The Feasibility of Blinding Intraoperative Electro-Auricular Acupuncture Under Neuraxial Anesthesia

Introduction

Perioperative acupuncture has shown promise for treating frequent postsurgical complaints, such as nausea, vomiting, and pain.^1,2 However, clinical studies on the effectiveness of acupuncture are frequently limited by the challenge of appropriate blinding and placebo control. To circumvent this challenge, researchers have tried administering sham acupuncture, which involves needling points at unrecognized acupuncture locations. It has since been recognized that sham acupuncture produces more analgesic effects than a true placebo, but less than those of verum acupuncture at appropriate acupoints. ³ An alternative placebo method is a placebo needle that does not penetrate the skin. Interestingly, placebo needles have also produced a more-substantial placebo response than that of placebo pills in pharmacologic studies. ⁴ Therefore, sham acupuncture indicates that these placebo interventions might not be as physiologically inert as a placebo should be.

The effectiveness of acupuncture on sedated or anesthetized patients remains controversial due to changes in patients' neural patterns under anesthesia. Functional magnetic resonance imaging (fMRI) has been used to evaluate the effects of acupuncture on connectivity networks of the brain. While there are no fMRI studies evaluating acupuncture on patients receiving sedation anesthesia, the existing data indicate that propofol sedation does not impair connectivity of the networks involved in responding to acupuncture.^5,6 There have been studies on acupuncture on awake patients versus acupuncture on patients under general anesthesia that found significant differences between the groups. ⁷

Adequate blinding of patients to study arms is a standard requirement for modern medical research as well as for acupuncture studies. However, inadequate blinding is surprisingly common in the literature and across disciplines, with only as high as 14% of studies testing for adequate blinding, and, within the chronic-pain literature, <5% of studies have tested for adequate blinding. ⁸ Acupuncture studies traditionally have difficulty with blinding. A meta-analysis from 2013 showed only 61% of studies with adequate blinding of verum acupuncture versus sham acupuncture groups. ⁹ A smaller study from 2015 demonstrated similar results. ¹⁰ Sham acupuncture can have up to 40%–50% of an analgesic effect; therefore, the results might not show a true difference. ¹¹ Studies investigating acupuncture applied solely during the intraoperative timeperiod are uncommon. However, there are data on positive effects on analgesia with intraoperative acupuncture. ¹² If patients can be blinded to a “no treatment” arm, then it might be possible to study the clinical effects of acupuncture more accurately.

Postoperative pain control continues to be a challenge for physicians, with ∼70% of surgical patients experiencing moderate-to-severe pain in the postoperative timperiod. ¹³ Systemic opioids remain a mainstay of postoperative pain management. ¹⁴ However, because of their side-effects and the ongoing effort to reduce opioid use in the setting of the opioid epidemic, complementary analgesic modalities may play a role in the management of surgical patients. Due to blinding difficulties and analgesic effects in previous sham-controlled studies, the current authors sought to assess the feasibility of patient blinding to intraoperative electro-auricular acupuncture (EAA) during anterior cruciate ligament (ACL) reconstruction without the use of sham acupuncture. This study utilized a novel approach of administering EAA intraoperatively to sedated patients after regional anesthesia for ACL reconstruction.

Materials and Methods

This randomized controlled trial (RCT) was approved by the institutional review board (IRB) of the Hospital for Special Surgery, in New York, NY, in 2018 (HSS IRB#: 2018-1478) and was registered in ClinicalTrials.gov (#: NCT03711734i Principal Investigator, Stephanie I. Cheng, MD, DABMA, October 2018). Written, informed participant assent and parental consent was obtained for each patient under age 18. Written, informed consent was obtained for all participants older than age 18.

Forty patients undergoing ACL reconstruction (bone-tendon-bone and hamstring autografts) under spinal anesthesia, were enrolled in this RCT between November 2018 and December 2019 (Fig. 1). Eligible patients were identified in EPIC electronic medical records (Epic Systems, Verona WI, USA) the night before their scheduled surgeries by a research assistant. Any American Society of Anesthesiology (ASA) status 1 or 2, English-speaking patients, ≥12 years old, undergoing ACL reconstruction under spinal anesthesia, were eligible for participation. Patients were ineligible for participation if they were under age 12, could not communicate in English, or could not understand or follow the study protocol. Additional exclusion criteria included patients planning to receive general anesthesia or a preoperative peripheral nerve block, patients with contraindications to the intraoperative protocol, patients with chronic pain, or patients who used opioids within 6 weeks prior to surgery, as well as patients with nonnative ear/previous scarring/surgical manipulation, gouges in their ears, or who refused to remove earrings/piercings prior to surgery.

OPEN IN VIEWER

FIG. 1.

CONSORT [CONsolidated Standards Of Reporting Trials] flow diagram. POD, postoperative day.

Subjects were randomly assigned to receive Acupuncture, or No Acupuncture based on a 1:1 schedule with blocks of 4 created prior to study initiation using SAS, version 9.4 (SAS Institute; Cary, NC, USA) by a member of the biostatistics core not otherwise involved in the trial. Treatment assignments were transcribed onto notecards and sealed in opaque, consecutively numbered envelopes by a research assistant not otherwise involved in the trial. After each patient consent and prior to anesthesia start, the research assistant involved in the trial handed the randomization envelope to the medical acupuncturist. All patients and research assistants involved with data collection were blinded to the group allocation. However, due to the intraoperative nature of the intervention, blinding of the anesthesiologist and surgeon was not feasible.

One board-certified medical acupuncturist* administered the EAA for this study. Patients in the Acupuncture group received Enhanced Auricular Trauma Protocol (ATP) acupuncture at 8 points bilaterally on clean ears. ¹⁵ Needling was performed at Hypothalamus, Amygdala, Hippocampus, Prefrontal Cortex, Point Zero, Shen Men, Insula, and Vagus bilaterally. Electrostimulation was applied using an ITO–130 ES (pulse width: 100 μsec; pulse shape: asymmetric biphasic square wave; pulse mode: continuous) at Shen Men and Hypothalamus at 30 Hz, at level 4 intensity, for 30 minutes. At Hypothalamus, Vagus, and Shen Men, SEIRIN^® 0.20 × 30 mm needles were used and the other points were needled with DBC^tm Detox-5 Acupuncture Needles (Lhasa OMS) 0.20 × 13 mm needles (Fig. 2). All needles were inserted so that they were secure in the ear. To maintain patient blinding, needle placement occurred after induction of anesthesia was complete and level 2 sedation was achieved. Needles were removed prior to emergence/awareness. Patients in the No Acupuncture group did not receive sham needles.

OPEN IN VIEWER

FIG. 2.

Model ear with Enhanced Auricular Trauma Protocol (Amygdala, Hippocampus, Prefrontal Cortex, Point Zero, Insula, inserted with DBC^tm Detox-5 Acupuncture Needles (Lhasa OMS), and Shen Men and Vagus inserted with SEIRIN^® needles.

All patients received spinal anesthesia (4 mL of 1.5% mepivacaine) and intravenous (IV) midazolam (up to 5 mg) upon arrival in the operating room. Ketamine (up to 50 mg), and propofol titrated to achieve level 2 sedation was administered for intraoperative sedation. 1000 mg of IV acetaminophen and up to 30 mg of IV ketorolac were given at the start of closure. No opioids or peripheral nerve blocks were given intraoperatively. A postoperative rescue nerve block could be administered if needed.

Regardless of group assignment, patients were assessed preoperatively for average and worst numeric rating scale (NRS) pain scores at rest and with movement in the 24 hours prior to surgery, and for prior experience with acupuncture. Postoperatively, patients were assessed in the post-anesthesia care unit (PACU) by a blinded research assistant every 30 minutes for 120 minutes for the average and worst NRS pain scores at rest and with movement, and for nausea and vomiting intensity. Administration of as-needed oral (PO) oxycodone or IV hydromorphone was at the discretion of the blinded nursing staff. Patients were also contacted on postoperative day (POD) 1 and were asked about their medication consumption and pain-management satisfaction in addition to the aforementioned assessments. Additionally, a blinding assessment was conducted in the PACU at least 60 minutes after admission, and then again on POD 1. All patients were asked: “Do you believe you received Acupuncture or No Acupuncture?” with acceptable responses being: “Acupuncture,” “No Acupuncture,” or “I Don't Know.” The success of blinding was analyzed using Bang's Blinding Index. Bang's Blinding Index is a method for calculating and assessing the possible unbinding in randomized clinical trials. It can be used to assess the possibility of unblinding in the treatment and control arms of a study independently. ¹⁶

Results

Twenty patients in the Acupuncture group and twenty patients in the No Acupuncture group received the allocated interventions and were analyzed. The two groups showed no difference in demographics or intraoperative characteristics (Table 1). One patient was lost to follow up on POD 1 and intention to treat principle was utilized for analysis. No missing data imputation was performed given the comparatively small sample size of this study.

This study confirmed the primary hypothesis that patients receiving acupuncture can be blinded adequately to whether they received acupuncture or the standard of care. Most patients stated that they did not know whether or not they had acupuncture, both in the recovery room and on POD 1, with Bang's Blinding index of 0.2 (95%CI: −0.02, 0.42) for the Acupuncture Group, and 0.11 (95% CI: −0.10, 0.31) for the No Acupuncture Group (Table 2).

There were no descriptive differences detected for nausea incidence, vomiting incidence, oral morphine equivalents (OME) 0–24 hours, or patient satisfaction (Table 3). Additionally, there were no descriptive differences detected for NRS pain scores at rest or with movement between groups (Table 4). Interestingly, despite no difference in OME, five patients in the No Acupuncture group received a rescue adductor canal nerve block 1–7.6 hours after surgery, while no patients in the Acupuncture group required a rescue block (Fisher's exact test: p = 0.047). See Table 3.

Given the small sample size, this study was not powered to detect any differences in efficacy between the 2 groups and therefore all secondary outcomes are reported descriptively. 90% of patients who were approached for participation consented to participate with only 4 eligible candidates declining to participate in this trial. The medical acupuncturist was available with the equipment and had sufficient time for needle placement and stimulation immediately after the patient reached level 2 sedation for 100% of the cases. Additionally, staff was available to remove needles in a timely manner for 100% of the cases. There were no adverse events reported. Patient sedation was sufficiently light that needling in the acupuncture group was stimulating enough to cause patient movement.

Discussion

Utilizing a novel approach by administering EAA intraoperatively to sedated patients after regional anesthesia for knee arthroscopy, this study proves the primary hypothesis where blinding is adequate between the acupuncture group and standard of care group. As far as the current authors are aware—at least in the literature published in English—this is the first randomized controlled study where adequate blinding has been achieved between acupuncture and standard of care without sham needling. This is important because blinding in complementary medicine studies can be suboptimal. A 2014 meta-analysis looking at complex medical interventions examined 12 randomized controlled trials (RCT), which were mostly acupuncture studies with patient reported outcomes, and showed that there was an exaggerated effect of 0.56 standard deviation between the 2 arms of the studies. ¹⁹ The authors concluded that the pronounced bias in results is due to the lack of appropriate blinding in complementary medicine RCTs, which allows for the possibility of the placebo effect. In the chronic pain literature, another meta-analysis showed that almost 95% of studies did not even assess patient blinding. The authors also sampled literature outside of pain research and found that it was also consistently low for blinding assessment (0%–14%). ⁸ This may contribute to the inherent bias against complementary medicine studies particularly negative outcome complementary medicine studies. ²⁰ The ability to appropriately blind intraoperatively when a patient is already sedated, as shown in our study, opens up many opportunities for high quality, high impact research in the field of acupuncture regarding intraoperative usage.

Acupuncture studies have traditionally had difficulty with blinding because of inadequate placebos and many efforts have been made to find an adequate placebo mostly in the form of sham acupuncture. The main limitation with sham acupuncture (needles placed at non acupuncture points) is that it may have an effect despite not being placed at traditional acupuncture points and therefore the results may not show a true difference between the two groups. Sham acupuncture can have up to a 40%–50% an analgesic effect. ²¹

There have been several attempts to develop technology to help with blinding using sham acupuncture for the placebo arm for studies. An ideal placebo is one that is clinically inert and able to blind in a research setting. The 3 most popular placebo acupuncture devices (PADs) used are the Takakura, ²² the Park, ²³ and the Streitberger. ²⁴ According to a recent systematic review of these devices, the Park and Streitberger do not successfully blind patients and the Takakura could possibly blind the acupuncturist but was inconsistent. ²⁵ Additionally, these studies do not test to see if blinding is adequate for the addition of electricity. Overall PADs have not been shown to adequately blind patients, and do not account for the increased effect of sham acupuncture, which are the biggest confounders with studies for acupuncture. For these reasons, our study chose to randomize between acupuncture and standard of care under sedation, with no third study arm of sham acupuncture.

Electro-auricular acupuncture was selected because of the proximity of the patient's head to the medical acupuncturist and the distance from the surgical site during surgery as well as the ability to allow for uniform stimulation (as opposed to manual stimulation by the practitioner). This allows for simultaneous insertion of the needles while the surgical team preps and drapes the surgical site causing no delays of Operating Room (OR) time. The Enhanced Auricular Trauma Protocol (ATP) was selected as the acupuncture protocol because, while most studies involving ATP are focused on patients with post-traumatic stress disorders and traumatic brain injuries, ¹⁵ the points used are also appropriate for pain syndromes. The International Association for the Study of Pain's Definition of Pain Task Force proposes that pain is defined as “An aversive sensory and emotional experience typically caused by, or resembling that caused by, actual or potential tissue injury” with accompanying notes stating, “is influenced to varying degrees by biological, psychological, and social factors.” ²⁶ This protocol in the present study was modified by adding the Vagus point for modulating the body's inflammatory responses as well as aiding in relieving anxiety as well as the point Insula, which is involved in perception, self-awareness, cognitive functioning and interpersonal experience. ²⁷ It is known that low-frequency (2–4 Hz) targets enkephalin, β-endorphin, and endomorphin, while higher frequencies (100 Hz and higher) release dynorphins. ²⁸ At the time of this writing, there has not been any recommendation for intraoperative electro auricular acupuncture stimulation frequency in the English literature. We decided to choose to use 30 Hz as an intermediary frequency. Though not specifically tailored to post-operative knee pain, this protocol can be used for non–site-specific pain ¹⁵ and adequate for the study's primary outcome of blinding.

Limitations of any trial include the inability to blind the medical acupuncturist, however, this should not have affected any of the current study's results. The treatment of postoperative pain with opioid medication was at the discretion of the nursing staff. Rescue nerve blocks were performed 1–7.6 hours after surgery in 5 patients who had intractable postoperative pain and were initiated by nursing staff and then evaluated by an anesthesiologist to determine the necessity. This study was also limited to patients who had spinal anesthesia and IV sedation. These patients theoretically had all native neurologic responses and therefore, these results may not be generalized to patients undergoing general anesthesia. As stated in the introduction, this is an on-going debate, and researchers continue to argue about the need to insert needles prior to induction of general anaesthesia.^29,30 However, based on fMRI studies, general anesthesia does cause differences in the somatosensory cortices, hypothalamus, posterior lateral thalamus, and contralateral head of the caudate. ⁷ General anesthetic agents can reduce transmitters released and alter the synaptic transmission. ³¹ It is unclear if acupuncture performed under general anesthesia, where the EEG of the brain is altered, would have any significant effect, as more studies would be needed. For patients under IV, there are no studies demonstrating a difference in fMRI and propofol sedation does not impair the networks associated with the effects of acupuncture. ⁶ It is necessary to note that it is unclear how the use of midazolam and ketamine would affect auricular acupuncture effectiveness. Insertion of the needles for this study reliably created enough stimulation for patient movement, as patients were all at a 2 level of sedation. Under general anesthesia, patients would not move during needle insertions.

Conclusions

Adequate blinding for patients receiving acupuncture can be achieved when patients are under sedation. Further research will be necessary to see what kind of effect acupuncture can have on patients in the perioperative setting with regard to a wide variety of concerns, such as pain control, anxiety, blood-pressure control, and range of motion. The proven ability to blind patients to acupuncture against standard of care under sedation intraoperatively opens the possibility of true randomization and can lead to high-quality research in a field that has traditionally been difficult to validate, as well as showing true promise for future use of intraoperative electro-auricular acupuncture as part of a multimodal strategy for postoperative pain control.