Article Text

other Versions


Acupuncture and standard emergency department care for pain and/or nausea and its impact on emergency care delivery: a feasibility study
  1. Anthony L Zhang1,
  2. Shefton J Parker1,
  3. De Villiers Smit2,
  4. David McD Taylor3,
  5. Charlie C L Xue1
  1. 1School of Health Sciences, RMIT University, Bundoora, Victoria, Australia
  2. 2Emergency and Trauma Centre, The Alfred Hospital, Melbourne, Victoria, Australia
  3. 3Department of Emergency, Austin Health, Heidelberg, Victoria, Australia
  1. Correspondence to Professor Charlie Xue, School of Health Sciences, RMIT University, 3083, Melbourne, VIC Australia; charlie.xue{at}


Objective To evaluate the feasibility of delivering acupuncture in an emergency department (ED) to patients presenting with pain and/or nausea.

Methods A feasibility study (with historical controls) undertaken at the Northern Hospital ED in Melbourne, Australia, involving people presenting to ED triage with pain (VAS 0–10) and/or nausea (Morrow Index 1–6) between January and August 2010 (n=400). The acupuncture group comprised 200 patients who received usual medical care and acupuncture; the usual care group comprised 200 patients with retrospective data closely matched from ED electronic health records.

Results Refusal rate was 31%, with ‘symptoms under control owing to medical treatment before acupuncture’ the most prevalent reason for refusal (n=36); 52.5% of participants responded ‘definitely yes’ for their willingness to repeat acupuncture, and a further 31.8% responded ‘probably yes’. Over half (57%) reported a satisfaction score of 10 for acupuncture treatment. Musculoskeletal conditions were the most common conditions treated n=117 (58.5%), followed by abdominal or flank pain n=49 (24.5%). Adverse events were rare (2%) and mild. Pain and nausea scores reduced from a mean±SD of 7.01±2.02 before acupuncture to 4.72±2.62 after acupuncture and from 2.6±2.19 to 1.42±1.86, respectively.

Conclusions Acupuncture in the ED appears safe and acceptable for patients with pain and/or nausea. Results suggest combined care may provide effective pain and nausea relief in ED patients. Further high-quality, sufficiently powered randomised studies evaluating the cost-effectiveness and efficacy of the add-on effect of acupuncture are recommended.


Statistics from


In a community setting among Australian GPs, complementary medicine for various health conditions is increasing in popularity.1 In a survey of GPs, 84% described acupuncture potentially as moderately or highly effective and 57% of GPs encourage their patients to use acupuncture.2

Over the last several decades, acupuncture research has identified a number of mechanisms of action related to clinical application, based on measurable responses triggered by acupuncture.3 Studies show that acupuncture enhances the post-stimulation spatial extent of resting brain networks to include anti-nociceptive, memory and affective brain regions, which might explain why acupuncture has an analgesic effect.4

Pharmacotherapy analgesia, although effective for painful conditions, has limitations to its use owing to side effects, allergies and drug interactions. The extended use of some analgesic drugs, such as opioids, places patients at higher risk of developing constipation, cardiovascular disease and/or potential dependence.5 ,6 Despite the short-term benefit of analgesic drugs, it is not uncommon for a patient's pain to return once the therapeutic effect has worn off.7 GPs are aware of the limits of pharmacotherapy and are increasingly suggesting complementary medicines such as acupuncture to patients who present with uncomplicated painful conditions such as carpal tunnel syndrome.2 ,8

In the emergency department (ED), the possible advantages of acupuncture compared with pharmacotherapy are relatively unknown and untested. In 2007, the Australian and New Zealand College of Anaesthetists, based on level 1 scientific evidence, recommended that “acupuncture may be effective in some acute pain settings… and acupuncture reduces postoperative pain as well as opioid-related adverse effects”.9 The National Institute of Health (NIH) has concluded that there is sufficient evidence of the value of acupuncture to expand its use into conventional medicine.10

This pilot study was designed to investigate the feasibility, potential efficacy, safety and acceptability to patient and medical staff of acupuncture in an ED. The results were intended to assist in the development of a large-scale RCT of acupuncture in the ED.11


Human research ethics approval was granted from the Northern Hospital and RMIT University and the protocol conformed to the provisions of the Declaration of Helsinki in 1995 (as revised in Seoul 2008). The study was conducted at the ED of the Northern Hospital, Victoria, Australia, between January and August 2010. Acupuncture was available to patients 10 hours a week, over 2 days, for 8 months. Figure 1 shows the path for patients who were screened and consented to participate in the pilot study.

Figure 1

Path for patients who were screened and consented to participate in the feasibility study.


Acupuncture group

Two hundred triaged patients, ≥18 years of age, with symptoms of pain and/or nausea, were identified via the ED information system (EDIS). Patients whose language prevented them providing informed consent, critically ill patients (determined by the treating physician or the consultant in charge) and pregnant women were screened and excluded before an approach for consent was made. Participation was voluntary. If the patient refused, a reason was documented without recording patient identifiers.

Usual care group

Two hundred historical patient records were screened and matched 1 week after each comparative case for gender and closest possible match to age, arrival day and time, presenting condition and discharge diagnosis to that of the acupuncture group participant. To limit selection bias, the usual care group patients were screened in the EDIS by a research assistant not associated with the care of patients. No identifying information was collected and no contact with matched patients was necessary. Usual care group data were collected to compare time to medical consultation and discharge with acupuncture group data, to investigate if acupuncture caused delays in patient care or extension of their length of stay.

Acupuncture protocol

Potential participants were screened chronologically from the earliest to the most recent triage via their EDIS triage notes for conditions of pain and/or nausea, then again by the treating physician or consultant in charge to ensure their suitability for acupuncture. After patient consent had been received, an acupuncturist undertook a short period of discussion with the patient and observation, obtaining a brief history of their condition, patient-reported pain score (VAS) and/or their reported nausea (Morrow Index) score.12 A pragmatic acupuncture treatment prescription was developed for each patient individually (using WHO acupuncture location recommendations),13 based on the acupuncturist's evaluation. Acupuncture was delivered by emergency medical physicians with medical acupuncture qualifications, registered acupuncturists with the Chinese medicine registration board of Victoria or final year RMIT university acupuncture students under supervision of registered acupuncturists. The supervising acupuncturist held an honorary ED position for the duration of the study. Funding for the equipment was provided by RMIT University with assistance from the Department of Health, Victoria. Medical acupuncturists were part of the ED staff and delivered acupuncture to patients using study equipment, in addition to their regular roles within the ED.

Manual acupuncture was performed at the bedside in an ED cubicle or treatment room using Chinese manufactured, copper-handled, stainless steel single-use needles, 0.25 mm gauge and either 30 mm or 40 mm in length. For each point, regardless of the treating acupuncturist, the needle was manipulated after insertion until the sensation of de qi (dull ache and possible grabbing sensation of the needle) was experienced. Needles were left for about 20 min before removal by an acupuncturist. Immediately after removal of needles, the acupuncturist recorded patient-reported pain and/or nausea scores, adverse events and the participant's acceptability of (a) the ED visit, (b) the acupuncture treatment and (c) their willingness to repeat acupuncture again in the future for their condition.

The ED doctor was free to assess and consult the patient at any stage before, during or after acupuncture, and pharmacotherapy was permitted as necessary.


Demographic data, triage presenting symptoms and discharge diagnosis were recorded as well as times for triage, initial medical consultation, approach for acupuncture, start of acupuncture, completion of acupuncture and discharge. Adverse events reported by patients, ED staff or acupuncturist were recorded after acupuncture treatment. Willingness to repeat acupuncture for a similar condition in the future and satisfaction with the acupuncture was measured using 10-point ordinal scale, and pain was measured (VAS 0–10) at triage, before acupuncture and immediately after acupuncture.14 If applicable, nausea was measured using the Morrow Index (range 1–6) at triage, before acupuncture and immediately after acupuncture. Staff time required to manage patients was calculated by subtracting the time from ‘triage to intervention’ (acupuncturist or physician consultation) from the ‘triage to discharge time’ (min).

Data analysis

No a priori power calculation was performed for sample size as this was a feasibility study, designed to guide the development of a full size RCT. The sample size was based on enrolling available cases over an 8-month period. Statistical Package for Social Sciences for Windows (SPSS) V.19.0 was used to analyse the data. The data for the two groups were compared using the independent t test, χ2 test and the Mann–Whitney U test, where applicable. A value of p ≤0.05 was considered statistically significant. Post-recruitment phase discussions were held between researchers and hospital management to evaluate and report the impacts of the pilot acupuncture study on the ED, grouping similar participant data and evaluating the various pathways by which participants were recruited, treated and discharged throughout the study.


A total of 290 ED patients were approached about acupuncture treatment, 200 (69.0%) consented and completed treatment. The acupuncture group consisted of two participant subgroups: acupuncture given before (n=55) or after (n=145) medical consultation (figure 2). Participants were not allocated to the groups; inclusion in a group was dependent on whether the treating physician or acupuncturist saw the patient first after triage. Common reasons for declining acupuncture were (1) symptoms were under control owing to medical treatment before acupuncture (n=36, 40%); (2) preference for medical treatment (n=13, 14.4%); (3); not interested in acupuncture (n=10, 11.1%) and (4) needle phobia (n=8, 8.9%).

Figure 2

Acupuncture treatment before or after medical consultation and historically matched controls.

Overall, 98.5% of respondents reported a satisfaction score of between 5 and 10. More than half (52.5%) of participants responded ‘definitely yes’ for their willingness to repeat acupuncture, and a further 31.8% responded ‘probably yes’. Over half (57%) reported a satisfaction score of 10.

Overall, 222 women and 178 men were included in the study. The acupuncture group was well matched for age and gender, and was representative of the ED usual care population in their diagnostic category and waiting time (table 1). More than half of the participants presented with musculoskeletal back or limb pain (58.5% and 59.5% for the acupuncture and control groups, respectively). Most participants had an Australasian Triage Scale (ATS) rating of ≥3 (1, most urgent; 5, least urgent).15

Table 1

Demographics, diagnosis, waiting time and staff time required of patient groups

Overall, there was no gender or age difference between patients who received acupuncture before or after medical consultation.

In the acupuncture group, there was significant difference between the patient-reported pain scores before (mean=7.01, SD=2.02) and after acupuncture (mean=4.72, SD=2.62) [t (193)=14.81, p<0.001]. Mean nausea reported scores before (mean=2.6, SD=2.19) and after (mean=1.42, SD=1.86) acupuncture treatments also differed significantly (p<0.001).

Waiting time to treatment and staff management time were not normally distributed. There was no significant difference in waiting time between the acupuncture (85±5 min, SE) and control groups (86±5mi n, SE) (p=0.94). However, those who received acupuncture before consultation with a medical doctor (N=55), had a significantly shorter waiting time to treatment (66±10 min, SE) than those receiving acupuncture after medical consultation (N=145, 134±4.95mi n, SE, p < 0.0001). Furthermore, the staff time required to manage patients who received acupuncture before (mean= 182, SD=99 min) and after (mean=273, SD=152 min) medical consultation differed considerably (p<0.0001).

Four patients (2%) in the acupuncturist group reported adverse events during the course of the treatment. These included two cases of slight bleeding and two cases of mild pain at the needling sites. No serious adverse events were reported.


If EDs are to maximise outcomes for patients, reduce healthcare resource use and develop strategies to deal with key management issues such as ‘access block’ and staff workload stress, then EDs should explore new interventions.16 ,17 A study in the USA explored acupressure in patients with minor trauma and concluded that it significantly reduced pain in prehospital ambulance care.18 A German study of 60 emergency medical service patients, predominantly exploring acupuncture effects on nausea (n=27/32) and pain (n=19/21) showed that acupuncture can alleviate these symptoms.19

To our knowledge, this is the first study of the impact of the acupuncture service on ED patients in an Australian hospital. A similar trial, conducted in the USA, showed efficacy for pain reduction. However, that study had a much smaller sample (n=20), a group with a narrower range of disease (acute musculoskeletal extremity pain due to non-penetrating injury only) and used different methodology (patients were withdrawn if analgesics were given before acupuncture).20

Our results suggest that acupuncture combined with usual care may reduce pain and nausea. However, a large-scale, adequately powered trial is recommended. As all participants were able to receive standard care during the study, the reduction of symptoms cannot solely be attributed to acupuncture and might be the sole or combined therapeutic effect of pharmacotherapy interventions before acupuncture. A multi-arm randomisation trial comparing standard care treatment, acupuncture alone and a combination of both would assist in calculating the efficacy of the acupuncture component.

The refusal rate to receive acupuncture was relatively low at 31% (n=90), with the primary reason for refusal being that their symptoms were under control owing to medical treatment before being approached for acupuncture (40% n=36). Only a small majority of the refusals (11.1% n=10) identified their primary reason as ‘not interested in acupuncture’, suggesting that most people presenting to the ED would be interested in receiving acupuncture if their symptoms were not under control.

Musculoskeletal conditions comprised 58.5% of conditions treated (n=117), thus this patient group would be most likely to achieve sample size targets in future acupuncture ED studies.

It was recognised that some patients were not suitable for acupuncture without first being assessed by an ED physician. These patients were identified as vulnerable or at risk by ED staff, requiring further consultation with a physician and testing before being approached for acupuncture. Reasons for acupuncture after physician assessment varied and included elderly age, complex medical history, atypical presenting signs and symptoms, triage category and symptom severity (ATS 1–3). In other instances patients first seen by a physician had inadequate pharmaceutical relief so acupuncture was performed.

Only two minor adverse events were reported, suggesting that acupuncture is probably safe in an ED. Interestingly, several patients stated that acupuncture enhanced the analgesic effect of drugs given earlier, and several reported that acupuncture ‘triggered’ the analgesic effect of drugs, which previously had provided little or no therapeutic relief.

It was noted that senior physicians were more likely than junior physicians to permit acupuncturists to approach patients before physical assessment. Staff knowledge of the study protocol and basic acupuncture theory was recognised as fundamental to the success of future acupuncture studies in the ED. Early and continuing education open to all ED staff prevented disruptions to ED workflow.

There was significant difference between the total time required to manage patients who had acupuncture after physician consultation and those who had acupuncture before consultation. The reason for the difference is unclear but may in part be because patients requiring prior physician assessment had more complex medical problems.

As EDs are very busy and introduction of acupuncture to an Australian ED is relatively recent, it was  quickly realised that implementing an RCT that would fulfil CONSORT statements guidelines on reporting21 would be difficult. The range of conditions of patients presenting with pain and or/nausea to the ED is wide, and stratifying patients accordingly for an RCT was not viable. Thus, this pragmatic feasibility study was developed to highlight trends in data and the prevalence of conditions, identify limiting factors and assist methodological development of a high-quality, large-sample RCT.11

Limitations and implications for future research

Although the process of usual care group selection was performed by an independent researcher, the possibility of patient selection bias remains. In addition, as the usual care patients were identified retrospectively, data extraction was limited to that available from the electronic records, thus preventing comparison of some potentially informative data not recorded, such as pain scores. Age, presenting time or day and other matching data between the groups were not identical. Randomisation would assist future ED acupuncture researchers to avoid confounders and potential bias.

The study only functioned when an acupuncturist was available 2 days a week (10 h total). Owing to the limited availability of an acupuncturist and the order in which patients were approached for consent (least recent to most recent triage time) there is a likelihood that many suitable patients were discharged before being approached. Some of these patients might have become comparator controls; however, this would not confound the results.

This pilot has attracted media and public attention. One patient presented to triage with written referral from their community GP to attend the hospital ED for acupuncture, highlighting the community's awareness of the study.

The passage of an ED patient from triage to discharge is a complex series of events, all affecting the eventual outcome. The time of day, level of experience of doctors on duty, daily level of nursing resources and distribution of ATS-classified patients,1–5 all affect a patient's length of stay. Given the pragmatic nature of this study and the complex environment in which it was delivered, the results for acupuncture delivery in an ED are encouraging. For future research, review of patient influx numbers is recommended with only health conditions with feasible recruitment targets being selected.

A dedicated ED area for acupuncture was not always available owing to space restrictions and sudden patient influx. Such an area is advantageous, though not essential, particularly for triaged patients awaiting bed allocation or further examination. If an acupuncture space is not available, reclinable chairs may be used as an alternative, as the participant needs to lie still in a prone or supine position for 15–30 min.

Further trials of acupuncture should examine the effect on health resources, staff management time and re-presentation rates, while controlling for as many variables as possible (condition, analgesic drug, etc). Randomisation, blinding of assessors and a larger sample size will improve the reliability of results and provide stronger evidence. Similarly a prospective rather than retrospective comparator study would allow the comparison of other potentially significant data, such as pain score reduction, among patients who did not receive acupuncture.


Acupuncture appears to be a safe and acceptable treatment for patients in an ED. Acupuncture, in combination with usual medical care, may reduce pain and nausea symptoms among ED patients. The complex ED environment requires acupuncture studies to be carefully planned to achieve recruitment targets; continuing ED staff education is recommended. Future research should develop specific inclusion/exclusion criteria to ensure a clinically similar sample, reduce potential confounders and aid reproducibility of results. High-quality RCTs, with a large sample size, investigating the cost-effectiveness of acupuncture for pain and/or nausea in ED patients are recommended. Patients with musculoskeletal conditions were most commonly suitable for acupuncture at this ED and it would be practical for these conditions to be selected for further ED acupuncture research.

Summary points

  • We explored the feasibility of providing acupuncture for pain and nausea in an emergency department.

  • Eighty nine per cent of patients were interested in receiving acupuncture.

  • Reduction of pain and nausea was significant.

  • Acupuncture did not delay conventional care.


The Department of Health, Victoria, provided financial assistance for the study, but had no role in the collection, analysis or interpretation of data. We thank Dr Robyn Parker from the Maroondah Hospital emergency department (ED) for the assistance she provided to ED staff at the Northern Hospital. We also thank Mr Thomas Keech for data collection and the staff and patients of the Northern Hospital ED who participated in the study.


View Abstract


  • Contributors All authors contributed to the planning and operation of the pilot study and were involved in drafting and writing this manuscript. SJP was the supervising acupuncturist, collecting data and administering acupuncture. ALZ analysed the results and provided statistical reporting.

  • Funding Department of Health, Victoria.

  • Competing interests ALZ, SJP, DVS, DMT and CCLX received support from RMIT University and the Department of Health, Victoria, for the submitted work but have no non-financial interests that might be relevant; DVS was employed in the previous 3 years by The Northern Hospital, which might have an interest in the submitted work.

  • Ethics approval RMIT and The Northern Hosiptal human research ethics committees.

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

  • Data sharing statement Any unpublished and additional data, such as description of the acupuncture points used can be requested from the corresponding author of this manuscript.

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.