Nausea and vomiting after an operation, which anaesthetists shorten to PONV, is the subject of a Cochrane review of a single acupuncture point that has been updated three times since 2004. The point is PC06 (Neiguan), on the inner forearm above the wrist. The 2025 update by Anna Lee and colleagues at the Chinese University of Hong Kong pooled 77 randomised trials with 9,847 participants, run between 1986 and 2022. Across the versions the question has moved. The 2015 version concluded that no further trials against sham were needed, and the 2025 version set out to compare PC6 stimulation with anti-sickness drugs and to measure what each adds to the other.
The point and how it is stimulated
The 2025 review places PC6 between the tendons of the palmaris longus and flexor carpi radialis muscles, 4 cm above the wrist crease. It sorted the trials into invasive techniques, such as needle acupuncture, and noninvasive ones, such as acupressure wristbands. The 2015 review lists the range in full: needle acupuncture, electroacupuncture, transcutaneous electrical stimulation through pads on the skin, laser, capsicum plaster, battery-powered stimulation devices and acupressure. Acupressure and Electroacupuncture describe the methods.
The 2025 review: each method against sham
The 2025 review is a network meta-analysis, which combines trials that compared different pairs of treatments so that every option can be set against a common reference, here sham. Sham meant a sham PC6 procedure, a saline injection in place of a drug, or both. A risk ratio of 0.67 means that 67 people had the outcome for every 100 who would have had it with sham. Certainty is the review’s GRADE rating of how far the estimate can be trusted. “Indirect only” marks an estimate that no trial measured head to head; it comes through the network.
| Outcome | Treatment | Risk ratio (95% CI) | Certainty |
|---|---|---|---|
| Nausea (63 trials, 7,534 people) | Invasive PC6 | 0.49 (0.38 to 0.64) | Low |
| Noninvasive PC6 | 0.67 (0.58 to 0.76) | Low | |
| Anti-sickness drugs | 0.73 (0.59 to 0.91) | Very low | |
| Invasive PC6 with drugs | 0.21 (0.10 to 0.45) | Moderate, indirect only | |
| Noninvasive PC6 with drugs | 0.65 (0.45 to 0.96) | Low, indirect | |
| Vomiting (75 trials, 8,627 people) | Invasive PC6 | 0.47 (0.34 to 0.64) | Low |
| Noninvasive PC6 | 0.58 (0.49 to 0.70) | Low | |
| Anti-sickness drugs | 0.62 (0.47 to 0.81) | Very low | |
| Invasive PC6 with drugs | 0.37 (0.18 to 0.76) | Low, indirect only | |
| Noninvasive PC6 with drugs | 0.52 (0.33 to 0.83) | Very low | |
| Rescue drugs needed (55 trials, 6,614 people) | Invasive PC6 | 0.61 (0.44 to 0.84) | Low |
| Noninvasive PC6 | 0.60 (0.51 to 0.70) | Low | |
| Anti-sickness drugs | 0.56 (0.40 to 0.80) | Very low | |
| Noninvasive PC6 with drugs | 0.44 (0.28 to 0.70) | Moderate, indirect |
The authors rated many trials at high risk of bias, mainly for selective reporting, in which a trial publishes some of its measured outcomes and not others. Funnel plots, which look for missing small trials, showed no sign of publication bias. None of the 77 trials reported a serious or lasting complication. Thirty-nine trials with 5,334 participants reported minor effects that passed, such as skin irritation, redness and swelling under a band, and the review rates that evidence as very uncertain.
The 2015 review: PC6 against the drugs
The 2015 version pooled 59 trials with 7,667 participants. Against sham, PC6 stimulation reduced nausea (risk ratio 0.68; 40 trials, 4,742 people), vomiting (0.60; 45 trials, 5,147 people) and the need for rescue drugs (0.64; 39 trials, 4,622 people), with the evidence rated low because the trials differed widely and had flaws. Trial sequential analysis, a method that asks whether the trials so far are large enough to settle a question, found that they were, which is why the authors wrote that no further sham-controlled trials were needed.
The same review set PC6 stimulation against six anti-sickness drugs: metoclopramide, cyclizine, prochlorperazine, droperidol, ondansetron and dexamethasone. It found no difference in nausea (risk ratio 0.91; 14 trials, 1,332 people) or vomiting (0.93; 19 trials, 1,708 people), at moderate quality, and judged further trials of the same comparison futile because they were unlikely to show a difference. PC6 stimulation added to a drug reduced vomiting compared with the drug alone (0.56; 9 trials, 687 people) but not nausea (0.79; 8 trials, 642 people), at very low quality.
Two trials in breast surgery
Two trials in women having breast surgery show how individual results vary. In a 2004 trial at Duke University, Tong Joo Gan and colleagues compared electrical stimulation at PC6, 4 mg of ondansetron and electrodes left switched off. At 24 hours, 73% of the stimulation group, 52% of the ondansetron group and 38% of the sham group had had no nausea, vomiting or rescue drug. The 2025 Cochrane review rates the trial as one of four at low risk of bias in every key domain. In a 2011 Danish trial, Majholm and Møller, at Copenhagen University Hospital, randomised 134 women to a PC6 acupressure wristband or a sham band; 112 finished. Nausea occurred in 35% and 43% and vomiting in 26% and 27%, differences the trial could not distinguish from chance, and about a third of the women reported redness, swelling or tenderness from the band.
Children
Eight of the 77 trials in the 2025 review recruited only children and three more recruited children and adults. The review found no difference in effect between children and adults in its subgroup analyses. A 2016 meta-analysis by Shin and colleagues in Korea looked at children after tonsillectomy and found 8 studies: 4 randomised trials, 3 prospective cohort studies and 1 pilot study. Fewer children had nausea or vomiting with PC6 acupuncture than in the comparison groups (risk ratio 0.77, 95% CI 0.63 to 0.94).
What guidelines say
The fourth consensus guidelines on PONV, published in 2020 by an international panel led by Tong Joo Gan under the American Society for Enhanced Recovery and the Society for Ambulatory Anesthesia, reviewed the literature to September 2019 and were endorsed by 23 professional societies. They cover risk scoring, drug prophylaxis, rescue treatment and enhanced recovery pathways for adults and children. The 2025 Cochrane review describes them as the latest consensus guidelines, as recommending drug prophylaxis or rescue treatment alone or combined with non-drug approaches, and as supporting PC6 stimulation, and notes that the technique has nonetheless not been widely taken up. The guideline’s full text is behind a paywall and this page reports its acupuncture position only as the Cochrane review describes it. Clinical guidelines sets out other bodies’ positions condition by condition.
What the reviews could not settle
Most estimates in the 2025 review carry low or very low certainty, and the largest effects, for needles combined with drugs, rest on indirect evidence alone. The review counted whether nausea or vomiting happened, not how bad it was. Its authors ask for larger multi-arm trials that combine stimulation techniques with drugs, particularly in children and in low-income countries, and say results from those settings could change the review’s conclusions. Both reviews covered surgery only; Nausea from chemotherapy reports the evidence after cancer treatment. Pain after surgery reports the trials that measured pain and opioid use instead, and Postoperative ileus those that timed the return of bowel function.
AcuiQ’s post-operative nausea and post-operative vomiting pages list the protocols individual studies prescribed, each with its citation.