due to significant heterogeneity
PONV	100 per 1000	34 per 1000 (70 to 16)	0.32[0.15 to 0.68]	556(3)	⨁⨁⨁⨁ High Upgraded due to very large magnitude of effect.	Fixed-effects estimate presented due to low heterogeneity
Complication	96 per 1000	75 per 1000 (156 to 35)	0.78[0.37 to 1.63]	2330(5)	⨁⨁◯◯ Low	Random-effects estimate presented due to significant heterogeneity
Satisfaction	885 per 1000	910 per 1000 (930 to 884)	1.31[0.99 to 1.73]	1826(3)	⨁⨁◯◯ Low	Random-effects estimate presented due to significant heterogeneity

The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; RR: Risk Ratio; MD: Mean Difference.

^aDowngraded one level for inconsistency: Significant statistical heterogeneity was observed (I²>75%).

^bUpgraded one level for large magnitude of effect: The reduction in LOS (1.56 days) is considered clinically substantial for ACDF.

^cBaseline risk: Estimated based on the average incidence in the conventional care groups of included studies.

^dUpgraded one level for large magnitude of effect: OR < 0.5 suggests a potent protective effect of ERAS on PONV.

^eDowngraded one level for imprecision: The 95% confidence interval crosses the null effect line (1.0).

^fObservational studies: Starting certainty is Low by default in the GRADE framework.

In summary, the meta-analysis robustly demonstrates that ERAS protocols significantly accelerate hospital discharge and reduce PONV without increasing the incidence of major or minor perioperative complications. These trends underscore the safety and short-term efficacy of ERAS in ACDF.

Discussion

Principal Findings and Clinical Efficacy

This meta-analysis, involving 2330 patients across diverse healthcare systems in China and France, confirms the safety and efficacy of ERAS protocols in ACDF. The most striking findings are the significant reduction in hospital length of stay (LOS) (MD: -1.56 days) and the marked decrease in postoperative nausea and vomiting (PONV) incidence compared to conventional care. By breaking down the ‘black box’ of ERAS, we identified that early mobilization and opioid-sparing analgesia were the most commonly shared elements, potentially serving as the primary drivers for the 1.56-day reduction in LOS. Importantly, these benefits were achieved without increasing overall complications or compromising patient satisfaction. Our findings reinforce the generalizability of ERAS for ACDF surgery, suggesting that standardized perioperative care can optimize recovery across distinct healthcare economies.^29-33

Mechanisms for Reducing LOS and Surgical Stress

The strategic reduction in LOS observed in the ERAS group reflects the operationalization of core principles: attenuating surgical stress and accelerating functional rehabilitation. ³⁴ By integrating preoperative comorbidity optimization, intraoperative neural preservation, and postoperative opioid-sparing analgesia, ERAS pathways optimize the perioperative trajectory.^35,36 Given that prolonged hospitalization imposes a significant economic burden and resource utilization, ¹⁵ the 1.56-day reduction found in our study highlights the potential for ERAS to enhance healthcare efficiency in spine surgery. Furthermore, early mobilization and protocol-driven drain removal-principles consistent with trauma and orthopedic recovery-are critical to preventing immobilization-related complications.^37-39 Specifically, among the included studies, the routine use of drains was either strictly prohibited (Debono 2019, 2021) or limited to early removal within 24-48 hours (Wang 2023, Leng 2022). Avoiding or early removal of drains not only facilitated same-day mobilization but also safely reduced LOS without increasing the risk of epidural hematoma.

Postoperative Nausea and Vomiting (PONV)

The significantly lower PONV incidence in the ERAS cohort (OR: 0.32) underscores the efficacy of multimodal interventions. This advantage is likely attributed to two synergistic mechanisms: first, the mitigation of surgical stress through optimized preoperative counseling and early mobilization; second, a standardized analgesic regimen that reduces the cumulative dosage of opioid analgesics.^17-19 These findings align with existing evidence that protocol-driven stress reduction and opioid-sparing strategies are pivotal in minimizing postoperative gastrointestinal adverse events.

Safety Profile and ACDF-specific Complications

Safety remains the paramount concern in cervical spine surgery. Postoperative neck hematoma is a rare (0.2%-5.6%) but potentially catastrophic complication due to acute airway compression.^30,40 While our study observed no cases of respiratory distress, the numerical decrease in hematoma incidence may be related to improved intraoperative hemostasis and early rehabilitation within ERAS frameworks. Similarly, dysphagia and dysphonia are common but transient after ACDF, with reported rates varying widely (1.7%-71%) due to inconsistent definitions. ⁴¹ Our analysis confirms that ERAS protocols do not exacerbate these soft-tissue-related morbidities, even in multilevel procedures requiring extended dissection.^42,43

General Complications and Surgical Risks

Regarding surgical site infections (SSI), our observed rate (0.37%-0.63%) was slightly higher than some previous reports (0.18%), ⁴⁴ yet showed no significant difference between cohorts. This underscores the importance of preoperative optimization, including potential hidden risks such as oral health screening, which has been shown to reduce infection rates in joint surgeries.^45-47 Furthermore, complications like hardware failure, pneumonia, and new neurological deficits remained comparable between groups. ⁴⁸ As new neurological deficits are often associated with operative skill or intraoperative neurophysiological monitoring rather than perioperative care, our findings confirm that ERAS does not compromise surgical safety.^49,50

Patient Satisfaction and Future Perspectives

The primary goals of ERAS are to reduce morbidity and improve satisfaction. Although the ERAS group showed a positive trend toward higher satisfaction, the difference lacked statistical significance (P=0.06), aligning with findings by Debono et al ³⁰ This may be due to the subjective nature of satisfaction metrics and the lack of standardized assessment tools. ⁸ Future trials should adopt multidimensional tools like the Neck Disability Index (NDI) or EQ-5D to better quantify patient-centric outcomes.

Critical Appraisal of ERAS Protocols and Healthcare Systems

It is essential to critically appraise the data within the context of different healthcare systems and the variability of ERAS protocols. The baseline LOS in conventional care varied significantly among the included studies, largely reflecting regional healthcare practices. For instance, studies conducted in China often report a longer baseline LOS compared to Western countries due to cultural expectations of in-hospital convalescence and different reimbursement structures. Conversely, outpatient ACDF is becoming increasingly common in the United States, a trend not fully captured in our dataset which exclusively comprised studies from France and China. Furthermore, the specific components of the ERAS protocols—such as the exact pharmacological agents used for analgesia or the criteria for drain removal—were not uniformly standardized across the studies. This variability in implementation strategies makes it challenging to isolate which specific ERAS component contributes most significantly to the observed outcomes, thereby affecting the overall reproducibility.

Heterogeneity and Limitations

The considerable heterogeneity observed in LOS and overall complications may be attributed to variations in surgical techniques, surgeon experience, and baseline patient characteristics across institutions. While the use of a random-effects model and subgroup analyses by ethnicity addressed some of this variability, the findings should be interpreted with caution. Limitations of this study include the relatively small number of eligible studies (n=5) and the inherent selection bias in non-randomized comparative studies. A critical limitation of the current literature is the exclusive focus on short-term perioperative metrics. None of the included studies reported long-term radiological outcomes, such as cervical fusion rates, nor did they employ validated long-term functional assessment tools (e.g., Neck Disability Index). Consequently, while ERAS safely accelerates early recovery, its impact on long-term surgical success and durable patient satisfaction remains to be established in future longitudinal trials.Finally, the geographic concentration of these studies in China and France may restrict the external validity of our findings, as these results may not be directly extrapolatable to other healthcare systems, such as that in the United States, where outpatient ACDF pathways are more prevalent.

Conclusions

Current evidence suggests that ERAS protocols for ACDF are safe and effective, significantly reducing hospital length of stay and the incidence of postoperative nausea and vomiting compared to conventional care. Importantly, these benefits are achieved without increasing overall complication rates or compromising patient satisfaction. While these findings highlight the potential clinical utility of ERAS pathways in cervical spine surgery, the routine standardization and broad implementation of these protocols should be approached with caution given the limitations of the current retrospective literature. Future high-quality randomized controlled trials with standardized outcome metrics are warranted to further optimize procedure-specific interventions and assess long-term clinical superiority.

ORCID iD

Lei Peng https://orcid.org/0000-0002-3654-1808

Ethical Considerations

As this study is a systematic review and meta-analysis of previously published literature, institutional review board (IRB) approval and ethical clearance were not required. This study does not involve direct contact with human participants or animals.

Consent to Participate

This study utilized anonymized data from previously published studies. Informed consent was obtained by the original investigators of the included studies in accordance with their respective institutional ethical standards.

Data Availability Statement

Via corresponding author:penglei23@mail2.sysu.edu.cn.