Patient safety incidents from acupuncture treatments: a review of reports to the National Patient Safety Agency
Acupuncture is frequently employed to treat chronic pain syndromes or other chronic conditions. Nevertheless, there is a growing literature on adverse events (AEs) from treatments including pneumothorax, cardiac tamponade and spinal cord injury. Acupuncture is provided in almost all NHS pain clinics and by an increasing number of GP's and physiotherapists. Considering acupuncture's popularity, its safety has become an important public health issue. To evaluate the harm caused to patients through acupuncture treatments within NHS organisations. The National Reporting and Learning System (NRLS) database was searched for incidents reported from 1st January 2009 to 31st December 2011. The free text fields of all reports received from all healthcare settings and specialties were searched for the keyword 'acupuncture'. All relevant incidents were reviewed to provide a qualitative theme of the harm to patients. Results: 468 patient safety incidents were identified; 325 met our inclusion criteria for analysis. Adverse events reported include retained needles (31%), dizziness (30%), loss of consciousness/unresponsive (19%), falls (4%), Bruising or soreness at needle site (2%), Pneumothorax (1%) and other adverse reactions (12%). The majority (95%) of the incidents were categorised as low or no harm. A number of AEs are recorded after acupuncture treatments in the NHS but the majority is not severe. However, miscategorisation and under-reporting may distort the overall picture. Acupuncture practitioners should be aware of, and be prepared to manage, any significant harm from treatments.
International Journal of Risk & Safety in Medicine 24 (2012) 163–169
Patient safety incidents from acupuncture
treatments: A review of reports to the
National Patient Safety Agency
, Taoﬁkat B. Agbabiaka
and Edzard Ernst
Patient Safety Division, National Patient Safety Agency, London, UK
Complementary Medicine, Peninsula Medical School, Plymouth, UK
Abstract. Background: Acupuncture is frequently employed to treat chronic pain syndromes or other chronic conditions.
Nevertheless, there is a growing literature on adverse events (AEs) from treatments including pneumothorax, cardiac tamponade
and spinal cord injury. Acupuncture is provided in almost all NHS pain clinics and by an increasing number of GP’s and
physiotherapists. Considering acupuncture’s popularity, its safety has become an important public health issue.
Objectives: To evaluate the harm caused to patients through acupuncture treatments within NHS organisations.
Methods: The National Reporting and Learning System (NRLS) database was searched for incidents reported from 1st January
2009 to 31st December 2011. The free text ﬁelds of all reports received from all healthcare settings and specialties were searched
for the keyword ‘acupuncture’. All relevant incidents were reviewed to provide a qualitative theme of the harm to patients.
Results: 468 patient safety incidents were identiﬁed; 325 met our inclusion criteria for analysis. Adverse events reported
include retained needles (31%), dizziness (30%), loss of consciousness/unresponsive (19%), falls (4%), Bruising or soreness at
needle site (2%), Pneumothorax (1%) and other adverse reactions (12%). The majority (95%) of the incidents were categorised
as low or no harm.
Conclusions: A number of AEs are recorded after acupuncture treatments in the NHS but the majority is not severe. However,
miscategorisation and under-reporting may distort the overall picture. Acupuncture practitioners should be aware of, and be
prepared to manage, any signiﬁcant harm from treatments.
Keywords: Acupuncture, patient safety, adverse event, pneumothorax
1. Introduction and background
Acupuncture is one of the best known forms of complementary and alternative medicine (CAM) and
is frequently employed to treat chronic pain syndromes or other chronic conditions . Acupuncture has
been used for millennia and some proponents claim that this long history suggests safety. However, there
is a growing literature on adverse events from acupuncture including pneumothorax, cardiac tamponade,
spinal cord injury and viral hepatitis [2–7].
The two main styles of acupuncture practised in the UK are Traditional Chinese and Western Medical
Acupuncture. As the majority of acupuncture treatments are provided outside the NHS it is difﬁcult to
Address for correspondence: Taoﬁkat B. Agbabiaka, Patient Safety Division, National Patient Safety Agency, 4-8 Maple
Street, W1T 5HD, London, UK. E-mail: taoﬁkat.firstname.lastname@example.org.
0924-6479/12/$27.50 © 2012 – IOS Press and the authors. All rights reserved
164 J. Wheway et al. / Patient safety incidents from acupuncture
estimate the number of patients who receive acupuncture treatments per year in the UK. Considering
acupuncture’s popularity, its safety has become an important public health issue.
1.1. Acupuncture in the NHS
The Royal London Hospital for Integrated Medicine, part of University College Hospital NHS Trust,
introduced acupuncture into the NHS in 1977. It is currently the largest provider of acupuncture services
to NHS patients, with several thousand patient sessions per year. Conventional doctors, nurses and phys-
iotherapists who are trained in acupuncture provide all treatments. Acupuncture is also offered as part of
treatment programs at the Glasgow Homeopathic Hospital . Furthermore, acupuncture is provided in
almost all NHS pain clinics and by an increasing number of GP’s and physiotherapists. Acupuncturists
are currently not regulated by statute but several acupuncture organisations exist of which practitioners
with certain qualiﬁcations can be members voluntarily.
1.2. Scale of the problem
Surveys to determine the frequency of acupuncture-related adverse events (AEs) have been conducted
in several countries, e.g. Germany [9, 10], the United Kingdom and Northern Ireland [11, 12] revealed
mild, transient adverse events with an incidence rate of 671 to 1510 per 10,000 consultations. The most
common AEs were local pain from needling and slight bleeding or haematoma. Other reports, however,
showed that acupuncture treatments could also lead to serious, sometimes life-threatening complications
[3, 6, 7, 13]. A recent systematic review of case reports, case series and other articles between 1965 and
2009 reported 86 deaths from complications arising after acupuncture treatments. Due to under-reporting,
this ﬁgure might merely depict the tip of a much bigger iceberg . There are currently no rigorous
systems for reporting AEs of acupuncture. Therefore, the number of AEs recorded in the literature is
likely to be considerably lower than the true ﬁgures.
1.3. Aims of study
This paper is aimed at evaluating the harm caused to patients through acupuncture treatments within
2.1. The National Reporting and Learning System (NRLS) database
The National Patient Safety Agency (NPSA) collects and analyses reports of patient safety incidents
where one or more patients receiving NHS care has been harmed or a near miss incident has occurred.
All NHS organisations in England and Wales report these incidents into the NRLS mainly via local
reporting mechanisms. The national Patient Safety function, which has responsibility for the NRLS, does
not investigate individual incidents but analyses data to identify common risks in order to enable national
learning from these incidents and improve patient safety in the NHS. The NRLS as at December 2011
held over 6.6 million patient safety incident reports and receives over 100,000 reports every month.
J. Wheway et al. / Patient safety incidents from acupuncture 165
Categories and severity of reported patient safety incidents from acupuncture
Type of AE Severity of AE Total
Severe Moderate Low No harm
Retained needles 0 2 12 86 100
Dizziness 0 0 30 69 99
Loss of consciousness/ 0 4 34 25 63
Falls 0 1 5 6 12
Bruising/soreness 0 0 6 1 7
at needle site
Pneumothorax 1 2 1 1 5
Other adverse reactions 0 5 16 18 39
Total 1 14 104 206 325
2.2. Search strategy
The NRLS database was searched for incidents in a two-year period (1st January 2009 to 31st December
2011). For this period, the NRLS received over 3.7 million patient safety incident reports. Each incident
reported to the NPSA has several free text ﬁelds in which the reporter is able to describe details of what
happened when the incident occurred, the apparent causes and actions that might be taken to prevent a
re-occurrence of the incident.
The free text ﬁelds were searched for “acupuncture” for all reports received into the NRLS from all
healthcare settings and specialties.
All relevant incidents were reviewed by a NPSA Clinical Reviewer in order to provide a qualitative
theme of the harm caused to patients as reported in the free text description of the incident. Reports that
were unclear or ambiguous were discussed with a second Clinical Reviewer to reach a consensus.
The search produced a total of 468 reported patient safety incidents of which 325 met the inclusion
criteria for classiﬁcation (Table 1). The incidents excluded contained “acupuncture” within the free text
sections of the report but did not describe harm caused to the patient during or immediately following
acupuncture treatment or was felt not to have been directly associated to the AE.
3.1. Retained needles
There were a total of 100 incidents where acupuncture needles were left in the patient longer than
prescribed. In 59 of these incidents, the patients (either on their way home or at home) found that a needle
had been inadvertently left in place after treatment. Needles that were found in items of clothing were
excluded from the review. For the remaining 41 incidents, the patient’s treatment period continued, in a
few cases for up to three hours longer than intended. Some of these incidents occurred in busy clinics, in
cases where several patients were treated simultaneously and when a timer reminding the clinician that
treatment was over was not heard or acted on.
166 J. Wheway et al. / Patient safety incidents from acupuncture
Twelve of these incidents recount that the member of staff treating the patient had left the department
for a clinical meeting, training session, lunch or even gone home at the end of the day leaving the patient
with needles in situ. Examples include a patient who was left with needles in situ after the department
had been locked up and everyone left for lunch, another patient was found with needles in situ in his
upper body over 30 minutes after all the clinical staff had left at the end of the day.
A total of 99 incidents described the patient feeling dizzy and/or faint without loss of consciousness.
There were several reports of patients feeling light headed, hot and nauseous during treatment. Other
patients complained of dizziness, feeling wobbly and chest tightness within a few minutes into the
3.3. Loss of consciousness/unresponsive
Sixty-three incidents related to patients temporary losing consciousness or found unresponsive by the
clinician providing the treatment. These incidents were described as “became unresponsive, stopped
breathing”, “became unwell and fainted”, “becoming unresponsive to voice for about 30 seconds” or
“slumped to side and fell onto ﬂoor . . . and lost consciousness and stopped breathing”. Twenty (32%)
of these incidents required accident and emergency (A&E)/ambulance assessment. Six of these patients
also appeared to have suffered a minor seizure. None of the reports described underlying conditions, not
related to the acupuncture treatment, as possible causes of the loss of consciousness and many stated that
the patient recovered quickly once the needles were removed.
There were 12 incidents relating to patients having a fall with one incident resulting in moderate harm.
Eight of these incidents described patients falling off the treatment couch during or just after treatment.
Three patients were reported to have fallen in the corridor and another patient fell down a step on the way
out of the clinic. There were several cases of falls reported in the previous category, where the patient
became unconscious or unresponsive and as a result fell onto the ﬂoor or slide out of the treatment chair.
3.5. Bruising/soreness at needle site
There were seven incidents where patients suffered bruising or soreness at the site of the needle entry.
A patient suffered small blisters over the lumbar spine and at the base of left and right thumbs after
Five incidents described pain and difﬁculty in breathing following acupuncture treatments. Two of
the incidents, reported as severe and moderate harm were diagnosed as pneumothorax attributable to
Deﬁned as incidents where either this diagnosis was speciﬁcally stated or there were clear symptoms indicative of this on
clinical review in combination with acupuncture site where pleural puncture was plausible.
J. Wheway et al. / Patient safety incidents from acupuncture 167
acupuncture. At least two of the other three incidents were very likely due to a pneumothorax but
remained undiagnosed at the time of the report.
3.7. Other adverse reactions
A further 39 incidents related to AEs other than those discussed above during or immediately after
treatment. Some of the reported symptoms included; patients experiencing hot ﬂushes or vomiting,
headache or other pains during treatment (not at the immediate needle site).
3.8. Degree of harm
The NRLS grades the degree of harm of patient safety incidents as no harm, low, moderate, severe and
death . The reporter of the AE classiﬁes the severity of harm. Therefore, there may be an element of
subjectivity as an incident reported as severe is one that “appears to have resulted in permanent harm”.
Categorisation of incidents by degree of harm in these data showed that 310 (95%) were classiﬁed as
no or low harm. Of the remaining incidents, 14 (4%) were classiﬁed as moderate harm and one case of
pneumothorax as severe harm. Incidents relating to loss of consciousness and the patient feeling unwell
e.g. stomach pains had a higher incidence of moderate harm. Only two of the incidents describing retained
needles were reported as causing moderate harm; one of which required surgical intervention to remove
the needle. There were no deaths reported in this search.
There were a total of 29 incidents where the patients required further assessment by ambulance crew
and/or were seen in the Emergency Department during or immediately after the acupuncture treatment
session. Eight incidents described problems with the emergency procedures in the clinical area in which
the acupuncture was provided (e.g. empty oxygen cylinder, call bell not working and lack of clarity
regarding emergency procedures). This was particularly apparent in the isolated locations in which some
acupuncture clinics are held.
To the best of our knowledge, this is the ﬁrst report describing AEs after acupuncture within NHS
organisations. A total of 325 AEs were noted within a three-year period. As there is no denominator, an
incidence rate cannot be calculated; presumably, it would be relatively low. Thus acupuncture, as practiced
in the NHS, seems to be a low harm treatment, considering that only one AE was reported as severe.
As the reporter did the classiﬁcation, it is possible that the severity was understated in some cases and
equally possible to have been overstated in others. For instance, there were ﬁve cases of pneumothorax,
a potentially life-threatening condition; only one was classiﬁed as severe and one was even classiﬁed as
causing “no harm”. Similarly, it is possible, even likely that many AEs were not reported at all. Under
reporting of AEs is a known problem and there are reasons to assume that it also affects acupuncture.
The nature of the reported AEs is similar as that from other investigations. A systematic review of 9
prospective surveys found that needle pain was experienced by 1–45% of patients, tiredness by 2–41%
and bleeding by 0.03–38% . Our investigation revealed only one AE reported as causing severe
harm. This is in contrast to case-reports which continue to disclose serious complications and deaths
after acupuncture . The apparent discrepancy might be due to the fact that, in our review, all the
acupuncturists worked in the NHS and had a ﬁrst healthcare qualiﬁcation as well as some acupuncture
training, that acupuncture in the NHS is usually practiced within a narrow ﬁeld of patient’s presenting
168 J. Wheway et al. / Patient safety incidents from acupuncture
problems or that the practitioner’s original clinical background provides at least a basic standard of
patient safety. Training programmes are also well established in the UK, the British Acupuncture Council
(BAcC) working closely with the British Acupuncture Accreditation Board (BAA) oversees acupuncture
education. Training programme includes 80 hours of basic training and formal assessments. However,
this level of training is not required by all NHS organisations. Acupuncturists outside the NHS or outside
the UK may not always be well trained and therefore serious AEs might occur more frequently. A review
of the Chinese literature, for instance, suggested that acupuncturists in rural China are often poorly trained
and therefore a risk factor .
Retained needles were the most frequent incident. Although the majority of these incidents were
reported as no harm, the distress or inconveniences caused to the patients who are unattended for a
couple of hours must be acknowledged. Moreover, only a few reports stated there was a policy on
counting needles in and out. We also identiﬁed a number of incidents where the emergency call bells
or buzzers for clinicians to alert others that they were in need of assistance; a serious shortfall in basic
AEs describing dizziness and a temporary loss of consciousness are commonly cited in the acupuncture
literature and usually attributed to “needle shock”, a vasovagal reaction to needle insertion . A qual-
itative review of these AEs highlighted a number of incidents in which the patient was described to have
been fully assessed and when experiencing an adverse reaction the care and treatment of the patient was
exemplary. However, there were also incident descriptions where the practitioner did not respond swiftly
to the onset of symptoms in contrast to others where the crash team or an ambulance was immediately
called for. Follow up actions also differed widely, ranging from arranging another acupuncture appoint-
ment to stopping treatment completely as acupunctured was no longer deemed appropriate following a
comparable adverse reaction.
A total of 325 patient safety incidents met our inclusion criteria for harm caused by acupuncture
treatment over a two year period. Although, only one incident was reported as severe, mis-categorisation
and under-reporting may distort the overall picture. Acupuncture practitioners should be prepared to
manage all aspects of the patient’s care and treatment during acupuncture, conduct full assessments of
patients at every session and respond readily to any adverse events following treatment. In particular, the
possibility of pneumothorax should be noted.
Practitioners and managers should be responsible for the safe and effective management of acupuncture
clinics with adequate consideration of the system within which the treatment is delivered. This can include
the emergency procedures available, maintenance of equipment, use of practice drills, timing and length
of treatment, counting needles in and out and management of patients within the clinic session. Proactive
hazard analysis techniques (such as Failure Modes and Effects Analysis) are useful in planning and
assessing the risks of services .
We would like to thank Dagmar Luettel for her contributions to the review.
J. Wheway et al. / Patient safety incidents from acupuncture 169
 Ernst G, Strzyz H, Hagmeister H. Incidence of adverse effects during acupuncture therapy—a multicentre survey. Com-
plement Ther Med. 2003;11(2):93-7.
 Ernst E, Sherman KJ. Is acupuncture a risk factor for hepatitis? Systematic review of epidemiological studies. J Gastroenterol
 Norheim AJ. Adverse effects of acupuncture: A study of the literature for the years 1981-1994. J Altern Complement Med.
 Peuker ET, White A, Ernst E, Pera F, Filler TJ. Traumatic complications of acupuncture. Therapists need to know human
anatomy. Arch Fam Med. 1999;8(6):553-8.
 White A. A cumulative review of the range and incidence of signiﬁcant adverse events associated with acupunture. Acupunct
 Yamashita H, Tsukayama, White A, Tanno Y, Sugishita C, Ernst E. Systematic review of adverse events following
acupuncture: The Japanese literature. Complement Ther Med. 2001;9(2):98-104.
 Zhang J, hang H, Gao X, Ernst E. Acupuncture-related adverse events: a systematic review of the Chinese literature. Bull
World Health Organ. 2010;88(12):915-21c.
 Roberts J, Moore D. Mapping the evidence base and use of acupunture within the NHS. Department of Public Health and
Epidemiology. West Midlands Health Technology Assessment Group. 2007; Report Number 59.
 Endres HG, Molsberger A, Lungenhausen M, Trampisch HJ. An internal standard for verifying the accuracy of serious
adverse event reporting: The example of an acupuncture study of 190,924 patients. Eur J Med Res. 2004;(9):545-51.
 Witt CM, Brinkhaus B, Pach D, et al. Safety of acupuncture: Results of a prospective observational study with 229,230
patients and introduction of a medical information and consent form. Forsch Komplementmed. 2009;16:91-7.
 Macpherson H, Scullion A, Thomas KJ, Walter S. Patient reports of adverse events associated with acupuncture treatment:
A prospective national survey. Qual Saf Health Care. 2004;13:349-55.
 White A, Hayhoe S, Hart A, Ernst E. Adverse events following acupuncture: Prospective survey of 32 000 consultations
with doctors and physiotherapists. BMJ. 2001; 323:485-6.
 Ernst E, Lee MS, Choi TY. Acupuncture: Does it alleviate pain and are there serious risks? A review of reviews. Pain.
 Ernst E. Deaths from acupunture: A systematic review. Int J Risk Saf Med. 2010;22(3):131-6.
 National Patient Safet Agency. Seven steps to patient safety.The full reference guide. London; 2004: p 100 [cited 21st
December 2011]. Available at http://www.nrls.npsa.nhs.uk/resources/?EntryId45=59787.
 Ernst E, White AR. Prospective studies of the safety of acupuncture: A systematic review. Am J Med. 2001;110(6):481-5.
 Weiner DK, Ernst E. Complementary and alternative approaches to the treatment of persistent musculoskeletal pain. Clin
J Pain. 2004;20(4):244-55.
 Ward JR, Clarkson PJ, Buckle P, Berman J, Lim R, Jun GT. Prospective hazard analysis: Tailoring prospective methods to
a healthcare context. Cambridge: University of Cambridge Engineering Design Centre; 2010. Research Project: PS/035.