Abstract
Objectives:
The objective of this study was to make non-Chinese readers better informed about the current safety situation of acupuncture in China.
Methods:
Four (4) major Chinese comprehensive databases were searched: Chinese Journal Full-text Database (1949–2010), China Biomedical Literature Database (1978–2010), Chinese Technology Journal Database (1989–2010), and Wanfang (1998–2010). Case reports of adverse events related to acupuncture therapy were included. Literature reviews, case–control study, translations, duplicate literatures in various databases, and duplicate published articles with duplicate data were excluded.
Results:
One hundred and sixty-seven (167) articles were included with 1038 cases, among which 35 patients died. The most frequent adverse events were syncope (468 cases), pneumothorax (307 cases), and subarachnoid hemorrhage (64 cases).
Conclusions:
Acupuncture-related adverse events were mainly caused by mental tension of the patient, improper operation by the doctor, and incompleteness of sterilization. Most of them can be avoided by standardizing teaching and clinical practices. Making the corresponding safety standards can greatly lower the risk of adverse events and protect patient safety to the greatest extent.
Introduction
Acupuncture therapy originated from China. It is used more widely in China than any other countries. Thus, there may be more occurrences of adverse events in China than in other countries. This article aims to systematically review the Chinese literature on acupuncture-related adverse events and inform non-Chinese national readers about the current safety situation of acupuncture in China.
Methods
Defining acupuncture-related adverse events and their range
The International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use, of which the U.S. Food and Drug Administration (FDA) and the World Health Organization are members, described an adverse event as “any untoward medical occurrence that may present during treatment with a pharmaceutical product but which does not necessarily have a causal relationship with this treatment.” 13 In accordance with its definition of an adverse event and the definition of acupuncture-related adverse events in the previous literature, 14 the following definition of acupuncture-related adverse events is proposed: “Any unfavorable and unintended sign, symptom, or disease that presents during or after treatment with acupuncture regardless of causal relationship.”
After preretrieval and repeated discussion, the range of acupuncture-related adverse events including syncope, bending of the needle, sticking of the needle, breaking of the needle, residual sense, burns, allergy, hemorrhage, hematoma, pain, infection, internal organ or tissue injury was finally determined.
Selecting databases
Four (4) major Chinese comprehensive databases were searched: Chinese Journal Full-text Database (1949–2010), China Biomedical Literature Database (1978–2010), Chinese Technology Journal Database (1989–2010), and Wanfang (1998–2010). The deadline of the retrieval is December 2010.
Formulating retrieval strategy
The following search terms were used:
Assessing the search results
Inclusion and exclusion criteria
Case reports about all kinds of acupuncture therapies including acupuncture, moxibustion, cupping, electro-acupuncture, laser acupuncture, auricular acupuncture, wrist and ankle acupuncture, acupuncture point injection, and acupressure were included. Literature reviews, case–control study, translations, duplicate literatures in various databases, and duplicate published articles with duplicate data were excluded. The search was limited to Chinese literature.
Data from all the articles that met the inclusion criteria were extracted by the first two authors independently including title of the articles, year published, type of the events, number of cases, applied acupoints, qualification of acupuncturists, primary disease, prognosis, and so on. The likelihood of causality between the event and acupuncture was assessed in each individual case. The causality assessment was based on the following criteria 15 : (1) time relationships between the acupuncture use and the adverse event, (2) pathophysiology of the adverse event, (3) competing causes for the adverse event, (4) response to dechallenge, and (5) response to rechallenge. The likelihood was classified as “certain,” “probable,” “possible,” “unlikely,” “unclassified,” and “unclassifiable.” 16 For controversial issues, the authors reached a consensus after discussion.
Results
One hundred and sixty-seven (167) reviewed papers, published between 1956 and 2010, identified 1038 cases of adverse events associated with acupuncture. The distribution of these adverse events over the 55 years is shown in Figure 1, which shows a significant upward trend in the 1980s and then a slight decrease in the following 20 years.
Distribution of cases of acupuncture-related adverse events reported from 1956 to 2010.
Acupuncture-related adverse events are classified into four categories: syncope, organ or tissue injury, infection, and others (Table 1). Syncope is the most common adverse event with 468 cases, accounting for 45% of the total number of the adverse events. The following sections present findings for each category of adverse events.
Syncope
Twenty-eight (28) articles 17 –44 reported a total of 468 cases of syncope associated with acupuncture including fainting of needling (453 cases), fainting of moxibustion (1 case), and fainting of cupping (14 cases). Ninety-eight percent (98%) of the syncope occurred during the therapeutic process, while only 10 patients fainted after the treatment. Syncope was mainly caused by the following reasons: accepting acupuncture therapy for the first time, mental tension, fatigue, hunger or improper position. Except for the 74 cases without mentioning the outcomes, all the fainting patients recovered after proper treatment without sequelae.
Organ or tissue injury
One hundred and six (106) articles reported a total of 451 cases of tissue or organ injury associated with acupuncture, including 26 cases of death (see Table 2 for tabulation of all reported deaths, case numbers 1–26). The specific cases are as follows.
Pneumothorax
Pneumothorax is the most frequently reported tissue or organ injury caused by acupuncture in 307 cases. 45 –96 Most of primary diseases of the patients who had pneumothorax included periarthritis of shoulder, cervical spondylosis, stiff neck, intercostal neuralgia, and so on. The acupoints involved were Feishu (BL13), Jianjing (GB21), Ganshu (BL18), Tiantu (CV22), Tianding (LI17), Jiuwei (CV15), Quyuan (SI13), Futu (LI18), Dingchuan (EX-B1), Quepen (ST12), Zhongfu (LU1), Fengmen (BL12), Gaohuang (BL43), Dabao (SP21), Shencang (KI25), and Ashi acupoints on the shoulders. The injuries were mainly caused by the following reasons: excessive inserting depth, improper acupoints selection or changing position during retaining needles. Two hundred fifty-two (252) of the 307 cases recovered completely; 49 not stated; and 6 died (Table 2, case numbers 1–6).
Subarachnoid hemorrhage
Subarachnoid hemorrhage was reported in 64 patients. 97 –113 The primary diseases of the patients with subarachnoid hemorrhage were headache, insomnia, neurasthenia, epilepsy, spasm of face, and the acupoints associated with such adverse events were Fengchi (GB20), Yamen (GV16), Anmian (EX-HN14), and Yiming (EX-HN13). Subarachnoid hemorrhage was mainly caused by inserting the needle too deeply. Of the 64 cases presented, 2 patients died (Table 2, case numbers 7–8), 50 recovered, and the outcome was not stated for 12 cases.
Hemorrhage and hematoma
Twenty-two (22) cases of hemorrhage and hematoma were reported. 114 –118 Nineteen (19) of the 22 cases were caused by puncturing the blood vessel and not pressing the acupuncture sites. Two (2) patients suffered such an adverse event by needling the acupoints Sanyinjiao (SP6) and Shenshu (BL23) during their menstrual period. The last one was confirmed with acquired hemophilia B, whereas the doctor neglected to ask the past history of the patient when needling. Eight (8) of the cases recovered. Outcomes for the remaining cases were not stated.
Nerve injury
A total of 10 nerve injury cases were all caused by using forceful strength during the insertion when needling. 119 Outcomes for all of the cases were not stated.
Death
Eight (8) cases of death caused by acupuncture were reported 120 –125 (Table 2, case numbers 9–16). The acupoints involved were Fengchi (GB20), Renying (ST9), Tiantu (CV22), and acupoints on the chest.
In addition to the above, epidural hematoma was reported in 7 cases, heart rupture and death in 3 cases, intestinal tube injury in 2 cases, peroneal nerve injury in 2 cases, hemothorax in 2 cases, each of the following was for 1 case: ileus, delayed pericardial tamponade, spinal cord puncture and death, heart puncture and death, medulla oblongata puncture and death, oculomotor nerve injury, retroperitoneal hematoma, phrenic nerve injury, muscle contracture, acute subdural hematoma, spinal cord injury, spinal cord injury and death, spinal epidural hematoma, brainstem hemorrhage, brainstem hemorrhage and death, brain abscess, rupture of spleen, chylothorax, upper limb function activities obstacles, sublingual hematoma, pericardial tamponade, medullary hemorrhage and death, liquid pneumothorax, greater occipital cistern hemorrhage, aortic arch hemorrhage and death 126 –150 (for death cases see Table 2, case numbers 17–26).
Infection
Thirty-eight (38) cases of infection were reported including 9 cases of death (see Table 2 for tabulation of all reported deaths, case numbers 27–35). The specific conditions are as follows:
Tetanus
Tetanus was the most common infection, with a total of 14 cases. 151 –156 Most of the cases of tetanus happened in the countryside of China and were manipulated by unqualified village practitioners. Eight (8) of the 14 died owing to poor sterilization procedure (Table 2, case numbers 27–34). A 2-year-old child who suffered malnutrition underwent acupuncture of the acupoints of Sifeng (EX-UE10) by an unlicensed practitioner with an unsterilized and rusty sewing needle; the child died of tetanus 7 days later.
Local infection
Local infection was reported in 12 patients. 157 –159 Eleven (11) of the 12 cases were caused by a failure to comply with the aseptic operation. All patients recovered after appropriate treatment. The last case was a 60-year-old woman patient who suffered from periarthritis of the shoulder acupunctured by Tiaokou (ST38) penetrating Chengshan (BL57). The doctor sterilized strictly and used disposable sterile acupuncture needles during the operation process; however, bacterial infection occurred in the lower limbs of the patient. The doctor found that the blood glucose level of the patient was as high as 14 mmol/L after a blood test. The infection of the lower limbs was under control after the treatments of reducing blood sugar and anti-infectious treatment.
Peritonitis
Seven (7) cases of peritonitis were reported, 160 –162 including 4 cases of localized peritonitis and 3 cases of diffuse peritonitis. Due to dysfunction of the gastrointestinal tract and decline of the body resistance, 1 patient had impairment of the intestinal tube when the abdomen was acupunctured, which resulted in localized peritonitis. Diffuse peritonitis was mainly caused by perforation of the gallbladder when acupunctured too deep or when there was enlargement of the gallbladder.
In addition to the above, there was 1 case of each of the following: sepsis, 163 intracranial infection, 164 deep vein inflammation, osteomyelitis, and facial neuritis, 123 respectively. Due to the white spherical objects in the alveolar socket being acupunctured with an unsterilized stainless steel needle, a child died from sepsis after 1 month (Table 2, case number 35). Intracranial infection was caused by the spread of folliculitis when the acupoints on the face and head were acupunctured. Lack of sterilization when taking an acupoint injection was the cause of deep vein inflammation, osteomyelitis, and facial neuritis.
Others
Other adverse events were reported with a total of 81 cases. 165 –183 The specific situation was as follows: retaining the needle, 19 cases; breaking of the needle, 15 cases; bending of the needle, 13 cases; sticking of the needle, 7 cases; allergy, 6 cases; epileptic seizure, fever, lame foot, stomachache, hyperventilation syndrome, allergic shock, cough, thirst, likeness of infusion reaction, abortion, scald of skin, hoarseness, blindness, retinal detachment, shock, subcutaneous hemorrhage, subcutaneous emphysema, subcutaneous scleroma, nettlerash, discomfort, fat liquefaction, 1 case for each.
Needle-left (forgetting to remove the needle after the treatment), was due to a poor sense of responsibility of the doctors. Bending of the needle is caused by improper manipulations. Sticking of the needle resulted from changing postures during the treatment. Poor quality of the needle led to breaking of the needle.
Allergies occurred in 6 cases, including 4 cases of moxibustion allergy and 2 cases of filiform needle allergy. The manifestations of allergy to moxibustion were redness and swelling, pruritus and papule, which would disappear after stopping the moxibustion and the anti-allergy treatment. After the electro-acupuncture treatment, the needling area on the patients showed symptoms of redness and swelling, pruritus, and blister. The symptoms disappeared after stopping the treatment. The patients were confirmed to be allergic to metal needles by means of allergy testing afterwards.
Discussion
Adverse events in medical practice are always of concern to the public and the medical profession. Prospective surveys conducted in the United Kingdom 184 and Germany 185 have shown that in extremely rare cases, acupuncture can lead to serious acupuncture-related adverse events, although most of them are mild and transient. The incidence of adverse events ranges from 6.71% to 8.6%, 184,185 and the rate of serious adverse events (death, organ trauma, or hospital admission) was about 0.001%. 185 Comparison of the rate of adverse events from acupuncture with those from drugs routinely prescribed in primary care suggests that acupuncture is a relatively safe treatment. 186
In this study, during the 55 years from 1956 to 2010, 1038 cases of adverse events associated with acupuncture were reported in the Chinese language literature, 35 of which (3.4%) were fatal (Table 2). Although no reliable data have been found on the total number of treatments during this period, it was estimated reliably that there were 2688 Traditional Chinese Medicine hospitals with total visits of 36,026,533 times in 2010. 187 If it is assumed that acupuncture treatment accounted for 1% of the total visits (as is known, it is quite conservative data), there would be 3,600,000 times of treatments by acupuncture in 2010. Conservatively estimating, there would be total visits of more than 1,500,000,000 times treatments by acupuncture during this period. Considering the large denominator, the incidence of acupuncture-related adverse events in China is extremely rare.
Time distribution of cases of acupuncture-related adverse events
The acupuncture-related adverse events increased remarkably during 1980s and then slightly decreased. It is highly likely that the sudden rise was due to increased ease of publication and improved reporting standards, which led to adverse events of the period 1955–1980 being significantly under-reported. With the publication of acupuncture textbooks and monographs on the prevention of acupuncture accidents in the late 1990s, people's safety consciousness was enhanced, which can lead to a reduction in adverse events. In addition, since various cases of adverse events had been reported before, repeated and similar reports seemed to be unnecessary and unwelcome from the perspective of journals.
Occurrence cause of reported acupuncture-related adverse events
Syncope during acupuncture is the most frequent adverse event in this study. Syncope occurred primarily in patients receiving acupuncture for the first time or in psychentonia, tiredness, hunger, or improper position during treatment. If the practitioners had given an adequate explanation about the advantage and disadvantage of acupuncture therapy and informed patients of the risk of syncope before treatment, there might not have been as many cases of syncope. In addition, acupuncture manipulation in China has a much stronger stimulus effect than in other countries, which can be a cause of the syncope.
Most organ or tissue injuries are caused by improper manipulation. Quite a number of the practitioners lack anatomical knowledge, which is crucial for the safety of needling manipulation. The safe depth of needle insertion must be given more attention.
Infections are caused by unsterile techniques. Most hospitals and clinics fail to use the disposable acupuncture needles and strict sterilization methods; however, fewer infections were found in this study than in the English language literature. 11 It is possible that acupuncture-related infections are under-reported in China, but there are much fewer possibilities of underreporting outside of China.
Acupuncture is a minimally invasive treatment; thus, some cases of adverse events such as hemorrhage, hematoma, or pain cannot be avoided completely. Allergies happen mainly due to individual variation. Practitioners should tell patients the risk of adverse events from acupuncture in advance, and patients should inform the acupuncturists of any past history. When adverse events occur, practitioners must remain calm and take corresponding measures to minimize the consequences.
Cause–effect relationship between adverse events and acupuncture
The causality between the event and acupuncture was assessed by 2 reviewers. Of the 1038 cases presented, there were 174 certain, 434 probable, 78 possible, 3 unlikely, 346 unclassified, and 3 unclassifiable. Thus, except for the 174 cases, the remaining cases could also be explained by concurrent diseases or other approaches. It can only be said that they were associated with acupuncture but not caused by acupuncture.
Limitations of this research
In order to avoid omission, the authors have formulated the search strategy as comprehensively as possible and preretrieved. However, the possibility still cannot be excluded that some reports were missed.
Most of the cases reported in China are not in standard form; some critical information such as qualifications of the acupuncturist, depth of the needling, and angle of the needling is unclear. The existence of a causal link between acupuncture and these adverse events is uncertain. Therefore, the authors recommend using the STRICTA Guidelines for case reporting.
At present in China, the relevant prospective investigations with multiple centers and large sample remain undone, and the accurate data about the frequency of acupuncture-related adverse events and severe cases are scarce. Whether the literature research is in accordance with the actual situation of clinic practice deserves further research.
Conclusions
Acupuncture-related adverse events are mainly caused by mental tension of the patient, improper operation of the doctor, and incompleteness of sterilization. Most of the adverse events can be avoided by standardizing teaching and clinical practices. Making the corresponding standards for safety such as standard of acupuncturist's qualifications, standardized operation procedure, sterilization standard, and quality standard of the product can greatly lower the risk of adverse events and protect patient safety to the greatest extent.
Footnotes
Acknowledgments
We would like to thank Li Zhongzheng for assistance with retrieval; and Yang Yi and Wang Zhankui for their help with the translation. This research was supported by the State Administration of Traditional Chinese Medicine, Beijing, China.
Disclosure Statement
No competing financial interests exist.