The effects of non-invasive brain stimulation on disorder of consciousness in patients with brain injury: A systematic review and meta-analysis of randomized controlled trial

Introduction: Disorders of consciousness (DOC) result from neural system injury and manifest as changes in arousal or awareness. This systematic review and meta -analysis aimed to investigate the therapeutic effects of non-invasive brain stimulation (NIBS) techniques, such as transcranial magnetic stimulation (TMS) and trans-cranial direct current stimulation (tDCS), on consciousness dysfunction in patients with brain disorders. Methods: Literature was systematically searched in Medline, Embase, Cochrane database, Web of Science, EBSCO from inception to May 2023. Only randomized controlled trial with NIBS as an intervention and participants with DOC were included. Results: A total of 7 studies with 313 participants were included for meta -analysis. Compared with sham-or placebo-stimulation, NIBS can improve the Coma Recovery Scale-Revised scores significantly (mean difference [MD] = 1.96, 95 % confidence interval [CI] = [1.49; 2.43], P < .0001). Conclusion: NIBS has a significant positive effect in enhancing the symptoms of DOC. Nevertheless, it is imperative for further investigations comprising high-quality research designs and larger sample sizes in order to comprehensively elucidate the effects of NIBS techniques on diverse targets of stimulation within the population of individuals suffering from DOC.


Introduction
Disorder of consciousness (DOC) is a condition characterized by cerebral dysfunction resulting from neural system injury, leading to notable changes in the levels of arousal or awareness (Kondziella et al., 2020).The categorization of DOC is contingent upon the extent of residual consciousness and includes three primary states: coma, unresponsive wakefulness syndrome (UWS), and minimally conscious state (MCS) (Thibaut et al., 2019).Patients in a comatose state are typically characterized by a complete absence of consciousness, marked by closed eyes and an unresponsiveness to external stimuli, encompassing auditory, tactile, and painful stimuli (Kondziella et al., 2020).MCS is a state of significantly disturbed consciousness in which a patient displays irregular yet unequivocal, replicable indications of recognizing either their own existence or their surroundings.Consciousness can be evidenced through specific behaviors such as obeying simple commands, sustained visual focus or tracking, manipulating objects, expressive language ability, verbal or gestural responses to stimuli, and other Abbreviations: DOC, Disorder of consciousness; UWS, Unresponsive Wakefulness Syndrome; MCS, Minimally Conscious State; VS, Vegetative State; TBI, Traumatic Brain Injury; CRS-R, Coma Recovery Scale-Revised; NIBS, non-invasive brain stimulation; TMS, Transcranial Magnetic Stimulation; tDCS, Transcranial Direct Current Stimulation; rTMS, Repetitive Transcranial Magnetic  E-mail addresses: myhuyu_0503@163.com(Y.Hu), 775292500@qq.com(L.Hu), 474657566@qq.com(Y.Wang), 78725881@qq.com(X.Luo), zhaoxinsykf@ 163.com (X.Zhao), helin_1204@163.com(L.He). 1 Yu Hu and Linzhe Hu have contributed equally to this work and share the first authorship.
selective reactions to environmental events that cannot be attributed to reflexes (De Luca et al., 2022).However, it's important to differentiate the MCS from the UWS, formerly known as the vegetative state (VS), as the nomenclature "VS" has been supplanted by the term "UWS" (De Luca et al., 2022).Unlike MCS, UWS lacks interaction with the environment, self-awareness, and the ability to engage with others (De Luca et al., 2022).The primary etiology leading to the occurrence of both UWS and the MCS is traumatic brain injury (TBI) (Giacino et al., 2014).Among adults, the non-traumatic origins of DOC predominantly include acute hypoxicischemic injury resulting from cardiac arrest, cerebral hemorrhage, and ischemic stroke (Edlow et al., 2021).A report documented an estimated incidence of approximately 56 to 140 cases of UWS per million individuals in the United States (PVS M-STFo, 1994).Moreover, a systematic review revealed that the prevalence of MCS is ten times higher than that of UWS (Lombardi et al., 2002).To facilitate the differential diagnosis of DOC, the Coma Recovery Scale-Revised (CRS-R) was introduced in 2004 as a comprehensive assessment tool for evaluating neurobehavioral functioning (Giacino et al., 2004).Additionally, empirical evidence suggests that the CRS-R exhibits satisfactory reliability and content validity, rendering it suitable for distinguishing between UWS and MCS (Løvstad et al., 2010).
Currently, treatment approaches for DOC primarily revolve around the utilization of amantadine to alleviate pronounced inhibition of the central thalamus (Gosseries et al., 2014).Nevertheless, a notable drawback of this treatment modality is the emergence of drug resistance among certain DOC patients.Consequently, in recent years, a growing body of research has increasingly focused on investigating the potential benefits of non-invasive brain stimulation (NIBS) techniques for individuals with DOC.Based on the concept of neuroplasticity, NIBS possesses the capacity to modify neuronal pathways by means of enhancing and inhibiting neural networks (Pisegna et al., 2016).Over the past few years, the utilization of NIBS in the field of neurorehabilitation has gained significant traction, leading to its increasing prevalence (Miniussi et al., 2013).Notably, NIBS has found extensive application in the rehabilitation of DOC, and a growing body of literature is being dedicated to exploring its potential in this domain.Among the various NIBS techniques employed in clinical research, transcranial magnetic stimulation (TMS) and transcranial direct current stimulation (tDCS) stand out as the most frequently utilized methods (Schulz et al., 2013).The TMS technique has the ability to induce depolarization in neurons and exert an impact on cortical excitability (Hummel and Cohen, 2005).Repetitive transcranial magnetic stimulation (rTMS), which operates at either high or low frequencies, has the potential to enhance or inhibit cortical activity, respectively, thereby exerting regulatory control over cortical excitability (Zhang et al., 2021).On the other hand, tDCS modulates neural excitability by administering direct current through various scalp electrodes, which subsequently excite neurons at the targeted region (Doruk et al., 2014).Specifically, stimulation at the anode leads to increased excitability by depolarizing the resting membrane potential of cells within the affected hemisphere, whereas cathodal stimulation produces an opposing effect (Flöel, 2014).
Both rTMS and tDCS are regarded as potentially safe NIBS techniques for the treatment of DOC.However, the therapeutic efficacy of NIBS is contingent upon several factors, including the frequency or mode of stimulation (Li et al., 2018).In general, patients with DOC exhibit compromised cranial nerve function, resulting in reduced excitability of the nerves and disruption of active neural pathways (Shou et al., 2021).As a result, the brain's capacity to integrate information and manifest consciousness becomes impaired.NIBS techniques have the potential to facilitate the recovery of consciousness by directly modulating the activity of neuronal circuits (Thibaut et al., 2019;Ragazzoni et al., 2017).Studies have investigated the distinct targets and intervention modes of tDCS and TMS in individuals diagnosed with DOC.The dorsolateral prefrontal cortex (DLPFC), recognized for its involvement in motor control, behavior, and decision-making networks, holds particular significance (Heekeren et al., 2006).Furthermore, the right DLPFC is implicated in the maintenance of sustained wakefulness and attention, factors of utmost importance in DOC patients.Consequently, NIBS techniques are frequently employed in the DLPFC region to facilitate comprehensive DOC treatment (Sturm and Willmes, 2001).Nonetheless, the body of randomized controlled trials (RCTs) examining the effects of NIBS on DOC remains relatively limited, with a need for further elucidation regarding its comprehensive impact.
A comprehensive review conducted in 2017 explored the effects of tDCS on the underlying pathophysiological mechanisms of DOC, as well as its potential to enhance levels of consciousness.Within the review, the authors identified eight pertinent studies.However, it is important to note that the selected studies encompassed varying study designs, and a meta-analysis was not conducted by the authors (Zhang and Song, 2018).Based on the aforementioned premises, we formulated a hypothesis postulating that NIBS exhibits favorable outcomes in facilitating the restoration of consciousness levels among individuals diagnosed with DOC.The primary objective of this study is to conduct a systematic review including a meta-analysis of the therapeutic effects associated with NIBS interventions in patients afflicted by DOC.Our intention is to provide substantiated evidence that can serve as a basis for formulating future clinical guidelines pertaining to the treatment of DOC.

Materials and methods
This systematic review and meta-analysis adhered to the guidelines outlined in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 (Page et al., 2021).To ensure transparency and protocol adherence, the protocol for this systematic review has been duly registered with the international prospective register of systematic reviews (PROSPERO), under the registration number CRD42023428425.

Search strategy
The literature search was conducted in five databases (Medline, Embase, Cochrane database, Web of Science, EBSCO).The target language was set as English and the publication time ranged from inception to 31 May 2023.Search was carried out following the combination of keywords "coma" OR "persistent vegetative state" OR "consciousness disorders" OR "vegetative state" AND "transcranial direct current stimulation" OR "transcranial magnetic stimulation" OR "non-invasive brain stimulation" AND "stroke*" OR "cerebral infarction*" OR "brain injuries*".The details of search strategy were provided in Supplement A.

Inclusion criteria
In accordance with the Population, Intervention, Comparison, Outcome, Settings (PICOS) framework (Amir-Behghadami and Janati, 2020), the inclusion criteria for this systematic review and meta-analysis were established as follows: 1) Participants: The study comprised adults (aged over 18 years old) who had been diagnosed with DOC; 2) Intervention: The interventions administered to the participants consisted of one of the NIBS techniques, namely TMS or tDCS; 3) Comparison: A sham or placebo stimulation was conducted in the control group, allowing for a comparative analysis of the effects of the active interventions; 4) Outcomes: The CRS-R was selected as the primary or secondary outcome measure to assess the presence and severity of DOC.Sufficient data pertaining to CRS-R assessments were collected for analysis; 5) Study design: The included studies were RCTs, ensuring a rigorous and controlled experimental design; 6) Language requirement: Only articles written in English were considered for inclusion in this review.

Exclusion criteria
1) Any studies including the combinations of NIBS and other treatments such as pharma-therapy as interventions were ineligible; 2) The studies with outcome data reported insufficiently were not included; 3) The RCTs with crossover design were not included; 4) The full texts could not be found.

Study selection
Two authors, namely YH and LZH, conducted a rigorous screening process utilizing Endnote X9 to assess the suitability of article titles and abstracts for potential eligibility (Hupe, 2019).Moreover, the complete texts of studies that were identified independently by both authors were obtained and thoroughly evaluated.In the course of the selection process, any inconsistencies arising from variations in study design and inclusion criteria were resolved through collaborative discussions between the authors.
Cohen's Kappa score was utilized to assess the degree of consistency between two independent reviewers during the literature screening process.The observed agreement (Po) is computed as 0.9646, and the expected agreement (Pe) is calculated as 0.8319.Consequently, Cohen's Kappa = (Po-Pe)/(1-Pe) ≈ 0.7906.Therefore, this indicates a relatively high level of agreement between the two assessors in the selection of literature.Generally, a Kappa score between 0.61 and 0.80 is considered to demonstrate substantial consistency (Sun, 2011).

Data extraction
Two authors, namely YH and LZH, conducted independent data extraction from the included studies, and any discrepancies were resolved through comprehensive discussions.The general characteristics that were collected from the included studies encompassed the following information: authors' names, year of publication, sample sizes for both the intervention and control groups, trial design, intervention methods, etiology, outcome measures, average age of the participants (including the corresponding standard deviation), the gender distribution within the participant pool (specifically, the ratio of males to females), and the detailed protocols followed during the intervention.

Risk bias and quality assessment
The quality assessment of each study was conducted independently by two reviewers (YH and LZH) using the Revised Cochrane Risk of Bias assessment tool (RoB) specifically designed for RCTs (Ma et al., 2020).The RoB encompasses five crucial aspects, which are as follows: 1) random sequence generation selection bias; 2) blinding of participants and personnel performance bias; 3) incomplete outcome data attrition bias; 4) blinding of outcome assessment detection bias; 5) selective reporting bias.

Quantitively synthesis and data analysis
The statistical analysis and publication bias analysis were conducted using the RStudio statistical software (Boston, USA) and Stata (version 16.0), respectively.Initially, we extracted all participant data from the included articles and computed pre-and post-intervention changes of the CRS-R scores by comparing baseline and end-of-intervention measurements.Subsequently, a meta-analysis was performed using the random effect model to pool the standard deviation (SD) of the CRS-R changes before and after the intervention.Hedges' g was employed to evaluate the magnitude of the pooling effect, as it is considered more suitable for correcting biases associated with small sample sizes compared to Cohen's d (Lin and Aloe, 2021).Additionally, the metaanalysis provided 95 % confidence intervals (CI) for the SMD values.The present meta-analysis utilized the I 2 statistic as a measure to evaluate the presence of heterogeneity within the included studies.Given the relatively small number of studies and the limited total sample size, it is important to note that the statistical power of the I 2 test is relatively modest.Consequently, a P value below 0.01 was deemed indicative of significant heterogeneity (Higgins and Thompson, 2002).To further elucidate the extent of heterogeneity, a categorization system was employed, classifying it as low (I 2 < 25 %), moderate (I 2 < 50 %), or high (I 2 > 75 %) (Patsopoulos et al., 2008).Furthermore, a subgroup analysis was conducted to explore the potential influences of different etiologies, various types of NIBS techniques, and distinct types of DOC on the outcomes.Notably, according to established guidelines, metaregression analysis is generally deemed unnecessary when the number of included studies is below 10 ( Thompson and Higgins, 2002).Consequently, given the limited number of studies incorporated in this analysis, meta-regression was not performed.To ascertain the stability of the meta-analysis, a sensitivity analysis was conducted using the leave-oneout approach, whereby each individual study was systematically excluded, and the effect size of the results was recalculated.This procedure served to evaluate the robustness and reliability of the findings.Finally, to assess the potential publication bias of the included studies, Egger's linear regression test was employed (Hayashino et al., 2005).This statistical method provided an examination of the presence of bias stemming from publication-related factors, which could impact the overall validity and generalizability of the meta-analysis findings.

Selection of studies
The PRISMA flow diagram, shown in Fig. 1, illustrates the process of study selection.Initially, a comprehensive search of databases yielded a total of 586 studies.Upon applying the inclusion criteria, it was determined that 65 articles held potential relevance to the current systematic review.Subsequently, a thorough examination of the full-text versions of these articles resulted in the exclusion of 58 studies for the following reasons: lack of relevance to conscious outcomes (n = 20), inappropriate research design (n = 25), and insufficiently reported data (n = 13).Eventually, a total of seven articles met the criteria for inclusion in the meta-analysis.

Study characteristics
Table 1 presents a comprehensive overview of the principal characteristics of the studies incorporated in this analysis.The 7 articles selected for inclusion were all RCTs, including a total of 313 subjects, with sample sizes varying between 15 and 99 participants.Of the total cohort, 138 individuals were allocated to the intervention group, while 137 were assigned to the control group.Among the included studies, (57.14 %) employed rTMS as the intervention modality, while the remaining studies utilized tDCS.The etiology of the participants primarily consisted of TBI, stroke, and anoxia.Notably, the target population exhibited predominant symptoms of UWS or MCS.Specifically, a total of 245 patients were classified as having UWS, constituting 78.27 % of the overall sample.The duration of consciousness impairment across all subjects ranged from 54 to 190 days.Furthermore, Table provides an elaborate outline of the NIBS protocol implemented within the scope of this study.The NIBS intervention encompassed 4 articles, employing high-frequency repetitive transcranial magnetic stimulation (HF-rTMS) as the primary methodology, with 2 of these articles utilizing a 20 Hz stimulation protocol.Each session entailed a stimulation duration of 20 min, targeting the left DLPFC (He et al., 2021;Fan et al., 2022).Moreover, 3 additional studies employed tDCS as the intervention protocol.In all 3 studies, the left DLPFC was targeted as the stimulation site, employing a current intensity of 2 mA and a single stimulation duration of 20 min (Yu et al., 2022;Zhang et al., 2017;Wu et al., 2019).In addition, the treatment sessions varied across the subjects, ranging from 10 to 24 sessions NIBS, non-invasive brain stimulation; RCT, randomized controlled trial; HF-rTMS, high-frequency repetitive transcranial magnetic stimulation; TBI, traumatic brain injury; VS, vegetative state; IG, intervention group; CG, control group; PG, placebo group; MCS, minimally conscious state; UWS, unresponsive wakefulness syndrome; CRS-R, Coma Recovery Scale-Revised; GCS, Glasgow Coma Scale; EEG, electroencephalograph; SSEP, spinal somatosensory evoked potential; tDCS, transcranial direct current stimulation.

Risk of bias within studies
All the included studies underwent a thorough assessment for ROB, and the corresponding results are illustrated in Figs. 2 and 3.It should be noted that all studies, with the exception of one, were categorized as randomized and thus deemed to have a low risk of randomization bias.However, one study was found to be associated with a high risk of bias due to inadequate randomization procedures.In terms of incomplete outcome data, a particular study presented a limitation wherein the authors solely provided the change values between pre-and post-intervention stages, while failing to present the actual post-intervention data.Consequently, concerns were raised regarding potential attrition bias within this study.Regarding performance bias, all studies were determined to exhibit a low risk in this aspect.Notably, 85.7 % of the studies implemented a double-blind approach involving both the subjects and the researchers, thereby minimizing potential performance bias.However, one study did not provide explicit information regarding blinding procedures, thus warranting concerns regarding detection bias within that particular study.Furthermore, it is worth highlighting that all studies demonstrated clear reporting of the primary measure, resulting in a low risk of reporting bias across the board.

Meta-analysis-effects of NIBS on DOC
The present meta-analysis employed the random effects model to examine the effects of NIBS by synthesizing data from a total of 7 studies (He et al., 2021;Fan et al., 2022;Yu et al., 2022;Zhang et al., 2017;Wu et al., 2019;Shen et al., 2023Xiong et al., 2023).Specifically, the changes observed before and after the administration of NIBS in these seven trials were assessed.In comparison to control groups (which received sham stimulation or placebo stimulation), NIBS yielded significantly higher scores on the CRS-R (mean difference [MD] = 1.96, 95 % confidence interval [CI] = [1.49;2.43], P <.0001) (Fig. 4).Given the presence of heterogeneity across these seven studies, sensitivity analysis and subgroup analysis were conducted in order to identify potential sources of this heterogeneity.

Sensitivity analysis and publication bias
Fig. 4 illustrates the presence of heterogeneity among the studies conducted (I 2 = 54 %, P =.04).A sensitivity analysis revealed that the heterogeneity stemmed from a single study included in the analysis   (Shen et al., 2023 (Table 2 and Fig. 5).This particular study was a three-arm RCT, consisting of a test group, a control group, and a placebo group.Notably, both the test and control groups employed rTMS as the intervention, which differed from the two-arm RCT design employed in the remaining included trials.Upon exclusion of this trial, a subsequent meta-analysis was conducted, resulting in a decrease of the pooled effect's I 2 value to 0 % (P =.54), thereby eliminating heterogeneity.Utilizing the fixed-effect model once again, the MD was calculated as 1.83 (95 % CI = [1.56;2.10], P <.0001) (Fig. 6).Furthermore, the sensitivity analysis presented in Table 2 further substantiates that the improvement in CRS-R scores due to NIBS remains statistically significant when compared to the control group, thus affirming the stability of the meta-analysis results.Publication bias was evaluated using Egger's test, with the P value serving as the criterion.The results, presented in  analysis.The P value obtained from Egger's test was found to be greater than the significance level of 0.05 (P = 0.189), indicating the absence of statistically significant publication bias.Fig. 7 depicts Egger's publication bias plot, which was constructed based on Egger's linear regression approach for detecting bias.

Subgroup analysis -Intervention
In a meta-analysis comprising four studies, it was observed that patients who underwent TMS intervention exhibited significantly greater improvement in CRS-R scores compared to the control group.The random-effects model yielded a MD of 2.81 (95 % CI = [1.28;4.35], P <.0001).However, substantial heterogeneity was detected among these studies (I 2 = 71 %, P =.02).A sensitivity analysis conducted on these four studies indicated that the heterogeneity could be attributed to a single study (Shen et al., 2023).Upon excluding this study, the heterogeneity vanished (I 2 = 0 %, P =.46).Regarding patients who received tDCS intervention, a meta-analysis involving three studies revealed a significantly greater enhancement in CRS-R scores in the NIBS group compared to the control group.No heterogeneity was reported in this analysis (I 2 = 17 %, P =.30).The fixed-effects model yielded an MD of 1.53 ([0.39; 2.68], P =.009) (Fig. 8).Moreover, the heterogeneity between the two subgroups did not indicate a statistically significant difference (P =.19).Therefore, the variations in intervention methods did not contribute to the heterogeneity observed in this meta-analysis.

Subgroup analysis -Symptoms
For the analysis of patients with UWS, a total of four studies were incorporated.The findings from these studies revealed a notable improvement in CRS-R scores when employing NIBS intervention in comparison to the control group.The application of a random-effects model yielded a MD of 2.62 (95 % CI = [-0.15;5.40], P = 0.02).Conversely, in the case of patients with MCS, only two studies with adequate data were available for subgroup analysis.The outcomes indicated that NIBS did not exhibit a significant impact on CRS-R scores as compared to the control group.The random-effects model yielded an MD of 5.84 (95 % CI = [-0.07;11.75], P = 0.05) (Fig. 9).Moreover, the heterogeneity between the two subgroups did not demonstrate a substantial difference (P = 0.33), suggesting that variations in clinical symptoms did not contribute to the observed heterogeneity within this meta-analysis.

Subgroup analysis -Etiology
patients diagnosed with TBI, our analysis considered three studies with substantial data and conducted a subgroup analysis.The findings of these studies indicated that NIBS yielded a statistically significant  improvement in CRS-R scores.Utilizing the random-effects model, the MD was calculated as 3.97 (95 % CI = [0.29;7.65], P =.02).Similarly, our investigation encompassed three studies for subgroup analysis among patients with cerebrovascular accident (CVA).The results revealed that NIBS demonstrated a noteworthy enhancement in CRS-R scores when compared to the control group.Employing the randomeffects model, the MD was estimated to be 1.53 (95 % CI = [0.44;2.62], P =.04) (Fig. 10).Importantly, Fig. 10 also presented evidence suggesting that the observed heterogeneity between these two subgroups was not statistically significant (P =.21).Consequently, the underlying etiology responsible for the DOC did not contribute to the heterogeneity observed in this meta-analysis.

Discussion
To the best of our knowledge, this study was the first systematic review and meta-analysis only including RCTs to examine the effects of NIBS techniques on DOC.This study identified a total of 7 studies including 313 participants.The findings derived from our meta-analysis unequivocally demonstrate that NIBS holds significant potential for improving the state of consciousness in individuals afflicted with DOC, as evidenced by notable enhancements in their CRS-R scores.The 7 studies we included all demonstrate a significant improvement in CRS-R total scores following NIBS intervention.Among them, the study by Shen et al. indicates a significant improvement in all items of CRS-R following NIBS (Shen et al., 2023.Specifically, Wu et al. pointed out that NIBS had the most significant impact on the CRS-Rvisual scores, and it resulted in slight improvements in the auditory, motor, and communicative function of MCS patients (Wu et al., 2019).This outcome assumes paramount importance in fostering an overall improvement in the level of consciousness.
In the present meta-analysis, four studies utilized HF-rTMS, while the remaining studies employed tDCS as an intervention.The findings revealed that the utilization of different NIBS techniques exhibited significant improvements in the CRS-R scores among patients with DOCIn comparison to a prior review conducted by Zhang et al. in 2017 (Zhang andSong, 2018), which specifically focused on tDCS for patients with DOC, the present meta-analysis encompasses one less article.Nevertheless, a notable distinction exists between our study and Zhang et al.'s review that the inclusion criteria of the latter encompassed diverse study types, including RCTs with crossover designs and case-control studies (Zhang and Song, 2018).Conversely, our meta-analysis exclusively incorporated RCTs, thereby enabling enhanced control of bias risk, ensuring consistency in data and statistical analyses, and bolstering the reliability of our findings.Furthermore, while Zhang et al.'s review solely examined the impact of tDCS on DOC, our study extends its scope to encompass the effects of rTMS as well.Consequently, our research provides additional empirical evidence to underpin the application of NIBS techniques in the treatment of individuals with DOC.
In the present meta-analysis, 4 studies utilized HF-rTMS, while the remaining studies employed tDCS as an intervention.The findings revealed that the utilization of different NIBS techniques exhibited significant improvements in the CRS-R scores among patients with DOC.As

Table 2
Summary of leave-one-out sensitivity analysis for total studies included in quantitative synthesis (n = 7).Fig. 5. Meta-analysis estimates, given named study is omitted for all studies.
Y. Hu et al.
previously mentioned, all studies included in the present meta-analysis have indicated a significant improvement in the total CRS-R score after receiving either rTMS or tDCS individually.However, there is limited research reporting changes in specific item scores.Currently, there are no studies that directly compare the effects of rTMS and tDCS on consciousness level improvement.We speculate that combining two or more NIBS techniques may lead to a greater improvement in CRS-R scores for patients and potentially enhance their level of consciousness.Both rTMS and tDCS are recognized for their ability to enhance synaptic plasticity and modulate the excitability of the cerebral cortex, thereby inducing enduring changes and reactivating the arousal pathway in patients (Terranova et al., 2019).Specifically, tDCS modulates cortical excitability through electrical stimulation, thereby influencing cortical plasticity and augmenting connectivity within the frontoparietal network in patients diagnosed with MCS and UWS (Martens et al., 2020).Studies have indicated that tDCS may possess neuroprotective effects, stimulate the release of brain-derived neurotrophic factors, and attenuate neuronal apoptosis (Guidetti et al., 2022).Moreover, HF-rTMS generates a magnetic field that facilitates the functional integration of neural networks in patients with impaired consciousness.It regulates synaptic plasticity, activates the thalamocortical system, and promotes neuronal regeneration and neurological recovery (Chen et al., 2022).Notably, rTMS can activate a larger population of neurons through repeated, continuous, and regular magnetic stimulation, exerting a sustained impact on neuronal activity (Xiong and Zheng, 2022).
Although the stimulation provided by tDCS is weaker compared to rTMS due to the attenuating effects of the scalp and skull, it exhibits higher current permeability (Priori et al., 2009).Conversely, the magnetic field generated by rTMS is not attenuated by other tissues but has limited penetration depth (O'Leary et al., 2021).As a result, the selection of treatment methods and stimulation locations should be tailored to the individual patient's specific condition, necessitating further analysis.
To further investigate the response of different states of consciousness to NIBS, as shown in Fig. 9, a subgroup analysis was conducted specifically targeting patients with MCS and UWS.Within this subgroup analysis, a total of four studies examined the effects of NIBS on UWS   patients, revealing a significant improvement in the CRS-R scores following NIBS intervention.Conversely, two studies included MCS patients, but did not find a significant impact of NIBS on the CRS-R scores of these patients.Interestingly, this contrasts with previous research, which has demonstrated that NIBS can significantly enhance consciousness in MCS patients (O'Neal et al., 2021;Formica et al., 2021).Previous studies stated that contrasting neural mechanisms underlying consciousness impairment in MCS patients when compared to UWS patients (Formica et al., 2021).Prior investigations have highlighted the capacity of NIBS to substantially ameliorate the consciousness disturbance observed in MCS patients, potentially stem from the preservation of a greater number of neural connections in this population (Formica et al., 2021).Consequently, the observed lack of significant effects of NIBS on CRS-R scores among MCS patients may be resulting from the relatively intact cortical remodeling and repair processes evident in this patient cohort.Moreover, it is worth noting that during NIBS interventions, MCS patients exhibited notable slow-wave electroencephalogram (EEG) activity, whereas UWS patients did not exhibit significant EEG activity (Formica et al., 2021).This discrepancy in EEG patterns suggests that MCS patients may undergo more favorable cortical remodeling and repair processes in response to NIBS interventions compared to UWS patients.Such findings further support that the divergent effects of NIBS on CRS-R scores observed in UWS and MCS patients may be related to inherent differences in the underlying neural mechanisms of these conditions.Furthermore, factors such as the duration of consciousness disturbance, location and extent of brain lesions, and specific NIBS protocols may also influence treatment outcomes.The accurate diagnosis of consciousness disturbance is influenced by multiple factors.The patient's overall health status and the severity of cognitive impairment can affect their responsiveness to commands and behavior.Moreover, underlying language impairments and neuromuscular disorders can simulate the true condition, introducing diagnostic biases in assessing the level of consciousness.It is important to acknowledge that these factors may have influenced the outcomes of the studies included in our analysis (Giacino et al., 2018).Therefore, in order to define more accurate conclusions, it is imperative to gather and analyze data from larger sample sizes.Subgroup analysis, as illustrated in Fig. 10, comprised 3 studies involving patients with TBI and 3 studies involving patients with CVA.Previous studies have indicated that NIBS can significantly enhance the  total CRS-R score.However, the 7 studies we included in our analysis did not provide information on the improvement in individual items of CRS-R for patients with DOC resulting from different etiologies.These results suggest that the beneficial effects of NIBS on DOC are not influenced by the underlying etiology.This observation can be attributed to the broad disruption of consciousness resulting from various etiologies, wherein common pathophysiological mechanisms involve the widespread cessation of excitatory synaptic activity in the cerebral cortex, diminished neuronal firing impulses, and structural impairments within the brain's ascending reticular activating system (Functional networks in disorders of consciousness, 2017).It is worth noting that the manifestation of consciousness disturbance does not exhibit characteristic alterations exclusive to particular diseases or specific regions of brain damage.Rather, it arises due to perturbations in the brain's network structure and abnormal functionality.Hence, the impact of the disease type itself on consciousness disturbance is relatively insignificant, as the crucial factor to be considered lies in the extent of damage inflicted upon the brain network associated with the ailment (Evers, 2022).In contrast, NIBS has the capacity to modulate the excitability of the cerebral cortex and stimulate synaptic activity (Functional networks in disorders of consciousness, 2017; Xia et al., 2018).Thus, this mechanism may explain its potential in enhancing disorders in consciousness.
In this meta-analysis, the majority of the 7 included studies focused on the left DLPFC as the targeted site for NIBS.Prior investigations have indicated that rTMS of the primary motor cortex (M1) has generally failed to yield significant clinical improvements in patients with DOC (Thibaut et al., 2019).The DLPFC, being a crucial component of the central executive network, plays a vital role in strengthening the connectivity between the prefrontal cortex and the thalamus.Additionally, it regulates the balance and connectivity between the central executive network and the default mode network (O'Neal et al., 2021), thereby exerting influence on higher cognitive functions and facilitating the recovery of consciousness.Consequently, it is plausible that the left DLPFC represents a more favorable stimulation site for enhancing DOC, which could potentially explain why all studies included in this metaanalysis adopted the left DLPFC as the targeted location for stimulation.Nevertheless, the therapeutic efficacy of stimulation protocols in the M1 region remains uncertain based on current literature, necessitating further investigations to explore its potential benefits.It is worth noting that there is a paucity of studies employing NIBS to stimulate the right DLPFC, and none of them met the inclusion criteria for this meta-analysis.Consequently, additional research is required to determine the optimal stimulation site for NIBS, thus providing more comprehensive insights into its therapeutic value.
In the past few decades, with continuous technological advancements, the field of rehabilitation therapy has witnessed the emergence of a series of novel therapeutic approaches.These new technologies have enhanced neuroplasticity, offering a multitude of treatment options for the rehabilitation process.Among these, widely employed techniques include brain stimulation, robotic training, and virtual reality, each possessing unique advantages and playing an increasingly significant role in the treatment of patients with consciousness disorders.For instance, robotic verticalization training (RVT) can activate a patient's proprioception, tactile sensations, and vestibular pathways, resulting in increased excitability in cortical regions associated with trunk and lower limb control, thereby enhancing brain plasticity.This can be employed to reduce intracranial pressure by adjusting posture, and improving neurological symptoms (De Luca et al., 2022).Furthermore, RVT training can also help prevent complications arising from orthostatic hypotension and prolonged bed rest.
On the other hand, Virtual Reality (VR) can deliver visual, auditory, and various sensory inputs to patients, enhancing their perception of the external environment (Bonanno et al., 2022).This contributes to increased neuroplasticity and neural network connectivity.Additionally, combining VR with a patient's past experiences and life contexts can provide emotional stimulation, fostering communication and interaction with the external world.The combined approach (NIBS plus VR or robotics) may have a positive effect on DOC.However, it is important to note that these new technologies are predominantly applied to improving functions related to the limbs and cognition.Their application strategies and effectiveness assessment in treating patients with consciousness disorders are still in their infancy.To determine how to maximize the utilization of these new technologies and formulate appropriate integrated treatment approaches, further studies are required through larger-scale, high-quality and long-term follow-up results.These efforts will provide valuable insights and directions for future clinical work and research.
This meta-analysis is subject to several limitations that should be acknowledged.Firstly, the inclusion of only seven studies in our analysis results in a relatively small overall sample size, which may restrict the generalizability of the findings.Moreover, it is worth noting that the majority of the included studies were authored by Chinese scholars, introducing a potential source of bias and limiting the external validity of the study's outcomes.Secondly, our analysis focused solely on the short-term effects of NIBS, as the included studies lacked long-term follow-up data.Consequently, the long-term impacts of NIBS on DOC could not be adequately assessed.Lastly, the outcome measures employed in this meta-analysis are relatively limited, with the CRS-R being the sole assessment tool used to evaluate the recovery levels of individuals with DOC.To enhance the robustness and validity of future research, it is recommended to incorporate supplementary examinations, such as imaging techniques and neurophysiological assessments.By incorporating a broader range of evaluation measures, a more comprehensive and reliable evaluation of the effects of NIBS on DOC can be achieved.

Conclusion
In conclusion, both rTMS and tDCS exhibit considerable benefits in facilitating the recovery of consciousness among patients diagnosed with UWS resulting from TBI or CVA.Nevertheless, it is imperative for further investigations comprising high-quality research designs and larger sample sizes in order to comprehensively elucidate the effects of NIBS techniques on diverse targets of stimulation within the population of individuals suffering from DOC.

Source of funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Fig. 2 .
Fig. 2. The risk of bias of included studies.

Fig. 3 .
Fig. 3.The risk bias summary of included studies.

Fig. 4 .
Fig. 4. Forest plot for meta-analysis of all studies; mean difference (95 % CI) between experimental and control groups for the change of CRS-R score.

Fig. 6 .
Fig. 6.Forest plot of meta-analysis after exclusion of highly heterogeneous studies (95 % CI) between experimental and control groups for the change of CRS-R score.

Table 1
The character of included studies.

Table 3
The result of Egger's test for publication bias of all studies.