A press needle is a very short needle fixed in the skin and left there, under tape, for a day to a week, so a point goes on being stimulated between visits. There are two kinds. The press-tack needle is a tiny pin on a round adhesive plaster, pressed straight into the skin; Seirin’s Pyonex, the one most trials use, is made in five lengths from 0.3 to 1.5 mm, with diameters from 0.11 to 0.20 mm, and is applied, in the maker’s words, to disinfected skin and pressed for proper adhesion. The intradermal needle is a short straight needle inserted into the outer layer of the skin and taped over. An article in Acupuncture Today credits the straight intradermal needle to the Japanese acupuncturist Kobei Akabane and the press tack to China. The press-tack needle can be compared with a placebo plaster that looks the same and has no needle at all. This page reports how long trials leave the needles in, how those placebos have performed, what the trials and reviews found, and the adverse events they counted.
How long they stay in
Seirin’s design page describes its tape as breathable and waterproof for long-term placement, without giving a number of days. Trials set their own times:
| Trial | Setting | Retention |
|---|---|---|
| Usichenko and colleagues, 2007 | Ear needles, knee arthroscopy | Inserted before surgery, removed the next morning |
| Mazda and colleagues, 2018 | Caesarean delivery | Applied the day before surgery, removed 48 hours after it |
| Usichenko and colleagues, 2005 | Ear needles, hip replacement | 3 days after surgery |
| Lee and colleagues, 2009 | Insomnia after stroke | 3 days |
| Kao and colleagues, 2022 | Chest injury | 4 days, following the manufacturer’s suggestion |
| Horike and Ukezono, 2024 | Chronic neck pain, self-applied | 7 days per application, for 3 weeks |
Placebo press needles
In a double-blind trial neither the patient nor the practitioner knows which treatment was given. Miyazaki and colleagues wrote in 2009 that the lack of a suitable needle had made such trials of acupuncture impossible, and tested the press-tack needle as one. Sham acupuncture describes the other devices built for the purpose and why none is inert. In their trial, Miyazaki and colleagues randomised 90 people who had never had acupuncture to a Pyonex needle or a placebo lacking the needle. The two groups did not differ in whether they felt the skin pierced, and among those with low back pain the real needle reduced pain scores more than the placebo. The authors concluded that the participants could not distinguish the two and that the pair made a double-blind trial possible.
Later trials have checked the blinding as they went. Kao’s 2022 trial used a placebo, supplied by Seirin, holding a small metal object under clear plastic; when asked to guess, patients were wrong 56% of the time and acupuncturists 41%. Horike and Ukezono used plasters whose needle tips had been removed during manufacture.
What the placebo-controlled trials found
The trials are small, and their results differ by outcome:
- Chest injury: Kao and colleagues randomised 72 patients with blunt chest trauma. A pain score was lower with the real needle immediately and on day 4; no other outcome differed, and the authors called the effect on pain inconclusive.
- Chronic neck pain: in Horike and Ukezono’s trial, 50 people applied the needles themselves. Pain at rest did not differ from placebo; pain on movement fell by 21.6 points against 8.1 with placebo.
- Delirium in intensive care: a 2025 pilot trial by Kao and colleagues of 24 critically ill patients reported a median of 5 delirium-free days with the real needle against 2 with sham.
- Itching after caesarean section: in Mazda’s trial of 30 women, 67% had itching from spinal morphine in each group, so the press needles made no difference.
- Pain after surgery, ear points: in Usichenko’s hip replacement trial, 54 patients who completed it used 37 mg of the opioid piritramide over 36 hours with real ear needles against 54 mg with sham, and pain scores were similar. In the knee arthroscopy trial, 120 patients took a median of 200 mg of ibuprofen against 600 mg in the control group, again with similar pain scores. Ear acupuncture covers ear points.
- Insomnia after stroke: Lee and colleagues placed intradermal needles at
HT07(Shenmen) andPC06(Neiguan) for 3 days in 52 patients; insomnia scores improved more than with sham.
What the reviews report
The reviews pool mostly Chinese trials that compare press needles with drugs, usual care or ordinary acupuncture rather than with a placebo plaster, so their results include the effect of being treated as well as the needle.
| Review | Condition | Trials (participants) | What it found |
|---|---|---|---|
| He and colleagues, 2025 | Neck pain | 7 (425); none sham-controlled | Pain scores 1.33 points lower than controls, low-certainty evidence (GRADE); disability scores lower, very low certainty. |
| Xu and colleagues, 2024 | Pain after anal surgery | 11 (895); none sham-controlled, all published in Chinese | Lower pain scores than drugs or usual care, very low certainty; fewer complications, moderate certainty. |
| Jing and Feng, 2021 | Insomnia | 45 | Better response rates and sleep scores than sham, placebo, drugs or herbs, and as an addition to them; evidence of very low to low certainty. |
| Du and colleagues, 2026 | Allergic rhinitis in children | 24 (1,625) | Added to medicines, a better response than medicines alone; press needles alone were not ahead of medicines alone. |
AcuiQ’s neck pain, insomnia and postoperative pain pages list the protocols individual studies prescribed, each with its citation.
Adverse events
Most trials report few or none. Kao’s chest trauma trial recorded no adverse events over 4 days, and of the 7 trials in He’s neck pain review, 4 reported none and 3 did not report on them. Horike and Ukezono counted them against placebo: after the first week, 36% of the needle group and 20% of the placebo group reported an event, mainly itching, rash and discomfort. Itching was as common under the placebo plaster, which had no needle, as under the real one.
Case reports describe harms from self-application. Kato and colleagues reported a 30-year-old table tennis player who put press tacks over a toe tendon herself during sport; a month later the tendon had ruptured and needed surgical repair, which the authors describe as the first such case reported. A 2021 case report by Lee and Park described a local skin infection after press needle self-acupuncture. Infection and clean needles covers skin preparation and the infections reported after acupuncture.
How AcuiQ records it
Each protocol on AcuiQ records how the points were treated, from a fixed list of modalities, and a press or intradermal needle study is recorded as “Press needles”, with its sessions and course length. A course of three 7-day applications is three sessions over three weeks. Reading a protocol explains each field. AcuiQ sells press needles, and What AcuiQ does not do sets out that commercial interest; the index records press-needle studies the same way as every other modality.
What the sources cannot settle
The needle-free placebo makes double-blind trials possible, and the trials that use it are small: the largest trial above has 120 patients, and the pilot in intensive care has 24. The reviews pool trials without a placebo, and He’s neck pain review notes that none of its trials had a sham or blank control. How long a needle should stay in has not been compared in trials; each sets its own time, from overnight to a week. The placebo itself is not inert: its plaster is pressed on a point and worn for days, and in Horike’s trial the placebo group reported itching too.