Article Text


Acupuncture in the emergency department: a systematic review of randomised controlled trials
  1. Kwan Leung Chia1,
  2. Rex Pui Kin Lam2,
  3. Chung Kwun Lam3,
  4. Sik Hon Tsui4
  1. 1 St Vincent’s Clinical School, University of New South Wales, Sydney, New South Wales, Australia
  2. 2 Emergency Medicine Unit, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Hong Kong SAR, China
  3. 3 Hong Kong Association for Integration of Chinese-Western Medicine, Hong Kong SAR, China
  4. 4 Department of Accident and Emergency, Queen Mary Hospital, Hong Kong SAR, China
  1. Correspondence to Dr Kwan Leung Chia, St Vincent’s Clinical School, University of New South Wales, Sydney, New South Wales 2052, Australia; kwanleung.chia{at}


Introduction A comprehensive review of both English and Chinese language literature to inform acupuncture practice in emergency department (ED) settings is lacking. Accordingly, we aimed to conduct a systematic review of English and Chinese randomised controlled trials (RCTs) of acupuncture use in the ED.

Methods Four English databases (Embase, PubMed, AMED and CENTRAL) and two Chinese databases (CNKI and Wanfang) were systematically searched using the keywords ‘acupuncture’ and ‘emergency department’, followed by a bibliographic search of references. The data were extracted and assessed by two independent authors. RCTs were selected based on pre-defined criteria. Data were extracted and a risk of bias assessment was performed using the Cochrane risk of bias tool. The quality of evidence was rated using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach.

Results In total, 1461 articles were screened and six RCTs involving 651 patients were included. For various acute pain conditions, acupuncture was superior to sham acupuncture, more effective than intravenous morphine, comparable to conventional ED treatment, and superior to standard ED care alone when used on an adjuvant basis; however, the overall level of evidence was low. Studies that applied acupuncture in hypertension and cardiac arrest were deemed to be at high risk of bias, and the level of evidence for these outcomes was very low. No major adverse events were reported in the included studies.

Conclusion There is a lack of high-quality evidence to support the use of acupuncture in the ED. Multicentre RCTs with rigorous designs are warranted.

  • acupuncture
  • electroacupuncture

Statistics from


Contrary to the general belief that acupuncture is useful only for chronic diseases or pain, it has also been used in medical emergencies in ancient and modern China.1 2 Acupuncture is the best known and studied modality of Traditional Chinese Medicine (TCM) with more than 500 randomised controlled trials (RCTs) published.3 It is widely used in 103 WHO member states with insurance coverage in 18 countries.4 Current evidence suggests benefits of acupuncture in a variety of clinical conditions, including chronic low back pain,5 tension-type headache,6 migraine,7 neck pain,8 osteoarthritis,9 postoperative nausea and vomiting,10 and fibromyalgia.11 When performed by trained practitioners, acupuncture is generally safe in adults,12 children,13 pregnant women14 and even in patients receiving anticoagulants.15 Serious adverse events, such as pneumothorax, are rare.16 17 Acupuncture has also been shown to be cost-effective for certain clinical conditions.18

Acupuncture has the potential to improve patient care in the emergency department (ED). However, the efficacy and safety of acupuncture in inpatient and outpatient practice environments cannot necessarily be extrapolated to the ED setting. The hectic ED environment is not conducive to the practice of acupuncture. Multiple adaptions are often required, including selection of fewer acupuncture points, shortening of the treatment session and flexible techniques, since access to certain regions of the body and patient repositioning are often limited by the presenting acute conditions.19 The efficacy and safety of acupuncture in the ED should be supported by high-quality evidence from clinical trials conducted in the ED setting.

So far, there has been a lack of evidence in the English language literature to support acupuncture use in the ED. Quasi-experimental trials have shown that acupuncture is better than sham acupuncture at relieving dizziness/vertigo and acute low back pain in the ED.20 21 In an observational study comparing acupuncture in 200 patients with pain and/or nausea with 200 matched historical controls who received usual care in an Australian ED, acupuncture reduced pain and nausea without prolonging waiting times.22 In a systematic review conducted in 2013, which identified only two small RCTs and two uncontrolled observational studies published in English, the authors concluded that there was insufficient evidence to make any recommendations because of the suboptimal methodological quality of the included studies.23

The paucity of evidence could have been the result of language bias in the previous review. Failure to search Chinese databases resulted in omission of many potentially eligible studies conducted in China, where acupuncture is most widely practised and studied. Besides, 3 years have passed since the last systematic review, during which time many new studies have been undertaken.24 25 It is not known whether any high-quality evidence that has emerged since then might change our view.

Knowledge gaps thus exist regarding the current evidence for the use of acupuncture in the ED. We therefore carried out this systematic review with the aim of including information from both the English and Chinese literature to determine the efficacy/effectiveness and safety of acupuncture, either alone or as an adjunct to standard treatment, in adult patients for a variety of acute clinical conditions in the ED, compared with sham acupuncture (or other placebos) or standard treatment.


This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.26 A review protocol was written and finalised before the literature search commenced, although the protocol was not registered on PROSPERO (International Prospective Register of Systematic Reviews).

Eligibility criteria

The inclusion and exclusion criteria were developed using the PICOS framework (participants, interventions, comparisons, outcomes, and study designs).

Studies were included for review if all of the following inclusion criteria were met:

  1. Population: adult, aged 18 or over, of either gender; ED patients.

  2. Intervention: acupuncture with needle insertion, irrespective of the styles (schools) or subsequent stimulation (manual vs electroacupuncture), either alone or as an adjunctive therapy; all interventions must have been administered and completed in the ED.

  3. Comparison: sham acupuncture or other placebos, and/or standard medical therapies in the ED.

  4. Outcome: clinical outcomes at least at the time of discharge from the ED setting, such as pain score or change in vital signs.

  5. Study design: RCT.

Studies were excluded if any of the following exclusion criteria were met:

  1. Population: mixed population studies in which data from adults and children, or ED and non-ED patients, were not separately presented.

  2. Intervention: acupuncture techniques that did not break the skin (ie, laser acupuncture or acupressure); acupuncture point injection of herbal medicine; combination of herbal medicine or other traditional medicine modalities of treatment (since their presence might confound the result).

  3. Comparison: no control group; inappropriate standard therapies or study interventions; studies comparing different styles of acupuncture only.

Search methods

Four English databases (Embase, PubMed, AMED and CENTRAL) and two Chinese databases (CNKI and Wanfang) were searched from their inception to 8 November 2016 as per the protocol. The keywords used included ‘acupuncture’ and ‘emergency department’, in English or Chinese whenever appropriate. A bibliographic search of references of previous studies was conducted to identify other relevant studies. Only English and Chinese articles were reviewed.

Two authors (KLC/RPKL) searched the databases to generate a list of articles. After removing duplicates, two authors (KLC/CKL) screened potentially eligible articles after reading the title and abstract of the identified studies to determine whether they met the inclusion or exclusion criteria above. Any inconsistency was resolved by consensus. Following the initial screening process, three authors (KLC/RPKL/CKL) conducted a full-text review to determine eligibility based on the pre-defined inclusion and exclusion criteria.

For eligible RCTs, data extraction was conducted independently by two authors (KLC/CKL) while a risk of bias assessment was performed independently by three authors (KLC/RPKL/CKL). The details of acupuncture treatments were extracted based on the revised Standards of Reporting Interventions in Clinical Trials of Acupuncture (STRICTA) items and information was presented across four domains (patients, interventions, outcome parameters, and outcomes).27 Data collected were based on the published findings of the included studies. Attempts were made to contact the corresponding authors of the eligible studies for further information and clarification of details.

Quality assessment

The Cochrane risk of bias tool was used by three authors (KLC/RPKL/CKL) independently to assess the internal validity of each study, covering the following domains: sequence generation, allocation concealment, blinding of participants and/or personnel, blinding of outcome assessment, incomplete outcome data, selective outcome reporting, and other source of bias. The risk of bias of each domain was judged as ‘low risk’, ‘high risk’ or ‘unclear’.28 Any inconsistency was discussed and resolved by consensus (KLC/RPKL/CKL).

Grading the quality of evidence

The quality of evidence for each individual outcome was rated using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach by two authors (KLC/CKL).28

Data analysis

A qualitative summary of results of the included studies was conducted. Where appropriate, we sought to conduct meta-analysis with statistical heterogeneity assessment using Review Manager 5.3.29 Trials were combined depending on their similarity in outcome parameters.


Search results

In total, 1773 publications were identified, including 1223 and 550 publications from the English and Chinese databases, respectively. The bibliographic search yielded 18 additional articles. The search result of each database is shown in online supplementary appendix 1. After removal of duplicates, the title and abstract of 1461 articles were screened for eligibility. The full text of 30 eligible articles were retrieved for detailed evaluation, including nine English and 21 Chinese articles. Six RCTs,30–35 involving 651 patients, were eligible and were included in the analysis. Because of the marked heterogeneity of the included studies in terms of disease spectrum, acupuncture technique and outcome measurement, quantitative synthesis of the study results (meta-analysis) could not be performed. The study selection process is summarised in figure 1. The list of excluded studies and the reasons for their exclusion are shown in online supplementary appendix 2.36–59

Supplementary file 1

Figure 1

Flow diagram showing study selection process. ED, emergency department; RCT, randomised controlled trial.

General characteristics

Three of the RCTs were conducted in China,32–34 one in the USA,30 one in Australia,31 and one in Tunisia.35 Conditions for which acupuncture was applied included acute pain,30–32 35 hypertension33 and cardiac arrest.34

Different acupuncture techniques were used in the included studies. Acupuncture based on TCM theory was used in two studies,31 35 balance acupuncture in three studies32–34 and auricular acupuncture in one study.30 The acupuncture protocol and points used were fixed in four studies.30 32–34 Individualised regimens based on predetermined traditional acupuncture points for different conditions were used in two studies.31 35 Manual manipulation was performed in five studies30 32–35 and electroacupuncture was given in one study.31 Regarding the needle response sought, de qi or a sensation of ache, numbness or distension was sought in three studies.31–33 One-time treatment was given in all six RCTs and the needle retention time ranged from seconds (when balance acupuncture was used)32–34 to a few days.30 Only three studies reported the background of the acupuncturists: TCM practitioner in one study;31 trained postgraduate(s) in another;32 and a trained ED doctor qualified in medical acupuncture in the study conducted in Tunisia.35

One of the studies compared acupuncture with sham acupuncture,32 two compared acupuncture as an adjunct to standard treatment with standard treatment alone,30 34 and three studies compared acupuncture with standard medical therapies.31 33 35 In three studies in which standard therapy was used in comparison, only two studies reported the details of standard treatment.33 35

Among the included studies, pain was the most frequently assessed outcome and was quantified using the visual analogue score (VAS) in three studies31 32 35 and verbal numerical rating scale (NRS) in two.30 35 The effective/success rate of treatment, based on the criteria defined by the authors in individual studies, was assessed in four studies.32–35 Time to recovery/symptom improvement was measured in two studies.34 35 Variability of vital signs were reported in three studies.31 33 35 Adverse effects were assessed in five studies.31–35Patient perspectives, such as patient satisfaction and willingness to undergo the same treatment in the future, was sought in one study.31 The impact of acupuncture on ED operations, such as medication use and length of stay, were reported in two studies.30 31

The characteristics of the included studies are summarised in table 1. The full description of each included study is detailed in online supplementary appendix 3.

Table 1

Characteristics of included studies

Methodological quality

Only five RCTs reported adequate details about random allocation, with four using computers to generate random numbers30 32 33 35 and one using random number tables.34 Liang et al 33 responded to our letter and provided us with the details of random allocation and concealment. Harkin et al 31 did not provide any details of randomisation. Allocation concealment was achieved using envelopes in five studies.30 32–35 Blinding of the participants and acupuncturists was either not done or not possible in all included studies, leading to a high risk of performance bias. Although some may argue that the survival rate was unlikely to be affected by a lack of blinding of patients in cardiac arrest, a lack of blinding of personnel still led to a high risk of performance bias in terms of the team’s effort in resuscitation and time to declare death.34

As for the outcome assessment, Goertz et al blinded the outcome assessors by placing an adhesive patch over the ear lobe after auricular acupuncture.30 Yang et al reported outcome assessment by independent blinded assessors who were not involved in the treatment. It is therefore unlikely that they could have recognised the group assignment since the needles were not retained.32 The risk of detection bias was high for the remaining RCTs,31 33 35 with the single exception of the cardiac arrest trial since the survival outcome was discrete and unlikely to have been affected by the way it was assessed.

Most included RCTs had no dropouts and were at low risk of attrition bias. Although Goertz et al had no dropouts at the time of discharge at ED, 13% of the participants were lost to contact at 24 hours follow-up.30 The risk of selective reporting of all included RCTs was unclear except for Grissa et al, who registered their trial with an apparently minor change in the primary outcome.35 A major imbalance of baseline characteristics between groups was seen in three RCTs on acute pain management.30 31 35 The paucity of eligible studies made it difficult to judge whether there was publication bias. The risk of bias summary and graph are shown in figure 2 and figure 3, respectively.

Figure 2

Risk of bias summary.

Effects of interventions

Acute pain management

Acupuncture versus sham acupuncture for acute pharyngitis

An RCT published in Chinese involving 74 patients with acute pharyngitis showed that balance acupuncture at Yantong on the side contralateral to the throat pain was superior to sham acupuncture at relieving pain at 1 min after completion of treatment, with a significantly higher ‘cure plus markedly effective rate’ as defined by the authors (acupuncture group 44.4% vs sham 10.5%, P<0.01).32 However, even for sham acupuncture, a significant reduction in the VAS score was observed at 1 min after treatment.32

Acupuncture versus standard ED care for acute pain

An RCT published in English on 300 haemodynamically stable adult patients with acute pain (<72  hours   of ED presentation) secondary to a variety of conditions, including musculoskeletal pain, renal colic, abdominal pain and headache, of moderate to severe intensity (VAS or NRS score ≥40) showed that acupuncture at predetermined traditional acupuncture points based on TCM theory was associated with more effective and faster pain control compared with intravenous morphine. The success rate of treatment at 60 min, defined as a reduction in VAS or NRS ≥50% of its baseline value, was significantly higher in the acupuncture group (92% vs 78%, P<0.001). The pain resolution time was also significantly shorter (acupuncture group 16±8 min vs morphine group 28±14 min, P<0.005).35 However, the mean absolute difference in pain score between the two groups was 7.7, which was lower than the threshold of clinical significance defined by Todd et al.60

Electroacupuncture versus standard ED care for acute pain

An RCT published in English involving 45 non-critical ED patients (triage categories—3: urgent; 4: semi-urgent; 5: non-urgent) with pain secondary to acute conditions such as migraine, tension headache, torticollis, muscular back pain, acute musculoskeletal pain and dysmenorrhoea, showed that electroacupuncture at LI4 (Hegu) and LR3 (Taichong) and supplementary points as determined by the acupuncturist was comparable to conventional ED care in reducing pain. A significant reduction in the mean VAS score was seen in both groups after treatment (acupuncture group 25.90±17.64, P<0.001; conventional ED care group 22.18±24.08, P=0.002), but the difference between the groups was not significant (P=0.422).31

Acupuncture as an adjunct to standard ED care for acute pain syndrome

One RCT published in English, which recruited 100 patients with acute pain in different body regions, showed that auricular acupuncture plus standard ED care performed better than standard ED care alone in immediate pain control (2.18 mean difference in pain reduction (NRS score) between groups, P<0.001). However, there was no significant difference between groups at the 24-hour follow-up time point (0.37 mean difference between groups, P=0.503).30


A small RCT published in Chinese with 68 participants showed that balance acupuncture at Jiangya and Toutong and sublingual captopril 25 mg had similar immediate antihypertensive effects at 10, 30, 60 and 120 min after treatment.33

Cardiac arrest

One RCT published in Chinese involving 64 cases of cardiac arrest, including both asystole and ventricular fibrillation, showed that balance acupuncture at GV26 plus standard cardiopulmonary resuscitation (CPR), according to the 2010 Advanced Cardiac Life Support (ACLS) guideline, was superior to standard CPR alone. The return of spontaneous circulation (ROSC) rate was significantly higher (acupuncture group 67.9% vs standard CPR group 29.0%, P=0.003) and the time to ROSC was significantly shorter (acupuncture group 10.47±4.23 min vs standard CPR group 16.00±4.12 min, P=0.004) in the acupuncture group. However, the survival rate did not differ significantly between the groups (acupuncture group 28.6% vs standard CPR group 12.9%, P=0.135).34

Adverse events

Adverse events were not reported in one RCT.30 No adverse event was reported in three RCTs of balance acupuncture, which involved 206 patients.32–34 One RCT reported six cases of mild pain at the needling sites in the electroacupuncture group and one case of adverse reaction in the conventional ED care group, but the difference between the groups was not significant (P=0.235). Another RCT comparing acupuncture with intravenous morphine showed a significantly lower rate of adverse events in the acupuncture group (2.6% vs 56.5%, P<0.001), with three cases of needle breakage and a case of fainting.35 No serious adverse events were reported in any of the included studies.


This systematic review shows that the number of RCTs on the use of acupuncture in the ED in the English databases had remained sparse since the last review in 2013,23 with only one new RCT being published since then.35 In the Chinese databases, more studies covering a wider spectrum of acute clinical conditions, such as cardiac arrest,44 47 acute coronary syndrome,53 shock,38 tachyarrhythmias,36 febrile convulsion,37 acute gastritis,56 59 renal colic48and  dysmenorrhoea,51 were identified. However, the credibility of the papers published in Chinese was weakened by a general lack of description of methodology, the use of inappropriate treatments as controls, and questionable trial results. Only three RCTs32–34 published in Chinese were eligible for analysis based on our pre-defined criteria.

The six RCTs included in this review are heterogeneous in terms of study design, patient characteristics, disease category, intervention, comparison and outcome assessment. Most of them are small RCTs with participant selection based on symptoms rather than a definitive biomedical or TCM diagnosis. While symptom-based inclusion of subjects is relevant to ED practice since a definitive diagnosis may not be available, the study samples became heterogeneous mixtures of patients with very different diseases. Also, acupuncture techniques, acupuncture point selection, and expertise of the acupuncturists varied considerably across different studies, making cross-study comparison difficult. Validated outcome measures, such as VAS for pain, were used in many studies, and were assessed consistently before and immediately after treatment. However, the marked clinical and methodological diversity across different studies precluded any meaningful meta-analysis.28

It is noteworthy that none of the included RCTs followed the Consolidated Standards of Reporting Trials (CONSORT) statement in reporting trial findings.61 Description of the details of the acupuncture intervention was in general inadequate according to the revised Standards for Reporting Interventions in Clinical Trials of Acupuncture (STRICTA) guideline.28 Every included study had a high or unclear risk of bias in at least two domains. A major weakness of all RCTs was a lack of blinding, leading to a high risk of performance and detection bias. The risk of reporting bias and other bias were unclear owing to inadequate description of the methodology, and they remained so despite our best efforts to seek further information from the corresponding authors. The results of trials with unclear biases should be interpreted cautiously since the impact of such biases on the findings is not known.

Quality of evidence

Using the GRADE approach, the quality of evidence for the use of acupuncture in the ED for acute pain, hypertension and cardiac arrest was assessed. For acute pain due to a variety of acute clinical conditions, acupuncture was reported to be superior to sham acupuncture for acute pharyngitis,32 more effective than intravenous morphine,35 comparable to conventional ED care,31 and better than standard ED care alone when used as an adjunct.30 However, the overall level of evidence for pain reduction was downgraded by two levels to ‘low’ because of limitations in study design and implementation, including a lack of blinding when pain score, a subjective outcome, was measured, imprecision of results owing to small sample size in three RCTs30–32 and a high probability of publication bias. A significant placebo effect of sham acupuncture was also noticed in one RCT, raising doubt about the true analgesic effect of balance acupuncture or the validity of sham acupuncture.32

For hypertension, the level of evidence was downgraded by three levels to ‘very low’ because of unclear risk of selection bias, a lack of blinding, small sample size with imprecise results, and a high likelihood of publication bias. The level of evidence for cardiac arrest was ‘very low’ in view of the serious methodological limitations, including pooling patients with very different prognoses without subgroup analysis (significantly higher ROSC and survival rates for ventricular fibrillation than asystole), unclear outcome definition (uncertainty whether or not the survival rate in the paper represented survival to discharge), failure to follow the Utstein-style template (which is commonly used in cardiac arrest research) in reporting study details and findings,62 and inadequate description of the details of the resuscitation given.

Despite our efforts to extend our research to Chinese databases that contain a larger volume of publications on acupuncture, the benefits of ED acupuncture remain undefined. Overall, acupuncture appeared to be safe when used in the ED setting with a low risk. However, one should take into account the small number of patients treated, the marked heterogeneity of the disease spectrum, the varied acupuncture techniques, and the different methodology of the included studies. Moreover, acupuncture is operator-dependent and the findings related to its safety might not be extrapolated across different centres with varying levels of expertise.

Strengths and limitations

To our knowledge, this is the first systematic review of acupuncture in ED settings to extend the literature search to Chinese databases, which were not covered by a previous systematic review.23 Another strength of this review is the systematic evaluation of available evidence using the Cochrane risk of bias tool and the GRADE approach. There are, however, several limitations of the present review. First, language bias remains since articles published in other languages, such as German or Korean, were not reviewed because our study team did not have resources to translate them into English. Second, not every variant of acupuncture was included, although we did cover major acupuncture techniques. Third, many RCTs conducted in China were excluded because of inappropriate standard treatment for comparison or questionable acupuncture practice/trial results. This inevitably reduced the number of studies included. The broad spectrum of ED application of acupuncture might not be fully revealed by the present review. However, most of the excluded RCTs had major methodological flaws and their exclusion is unlikely to change the overall conclusion.

Implications for practice

There is insufficient evidence to support or refute the use of acupuncture in the ED, either alone or as an adjunct to standard ED treatment. The level of evidence is low for acute pain management, and very low for hypertension and cardiac arrest. Acupuncture may have a role in acute pain management as an alternative to pharmacological treatment in the presence of appropriate expertise. Yet, the optimal patient group, disease category, acupuncture techniques, acupuncture point selection, needle stimulation, and frequency and duration of treatment sessions are yet to be determined. High-quality evidence addressing the clinical effectiveness, safety and cost-effectiveness of acupuncture is needed before it can be integrated into routine ED practice.

Implications for research

The current lack of high-quality evidence should be addressed by rigorous RCTs with sufficient sample sizes. Inclusion criteria should be disease-specific, taking into account the TCM diagnosis and syndrome differentiation criteria when a TCM theoretical framework is used. Multicentre trials involving different ethnic groups with different acceptability of acupuncture may improve external validity. The CONSORT statement and the STRICTA guideline should be strictly followed in reporting trial findings. Cost-effectiveness analysis should be conducted to evaluate whether acupuncture is justified among different competing treatment options in the ED.


Overall, there is insufficient evidence to support or refute the use of acupuncture in the ED, either alone or as an adjunct to standard ED treatment. The current lack of high-quality evidence should be addressed by rigorous RCTs with sufficient sample sizes, disease-specific inclusion criteria, and reporting that follows the current best practice for therapeutic trials. Further studies evaluating the clinical effectiveness, safety and cost-effectiveness of acupuncture in the ED are warranted.


View Abstract


  • Contributors KLC and RPKL conceived the study and designed the study. SHT revised the study protocol and approved the project. KLC, RPKL, and SHT supervised data collection. KLC and RPKL recruited and managed the data, including quality control. KLC, RPKL, and CKL provided statistical advice on study design and analysed the data. KLC and RPKL drafted the manuscript. KLC, RPKL, CKL, and SHT contributed substantially to its revision. KLC takes overall responsibility for the paper. All authors read and approved the final version accepted for publication.

  • Competing interests The second author (PKL) has written a master thesis on a similar topic but using a different search strategy and different eligibility criteria. The other authors have nothing to declare.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.