A needling prescription names points, but the treatment also depends on choices a list of points leaves out: which needle, at what angle and to what depth, for how long, and how it was moved. The classical texts tie all of these to de qi, a sensation said to mark the arrival of qi at the needle. The STRICTA reporting guideline, first published in 2001 and revised in 2010 by Hugh MacPherson and colleagues to make trial reports describe their interventions completely enough to be repeated, asks every acupuncture trial to report each of these choices: the number of insertions, the depth, the response sought, how the needle was stimulated, how long it stayed in, and the needle’s diameter, length and maker. This page takes them in that order and closes on the trials that set out to produce de qi in one group and avoid it in another.

The filiform needle

The Lingshu, part of the Huangdi Neijing, describes nine needles of different shapes. The Wellcome Collection’s catalogue entry for a 17th-century woodblock of them records that the Lingshu gives the filiform needle, hao zhen, a length of 3.6 or 1.6 cun depending on the chapter and a body as fine as a hair, and that it was the needle most used in practice. Another of the nine, the lance or three-edged needle, was used to draw blood and survives in bloodletting.

In Japan, a 2017 history of needling by Yu-Ming Wong records that early needles were made of bone, copper and iron and were thick enough that insertion stayed shallow; that Isai Misono (1557 to 1616) developed fine, flexible needles of gold and silver; that Waichi Sugiyama (1614 to 1694) made the first insertion guide tube from bamboo reed; and that stainless steel needles followed after the Second World War. Wong describes the guide tube, which surrounds the needle as it goes in, as reducing the discomfort of insertion and holding the needle at the intended angle.

ISO 17218:2014 sets the international requirements for sterile single-use filiform needles; the British Standards Institution lists it as current and under review. Sizes are given as diameter by length in millimetres: STRICTA’s examples of good reporting include 0.25 × 40 mm stainless steel needles and, in the older gauge numbering, Seirin 36-gauge needles. The Clean Needle Technique Manual of the US Council of Colleges of Acupuncture and Oriental Medicine (seventh edition, revised 2017) calls for single-use sterile filiform needles, and says that a thin needle, such as one 0.15 mm wide, or one longer than 25 mm may need its shaft supported during insertion, and that the fingers must then touch the shaft only through sterile gauze or cotton. Infection and clean needles covers the rest of that procedure.

Angle and depth

Papers describe insertion as perpendicular to the skin, oblique, or transverse, meaning almost flat along it. STRICTA asks for depth as a measured distance, a number of cun (the proportional body inch) or a named tissue layer such as fascia or muscle, and for the angle and direction where the protocol specified them; one of its examples needled only 1 to 2 mm deep at 10 to 20 degrees.

A 2012 MRI study by Cheng and Dong scanned 10 healthy adults at CV15 (Jiuwei), on the midline just below the breastbone, and measured the safe needling depth at several angles. The average safe depth was about 17 mm for a perpendicular needle, 23 mm for one angled 45 degrees downward and 53 mm for one angled 15 degrees downward, and body mass and body mass index changed it. A 2013 review of needling depth by Jaung-Geng Lin and colleagues gathered 47 studies that measured safe depths by MRI, CT, ultrasound, dissection or clinical observation, and found great inconsistency in the depth reported for the same point across subject groups and tools. Its authors called for a standard definition of safe depth. Needling over the chest reports what happens when a needle goes too deep over the lung.

Retention

Retention is the time a needle stays in after it has been placed and stimulated. A 2023 review of manual stimulation parameters by Bo-Gang Wang and colleagues reports that in clinical studies the needling itself commonly lasts 30 seconds to 5 minutes, retention 20 minutes to an hour, and a course 2 weeks to 3 months. STRICTA asks for retention to be reported separately from total treatment time, and quotes one trial protocol in which needles placed to tonify were withdrawn at once while those placed with even technique stayed in for up to 20 minutes.

Manipulation

Manual manipulation has two basic movements, which Wang’s review describes as the thumb and index finger working the handle: lifting and thrusting, which moves the needle up and down along its own axis, and twirling, which rotates it back and forth, commonly through 180 to 360 degrees.

The classical texts use these movements to tonify (reinforce) or to reduce (drain). A 2022 comparison of manipulation in East Asian and Western medical acupuncture by Da-Eun Yoon, In-Seon Lee and Younbyoung Chae reports that the Huangdi Neijing recommends quick lifting and slow thrusting to tonify and slow lifting and quick thrusting to reduce, and that the same tradition varies the speed and direction of rotation for the same purpose. Tonifying is used for deficiency and reducing for excess, terms Patterns, symptoms and what you can look up explains. The authors read the two techniques as two doses: tonifying is mild stimulation and reducing is intense, and they cite measurements in which the de qi sensation from reducing was the stronger.

De qi

De qi translates as obtaining qi. A 2013 systematic review by Shuo Zhang and colleagues drew on 120 articles about de qi in the classical texts: the Lingshu says needling takes effect only once qi arrives and distinguishes skilled practitioners by their ability to feel it, and ancient texts read the quality of the sensation as a sign of the disease’s nature and course. A 2013 trial in the Canadian Medical Association Journal defines it as a compound sensation of soreness, tingling, fullness, aching, coolness, warmth and heaviness, with a spreading sensation around the point.

Researchers measure it by asking the patient. Jian Kong and colleagues at Massachusetts General Hospital introduced the MGH Acupuncture Sensation Scale (MASS) in 2007, a questionnaire that rates sensations as the patient alone perceives them; they note that the classical texts mention de qi but that detailed descriptions of it date only from the last hundred years. In a 2013 trial of 30 people with knee osteoarthritis by Rosa Spaeth and colleagues, patients rated 11 of the 12 MASS sensations consistently from one session to the next, heaviness the most consistently, and real needling produced stronger ratings than sham needling. STRICTA asks trials to report the response the protocol sought, such as de qi or a muscle twitch, separately from the response patients reported.

Trials of de qi against no de qi

Studies that compared needling designed to produce de qi with needling designed to avoid it
StudyWho and how manyHow the groups differedWhat it found
Zhang and colleagues, systematic review, 20131 cohort study and 7 controlled trials, in high blood pressure, Bell’s palsy and painVaried by studyConflicting results; the evidence was not enough to say whether de qi predicts or improves the outcome
Xu and colleagues, CMAJ, 2013338 people with Bell’s palsy, all also given prednisoneNeedles manipulated until de qi, or inserted with no manipulationBetter facial nerve function at 6 months in the de qi group (adjusted odds ratio 4.16, 95% CI 2.23 to 7.78)
Zhao and colleagues, Acupuncture in Medicine, 201788 people with primary dysmenorrhoea, one session at SP06 (Sanyinjiao)Deep needling with a thick needle and manipulation, or shallow needling with a thin needle and noneNo difference in pain between the groups
Liu and colleagues, network meta-analysis, 202114 trials, 1,868 people with knee osteoarthritisManual acupuncture with de qi, without it, or conventional drugsNo significant effect of de qi on the WOMAC osteoarthritis score (SMD −0.23, 95% CI −0.81 to 0.35)

The sources are Zhang 2013, Xu 2013, Zhao 2017 and Liu 2021. In Zhao’s trial, every patient in the de qi group reported de qi, but so did 25 of the 45 in the group designed to avoid it. When the authors regrouped patients by what they felt rather than by the group they were assigned to, those who felt de qi had a larger fall in pain, 28 mm against 15 mm on a visual analogue pain scale; that comparison is no longer randomised, so it cannot separate the sensation from whatever else differed between the people who felt it. Xu’s Bell’s palsy trial found a stronger effect in the group with the stronger stimulus, and Zhang’s review rated its evidence as high quality.

What the sources cannot tell you

In each comparison above, the group meant to have no de qi also had a different needle, depth or amount of manipulation, so none of these trials isolates the sensation from the stimulus that produces it. A “no de qi” group is not a sham: its needles still go through the skin, and Sham acupuncture sets out why no control of that kind is inert. The depth studies disagree with one another for the same point, and the parameters in Wang’s review describe what trials used, not what works best.

A protocol entry on AcuiQ names how the study delivered it, with needling as one modality among several, alongside the number of sessions and the length of the course; it does not record needle size, depth, retention or whether de qi was sought. Reading a protocol explains each field. For those details, the cited paper is the source.