Incidence of adverse effects during acupuncture therapy

Ernst G, Strzyz H, Hagmeister H
Humboldt University, Charité,
Campus Virchow Clinic, Anesthesiology;
Berlin / Germany

Abstract:
Introduction: Acupuncture is frequently used to treat chronic pain syndromes or other chronic diseases. There exist several hundred casuistic reports about partially life-threatening adverse effects. Aim of our study was to assess adverse effects of acupunture during a normal treatment routine.
Methods: Acupuncture patients of 13 general practitioners and outpatient clinics and of 16 healers were observed. By questionnaires gender, age, indications, way of treatment and adverse effects were documented by the therapists.
Results: 409 patients with 3535 acupuncture treatment were included. 402 treatments (11,37%, 153 different patients) with adverse effects were observed. Main side effects were small bleedings (2,9%), hematoma (2,2%), dizziness (1 %) and other vegetative symptoms (2,74%). Other mentioned side effects (below 1%) were fainting, nausea, prolonged De-Qi-effect (paresthesia) and increase of pain. One patient with an hour lasting aphasia after acupuncture treatment was reported.
Discussion: Assessing adverse effects by the therapists themselves contains a high risk of underestimation. A bias can be supposed regarding life threatening adverse effects like pneumothoraces. The other reported side effects however can be assumed not to have this bias. More than one third of acupuncture patients experience adverse effects while being treated with acupuncture. Most of them are harmless. Small bleedings and hematoma are common. No local infection was reported despite of several therapists, who used sterile needles without skin desinfection. Acupuncture is a safe method but has like any therapistic approach adverse effects.

Keywords
Acupuncture, adverse effects, side effects, hematoma, de-qi-sensation, bleeding, infection
--------------------------------------------------------------------------------
Introduction
Acupuncture is one of several therapy methods of the Traditional Chinese Medicine. In western medicine it is frequently used to treat chronic pain syndromes or other chronic diseases. It is claimed, that acupuncture has no adverse effects [1]. Complementary therapy methods are recommended as natural [2] and „natural" is often used as synonym for harmless or without adverse effects. In contradiction since the 70ies there exist several hundred casuistic reports about partially life-threatening adverse effects (recent reviews in [3], complete review in [4]). However no attempt to quantify these complications was ever made. Systematic studies exist only in some attempts [5-7], incidences numbers are normally estimated.
The issue is complicated by a broad variability of acupuncture methods differing from country to country. In Japan the usage of permanent needling methods is more frequent than in Europe causing serious complications. Some adverse effects are typical for ear acupuncture (e.g. local infections) but not for body acupuncture. Some special effects are inherent with associated techniques like moxa-therapy (burning cone or sticks close to the skin or in contact with a acupuncture needle) causing burns.
Aim of our study was to asses adverse effects of acupunture during a normal treatment routine (normally 10 to 15 insertion of about 10 needles) in daily practice. In addition potential correlating factors were examined regarding their possible enhancing role.

Material and Methods
Acupuncture patients of 13 general practitioners and outpatient clinics and of 16 healers were observed. Two questionnaires were used. On the first gender and age of the patients, indication for acupuncture treatment, acupuncture points, treatment method and accompanying illnesses were documented. The second was used after every treatment documenting adverse effects. Mentioned were bleedings, hematoma, infections, neurologic abnormalities, faintings, vestibular symptoms, nausea, prolonged deqi-effect and increase of pain. There was free space for other observed adverse effects. Both questionnaires were filled out by the therapists. Informed consent of patients for a „observational study on acupuncture effects" were obtained.

Statistics
Data were processed descriptively including 95% confidence intervalls. In addition correlations between nominal structured data were processed with Chi square test, in case of low numbers with Fisher’s exact test.. A combination of nominal and metrical structured data was processed by Student’s t-test, in case of non-normal distribution with Mann Whitney-U-Test. Because most patients had several treatments, data analysis was carried out both with numbers of patients and numbers oft treatments. Correlations are reported if they are existing on both levels. Level of significance is p < 0.01. Because of the explorative character of the study significant results are only one measure for the strength of the association, not a proof.

Ethics
Study method, questionnaires, patient information sheet and informed consent sheet were reviewed by the local ethic comitee.

Results
The study took place december 1995 – december 1996. At all 409 patients with 3535 acupuncture treatments were included. 402 treatments with 153 different patients adverse effects were observed. Descriptives are shown in table 1.
 

Table 1: Adverse effects (Number of treatments and afflicted patients)

Most frequent other vegetative side effects were fatigue and generalized sweating. Other side effects were paresthesias and increase of symptoms after removing of the needle.
Small bleedings and hematoma were often observed together and there exists a correlation with the number of used needles. Small bleedings were observed less often at female patients. Hematoma were observed less often in young patients, on the other hand they expressed more pain. (All p < 0.01). As unprecedented side effect a patient with an hour lasting aphasia after acupuncture treatment was reported by one therapist.


Discussion
Assessing adverse effects by the therapists themselves contains a high risk of underestimation. A bias can be supposed, the reported numbers might be too low. This regards specially serious life threatening adverse effects like pneumothoraces or other acupuncture-associated trauma where a therapist might be afraid of possible legal consequences. Other reported side effects however can be assumed not to have this bias. There was no monitoring to test the exactness of the questionaires, this might lead to improper documentation. On the other hand due to the multicenter character of the study some of this subjective factors might be eliminated. But besides of this arguments the lack of serious side effects might be not only because of methodologic reasons, but also because of their low incidence rates.
More than one third of acupuncture patient experience adverse effects while being treated with acupuncture. Most of them are harmless. The most frequent side effects were small bleedings and hematoma. Vegetativ symptoms like fatigue, but also awakeness, dizziness, changes in digestion and sleeping manners were not seldom. Because of the subjective nature of those phenomena they should be tested in a further study because of possible associations to acupuncture point combinations or diagnosis. Yet 1.5% of patients reported symptom of prolonged DeQi. Normally after needle insertion a certain feeling of dumb pain and paresthesia lasting about one minute or as long as stimulation occurs is reported by patients. Prolonged DeQi means this effect lasts more than five minutes up to hours. There is no report about permanent neurological symptoms neither it was observed in this study. Patients should be informed about this possible effect and it’s harmlessness.
Most interesting are adverse effects which did not appear. No local infection was reported despite of several therapists, who used sterile needles without skin desinfection before inserting them. This corresponds to widespread experience of acupuncture practising therapists, but was not yet reported in a systematic form. Reported hepatitis-B infections in acupuncture patients [8] occur because of unsufficiently sterilizing procedures of multiple used needles. There is no evidence, however, that one way needles provide advantages if multiple used needles are processed according to usual hygienic rules (in desinfection solution after use, cleaning and controlling possible needle damages, sterilizing with heat or autoclaving before use). Despite the results of our study we recommend skin desinfection before ear acupuncture, because of the bradytrophic character of the ear tissue. Also small infections there can have a chronic course and might cause strong cosmetic changes.
Unfortunately we are not able to answer the very important question of possible acupuncture effects on pregnant patients. Most of the therapists tend to avoid several acupuncture points and to be very careful with this patients [9] although observations are published which could not observe any side effects.
We did not analyze the risk of symptomatic treatment by acupuncture therapists causing (fatal) delays in possible causal treatment approaches. This might be the most dangerous side effect of any complementary treatment method. Acupuncture treatment is only acceptable after adequate diagnostic procedures. Specially healers should be aware of this aspect and legal consequences should be drawn if therapists act not careful in this point. Obligatory additional anamnesis before acupuncture treatment should at least cover bleeding or wound heeling disorders, immune deficiencies, possible heart diseases (risk of bacteremias), possible pregnancy, and existing anatomical abnormalities.

Conclusion
Acupuncture is a safe method but has like any therapistic approach adverse effects. Few serious adverse effects were published in casuistic form but could not be observed in our study. Bleeding, hematomas and vegetative syndromes are common. Using sterile needles is obligatory, skin desinfection however might not be mandatory in healthy, not immunsupprimized patients.


Literature
[1] Bischko J.: Akupunktur ; Schweiz Rundschau Med (Praxis) 1979 ; 68: 987.
[2] Ernst E.: Risiken komplementärer Therapieverfahren ; Fortschritte der Medizin 1998; 1-2: 28-30.
[3] Norheim A.J.: Komplikasjoner ved akupunkturbehandling. Literaturstudie for arene 1981-92; Tidsskr Nor Laegeforen 1994; 114(10): 1192-1194.
[4] Strzyz H., Ernst G.: Nebenwirkungen bei der Akupunktur; Der Schmerz Februar 1997 ; 11: 13-19.
[5] Norheim A.J.: Adverse effects of acupuncture ; Lancet 1995; 345: 1576.
[6] Andersson A., Carlsson C.A., Eriksson M.: Akupunktur-Frän Tro Till Vetenskap; Liber Malmoe 1984: 124-126.
[7] List T., Helkimo M., Andersson S., Carlsson G.E.: Acupuncture and occlusal splint therapy in the treatment of craniomandibular disorders: Part 1 . A comparative study ; Swed Dent J 1992 ;16.
[8] Kent G.P., Brondum J., Keenlyside R.A., LaFazia L.M., Scott D.: A large outbreak of acupuncture-associated hepatitis B; Am J Epidemiol 1988; 127: 591-598.
[9] Bahr F.R.: Akupunktur in der Gynäkologie und Geburtshilfe ; Gynäkologe 1994; 27(6): 369-374.