A study was recently published investigating the possible effects of “electroacupuncture” on nausea and vomiting induced by morphine in dogs. This study illustrates some of the challenges of evaluating acupuncture in general, and it shows how the ambiguity in study results can allow for positive or negative conclusions depending on one’s point of view.
Koh RB, Isaza N, Huisheng X, Cooke K, Robertson S. Effects of maropitant, acepromazine, and electroacupuncture on vomiting associated with administration of morphine in dogs. J Amer Vet Med Assoc 2014;244(7):820-29.
222 dogs who were going to be neutered were included in the study, 37 in each of six groups:
saline injection (placebo control)
maropitant injection (anti-emetic medication)
acepromazine injection (anti-emetic/sedative medication)
electrical stimulation/needling at 1 acupuncture point
electrical stimulation/needling at 5 acupuncture points
electrical stimulation/needling at location not designated as acupuncture point
Twenty minutes after one of the treatments above, the dogs were given an injection of morphine, a narcotic pain reliever commonly used for surgical patients. Morphine frequently induces nausea and vomiting shortly after injection, so the authors evaluated the dogs for the occurrence of vomiting or retching, the number of times dogs vomited or retched, the time it took for vomiting or retching to stop, and signs of nausea evaluated on a subjective numerical nausea assessment scale.
The dogs were randomly assigned to the treatment groups, and there appeared to be no significant differences between the dogs in the various groups. Assessment of vomiting and retching were made by an investigator aware of the treatment used in each dog. Assessment of nausea was made using a video recording assessment by a blinded investigator.
In terms of the objective sign of response to treatment, the occurrence of vomiting and retching, these events occurred significantly less in the medication groups than in the placebo and acupuncture groups, which did not differ from each other. Out of the 37 dogs in each group, the number (percentage) experiencing vomiting or retching were as follows:
Saline- 28 (75.7%)
maropitant- 14 (37.8%)
acepromazine- 17 (45.9%)
1 acupuncture point- 24 (64.8%)
5 acupuncture points- 26 (70.3%)
location not designated as acupuncture point- 32 (86.5%)
When the number of vomiting or retching episodes in total were counted, the pattern was similar, but the two acupuncture groups appeared to have fewer episodes than the control groups:
1 acupuncture point- 35
5 acupuncture points- 34
location not designated as acupuncture point- 109
In terms of the number (percentage) of dogs showing signs of nausea, the results were as follows:
Saline- 11 (29.7%)
maropitant- 12 (32.4%)
acepromazine- 3 (8.1%)
1 acupuncture point- 7 (18.9%)
5 acupuncture points- 4 (10.8%)
location not designated as acupuncture point- 15 (40.5%)
Nausea scores were calculated on a 4-point scale 10min, 15min, and 20min after the morphine was given. The average scores ranged from 1.1 to 1.6, so there was little difference between groups. Statistically, the scores for maropitant, saline, and sham acupuncture were higher than before the morphine at 2 of these 3 time points while the score did not change after morphine for the other three groups.
What Does It Mean?
In humans, there is some evidence acupuncture may be effective at reducing nausea and vomiting in a variety of situations, though some reviews for some conditions find no benefit. The tricky part in assessing this is that nausea and vomiting are very subjective and influenced by mental states which are highly subject to placebo effects and bias. It is challenging to blind patients to whether or not they are receiving acupuncture, and it is impossible to blind the people giving the acupuncture therapy, so some residual bias exists in all acupuncture studies. This is problematic because the attitude and demeanor of the acupuncturist has a significant effect on outcome in acupuncture studies, so unblinded therapists, patients, and investigators make the results of studies looking at highly subjective symptoms difficult to rely on.
In this study, the sham control for the acupuncture treatment was application of the same therapy (needling and electrical stimulation) at a location not considered an acupuncture ;point. The biggest problem with this is that there is no such thing as an acupuncture point. The locations for needling in dogs are determined subjectively based on interpretation of historical texts for humans which were themselves based on mystical and supernatural criteria, and the localization of acupuncture points is highly subjective and inconsistent depending on the individual acupuncturist and the school of acupuncture in which they were trained. They have not been consistently correlated with identifiable anatomic structures, and they cannot be consistently and repeatedly detected by acupuncturists tested under controlled conditions.
More challenging still is the question of whether the “electroacupuncture” tested in this study is really acupuncture at all. Sticking needles in the skin and running electrical current through them undoubtedly has physiologic effects. Transcutaneous electrical nerve stimulation (TENS) is a conventional therapy that uses this process to treat pain and nausea. The only thing that qualifies the therapy used in this study as acupuncture is the choice of needling location based on meridians and acupuncture points. However, since these don’t really exist, it is not clear that the sham treatment was really sham or the “real” acupuncture treatment real or distinct from any other method of using needles and electrical current.
As for the other treatments used, maropitant is a widely used and well-studied anti-emetic which has good evidentiary support for its effectiveness. Acepromazine is a sedative which has been reported to have anti-nausea effects, though it is less widely used for this purposes because of its sedating effects. And certainly, morphine is well-known to cause nausea and vomiting in dogs. So in general the study design was reasonable apart from the question of how acupuncture or “electroacupuncture” are defined.
The investigator assessing vomiting and retching was not blind to treatment, so even though these are pretty objective measures, it is possible for there to be some bias in this assessment. The results show pretty clearly that dogs given maropitant or acepromazine were less likely to vomit than those given placebo and that acupuncture had no more effect than the placebo. So for the most obviously important and objective measure, the acupuncture did not seem to work and the medications did.
When the total number of vomiting and retching episodes were counted, it appeared that the acupuncture was almost as effective as the maropitant. However, this measure by definition only included dogs who vomited. So if this is a true representation of the effects of these treatments, it would mean that maropitant and acepromazine prevents vomiting and acupuncture does not, however if the patient does vomit acupuncture is about as good as acepromazine, though not as good as maropitant, in reducing the number of times the dog vomits. This is a possible interpretation, however it is a bit convoluted and raises the question of whether using acupuncture instead of maropitant makes sense since it would seem preventing vomiting is better than simply reducing the number of times a dog vomits after getting morphine.
The blinding was appropriate for the assessment of signs of nausea, though this measure is inherently more subjective, and the nausea scale is not a validated instrument. However, the results for this measure make little sense. The number of dogs showing signs of nausea at all was lower for the acupuncture treatment groups and the acepromazine group than for the sham acupuncture, placebo and maropitant groups. For this to be true result, it would mean that acupuncture was better at reducing nausea than maropitant even though maropitant is better and preventing vomiting. This is possible, of course, but it seems the more plausible interpretation is that the results don’t make sense because this measure is not reliable.
The results of the numerical nausea scores are similarly dubious. Again, the score increased after morphine for the saline, sham acupuncture, and maropitant but not for the acepromazine and acupuncture treatments. This might mean that the acupuncture reduced nausea and the maropitant didn’t, but again this makes little sense given the extensive evidence for the anti-emetic effects of maropitant and the significantly lower number of dogs given this drug who vomited compared with those treated with placebo or with acupuncture. It is also hard to believe that a drug so effective in preventing vomiting was worse than saline placebo in preventing nausea. It seems more likely that the unvalidated nausea scale simply wasn’t a useful measurement tool.
It is also not clear that the differences in scores seen, even if they are real, are meaningful. The largest change for the maropitant seen was an increase from 1.0 to 1.5 on a 4-point scale. In humans evaluated with nausea scales, the smallest change in nausea considered even marginally significant is a change of 15%. The largest change measured in this study was only 15% for the sham acupuncture, and the change for the saline placebo was no more than 10%.
The most positive interpretation possible for this study is that sticking needles into patients and running electrical current through them might reduce morphine-induced nausea marginally but does not prevent vomiting compared with maropitant. The question of whether or not this procedure can be legitimately called acupuncture, rather than TENS, hinges on whether or not the locations selected by these particular acupuncturists yield a better outcome than other locations.
While this study suggests they might, the bulk of the literature in acupuncture suggests needling locations don’t really matter for the effects of needling. Different acupuncture schools use so many different locations, ranging from locations on the entire body to just those on the ear or the hand, that any point at all would be considered an acupuncture point by someone. Acupucnturists are inconsistent in where they locate specific points, and no consistent detectable anatomic or physiologic structure has been associated with all the points claimed to be special. So it seems unlikely that using any particular traditional scheme for selecting needling location would be beneficial over any other.
In any case, the amount of change in nausea was not only clinically insignificant, but it requires us to accept that a drug which clearly prevents vomiting is less effective than a saline placebo in preventing nausea, which makes little sense.
The results also suggest that the nausea scale employed is not a very useful instrument. It would be worthwhile to have a validated measurement tool for this problem in veterinary patients, but developing it beginning with testing of a controversial intervention such as “electroacupuncture” is probably not the best way to proceed. Once such an instrument has been validated with more evidence-based therapies, using it to evaluate acupuncture would be reasonable.
This study provides little evidence for the value of acupuncture in treating morphine-induced nausea. Replication with a valid instrument would be necessary to document that there is, in fact, an effect of acupuncture on nausea and vomiting in dogs, but even if this were done this study suggests such effects are unlikely to be clinically significant, so there is little reason to substitute acupuncture for medications with clear efficacy, or even to add acupuncture to these treatments.
Pingback: Finally, A Journalist Takes a Skeptical View of Claims for Veterinary Acupuncture | The SkeptVet Blog