Point application (穴位贴敷, xuewei tiefu) covers an acupuncture point with a paste of powdered herbs, holds it on with tape, and leaves it in place for hours. No needle is used. The paste is made to irritate the skin, and in its best-known form, sanfu application (三伏贴, sanfu tie), a team at the University of Hong Kong described the aim as making the skin form blisters, flush red and sometimes suppurate. Its research comes almost entirely from China. A 2015 bibliometric analysis by Zhou and colleagues found 937 clinical studies of summer application published between 1977 and 2015, 404 of them randomised trials; 936 came from China. Asthma was the subject of 401, chronic bronchitis of 146, allergic rhinitis of 117 and chronic obstructive pulmonary disease (COPD) of 73. Asthma sets the asthma trials beside the trials of needling and the guidelines. This page reports what the paste contains, when it is applied, what reviews and placebo-controlled trials found, and the skin reactions they counted.
What goes in the paste
A 2015 review by Wen and colleagues gives the basic sanfu formula as four herbs: Bai Jie Zi, the seed of white mustard (Sinapis alba); Xi Xin, the root of wild ginger (Asarum); Gan Sui, the root of Euphorbia kansui; and Yan Hu Suo, the tuber of Corydalis. The powder is mixed with ginger juice or honey and shaped into cubes of 3 to 5 g, and practitioners substitute other herbs according to the patient’s pattern. In the 11 trials in children that Wei and colleagues reviewed in 2020, white mustard seed, corydalis and kansui were the herbs used most often.
The plasters go on the back and chest. Wen’s review found BL13 (Feishu), the back-shu point of the lung, the point used most for allergic rhinitis. For asthma the most common were BL13, GV14 (Dazhui), XB01B (Dingchuan, written EX-B1 in the WHO standard), BL23 (Shenshu) and BL20 (Pishu). Back-shu and front-mu points covers the back-shu points, which sit on the bladder channel beside the spine.
Sanfu: the hottest days of summer
Sanfu means the three fu periods, the dog days. Wen’s review describes them as the three ten-day hottest periods of the year, reckoned in the traditional calendar from the summer solstice and the beginning of autumn. Patients in the trials it reviewed were treated once in each period, three times a year, or four in years with a fourth period. The review gives the time the paste stays on as 4 to 6 hours for adults in principle, depending on the herbs’ strength and how the skin tolerates them, and usually less than 2 hours for children, to avoid blisters. Many patients come back each summer. In a 2016 survey of 949 users at three hospitals in northern China, aged 2 to 96, 62% were women, and 380 of the 587 regular users had been treated for three or more consecutive years.
What the reviews report
Most trials add the plaster to the patient’s usual medicines and compare the result with the medicines alone, which shows whether the combination did better but cannot separate the herbs from the attention and expectation that come with them; a placebo plaster made to look like the real one is the comparison that can. Reading an acupuncture trial sets out why the control matters. The table gives the comparison for each review.
| Review | Condition | Trials (participants) | Comparison | What it found |
|---|---|---|---|---|
| Wu and colleagues, 2022 | Asthma | 16 (1,730), all double-blind and placebo-controlled | Placebo plaster, with or without usual medicines | Short-term use: a mean of 1.84 fewer acute attacks and better quality-of-life scores than placebo, moderate-certainty evidence (GRADE). Better lung function, low-certainty evidence. |
| Zhou and colleagues, 2017 | Stable asthma | 34 (3,313) | Usual medicines alone; sham plaster in 1 trial | Added to medicines, about 1.4 fewer attacks a year (7 trials). The trials were at high risk of bias overall. |
| Wei and colleagues, 2020 | Asthma in children | 11 (882) | Usual medicines alone | About 1.6 fewer attacks a year, with better peak flow; the authors ask for caution because of the trials’ methodological flaws. |
| Mengxia and colleagues, 2022 | Allergic rhinitis | 28 (3,282) | Placebo, Western medicine, acupuncture, oral Chinese medicine | Lower nasal symptom scores than placebo after treatment, and fewer relapses than Western medicine at 6 and 12 months. |
| Pang and colleagues, 2020 | Stable COPD | 28 (1,615) | Usual medicines alone, or with placebo | Better quality of life when added to medicines; no clear effect on lung function. |
| Wen and colleagues, 2015 | Asthma and allergic rhinitis | 18 (1,785), all single-centre trials in China | Various | Every trial reported the plaster ahead of its control; the trials differed too much to pool, and most were rated low quality. |
The placebo-controlled trials
Three trials with a placebo plaster show how the result moves with the outcome measured. In a 2010 trial in Hong Kong, published in 2016, Zhu and colleagues randomised 323 people over 13 with asthma in remission to herbal or placebo plasters on three sanfu days, on top of their usual treatment. Lung function, the primary outcome, did not differ between the groups, and neither did the questionnaire scores at the end of treatment; a sub-analysis of the follow-up questionnaires reported less need for asthma medicine in the herbal group. In a 2019 Hong Kong trial of 138 people with allergic rhinitis, Dai and colleagues applied patches for 2 hours a week for 4 weeks over GV14, BL13 and BL23. The total nasal symptom score fell from baseline but did not differ from placebo; nasal blockage and quality-of-life scores did.
The largest is a multicentre trial published in 2026 by Xu and colleagues. Everyone received placebo plasters in the first summer to set a baseline, and 835 patients with asthma were then randomised to real or placebo plasters for the next two summers. Over 24 months the real plaster reduced the average number of asthma attacks by 1.42 compared with placebo, and one lung-function measure, forced expiratory volume as a share of the predicted value, was higher at 24 months. Peak flow and the Asthma Control Test score did not differ at 12 or 24 months.
Skin reactions
A 2021 safety review by Zhao and colleagues pooled 9 trials with 2,119 patients and put the rate of adverse reactions at about 9% in both the treatment and control groups; across all patients the odds of a reaction were higher with the herbal plaster. The reactions were mainly skin ulcers and blisters, and the authors judged them mild. Reporting is thin: of the 18 trials in Wen’s review, 12 did not mention adverse events, 4 reported none, one reported a mild skin allergy and one reported swelling and blisters in two patients. Wen’s review advises removing the paste at once if the pain is more than a patient can stand and warns that scratching the skin, or leaving the paste on too long, can scar. Mengxia’s review notes that severe skin reactions lead patients to stop. In the 2026 multicentre trial, treatment-related adverse events were mild, no serious events were reported, and events were more common with the real plaster than with placebo.
Outside trials, Sixth Tone reported in July 2019 that the Jiangxi Provincial Children’s Hospital suspended its sanfu programme after 92 of nearly 900 children treated were reported to have blisters, itching or burns. The hospital attributed the reactions to old ginger in the paste and an alcohol content of 62%. Burns from moxibustion and cupping covers the related injuries from heat and suction.
How AcuiQ records it
Each protocol on AcuiQ records how the points were treated, from the fixed list of modalities the site uses; a plaster study is recorded as “Herbal plaster on points”, alongside its sessions and course length. Reading a protocol explains each field. The entry lists the points the plaster covered; the formula is in the cited paper. Entries added before the modality field existed carry none, and Nothing in this literature ever fails describes one such entry, a plaster study that reads as needling. AcuiQ’s asthma, allergic rhinitis and COPD pages list the protocols individual studies prescribed, each with its citation.
What the sources cannot settle
Almost all of this research comes from China, which Where the trials come from puts in context, and the formula, the points and the time on the skin differ from trial to trial, so a pooled result describes a family of plasters rather than one recipe. None of the reviews above compares sanfu application with the same plaster given on other days, and the protocol for a 2024 review by Xu and colleagues sets out to assess application outside the sanfu and winter sanjiu days separately. The placebo-controlled trials report fewer attacks or better symptom scores with the real plaster in some measures and not others, and they disagree on lung function. Sham acupuncture covers what a placebo control can and cannot hide from a patient.