Three methods used alongside acupuncture put heat or suction on the skin: moxibustion, which burns dried mugwort (moxa) on or above a point; cupping, which draws the skin up into a cup; and heat lamps. Most of the sources on this page are case reports and series from hospital burn units, which describe injuries without saying how often they happen; a few clinic surveys counted every treatment and supply a rate. Moxibustion and Cupping describe the methods; this page reports the injuries.
Moxibustion: what the reviews found
A 2010 systematic review by Park and colleagues at the Korea Institute of Oriental Medicine searched 14 databases and found adverse events from moxibustion in 4 randomised trials, 1 controlled trial, 2 uncontrolled observational studies, 13 case reports and 1 prospective study. The events reported most were allergic reactions, burns, and infections. The trials recorded redness, blistering, itching, discomfort from the smoke, fatigue, headache and burns. The authors concluded that the incidence of these events is not known.
A 2014 review of case reports by Xu and colleagues gathered 64 cases from 24 articles in six countries: China, the US, South Korea, Spain, Japan and Israel. Burns made up 43 of the 64, from 6 articles. Most of the burned patients were newborns, children or people with reduced sensation, and the injuries ran from pain and blisters to ulcers and secondary infection. The more severe burns left permanent scars and were associated with direct moxibustion, where the moxa burns on the skin. Among the cases:
- 37 newborns burned by spark moxibustion, a folk method for neonatal jaundice in which a burning rush is tapped on the skin, between 1990 and 1997: 34 mild burns and 3 moderate, and 8 of them developed infections that spread through the body;
- an 18-year-old with a paralysed leg who could not feel the heat and was burned at
GB34(Yanglingquan) during indirect moxibustion, where the moxa is held above the skin; the wound took 40 days to heal; - a basal cell carcinoma, a skin cancer, that arose in a burn scar on the abdomen of a man who had burned moxa on the same spot for ten years.
Xu and colleagues also raise the question of whether a moxibustion burn is always an adverse event. Scarring moxibustion deliberately burns the skin to produce a blister and a scar, and some patients in the reports accepted the lesions as part of the treatment. The World Health Organization’s 1999 safety guidelines say scarring moxibustion should be used only with the patient’s full knowledge and prior consent, and that the skin should not be burned in indirect moxibustion. They rule out direct moxibustion on the face and over tendons or large blood vessels, rule out scarring moxibustion near a joint because moving the joint may make the wound hard to close, and ask for special care with patients who have reduced consciousness, sensory disturbance or poor circulation.
Moxibustion: what burn units saw
Two hospital burn units published series of patients burned by moxa. The Jilin University burn unit in Changchun, China, treated 82 patients between 2009 and 2017. Of these, 75 were burned by indirect moxibustion, held on average 3.4 cm from the skin, and 7 by direct moxibustion. Most burns, 69, were under 2 cm across, but 59 were deep partial-thickness and 7 full-thickness, and the lower limbs were both the commonest site and the likeliest to burn deep. A Korean study of 59 patients treated in 2014 and 2015 found deep second- or third-degree burns at 44 of 68 burn sites. Patients had applied the moxa themselves in 27 cases and an Oriental medicine practitioner, the paper’s term for a Korean medicine practitioner, in 23. Burns away from the abdomen were 9.4 times more likely to be deep than burns on it.
How often clinic surveys counted a burn
| Study | Setting | Burns |
|---|---|---|
| Yamashita et al. 1999 | Acupuncture and moxibustion at a national college clinic in Tsukuba, Japan, over six years | 7 burn injuries in 65,482 treatments |
| White et al. 2001 | 78 UK doctors and physiotherapists | 1 blister after moxibustion in 31,822 consultations |
| Uehara et al. 2021 | Four Japanese acupuncture and moxibustion teaching clinics over six months | 16 burn injuries in 4,369 sessions |
None of these surveys gives the number of treatments that used moxibustion, so each rate is per treatment of any kind rather than per moxibustion session. The large prospective needling studies summarised on After a treatment left moxibustion on its own out of their counts.
Cupping: what a mark is
A 2017 review by Duane Lowe of the US Army’s Madigan Army Medical Center describes what produces the mark a cup leaves: the negative pressure stretches the skin and the tissue under it and widens the capillaries, and the increased blood flow eventually ruptures capillaries and leaves an ecchymosis, blood under the skin of the kind seen in a bruise. The US National Center for Complementary and Integrative Health says the marks are temporary and advises explaining them to health care providers so they are not mistaken for signs of physical abuse. The same page lists persistent skin discoloration, scars, burns and infections as side effects, and says cupping may worsen eczema or psoriasis.
A 2012 review of 135 randomised trials of cupping, published from 1992 to 2010, found no serious adverse effects reported in any of them.
Cupping: burns from fire
In fire cupping, NCCIH says, a flame is applied to the cup to remove oxygen before it goes on the skin; the alternative is a suction device attached to the cup once it is in place. The Victorian Adult Burns Service in Melbourne searched the Burns Registry of Australia and New Zealand for 2009 to 2016 and found 20 patients with cupping burns among 18,703 burn patients, 0.11%. Eighteen were flame burns, and methylated spirits were the accelerant in 17. Sixteen of the burns happened at home and four at a professional service. Seventeen patients needed surgical cleaning of the wound under anaesthetic, eight needed skin grafts, and the mean stay in hospital was 10.9 days. The index case was a man whose spilled methylated spirits caught fire across his arm and flank. The authors say a registry of admissions is likely to miss milder burns.
The Jilin burn unit also reviewed 14 patients burned by cupping. Most injuries were mild to moderate and healed without surgery, and wet cupping, in which the skin is pierced so that blood flows into the cup, was more common among them than dry cupping. The 1999 WHO guidelines recommend glass rather than rubber or plastic cups, because glass withstands the temperatures needed to sterilise it.
Heat lamps
Some acupuncture services place a heat lamp over the treated area during a session. One type is the TDP lamp, named from the Chinese Teding Diancibo Pu; at a US hospital’s IVF acupuncture service, a TDP lamp was used in 83% of 1,896 treatments. Published reports of lamp burns are few. A 2005 letter from the plastic surgery department of Cork University Hospital in Ireland reported burns from an infrared lamp after acupuncture. A 2021 case report from Kepala Batas Hospital in Malaysia describes a 39-year-old man who used a wheelchair and had diabetes. He developed a blistering second-degree burn on his leg a few hours after a 30-minute electroacupuncture session, under the spot where an infrared lamp with a timer had been placed. He had felt no pain during the session. The authors name reduced sensation, as in paraplegia, diabetes and vascular disease, as the factor that raised his risk, and the burn healed with darker skin left behind.
What the sources cannot tell you
No study on this page gives a burn rate per moxibustion session, per cupping session or per lamp session, because none counted how many of each were given. Case reports and burn-unit series record the injuries severe enough to be written up or admitted, and the Australian registry included patients who did their own cupping at home alongside those treated by a practitioner. The Korean series did the same for moxa. Xu and colleagues argue that whether a moxibustion burn counts as an adverse event depends on what the practitioner intended and what the patient accepted, which a count of burns does not capture. AcuiQ’s protocol entries record the modality a study used, such as moxibustion or cupping, but not any injury it caused; What AcuiQ does not do sets out that limit.