Transcutaneous auricular vagus nerve stimulation for postoperative pain: a protocol for a systematic review and meta-analysis

Abstract

Introduction

Postoperative pain is common after surgery, with a high incidence and risk of becoming chronic. Current multimodal analgesia has drawbacks, including limited efficacy from single agents and opioid side effects and addiction risk. These issues have led to opioid-sparing multimodal analgesia. Transcutaneous auricular vagus nerve stimulation (taVNS) is non-invasive and convenient. Studies have shown it can reduce postoperative pain, improve mood and lower adverse events. However, taVNS lacks a comprehensive evaluation and standardised protocols, so further research is needed to provide reliable evidence.

Methods and analysis

This study strictly adheres to the guidelines of the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols. To identify suitable randomised controlled trials (RCTs), eight credible databases will be searched, including four English databases (Web of Science, PubMed, Cochrane Central Register of Controlled Trials, EMBASE) and four Chinese databases (China National Knowledge Infrastructure, VIP Database for Chinese Technical Periodicals, Wanfang Database, Chinese Biomedical Literature Database). RevMan V.5.3 will be employed to integrate the retrieved data and conduct meta-analyses. The methodological quality of included RCTs will be evaluated using the Cochrane Risk of Bias Assessment 2.0 tool. Additionally, the Grading of Recommendations, Assessment, Development and Evaluation system will be applied to assess the strength and certainty of the evidence. We will also conduct publication bias analyses, sensitivity analyses and subgroup analyses.

Ethics and dissemination

No ethical review is required as no private or confidential patient data will be included. Results of this study will be disseminated through a peer-reviewed journal.

PROSPERO registration number

CRD420251207651.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This study will include multiple outcomes regarding the condition of postoperative pain.

• This study will employ subgroup analyses and meta-regression to address heterogeneity arising from variations in transcutaneous auricular vagus nerve stimulation parameters and study populations.

• This study will limit the search scope to Chinese and English databases, which may introduce language bias.

• Small-sample trials may limit the statistical power of this research, thereby affecting the robustness of the results.

Introduction

Postoperative pain, as one of the most common complications of surgical operations, is a complex physiological and psychological response of the body to surgical trauma.¹ Postoperative pain is characterised by a high incidence rate and a risk of chronic transition. Acute pain occurring after surgery is widespread, and approximately 10% of cases progress to chronic postoperative pain.² Long-term pain distress triggers anxiety in patients and may lead to pathological changes in multiple systems, such as the central nervous system, circulatory system and endocrine system, significantly delaying the postoperative rehabilitation process.³ It is worth noting that the risk of chronic transition of postoperative pain is influenced by multiple factors, including age, underlying diseases, surgical type and pain management protocols.⁴ This underscores the necessity of multimodal analgesic interventions.⁵

Currently, postoperative pain management mainly involves the combined use of drugs with different mechanisms of action, such as non-steroidal anti-inflammatory drugs (NSAIDs), local anaesthetics and opioid drugs, to block pain signal transmission at multiple targets, thereby achieving better analgesic effects.^{6 7} However, this type of treatment still has significant limitations: first, the analgesic effect of a single drug is relatively limited, making it difficult to meet the needs of controlling high-intensity pain⁸; second, although opioid drugs have a significant analgesic effect, the adverse reactions they cause, such as nausea and respiratory depression, as well as the potential risk of addiction, seriously affect patient safety and rehabilitation quality⁹; third, some patients have a poor response to existing treatment regimens, resulting in unsatisfactory pain control effects.¹⁰ Based on this current situation, the innovative concept of opioid-sparing multimodal analgesia has been proposed. This concept suggests that by integrating a variety of analgesic methods to systematically interfere with multiple transmission pathways of pain signals, it can effectively relieve the patient’s pain symptoms and significantly reduce the dosage of opioid drugs, thereby reducing the risk of side effects that may be caused by them.¹¹

Transcutaneous auricular vagus nerve stimulation (taVNS) is an intervention method that treats diseases by electrically stimulating the skin area innervated by the auricular branch of the vagus nerve in the external ear.¹² Due to its convenience and non-invasiveness, it is widely used in clinical practice.^{13 14} Studies have shown that taVNS exerts analgesic effects through multiple pathways, including regulating autonomic nervous function, activating anti-inflammatory pathways and modulating cortical spreading depression. Moreover, taVNS exhibits comparable efficacy to traditional invasive vagus nerve stimulation.¹⁵ Several randomised controlled trials (RCTs) have investigated the therapeutic effect of taVNS on pain after different types of surgeries. The findings revealed that taVNS provides good relief for postoperative pain of various types, alleviates patients’ anxiety to a certain extent, reduces the incidence of other postoperative adverse events and plays a positive role in postoperative recovery.16,19 There are certain variations in the selection of treatment parameters (eg, frequency, pulse width) among these studies. To date, no research has yet summarised the optimal parameters of taVNS for postoperative pain management.

Existing studies have shown that taVNS exerts favourable therapeutic effects on both acute and chronic pain.20,22 A systematic review and meta-analysis indicated that there are significant variations in the therapeutic effect of taVNS among patients with acute pain, which mainly depend on the combination of medications and the timing of postoperative assessment. For patients with chronic pain, studies have shown that taVNS can sustainably alleviate pain over a longer period, demonstrating its good tolerability.²³ Although systematic reviews and meta-analyses on taVNS for acute and chronic pain have been published, the complexity of postoperative pain warrants further in-depth investigation. Current clinical studies generally support the efficacy of taVNS; however, there is a lack of comprehensive evaluation and standardised protocols. Therefore, further research is needed to determine the appropriate parameter selection and provide reliable evidence for multimodal analgesia aimed at reducing opioid use.

Methods and analysis

Objectives

This article is a protocol developed for a subsequent systematic review and meta-analysis, which aims to synthesise the current evidence regarding the efficacy and optimal parameters of taVNS in the treatment of postoperative pain.

Methods and analysis

To enhance the openness and transparency of this study, the research strictly adheres to the guidelines of the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols statement²⁴ and has completed registration in the International Prospective Register of Systematic Reviews (PROSPERO) with registration number CRD420251207651. This systematic review and meta-analysis is scheduled to commence in March 2026, with the draft paper to be completed in May 2026.

Patient and public involvement

Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

Eligibility criteria

The PICOS (Participants, Interventions, Comparisons, Outcomes, and Study design) framework will be applied to establish trial eligibility in this systematic review and meta-analysis.

Study designs

Our systematic review and meta-analysis will only include RCTs designed to investigate the efficacy of taVNS in the treatment of postoperative pain. The languages of the included studies will be limited to Chinese and English. This systematic review and meta-analysis will exclude non-RCTs, case series studies, reviews and other study types that do not meet the inclusion criteria.

Participants

Eligible participants will be patients with postoperative pain who meet the following criteria: (1) aged 18 years or older; (2) have undergone surgical procedures; (3) experience different degrees of pain after surgery; and (4) have a physical status classification of I–III according to the American Society of Anesthesiologists. Patients will be eligible regardless of their age, gender, country, ethnicity or source of cases.

Exclusion criteria include preoperative chronic pain, neurodermatitis of the ear, neuromuscular disorders, history of substance abuse (including opioids) and severe hepatic or renal insufficiency.

Interventions

The main intervention for the treatment group is taVNS, with no specific requirements imposed on the frequency, pulse width, intensity or stimulation duration of taVNS. The treatment group may use taVNS alone or in combination with medications administered in the control group.

Comparators

1. Treatment with analgesic medications (eg, NSAIDs, opioids).

2. Placebo control: sham taVNS, placebo medications.

3. Waiting list: no intervention.

Outcomes

After reviewing relevant references^{23 25} to identify outcome measures relevant to the methodology of this study, eligible trials should report at least one of the following outcome measures.

Primary outcomes

Pain intensity measured by any pain assessment scale (eg, Visual Analog Scale, Numerical Rating Scale, Faces Pain Scale-Revised), with a focus on score data at key time points: 24 hours postoperatively, 72 hours postoperatively, at discharge and 3 months postoperatively.

Secondary outcomes

1. Cumulative consumption of analgesic medications within a specified time frame.

2. Anxiety-related and depression-related scale scores (eg, Self-Rating Anxiety Scale,⁷ Self-Rating Depression Scale, Hamilton Depression Rating Scale, Beck Depression Inventory).

3. Sleep quality scores (eg, Pittsburgh Sleep Quality Index, Athens Insomnia Scale).

4. Patient satisfaction at discharge (categorised as satisfied, basically satisfied or dissatisfied).

5. Incidence rate of postoperative adverse events (eg, nausea and vomiting, pruritus, headache).

Information sources

All potentially eligible RCTs will be comprehensively searched in the following eight databases, including four English databases (Web of Science, PubMed, Cochrane Central Register of Controlled Trials, EMBASE) and four Chinese databases (China National Knowledge Infrastructure, VIP Database for Chinese Technical Periodicals, Wanfang Database, Chinese Biomedical Literature Database). The search period will cover from the inception of each database to November 2025.

Search strategy

The search strategy will include three basic elements: interventions (eg, transcutaneous auricular vagus nerve stimulation, taVNS); participants (eg, postoperative pain, postsurgical pain); and study type (randomised controlled trial, RCT). We will use the aforementioned English search terms to search English databases either individually or in combination; for Chinese databases, these English search terms will be replaced with their corresponding Chinese synonyms. To ensure the comprehensiveness of database searches and cover a wide range of potentially relevant studies, we will adopt a combined search approach of ‘controlled terms (eg, Medical Subject Headings [MeSH]) + free text terms’. To improve search sensitivity, MeSH can be replaced with corresponding adapted controlled terms for each database, which will be combined with appropriate free-text terms for searching. Corresponding adjustments to the search strategy will be made in accordance with the search requirements of different databases. Among these, the search strategies for PubMed, EMBASE and Web of Science databases are detailed in tables13.

Table 1. Search strategies for PubMed.

No	Search items
1	Randomized controlled trial[pt]
2	Controlled clinical trial[pt]
3	Randomized OR Randomised[Title/Abstract]
4	“Clinical Trials as Topic”[MeSH]
5	Randomly[Title/Abstract]
6	Trial[Title/Abstract]
7	#1 OR #2 OR #3 OR #4 OR #5 OR #6
8	“Pain, Postoperative”[Mesh]
9	(Postoperative Pain OR postsurgical pain OR pain after surgery OR postoperation pain OR postsurgery pain)[Title/Abstract]
10	#8 OR #9
11	(Transcutaneous auricular vagus nerve stimulation OR taVNS OR auricular vagus nerve stimulation OR aVNS) [Title/Abstract]
12	(ear OR auricular [Title/Abstract]) AND (vagus nerve stimulation[Title/Abstract])
13	#11 OR #12
14	#7 AND #10 AND #13

Table 3. Search strategies for Web of Science.

No	Search items
1	TS = (“Randomized controlled trial” OR “Controlled clinical trial” OR “Randomized” OR “Randomised” OR “Clinical trials” OR “Randomly” OR “Trial”)
2	TS = (“Postoperative Pain” OR “postsurgical pain” OR “pain after surgery” OR “postoperation pain” OR “postsurgery pain”)
3	TS = (“Transcutaneous auricular vagus nerve stimulation” OR “taVNS” OR “auricular vagus nerve stimulation” OR “aVNS” OR ((“ear” OR “auricular”) AND “vagus nerve stimulation”)))
4	#1 AND #2 AND #3

Table 2. Search strategies for EMBASE.

No	Search items
1	'Randomized controlled trial':ti,ab,kw
2	'Controlled clinical trial':ti,ab,kw
3	'Randomized' OR 'Randomised':ti,ab,kw
4	'clinical trial (topic)'/exp
5	'Randomly':ti,ab,kw
6	'Trial':ti,ab,kw
7	#1 OR #2 OR #3 OR #4 OR #5 OR #6
8	'postoperative pain'/exp
9	('Postoperative Pain' OR 'postsurgical pain' OR 'pain after surgery' OR 'postoperation pain' OR 'postsurgery pain'):ti,ab,kw
10	#8 OR #9
11	('Transcutaneous auricular vagus nerve stimulation' OR 'taVNS' OR 'auricular vagus nerve stimulation' OR 'aVNS'):ti,ab,kw
12	('ear' OR 'auricular'):ti,ab,kw AND ('vagus nerve stimulation'):ti,ab,kw
13	#11 OR #12
14	#7 AND #10 AND #13

Furthermore, we will also carefully check the references of systematic reviews and meta-analyses related to postoperative pain. To facilitate the subsequent update of this review, we will search the Chinese Clinical Trial Registry (http://www.chictr.org.cn), the WHO International Clinical Trials Registry Platform (https://www.who.int/clinical-trials-registry-platform) and the Clinical Trials Registry (https://clinicaltrials.gov/) to avoid missing any ongoing eligible RCTs.

Study selection process

EndNote V.X9 will be used to record literature entries retrieved from all databases to remove duplicate studies. During the initial literature screening stage, two independent reviewers (YL and HL) will identify and screen eligible RCTs by reviewing titles and abstracts in accordance with the aforementioned inclusion/exclusion criteria. To further confirm the eligibility of the included studies, full-text assessment will be conducted for potentially eligible literature. If the two reviewers have discrepancies regarding the inclusion of a specific study, a third arbiter (YS) will intervene to reach a final decision through discussion. The summary of the literature screening process for this study is detailed in online supplemental figure S1.

Data extraction and data items

For the purpose of gathering pertinent data, following the identification of all eligible RCTs, two independent reviewers (ZB and RL) will use a pre-established form (specified in online supplemental table S2) to extract and collate the following details: author names, year of publication, study design, participant demographics, sample size, as well as the interventions (including the number and duration of interventions) and outcome measures for both the experimental and control groups.

For continuous outcome measures, the mean and SD will be extracted; if continuous outcome measures are presented in other formats (eg, mean (95% CI), median (IQR)), they will be converted to means (SDs) using the methods recommended in the Cochrane Handbook for the Assessment of Intervention System.²⁶ For dichotomous outcome measures, the total number of participants and the number of responders in each group will be obtained. If the original studies lack the required relevant data, the reviewers will contact the authors via email to obtain the missing information. During the data extraction process, if there are discrepancies between the reviewers, a senior reviewer (J-QF) will make a final decision to resolve them.

Prior to the complete collection of data from all included studies, we will first extract data from 10 randomly selected studies and use the kappa coefficient to assess inter-reviewer’s agreement regarding the accuracy of data extraction. If there is insufficient accuracy and agreement between the reviewers, additional training on data extraction will be provided to them.

Methodological quality assessment

Two independent assessors (YL and HL) will assess the methodological quality of all included RCTs following the Risk of Bias 2.0 (RoB 2.0) tool recommended by Cochrane.²⁷ The RoB of each included study will be evaluated according to five key criteria: (1) randomisation method; (2) deviation from the planned intervention; (3) incomplete outcome data; (4) outcome measurement; and (5) selection of reported results. Each domain will be assigned one of three ratings: ‘low’, ‘unclear’ or ‘high’. Furthermore, reviewers will determine an overall RoB rating for each trial, which can be ‘low’ (meaning low RoB across all domains), ‘unclear’ (indicating concerns in one or more domains) or ‘high’ (signifying high RoB in one or more domains, or concerns in another domain). Any discrepancies between reviewers will be resolved by the senior arbitrator (J-QF) through discussion.

Data synthesis and statistical analysis

The Cochrane Handbook for Systematic Reviews of Interventions and Review Manager (RevMan) software (V.5.30, Cochrane Collaboration, Oxford, UK) will serve as references for conducting the risk of bias assessment of each included publication. Relative risk will be adopted to present dichotomous data, while standardised mean difference with 95% CI will be used to report continuous variables. Statistical heterogeneity across trials will be evaluated via the I² statistic and classified into three levels: low (I²<50%), medium (I²=50–75%) and high (I²>75%).

Two modelling approaches—the fixed-effect model and the random-effects model—both calculate a single effect size of interest. The fixed-effect meta-analysis presumes that all studies share one common effect; consequently, all variance in the observed effect sizes is caused by sampling error. In contrast, the random-effects meta-analysis estimates the mean value of an effect distribution, thus assuming that effect sizes differ across individual studies. Under this model, variance in the observed effect sizes stems from two sources: sampling error (within-study variance) and statistical heterogeneity (between-study variance). The most commonly used meta-analyses employ a weighted average to integrate effect sizes at the study level. To enhance the robustness and reliability of meta-analysis results, different effect models will be selected based on varying heterogeneity levels: when heterogeneity is rated as low, a fixed-effects model will be adopted; when heterogeneity is moderate, a random-effects model will be used for combined analyses; and when heterogeneity is high, discussions with the review panel will be conducted to explore clinical or methodological heterogeneity. If a meta-analysis is not feasible, narrative analyses will be carried out instead.

Sensitivity analysis

To mitigate the potential impact of small-sample studies on the statistical power, a sensitivity analysis will be performed. This analysis will evaluate the influence of small-sample trials on the pooled effect sizes and heterogeneity of the meta-analysis, by excluding such studies and comparing the resultant outcomes with those of the primary analysis that includes all eligible RCTs. This analysis will help assess the robustness of the conclusions regarding the efficacy and safety of taVNS for postoperative pain.

Subgroup analysis and meta-regression

If a sufficient number of studies are included, we plan to conduct several subgroup analyses.

First, according to preliminary search results, taVNS parameters vary substantially across studies, and there is currently no standardised classification framework for parameter categorisation.15,19 To preliminarily explore potential parameter-response patterns, we plan to dichotomise stimulation frequency and pulse width based on the median values reported in the included studies. Studies will be divided into the low or high-frequency subgroups, and narrow and the wide pulse width subgroups based on the median of the reported values in the included studies. Regarding stimulation intensity, most studies describe stimulation intensity based on participants’ subjective perception without reporting specific quantitative values.16,19 Therefore, studies will be divided into the subsensory threshold or suprasensory threshold subgroups based on the descriptions of participants’ subjective perception reported in the included studies.

Second, given that different types of control conditions may influence the estimation of treatment effects, we will conduct a subgroup analysis based on whether the control group is an active control or a placebo control.

Third, postoperative pain can be categorised into acute and chronic phases, which differ substantially in pathophysiological mechanisms and may thus exhibit different responses to taVNS. Analysing these two phases separately will clarify the efficacy of this intervention across distinct pain trajectories, providing targeted evidence for clinical practice.

Fourth, because placebo effects are frequently pronounced in pain outcomes, and adequate blinding is a key methodological approach to reducing placebo-related bias, a subgroup analysis will be performed according to the quality of blinding.^{28 29} Specifically, based on domains 2 and 4 of the RoB 2.0 tool, which evaluate bias related to blinding, studies will be categorised into low-risk, some concerns and high-risk groups for subgroup analyses.

Finally, some studies suggest that the effects of taVNS may differ across populations with different age distributions.^{30 31} Given the potential variability in participant age across studies, and the lack of age-stratified outcome data in most trials, subgroup analyses based on age are not feasible. Meta-regression will therefore be conducted to explore the potential impact of age on treatment effects.

Assessment of the quality of evidence

The quality and reliability of the evidence will be assessed in line with the Grading of Recommendations, Assessment, Development and Evaluation approach advised by Cochrane. The evidence’s level of certainty and robustness will be summarised through general ‘confidence in evidence’ ratings, which are classified into four levels: high, medium, low and very low.³²

Publication bias

When the meta-analysis includes 10 or more RCTs, Egger’s test and funnel plots will be employed to evaluate publication bias.

Discussion

Postoperative pain is a common and clinically significant consequence of surgery. If not adequately and promptly managed, its impact extends beyond local nociception and may involve systemic responses through the neuroendocrine-immune network. This can contribute to a vicious cycle characterised by delayed recovery and psychological distress, ultimately exerting detrimental effects on both short-term rehabilitation and long-term quality of life.¹⁰ The vagus nerve is essential for endogenous pain modulation, which exerts a top-down endogenous analgesic effect mainly through three core mechanisms: inhibiting pain signal transmission,³³ regulating neurotransmitter release^{34 35} and exerting anti-inflammatory effects.³⁶ As an emerging non-invasive neuromodulation technology, current studies have shown that the analgesic mechanism of taVNS involves multiple levels, such as the endogenous pain regulation system,³⁷ regulation of the autonomic nervous system,³⁸ anti-inflammatory effects,³⁹ modulation of brain network connectivity^{37 40} and transmission of pain information.⁴¹ Collectively, these mechanisms support the potential of taVNS as a therapeutic approach for pain management. At present, both clinical trials and animal studies^{42 43} have reported the positive effect of taVNS on acute and chronic postoperative pain. However, the detailed mechanism of taVNS in treating postoperative pain and the impact of specific parameters on its efficacy require further research and verification.

The strengths of this study are discussed below. First, existing evidence suggests that taVNS is effective in alleviating postoperative pain, which may consequently reduce the need for analgesic medications and decrease the incidence of related postoperative complications, such as nausea and vomiting.^{18 19} However, systematic reviews and meta-analyses regarding the application of taVNS in postoperative pain management remain lacking. To address this research gap, the present study aims to conduct a rigorous systematic review to evaluate the effects of taVNS on multiple postoperative pain-related outcomes—including pain intensity, analgesic consumption, associated emotional disturbances and sleep disorders—in order to comprehensively assess its efficacy and safety, and thereby to provide evidence to support clinical decision-making. Second, this review will strictly adhere to the guidelines established by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses.²⁴ During the study process, we will conduct comprehensive and systematic searches across eight major Chinese and English databases to screen relevant literature, while implementing strict inclusion criteria for relevant studies. Third, this study will adhere to the current guidelines included in the Cochrane Collaboration Handbook. This handbook provides a standardised framework for the meta-analysis process and proactively ensures the quality of subsequent meta-analysis results. In addition, this study will also conduct sensitivity analysis, publication bias analysis and subgroup analysis to further enhance the integrity of the research protocol.

While this study holds certain strengths, there are also limitations that require clarification. First, limiting the search scope to English and Chinese databases may introduce language bias, which could result in eligible RCTs published in other languages being overlooked. Second, although subgroup analyses and meta-regression will be conducted to explore the effects of different taVNS parameters and participant characteristics, significant clinical heterogeneity may still persist due to variations in treatment protocols and the complexity of postoperative care. Therefore, the findings of this study should be interpreted with caution. Third, as a relatively novel therapeutic modality, taVNS has primarily been evaluated in small-scale clinical studies, and the inclusion of these trials may influence the statistical power and robustness of the pooled results. Future studies with larger sample sizes and high methodological quality are needed to validate the therapeutic effects of taVNS and to determine the optimal stimulation parameters for postoperative pain management.

In conclusion, this raises challenging questions for the current research: whether taVNS can effectively alleviate postoperative pain and serve as a non-pharmacological option for multimodal analgesia, and how to more effectively use taVNS in clinical treatment.