Electroacupuncture Modestly Cuts Delirium Risk in Older Surgery Patients

Older patients undergoing major surgery who received electroacupuncture before and during their operations developed postoperative delirium less often than patients who received a convincing sham version of the same treatment, according to a large randomized trial conducted at eight hospitals in China. The study, published in Anesthesiology and Perioperative Science, enrolled 878 patients aged 65 to 90 years between September 2018 and October 2022 and randomly assigned them in equal numbers to real electroacupuncture or a sham control designed to look and feel identical to patients and operating-room staff. Postoperative delirium, an acute and fluctuating disturbance of attention and cognition that appears hours to days after surgery, occurred in 53 of 439 patients (12.1 percent) in the electroacupuncture group compared with 77 of 439 patients (17.5 percent) in the sham group.

In the primary unadjusted analysis, that difference translated into an absolute risk reduction of 5.5 percentage points, with a 95 percent confidence interval spanning 0.8 to 10.2 percentage points. The relative risk was 0.69 (95 percent confidence interval, 0.50 to 0.95), the odds ratio was 0.65 (95 percent confidence interval, 0.44 to 0.94; P = 0.023), and the estimated number needed to treat was 19, meaning roughly nineteen patients would need the intervention for one to be spared an episode of delirium. Supportive logistic regression models that adjusted for baseline covariates, or that treated study center as a random effect in a mixed-effects model, produced nearly identical estimates, with odds ratios of 0.66 in both cases.

The trial’s design addressed many of the weaknesses that typically plague acupuncture research. Participants were randomly assigned in a 1:1 ratio using a sequence generated by independent staff and stratified by study center, with allocation concealed in sequentially numbered opaque sealed envelopes. A single unblinded acupuncturist performed all intervention-specific procedures in a separate pre-anesthesia room, while anesthesiologists, surgeons, and outcome assessors remained blinded throughout. Patients in both groups entered the operating room with the same number of needles and leads connected to externally identical devices with illuminated indicators; in the sham group, the device’s internal output circuit had been disabled. Patients, statisticians, data collectors, and outcome assessors were all blinded to group assignment until the trial was complete.

The active intervention was precisely specified. Needles were placed at Baihui (GV20) on the scalp, Yintang (EX-HN3) between the eyebrows, and bilateral Shenmen (HT7) and Neiguan (PC6) on the wrists. After manual manipulation to elicit the traditional sensation of de qi, the needles were connected to an electroacupuncture stimulator delivering a disperse-dense waveform alternating between 2 and 100 Hz, with a pulse width of 0.25 milliseconds and current intensity between 1 and 5 milliamperes, individually titrated to the strongest comfortably tolerated level while patients were conscious. Stimulation began 30 minutes before induction of anesthesia and continued until the end of surgery. Sham patients received blunt-tipped placebo needles at non-acupoint sites roughly 20 millimeters lateral to the corresponding acupoints, with no skin penetration and no electrical output.

Delirium was assessed with the validated simplified Chinese version of the Confusion Assessment Method for the Intensive Care Unit, applied twice daily by blinded, trained physicians between postoperative days 1 and 5, preceded in each assessment by a Richmond Agitation and Sedation Scale evaluation. Of 8,780 scheduled assessments, 98.4 percent were completed. A patient was classified as having delirium if at least one CAM-ICU assessment was positive during the five-day window. Eligible patients were scheduled for gastrointestinal tumor, bile duct, thoracic, or orthopedic surgery under general anesthesia lasting at least two hours, with preoperative cognitive screening excluding anyone scoring below 24 on the mini-mental state examination.

However, the robustness of the primary result depends heavily on how missing outcomes were handled. Thirty-seven participants with incomplete assessment windows and no documented delirium, 20 in the electroacupuncture group and 17 in the sham group, were initially classified as delirium-negative, a rule the authors acknowledge was not explicitly prespecified in the original protocol and therefore constitutes a post hoc missing-outcome assumption. When all indeterminate outcomes were instead classified as delirium, the association was attenuated and no longer statistically significant (odds ratio 0.73, 95 percent confidence interval 0.52 to 1.03; P = 0.072), and under an extreme worst-case assumption for the electroacupuncture group the effect disappeared entirely. Analyses restricted to participants with ascertainable outcomes or complete assessment windows, by contrast, yielded estimates consistent with the primary analysis.

The secondary outcomes told a more restrained story. After adjustment for multiple comparisons using the Benjamini-Hochberg false discovery rate procedure, applied post hoc because the protocol had not specified the adjustment families, there were no significant between-group differences in intensive care unit or hospital length of stay, postoperative Acute Physiology and Chronic Health Evaluation II scores, postoperative nausea and vomiting, 30-day cognitive function measured with the modified Telephone Interview for Cognitive Status, or 30-day quality of life measured with the World Health Organization Quality of Life-BREF questionnaire. Repeated-measures mixed-effects models did show slightly lower pain scores in the electroacupuncture group at 2 hours after surgery and on postoperative days 1 and 2, and lower subjective sleep scores on the day of surgery and the first postoperative morning, though these differences were small and not sustained at later time points. Adverse events were infrequent and similar between groups, and 30-day mortality was low, with three deaths in the electroacupuncture group and two in the sham group.

Several perioperative observations may hint at mechanisms, though none establishes one. In exploratory comparisons, the electroacupuncture group received less intraoperative remifentanil and showed slightly higher bispectral index values, suggesting less opioid exposure and less deep anesthesia, both of which have been associated with delirium risk in prior research. An exploratory analysis stratified by bispectral index found a nominal interaction: among patients with intraoperative bispectral index values below 50, the odds of delirium were markedly lower with electroacupuncture (odds ratio 0.46, 95 percent confidence interval 0.28 to 0.76), whereas no difference appeared among patients with values of 50 or above. Because these variables were measured after randomization and may themselves have been influenced by the intervention, the authors caution that such analyses are hypothesis-generating and vulnerable to selection or collider bias rather than evidence of a clinically actionable subgroup effect.

The observed delirium rate in the sham group, 17.5 percent, fell well below the 26 percent incidence assumed in the sample size calculation, which was based on prior estimates for older adults after major surgery. The authors suggest this may reflect the characteristics of the enrolled population, contemporary perioperative care, differences in surgical case mix, or incomplete detection of brief, hypoactive delirium episodes that twice-daily assessments could miss. Because assessment schedules were identical in both groups and assessors were blinded, any under-detection would be expected to affect the groups similarly, but it may still have reduced the observed event rate and the statistical precision of the treatment estimate. The trial had been powered on the assumption that electroacupuncture would reduce delirium incidence by one third, from 26 percent to 17.3 percent, requiring 878 participants for 80 percent power at a two-sided alpha of 0.05.

The authors conclude that perioperative electroacupuncture was associated with a modest absolute reduction in postoperative delirium among older patients undergoing major surgery, without an increase in adverse events, but they emphasize that the finding was sensitive to assumptions about indeterminate delirium assessments and that the improvements in pain and sleep were small secondary observations. The trial cannot determine whether the effect operates through modulation of inflammatory and stress responses, through reduced pain, better early sleep, or lower opioid exposure, or through some combination of these pathways, and no biomarker or neurophysiological data were collected to distinguish them. Animal studies have shown electroacupuncture activating a vagal-adrenal anti-inflammatory pathway and attenuating delirium-like behavior after surgery in mice, but those findings used different acupoints and models and provide biological plausibility rather than direct evidence for this protocol. Larger studies with longer follow-up, the authors write, will be needed to confirm the effect and to determine whether the reduction in delirium translates into better functional recovery, lower health-care use, or lasting cognitive benefit.

Subject of Research: A randomized sham-controlled trial of perioperative electroacupuncture for preventing postoperative delirium in older patients undergoing major surgery

Article Title: Effect of electroacupuncture on postoperative delirium in older patients undergoing major surgery: a prospective, multicenter, randomized sham-controlled trial

Article References: Shi, J., Huang, Z., Li, H., Zhang, S., Lin, Q., Feng, P., Qi, Q., Gao, J., He, Y., Gao, Y., Zhao, Z., He, S., Zhao, Z., Han, X., Li, Y., Liu, S., Fan, J., & Yu, J. (2026). Effect of electroacupuncture on postoperative delirium in older patients undergoing major surgery: a prospective, multicenter, randomized sham-controlled trial. Anesthesiology and Perioperative Science, 4(1), Article 48. https://doi.org/10.1007/s44254-026-00191-0

Image Credits: AI Generated

DOI: 10.1007/s44254-026-00191-0

Keywords: electroacupuncture, postoperative delirium, older patients, major surgery, randomized controlled trial, sham control, CAM-ICU, anesthesia, pain, sleep quality, perioperative care, acupuncture