Acupuncture to Promote Recovery of Disorder of Consciousness after Traumatic Brain Injury: A Systematic Review and Meta-Analysis

Traumatic brain injury (TBI) has become an economic and social burden for patients and their families. While acupuncture is an effective tool for promoting recovery of disorder of consciousness (DOC) following TBI, there have been no comprehensive meta-analyses and/or systematic reviews addressing this topic. The present systematic review and meta-analysis aimed to assess the therapeutic efficacy of acupuncture for DOC after TBI. All randomized controlled trials (RCTs) incorporating acupuncture, or acupuncture combined with other interventions for DOC after TBI, were included and assessed by two independent investigators. Six outcome indicators were assessed: Glasgow Coma Scale (GCS); Glasgow Outcome Scale (GOS); mortality; efficacy rate; activities of daily living (ADL); and functional comprehensive assessment. Direct comparisons were performed using RevMan 5.3.0 software, with results presented as mean difference (MD) for continuous outcomes and relative risk (RR) for binary outcomes. A total of 3511 patients from 49 trials were included. Pooled analyses indicated that acupuncture may have a superior effect on GCS score (MD=2.03, 95% CI :1.92 2.43, Z=16.54, and P<0.00001); GOS score (RR=1.23, 95%CI: 1.18 1.35, Z=6.65, and P<0.00001); efficacy rate (RR=1.48, 95%CI: 1.40 1.56, Z=13.49, and P<0.00001); ADL (MD=9.20, 95% CI:8.19 10.21, Z=17.84, and P<0.00001); and mortality (RR=0.50, 95% CI:0.38 0.67, Z=4.70, and P<0.00001). The results demonstrated that the acupuncture group fared better than the control group in the treatment of DOC after TBI. However, studies were generally of poor quality, and publication bias favoring positive studies was obvious. Therefore, rigorous evaluation standards and well-designed studies are necessary in future studies.


Introduction
Traumatic brain injury (TBI) is a global public health problem and one of the major causes of death and disability [1]. According to a survey by the World Health Organization, TBI will surpass many other factors as a main cause of death and disability by the year 2020 [2]. According to statistics, more than 50 million individuals experience TBI each year, and the total number of new TBI cases has been reported to approach 3.5 million per year in the United States [3], with 30-40% of deaths related to the injury. The mortality rate of TBI in China was 12.99/100,000 in 2013 and 17.1/100,000 in the United States, with traffic accidents and falls as the main causes [4][5][6]. TBI has become an enormous economic burden to families and society as a whole. TBI-related costs were estimated to be as high as €33 billion in Europe in 2010, of which indirect costs accounted for 59% and direct costs accounted for 41% [7,8]. In the United States, it has been reported that aggregate direct and indirect costs increased from USD$60.4 billion in 2000 to USD$221 billion in 2009 [9].
The rapid development of emergency medicine and intensive care technology in recent years has clearly improved the survival of individuals who sustain TBI [10]. However, regarding severe brain damage, patients also experienced varying degrees of disorders of consciousness (DOC) after 2 Evidence-Based Complementary and Alternative Medicine TBI, which seriously affects an individual's ability to perform the activities of daily living (ADL) and reduces the quality of life [11][12][13]. The early recovery of consciousness is closely associated with recovery in other functional domains, and the duration of DOC is an important prognostic factor in patients who experience TBI [13][14][15].
The annual cost for treatment and rehabilitation of TBI patients is nearly $400 billion, which is associated with tremendous inconvenience and economic pressure on patients, their families, and society [16,17]. Therefore, studies investigating the recovery of patients' consciousness after TBI are important in the field of neurologic rehabilitation and have a great economic and social impact. There are many important treatment approaches for recovering consciousness, including pharmacological interventions, surgery, rehabilitation, and alternative medicine treatments such as acupuncture. Acupuncture has been shown to be an effective method when applied to traumatic injuries. A systematic review by Ma concluded that the use of acupuncture to treat spinal cord injury may have a beneficial effect on neurological and functional recovery. [18] Another review suggested that acupuncture is efficacious in acute TBI [19]. In China, acupuncture has been used for centuries to promote the recovery of consciousness after coma or other DOCs by increasing the excitability of nerve cells [20][21][22] and the supply of oxygen and blood flow in the traumatized area of the brain, which may be a potential mechanism for its rousing effects [23][24][25].
Although many studies investigating acupuncture for treatment of DOC after TBI have been published, they have not been systematically reviewed. Thus, the aim of the present systematic literature review was to assess the rousing effect of acupuncture on DOC after TBI.

Protocol and Registration.
We had published a protocol of this systematic review in the PROSPERO database (identification number: CRD42018091226).

Data Source and Search Strategy.
A systematic search of the literature was conducted in PubMed, Cochrane Library, Chinese Biomedical Literature Database (CBM), VIP, WanFang Database, and Chinese National Knowledge Infrastructure (CNKI) databases for articles published up to February 28, 2018. The searching terms include "coma", "disorder of consciousness", "minimally consciousness state", "vegetative state", "locked-in state", "traumatic brain injury", "brain injury", "head injury", "acupuncture", and "electroacupuncture". The search was limited to studies published in Chinese and English.

Inclusion Criteria.
All trials in which the intervention involved acupuncture or electroacupuncture separately, or acupuncture combined with other interventions, such as hyperbaric oxygen (HPO), traditional Chinese medicine (TCM) rehabilitation, and electrical stimulation, were included. There was no distinction for acupoints, manipulation, stimulation intensity, or the course of treatment. It is available for basic treatment or combined with rehabilitation training in the control group. The trials' outcome indicators had to include at least one of the following: Glasgow Coma Scale (GCS) or Glasgow Outcome Scale (GOS).

Exclusion Criteria.
Trials that used transcutaneous electrical nerve stimulation and/or manual acupressure and DOC results from nontraumatic diseases, such as cerebrovascular disease, tumor, or toxicosis, were excluded. In addition, trials without a nonacupuncture group were also excluded.

Data
Extraction. Data were extracted independently by two authors (LT and LZ) using a specifically designed data extraction form. Disagreements were resolved by assistance from the third author (NW). Data extracted from each study were summarized in a data extraction table. Items noted included the following: first author and year of publication; type of intervention; sample size; sex ratios; years; TBI's degree, duration of treatment, and main outcome indicators in each trial. For studies in which outcomes were assessed at several different time points, the measurement after the final treatment was selected. Meanwhile, if the duration of treatment described as applying according to the condition of disease, we attributed it to an unclear duration of treatment.
2.6. Quality Assessment. Two authors (MJD and TIM) evaluated the quality of included trials. The risk of bias was assessed using the risk of bias tool from the Cochrane Handbook of Systematic Reviews of Interventions [26] according to the following criteria: random sequence generation allocation concealment; blinding of participants and personnel; blinding of outcome assessment; incomplete outcome data; and other biases and descriptions of the risk of bias.

Data
Analysis. RevMan 5.3 software was used for data analysis and Stata version 13.0 (StataCorp LP, College Station, TX, USA) was used for sensitivity analysis. Different outcome indicators were classified and analyzed. Relative risk (RR) was evaluated for dichotomous outcomes, while mean difference (MD) was used for continuous outcomes with 95% confidence interval (CI). Because there was significant clinical heterogeneity among the trials (P<0.1), a randomeffects model was used in the meta-analysis, and a fixedeffects model was used for those with low heterogeneity (P>0.1). Potential publication bias was assessed using funnel plots.

Results
A flowchart depicting the search process and study selection is shown in Figure 1. Forty-nine studies including 3511 patients were ultimately included. There were 1800 participants in the acupuncture group and 1711 in the control group.
3.1. The Basic Characteristics. All trials were conducted in China and published in Chinese. The acupuncture and control groups were compared statistically according to sex, age, duration, and level of consciousness; baseline data were comparable. Regarding the intervention in the treatment groups, acupuncture was applied in 18 RCTs, electroacupuncture in   Evidence-Based Complementary and Alternative Medicine 16, acupuncture combined with HPO in 6, acupuncture combined with TCM in 7, electroacupuncture combined TCM and HPO in 1, and acupuncture combined TCM and laser intravascular irradiation in 1. Thirty-seven studies reported GCS data before treatment, with sample sizes ranging from 29 to 126. Only 10 RCTs enrolled patients and reported the duration of TBI. The characteristics of the included trials are summarized in Table 1.

Quality
Evaluation. The majority of the included studies were assessed to be of generally poor methodological quality according to the Cochrane Handbook of Systematic Reviews of Interventions. Although "random allocation" was mentioned in all studies, only 12 RCTs reported the details of random sequence generation using a random numbers table. [21][22][23][27][28][29][30][31][32][33][34][35]. There was no clear information regarding allocation concealment, and only 2 RCTs reported the blinding method [22,32]. For incomplete outcome data, 1 study reported drop-out and exit of patients without any effect on the results, and 1 reported transfer from one hospital to another [29,32]. Selective reporting was generally unclear due to inaccessibility of trial protocols. Only 7 trials reported funding support [25,30,32,[34][35][36][37]. The methodological quality of the included RCTs is shown in Figures 2 and 3.

ADL.
Two trials [27,39] used ADL to assess the curative effect and the continuous data were analyzed. Meta-analysis revealed that the acupuncture group demonstrated significantly improved ADL, with moderate heterogeneity (I 2 = 61%, P=0.11). The results indicated that improvement of ADL was 9.20 times higher than the control group (MD=9.20, 95% CI: 8.19 10.21, Z=17.84, and P<0.00001) (Figure 10).

Functional Comprehensive Assessment(FCA).
Among all included studies, FCA was used as an evaluation of neural function recovery in 2 investigations [23,42]. Both demonstrated a better effect in the acupuncture group compared with the control group; however, the effect was not combined with high heterogeneity (I 2 = 97%, P<0.00001).

Discussion
The results of the present review suggest that acupuncture may have a beneficial effect on the recovery of consciousness after TBI. According to these findings, acupuncture may improve consciousness levels evaluated using GCS or GOS scores, clinical efficacy, and lower mortality after TBI. Acupuncture may also help in rehabilitation. The mechanism for the beneficial effects of acupuncture in promoting arousal in TBI patients has been investigated in many animal experiments [70][71][72]. Cheng et al. [36] found that lowering the level of plasma D-dimer may be one of the therapeutic mechanisms of electroacupuncture in recovering consciousness.
However, these positive findings should be interpreted cautiously due to the high risk of bias in all of the included studies, the quality of which was poor overall. Among the 49 studies, the method of randomization was completely described in only 12 trials and incompletely described in 6, whereas the remainder merely only mentioned that "patients were randomized into two groups". Without appropriate randomization, there is a strong possibility that some of these studies were not actually RCTs. Of all RCTs, only 2 reported blinding and none of allocation, which demonstrated exaggerated treatment effects and limited the reliability of the Study omitted Meta-analysis fixed-effects estimates (exponential form) study results. In addition, only 2 trials reported dropouts and some trials did not report the missing data that led to the exaggerated treatment effects of acupuncture.
Despite the positive results, there were several obvious problems with the trials we analyzed. First, the sample size in many of the studies was too small to draw firm conclusions, and none of the included RCTs reported calculation of sample size. Second, treatment in the control groups could be a major source of bias, especially for clinical efficacy rate, which was subjective. All RCTs compared acupuncture versus basic treatment, and no sham acupuncture or placebo was used; moreover, the patients in these studies were not blinded as to whether they were receiving acupuncture or nonacupuncture. If participants had a preference for acupuncture relative to control treatment, it may have resulted in a greater placebo effect than basic treatment. Third, some trials used acupuncture combined with other interventions in the control group, which resulted in mixed effects and, in turn, made it difficult to evaluate the correlation between differences in acupuncture treatment and its therapeutic effectiveness. Thus, analysis of acupuncture alone or combined with other interventions in the control or experimental groups may increase reliability. Meanwhile, the duration and frequency of acupuncture, acupoint selection and stimulation parameters, and time of electroacupuncture varied among the RCTs, which may have contributed to the heterogeneity among the included trials. Fourth, the duration and grade of TBIs were different in the included trials and not reported in many RCTs, which may have led to different outcomes in clinical efficacy rate, GOS score, and mortality and was an additional source of heterogeneity in the metaanalysis. It has been established that high GCS scores are likely to result in good outcomes, and patients with DOC may regain consciousness more quickly. Fifth, the funnel plot demonstrated asymmetry, indicating potential publication bias against "negative" studies due to journal' preferences for positive research. Moreover, it is difficult to judge whether there were unreported negative research results given the inaccessibility of the study protocols. Adverse events and safety profiles of acupuncture were not reported.

Conclusion
To our knowledge, this was the first systematic review and meta-analysis to investigate acupuncture for DOC after TBI. Although the results suggest that acupuncture produced superior effects on the recovery of consciousness in the included trials, the above limitations make this effect questionable and difficult in drawing definitive conclusions. In the future, there is a need for more high-quality studies  46  33  37  37  38  13  20  32  30  32  14  40  11  37  20  15  31  30  21  26  10  42  35  26  29  25  27  41  18  13  48  18  24  69   988   Total   54  36  39  40  44  16  23  61  33  40  16  50  15  42  25  20  37  43  29  30  15  51  50  50  32  31  36  50  20  15  50  22  30  69   1214   Events   32  25  31  30  25  7  13  17  21  20  8  30  4  28  13  4  26  15  10  18  5  29  16  12  24  20  22  26  12  7  30  12  14  30   636   Total   51  36  38  40  40  16  23  61  30  40  14  50  14  43  25  20  37  42  27  30  15  51  50  50  32  29  38  50  20  15  40  20  30  Heterogeneity: ChＣ 2 = 42.30, df = 33 (P = 0.13); ) 2 = 22% designed to investigate the rousing effect of acupuncture on patients with TBI. In particular, more effort and attention should be dedicated to improving the quality of studies to enhance internal validity. For example, inclusion criteria, sufficient sample size, truly randomized methods and adequate concealment of allocation, blinding of investigators and participants, and outcome assessors should be unified to minimize performance and assessment bias. Moreover, widely recognized and standardized outcome measurements, such as GCS and GOS, should be adopted to ensure reliability, comparability, and validity and to present results clearly in sufficient detail. Additionally, study protocols and clinical trial registration are an essential part of research. Researchers should be encouraged to devote special attention to monitoring and reporting adverse events regarding acupuncture intervention, quality of life, long-term efficacy, and objective outcomes (such as electroencephalography and functional magnetic resonance imaging). Even if acupuncture demonstrated an effect superior to control in our study, more evidence as to whether it has significant advantages over sham acupuncture at the same acupoints is needed.

Data Availability
The data supporting this systematic review are from previous studies and datasets, which have been cited. The processed  Heterogeneity: ChＣ 2 = 2.56, df = 1 (P = 0.11); ) 2 = 61% data are available from the corresponding author upon request.

Conflicts of Interest
No financial conflicts of interest exist.