Clinical Efficacy and Postoperative Nerve Function Recovery of Buyang Huanwu Decoction for Spinal Cord Injury: A Meta-Analysis

Summary

Guided by PRISMA standards, this meta-analysis of nine Randomized Controlled Trials (RCTs) evaluates Buyang Huanwu Decoction as a surgical adjunct. Results confirm its clinical efficacy in enhancing sensory-motor recovery and neurological outcomes for postoperative spinal cord injury patients.

Abstract

This meta-analysis evaluated whether adding Buyang Huanwu Decoction (BYHWD) to standard surgery improves spinal-cord nerve recovery after traumatic spinal-cord injury (SCI). Following PRISMA and PROSPERO protocol CRD42024612859, we systematically searched PubMed, Cochrane, Embase, Web of Science, China National Knowledge Infrastructure (CNKI), Wanfang and VIP to January 2025 for randomized controlled trials (RCTs) comparing surgery plus usual care with the same regimen plus BYHWD. Data were pooled with RevMan 5.3.

Nine Chinese RCTs (614 patients, 306 BYHWD, 308 control) were included. Relative to controls, BYHWD significantly raised post-operative American Spinal Injury Association(ASIA) sensory score (MD 7.51; 95% CI 5.08–9.93; P <0.01), ASIA motor score (MD 7.65; 5.81–9.49; P <0.01), somatosensory evoked potential amplitude (MD 0.37 µV; 0.18–0.55; P <0.01), motor evoked potential amplitude (MD 0.43 µV; 0.19–0.68; P <0.01) and the proportion of patients achieving ≥1 ASIA grade improvement (RR 1.23; 1.09–1.38; P <0.01). Heterogeneity was low for most outcomes. No serious herb-related adverse events were reported. Adjunctive BYHWD appears to accelerate short-term neurological recovery after surgical decompression for SCI. However, all trials were Chinese, most lacked allocation concealment or blinding, follow-up was brief and outcome sets varied. Large-scale, rigorously blinded, multicentre RCTs with long-term functional endpoints are needed before routine clinical adoption.

Introduction

Spinal Cord Injury (SCI) is a devastating central nervous system trauma typically caused by external mechanical forces, resulting in partial or complete loss of sensory, motor, and autonomic functions^1,2. Globally, the incidence and prevalence of SCI are increasing, with an estimated 250,000 to 500,000 new cases reported annually^3,4. In China specifically, the burden is substantial, with approximately 759,300 cumulative cases and 66,374 newly diagnosed instances each year⁵. The average age of onset is 31.7 years, predominantly affecting individuals aged 15–25 years and males (male to female ratio of 4:1). Notably, traffic accidents (36%–48%) are the leading cause, followed by violence (5%­­­­­­­­­­­–29%), falls (17%­–21%), and recreational activities (7%–16%)⁶. SCI imposes a dual burden: a significant physical and psychological toll on patients and a substantial economic strain on society. The Need for Adjuncts Current standard management for SCI primarily involves surgical decompression to restore spinal stability, pharmacotherapy (e.g., corticosteroids) to mitigate secondary inflammation, and comprehensive rehabilitation to maximize functional independence⁷. While surgery is critical for relieving compression, it does not directly facilitate neural regeneration. Traditional Chinese Medicine (TCM) offers potential complementary strategies. Buyang Huanwu Decoction (BYHWD), a classic TCM formulation known for its qi-supplementing and blood-activating properties, is theorized to alleviate the "stasis" and "blockage" associated with SCI in TCM pathology. Previous research indicates that BYHWD can lower post-operative plasma C-reactive protein (CRP) levels, improve neurological scores, shorten hospital stays, and reduce complications⁸.

However, despite its widespread clinical use in China, the current evidence base remains fragmented. While individual randomized controlled trials (RCTs) have reported benefits, there is a lack of high-level evidence quantitatively synthesizing the efficacy of BYHWD specifically as an adjunct to surgical intervention. To address this gap, this study presents the first systematic review and meta-analysis to strictly evaluate the synergistic effect of BYHWD combined with surgical decompression. We hypothesized that adding BYHWD to standard surgical care would significantly improve neurological recovery, specifically ASIA sensory and motor scores and evoked potentials (SEP/MEP)—compared with surgery alone. By strictly adhering to PRISMA guidelines, this study aims to resolve inconsistencies in existing small-scale trials and provide robust, evidence-based recommendations for the integration of TCM into standardized postoperative SCI management.In traditional Chinese medicine (TCM), spinal cord injury is typically categorized as “bi syndrome” or “wei syndrome.” The pathogenesis is primarily associated with trauma-induced damage to the Du meridian, which governs the body’s yang energy. Once damaged, the yang energy cannot be properly distributed, leading to extravasation of blood outside the vessels and the formation of stasis, which obstructs the meridians and deprives the sinews and vessels of nourishment, thereby causing disease. TCM posits that the treatment of spinal cord injury should focus on replenishing and regulating qi, as well as activating blood circulation to resolve stasis, to restore the normal function of the Du meridian and ensure the smooth flow of qi and blood⁹.

Protocol

This systematic review and meta-analysis were conducted in strict accordance with the PRISMA 2020 guidelines. As this study utilizes secondary data from previously published trials, institutional review board (IRB) approval was not required; however, all included original studies adhered to ethical standards involving human subjects. The protocol was prospectively registered with the PROSPERO database immediately upon study initiation (Registration number: CRD42025637800).

Search Strategy and Selection Criteria Two independent researchers performed a comprehensive systematic search for Randomized Controlled Trials (RCTs) investigating Buyang Huanwu Decoction (BYHWD) for spinal cord injury (SCI). The search spanned from the inception of each database to January 2025. Electronic databases included PubMed, Embase, The Cochrane Library, Web of Science, CNKI, VIP, Wanfang, and CBM. The search strategy utilized a combination of controlled vocabulary (e.g., MeSH Terms in PubMed, Emtree in Embase) and free-text terms (Title/Abstract fields). Specifically, the boolean operators AND and OR were used to combine terms related to Spinal Cord Injury (e.g., Traumatic SCI, Spinal Injury), Buyang Huanwu Decoction (e.g., Buyang Huanwu Tang), and Randomized Controlled Trial. Taking PubMed as an example:
#1: ((Spinal Cord Injuries[Mesh] OR Spinal Cord Injury[Mesh] OR Spinal Cord Injuries, Traumatic[Mesh] OR Spinal Cord Injury, Traumatic[Mesh]) OR (Spinal Cord Injury OR SCI OR Spinal Injury))
#2: ((Buyang Huanwu Tang[Mesh] OR Buyang Huanwu Decoction[Mesh]) OR (Buyang Huanwu Tang OR Buyang Huanwu Decoction))
#3: (randomized controlled trial[pt] OR controlled clinical trial[pt] OR randomized[tiab] OR placebo[tiab] OR drug therapy[sh] OR randomly[tiab] OR trial[tiab] OR groups[tiab]) NOT (animals[mh] NOT humans[mh])
#4: #1 AND #2 AND #3

Filters and Limits: The Humans filter was activated where available. No language restrictions were imposed. Export Formats: All retrieved records were exported in standard citation formats (e.g., .nbib, .ris, or .ciw) and imported into reference management software (Table of Materials). Duplicates were removed using EndNote’s automatic deduplication function followed by manual verification.

Study Selection and Data Extraction Outcomes were screened based on strict inclusion criteria: (1) RCT study design; (2) participants diagnosed with SCI undergoing surgical treatment; (3) an experimental group receiving BYHWD plus routine surgical care versus a control group receiving routine surgical care alone; and (4) reporting of quantitative outcome measures including ASIA scores, evoked potentials (SEP/MEP), or clinical efficacy rates. Exclusion criteria encompassed animal studies, non-RCTs, duplicates, and studies with incomplete data. Two researchers independently screened titles, abstracts, and full texts. Disagreements regarding study eligibility were resolved operationally through immediate discussion; if consensus could not be reached, a third senior reviewer (Z.P.W.) made the final decision based on the full-text review. Data extraction was performed using a standardized spreadsheet form (see Table of Materials), capturing first author, year, demographics, interventions, and outcomes.

Risk of Bias Assessment The methodological quality of included RCTs was assessed independently by two reviewers using the Cochrane Risk of Bias Tool (RoB 2). For each included study, detailed justifications were provided for judgments across five domains: (1) Randomization process—sequence generation was considered low risk if computer-generated random numbers or randomization tables were used, and high risk if alternation, date of birth, or case record number was used; allocation concealment was judged low risk if centralized allocation, sealed opaque envelopes, or pharmacy-controlled randomization was employed, and high risk if open allocation schedule or unconcealed envelopes were used; (2) Deviations from intended interventions—blinding of participants and personnel was assessed as low risk for double-blind studies, some concerns for single-blind or objective outcomes, and high risk for open-label studies with subjective outcomes; (3) Missing outcome data—judged as low risk if <5% missing data with intention-to-treat analysis, moderate risk if 5%–20% missing, and high risk if >20% missing or per-protocol analysis with significant attrition (>15% difference between groups); (4) Measurement of the outcome—assessed as low risk if outcome assessors were blinded or outcomes were objective (e.g., mortality), and high risk if assessors were aware of group allocation and outcomes were subjective; (5) Selection of the reported result—evaluated by comparing protocols (if available) or methods sections against reported results, rated high risk if primary outcome was changed or selectively reported. Discrepancies between reviewers were resolved through discussion or consultation with a third reviewer (Sun).

Evidence quality was graded using the GRADE system across four levels: High quality (further research is very unlikely to change our confidence in the effect estimate; started as high for RCTs and rated down for limitations); Moderate quality (further research is likely to have an important impact on confidence and may change the estimate); Low quality (further research is very likely to have an important impact and is likely to change the estimate); and Very low quality (any estimate of effect is very uncertain). Quality was rated down from high for: risk of bias (serious limitations in randomization or blinding in >25% weight of studies), inconsistency (I² > 50% or non-overlapping confidence intervals), indirectness (differences in population, intervention, comparator, or outcome measures), imprecision (optimal information size not met [<300 events for dichotomous outcomes or <400 participants for continuous outcomes] or 95% CI includes both meaningful benefit and harm), and publication bias (evident from funnel plot asymmetry, Egger's test p < 0.10, or <10 included studies with apparent small-study effects).

Statistical Analysis and Software Workflow Meta-analysis was performed using meta-analysis software (Table of Materials). Continuous data (e.g., ASIA scores, SEP/MEP amplitude) were entered as mean and standard deviation (SD) to calculate Mean Differences (MD) with 95% Confidence Intervals (95% CI). For dichotomous outcomes (e.g., efficacy rates), data were entered as event counts (number of events/total sample size) to calculate Risk Ratios (RR) with 95% Confidence Intervals (CI) using the Mantel-Haenszel method; for continuous outcomes (e.g., ASIA scores), the Inverse Variance method was employed to calculate Mean Differences (MD) with 95% CI. All analyses were initially performed using a fixed-effect model in the meta-analysis software (Table of Materials).

Heterogeneity assessment and model selection: Statistical heterogeneity across studies was quantified using the I² statistic and Cochran’s Q test. The pre-specified threshold for significant heterogeneity was set at I² > 50% or P < 0.10. If heterogeneity exceeded these thresholds, indicating that true effect sizes likely varied between studies rather than arising from chance alone, the analysis model was switched to a random-effects model (Der Simonian-Laird method) to provide a more conservative pooled estimate that accounts for between-study variance.

Sensitivity analysis: To test the robustness of findings, particularly for outcomes exhibiting high heterogeneity, sensitivity analyses were conducted by sequentially omitting one study at a time (leave-one-out method) and recalculating the pooled effect estimate for the remaining studies. If the omission of a single study substantially altered the direction or statistical significance of the overall effect (e.g., causing the 95% CI to cross the null value of 1.0 for RR or 0 for MD), that study was considered influential, and the stability of the results was interpreted with caution.

Publication bias assessment: Funnel plots were generated to visually assess publication bias only when at least 10 studies were included in the meta-analysis, as fewer studies result in insufficient power to distinguish true asymmetry from random scatter. When applicable, funnel plot asymmetry was interpreted considering the involvement of small-study effects; for analyses with fewer than 10 studies, publication bias assessment was not performed due to inadequate statistical power.

Results

The final meta-analysis comprised nine randomized controlled trials with 614 well-characterized patients, providing a robust evidence base to test our hypothesis^{8,9,10,11,12,13,14,15,16}. The pooled sample demonstrated bala...

Discussion

Postoperative Assessment of Spinal Cord Injury
Spinal Cord Injury is a severe central nervous system disorder characterized by diverse and complex treatment and assessment methods. In terms of therapeutic approaches, conservative management encompasses pharmacological interventions, such as the use of corticosteroids to alleviate edema and neurotrophic agents to promote nerve repair^7,18. Rehabilitation therapies include physical, occupation...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Chinese National Knowledge Infrastructure (CNKI)	Tongfang Knowledge Network	Database	https://www.cnki.net
Cochrane Library	Wiley	Database	https://www.cochranelibrary.com
Embase	Elsevier	Database	https://www.embase.com
EndNote X9	Clarivate Analytics	Software	https://endnote.com
GRADEpro system tool	Evidence Prime Inc.	Software	https://www.gradepro.org/product
PubMed	National Library of Medicine	Database	https://pubmed.ncbi.nlm.nih.gov
R Software (Version 4.5.2)	R Foundation for Statistical Computing	Software	https://www.r-project.org
Review Manager (RevMan) 5.3	The Cochrane Collaboration	Software	https://training.cochrane.org
Web of Science	Clarivate Analytics	Database	https://www.webofscience.com