Proponents of the theatrical placebo known as acupuncture continue to push the limits of my ability to maintain sanity in these uncertain times. In December of last year, researchers at hospitals in Australia, Canada, and Malaysia ran a “multicentre randomized controlled trial” designed to test the efficacy of ear acupuncture in the management of procedural pain in premature infants. The specific procedure in this case was an eye exam performed to evaluate for a condition called retinopathy of prematurity, and it is potentially uncomfortable. It’s more than a little bit like that scene from A Clockwork Orange. You know the one.

Allow me to explain.

What is retinopathy of prematurity?

Retinopathy of prematurity, or ROP for short, is a common cause of visual impairment in former premature infants. Though typically affecting babies born at less than 32 weeks of competed gestation, it is most likely to cause severe disease, including blindness, in those born much earlier. The eyes of prematurely born infants have important differences when compared to babies born closer to term, particularly when it comes to the development of a blood supply. Vascularization of the eye occurs in a certain sequence, with retinal blood vessels normally coming on line at around 15 to 18 weeks of gestation. They begin where the optic nerve enters to back of the eye and grow peripherally, usually completing the process by 40 weeks unless the baby is born very early.

ROP occurs when the healthy sequence of vascularization becomes derailed in two stages after delivery of an at risk preterm infant. The first is when there is injury to the developing retinal blood vessels, as often occurs in the setting of low systemic blood pressures or inadequate levels of oxygen. Another potential cause of injury occurs when there is too much oxygen in the blood, resulting in the development of damaging free radicals.

This state of hyperoxia, an excessive amount of oxygen in the blood, tends to happen during our attempts to resuscitate and/or stabilize premature infants. Older protocols involved the use of 100% oxygen for all newborns and we used to cause a lot of blindness. Over the years, as our understanding of the pathophysiology of ROP improved, we have refined these protocols and they can be quite stingy at times when it comes to how much we crank up the oxygen. These days oxygen is treated much more like a medication with potential serious adverse effects than in the past.

The second stage of ROP occurs during the healing process after the initial injury when new blood vessels can begin to grow abnormally. Instead of peripheral growth, they can actually start to grow into the vitreous, the gel that fills the eye between the lens and the retina. These new vessels are prone to leaking, which can cause swelling and bleeding. Abnormal fibrous tissue can also form which pulls on the retina, potentially causing distortion or even detachment.

Thankfully, in many cases this abnormal growth of blood vessels regresses without intervention. Although we have learned a lot about the underlying causes of ROP, we still don’t know what determines whether or not these abnormal retinal blood vessels will resume healthy growth or progress and disrupt vision permanently without intervention. There are treatments, including laser photocoagulation of abnormally developing blood vessels or injections of a medication that arrests blood vessel growth into the vitreous, that can improve outcomes.

Despite our improved understanding of the risk factors for the development of ROP, it continues to be a common issue faced by former premature infants. Iatrogenic hyperoxia can’t always be avoided because many of these babies have severely immature lungs and require 100% oxygen for resuscitation/stabilization. So even in resource rich regions such as the United States, ROP is not going away any time soon. In fact, the incidence is actually on the rise. As our ability to keep babies alive even when born at the extremes of prematurity, we are seeing more of them live long enough to develop ROP despite the overall lower risk in individuals because of more judicious use of oxygen.

Babies at risk for ROP require screening by pediatric ophthalmologists trained in the detection of abnormal retinal vascularization. The initial screen takes place a few weeks after birth and the exam is repeated every 1-3 weeks depending on the severity of disease in order to track progression. Prior to screening, a baby’s eyes are dilated and often a topical anesthetic is administered. But because of the need to use a speculum to force the eyes open, it can be uncomfortable and maybe even outright painful.

Which brings us back to the study of ear acupuncture as a means of reducing procedural pain during ROP evaluation.

Magnetic Auricular Acupuncture (MAA) for infant pain: Introductions are in order

Before I get into the specific details of the study, I feel compelled to point out that while the authors frequently use the word acupuncture, I don’t think it means what they think it does. At least it’s decidedly unfair to make claims regarding the pre-test plausibility of MAA based on the history of traditional acupuncture involving the insertion of needles:

Acupuncture, a field of traditional Chinese Medicine, has been used for thousands of years to provide analgesia for a gamut of illnesses in adults and older children…

Readers of SBM should be quick to point out that this is both historically and medically inaccurate. Acupuncture as commonly understood today has been around for several decades and is not supported by high quality evidence as a treatment for any condition. They go on to point out, however, that shoving needles into premature infants is impractical, particularly when the patients are in a neonatal intensive care unit. We are then reassured that “acupuncture effects” can be achieved through the use of other forms of stimulation, such as heat, pressure, and magnets.

Here is where it really goes off the rails, and we aren’t even done with the study introduction. To support the acupuncture needle and magnetic bead switcheroo they are about to propose, they cite their own previous study:

…we demonstrated that magnetic auricular acupuncture (MAA), where small magnets are placed around specific points around the ear, significantly decreased procedural pain perception in 40 infants undergoing routine heel pricks.

Not so fast! I was actually familiar with the earlier study and almost wrote about it back in 2017. My twitter buddy Michael Narvey (@NICU_Musings) did write a brief post about it on his All Things Neonatal blog and I agree with his take home message:

The second and perhaps biggest issue I have here is that although the primary outcome was reached it does seem that there was some fishing going on here. By that I mean there were three PIPP scores examined (before, during and after) and one barely reached statistical significance. My hunch is that indeed this was reached by chance rather than it being a real difference.

The last concern is that while the intervention was done in a blinded and randomized fashion, the evidence supporting the use of this in the first place is not strong. Taking this into account and adding the previous concern in as well and I have strong doubts that this is indeed “for real”. I doubt this will be the last we will hear about it and while my skepticism continues I have to admit if a larger study is produced I will be willing and interested to read it.

I think that Michael pulled his punches a bit there. Basing a conclusion of efficacy on the PIPP (Premature Infant Pain Profile) in the way that the authors did is highly problematic and for more than just cherry picking outcomes. Not to mention the trial was tiny and billed as a pilot study. But the particular problem with using the PIPP also applies to the latest study and I’ll get to it shortly. It’s pretty damning though. Kudos to Michael for his prescience regarding future studies. Of course, in the world of irregular medicine there must always be more studies.

In the more recent study, the authors actually get off on a very good foot in their overview of pain in newborns. As a newborn hospitalist, I worry about pain in my patients a lot. As discussed in their introduction, it’s true that some amount of pain is unavoidable in the care of premature infants in the NICU setting, and that there are well-established acute and chronic adverse outcomes that can occur when it is poorly managed. The situation is further complicated by the fact that pharmaceutical interventions expose these babies to some pretty serious potential risks.

Newborn pain management is inconsistent. It largely depends on the facility, and the often-unique protocols in place that can vary from place to place, but also on which nurse, physician, or therapist happens to be running the show at any given moment. As I’ll explain shortly, this study is a perfect example of this because it sure looks like one or more of the involved facilities did not maximize proven safe and effective non-pharmaceutical approaches to reducing pain in babies. This is ironic considering that it’s the entire point of what they were trying to do with this study.

Those pesky details…

This was a study that was coordinated across hospitals in Australia, Canada, and Malaysia that was randomized, blinded, and placebo-controlled. It involved the randomization of infants requiring an initial ROP screening into either a group that had magnetic acupuncture beads taped to specific acupuncture points on the ear using a small round sticker or a placebo group that only got the sticker. Of note, the acupuncture points used in this study were completely different from the ones used in the earlier trial that the authors cited as justification, again raising concerns of applicability. The 100 babies that were eventually signed up were appropriately matched.

In order to prevent the caregivers, and more importantly the nurses who would be assessing the study subjects for pain responses, from knowing which babies were in the active treatment group a dab of white correction fluid was applied. There is a picture provided in the paper and I encourage you all to check this for yourselves; in my opinion, it looks like the magnetic bead is visible under the concealer and would also be palpable. As a result, I’m not confident that blinding was maintained, and that would render this study worthless.

Prior to each ROP exam, the babies were all given a dose of oral sucrose, a very common and well supported means of treating minor to moderate pain, or at least distracting from it, in the first several weeks of life. A topical anesthetic was also applied to the eyes of each baby prior to insertion of the speculum to hold the eyes open. Additional doses of sucrose, use of a pacifier, and even swaddling during the procedure were considered to be additional analgesia and only provided if it was determined to be clinically indicated.

The primary outcome in the study was the difference in pain perception using the PIPP mentioned earlier. Scores were calculated prior to the procedure, during, and 60 minutes afterwards by a “blinded” nurse trained in using the tool. Secondary outcomes were changes in heart rate, arterial oxygen saturation, and the need for additional pain control measures during and 60 minutes after completion of the exam. Heart rate and oxygen levels would be expected to rise and fall respectively when a premature infant is distressed by pain, but they are already a part of the PIPP so I’m not sure why they double dipped like that. It doesn’t add anything.

The big reveal…

Compared to babies who were in the active treatment group, those in the placebo group were found to have a statistically significant increase in PIPP scoring during the ROP exam and at the one hour post procedure mark. The differences, expressed in average PIPP scores were 1.6 and 1.5 respectively. To provide another layer here, the average PIPP score during the ROP exams were 13.5 in the placebo group and 11.9 in the active treatment group. At 60 minutes, the two groups had decreased to average scores of 3.1 and 1.6.

Mean heart rate was significantly lower in the treatment group, 172 compared to 184 beats per minute, during the procedure. Oxygen levels were also higher in the treatment group, 93.8% compared to 91.7%, during the exams. There were no differences in the use of additional measures to treat pain and there were no safety concerns.

The authors of the paper concluded that “magnetic auricular acupuncture significantly reduced scores associated with pain, discomfort and distress in preterm infants during and up to 1 hour after the ROP examination”. They also claimed that the differences in heart rate and oxygen saturation were signs of an improved sympathetic pain response. They believe that the magnetic beads are cheap, easy to apply, and do not interfere with routine NICU care. They wondered if acupuncture involving needles, lasers, or heat might be even more effective and called for further research to answer that question. They also questioned whether the beneficial effect seen in the study came from the pressure applied by the tape against the bead or the magnetic stimulation from the 100 Gauss magnet.

Magnets, how do they not work?

So according to the authors of the study, we’ve now got a safe and effectiveness new tool in our analgesia toolbox when it comes to managing pain in premature infants, but if that sounds a bit too good to be true it’s because it most certainly is.

Where to begin?

How about the extremely low prior plausibility? There just isn’t any reasonable basic science or good clinical data to support an expectation that extremely light pressure on the surface of the ear would have cause any meaningful changes in baby physiology. And there is even less reason to believe that a 100 Gauss magnet, which is the equivalent of a standard refrigerator magnet, would have any effect either. Auricular acupuncture of any kind is pure nonsense. These particular authors were inspired by tales of “battlefield acupuncture”, a topic well known to long time readers of SBM.

With such low prior plausibility, the results of this study would need to be extremely robust in order to tilt the needle towards acceptance. The authors sure seemed impressed. But there are some significant issues with their interpretation. First off, as I previously mentioned, I think that the blinding in this study was well-intentioned but likely not particularly effective. The study did not include an exit interview to determine if blinding broke down, which would have been very helpful.

Assessing pain in children is hard. It’s particularly challenging in infants and even more prone to error in premature infants. The PIPP is a validated tool but it is still very subjective, even with the inclusion of seemingly objective measurements such as heart rate and pulse oximetry. Both are not always reliable when obtaining a one-time measurement and clinical usefulness improves with repeated measurements over longer periods of time.

Even if we could be confident that blinding held and the PIPP scores obtained were free of potential bias, the differences found in this study were not clinically meaningful. As we have written many times on the pages of this blog, statistical significance does not necessarily mean the same thing as clinical significance. On the PIPP, a score of 0-6 is generally interpreted as minimal or no pain. Infants at the 1-hour post procedure check in both treatment and placebo groups had scores in this range. It’s not fair to count an average difference of 1.5 as a success.

PIPP scores of 7-12 are typically interpreted as slight to moderate pain and scores from 12 to the maximum of 21 may indicated severe pain. During the ROP exam, babies in both groups had an average PIPP score hovering around 12. Yes, the babies in the placebo group were a little higher than in the treatment group, but the same issue of clinical versus statistical significance applies. Again, a difference of 1.6 is not very meaningful even if we could rule out bias.

Conclusion: Why this matters

I’m obviously not convinced in the slightest that magnets or magnetic ear acupuncture should play any role whatsoever in the management of pain in the NICU. And though it is almost certainly safe, it is a waste of time and resources. And the introduction of this kind of pseudoscience into medicine helps to further erode the already crumbling bulkhead separating science-based medicine from prescientific belief.

But that’s not what bothered me the most about this study. What made me downright angry was something that it might have unintentionally revealed, that focusing on nonsense like ear magnets and acupuncture beads might actually have resulted in these babies experiencing more pain than if they hadn’t been enrolled in this ridiculous study. Let me explain.

One of the secondary outcomes that the authors decided to focus on was the use of additional pain management interventions after a one-time dose of oral sucrose drops given prior to the procedure. And I mentioned that there was no difference found in this outcome between the treatment and placebo groups. In fact, less than a quarter of babies in either group were given sucrose during the procedure when the average PIPP scoring showed that they were at least experiencing mild if not moderate pain. And not even a third of the babies in each group were allowed a pacifier. 75% in each group weren’t even swaddled!

Sucrose, pacifiers, and swaddling are safe and proven non-pharmaceutical interventions used to reduce pain. They should have been used in 100% of these babies during the procedure. So I’m left wondering why they weren’t. Was it perhaps because these babies were in a trial and there was subconscious bias against their use? Maybe these institutions have protocols that limit use of these methods? I don’t know. But I found it ironic that they were focusing on a bogus intervention in the hopes of reducing the need for medications while ignoring the ways we already do that.



  • Clay Jones, M.D. is a pediatrician and a regular contributor to the Science-Based Medicine blog. He primarily cares for healthy newborns and hospitalized children, and devotes his full time to educating pediatric residents and medical students. Dr. Jones first became aware of and interested in the incursion of pseudoscience into his chosen profession while completing his pediatric residency at Vanderbilt Children’s Hospital a decade ago. He has since focused his efforts on teaching the application of critical thinking and scientific skepticism to the practice of pediatric medicine. Dr. Jones has no conflicts of interest to disclose and no ties to the pharmaceutical industry. He can be found on Twitter as @SBMPediatrics and is the co-host of The Prism Podcast with fellow SBM contributor Grant Ritchey. The comments expressed by Dr. Jones are his own and do not represent the views or opinions of Newton-Wellesley Hospital or its administration.

Posted by Clay Jones

Clay Jones, M.D. is a pediatrician and a regular contributor to the Science-Based Medicine blog. He primarily cares for healthy newborns and hospitalized children, and devotes his full time to educating pediatric residents and medical students. Dr. Jones first became aware of and interested in the incursion of pseudoscience into his chosen profession while completing his pediatric residency at Vanderbilt Children’s Hospital a decade ago. He has since focused his efforts on teaching the application of critical thinking and scientific skepticism to the practice of pediatric medicine. Dr. Jones has no conflicts of interest to disclose and no ties to the pharmaceutical industry. He can be found on Twitter as @SBMPediatrics and is the co-host of The Prism Podcast with fellow SBM contributor Grant Ritchey. The comments expressed by Dr. Jones are his own and do not represent the views or opinions of Newton-Wellesley Hospital or its administration.