Introduction
In 2010 I first evaluated the Chinese herbal remedy known as Yunnan Baiyao, and I have reviewed additional evidence repeatedly since (1, 2, 3). So far, I have found little reason to believe claims that this remedy can stop bleeding in veterinary patients.

The ingredients are poorly described and not regulated or standardized, as is always the case with Traditional Chinese Medicine (TCM) remedies, and it is unclear if they have all been disclosed since the recipe is consider a commercial secret. TCM remedies have frequently been found to contain toxins and unreported pharmaceuticals, and some YB tested has been found to contain low levels of toxic substances. The TCM theory for how it works, involving movement of blood and mystical and ill-defined entities such as Qi, is inconsistent with established scientific understanding of physiology and blood clotting mechanisms. There are a number of more scientific theories for how the many chemicals in the product might affect bleeding, but none have been properly validated, and it is recommended primarily on the basis of tradition and anecdote.

Of course, despite the implausibility and lack of a clear mechanism, some clinical research has been done. In humans, the most recent review (from a source known to be biased in favor of TCM treatments, c.f. 3, 4) found some evidence of effect but also found that 1) most of the published research was of low quality and high risk of bias, 2) for some conditions the apparent effect disappears when lower quality studies are excluded, and 3) there is evidence of publication bias, in which negative studies remain unpublished creating an inaccurate impression of the true state of the evidence.

Such weak evidence might justify further research, ideally beginning with full identification of components, basic physiologic and pharmacologic studies, and then progressing through pre-clinical studies before clinical trials, as is the appropriate and expected course of investigating new medications. However, the current evidence does not really justify routine clinical use.

There is also some research into Yunnan Baiyao in veterinary species, and I have reviewed this in my previous posts. A couple of new small studies were recently presented at the American College of Veterinary Internal Medicine (ACVIM) annual forum, so I thought it would be worth reviewing those and summarizing the evidence to date.

Recent Studies
MacRae R. Carr A. The Effect of Yunnan Baiyao on the Kinetics of Hemostasis in Healthy Dogs. ACVIM  Forum, National Harbor, MD, 2017.

The goal of this study was to evaluate the effect, if any, of Yunnan Baiyao on laboratory measures of blood clotting and to look for any obvious, short-term adverse effects. Six laboratory dogs were given the product once a day for six days and clotting measures compared before and during the treatment. No harmful effects were reported, and there was no change in any measure of blood clotting.

Adelman L. Olin S. Egger CM. Stokes JE. Effect of Oral Yunnan Baiyao on Periprocedural Hemorrhage and Coagulation in Dogs Undergoing Nasal Biopsy. ACVIM Forum, National Harbor, MD, 2017.

The abstract begins with the statement, “The hemostatic efficacy and safety of Yunnan baiyao (YB) has been demonstrated across multiple species.” This isn’t actually accurate, and it reveals a pretty clear bias on the part of the investigators, which is relevant when looking at the methods and conclusions of the study.

Nineteen dogs having nasal biopsies were randomized to Yunnan baiyao or placebo before having nasal biopsies taken, which commonly causes some bleeding. A variety of outcome measures were evaluated, including laboratory values associated with clotting and assessments of how much blood was lost and how long it took the dogs to clot after biopsy. The report indicates appropriate blinding of investigators, caretakers, and statisticians. However, not all dogs were assessed with the same measures, which introduces some inconsistency and variability into the results.

Time for bleeding to stop- Exactly how this was measured is not described. The YB group ha d a slightly shorter BMBT (300+/- 12 seconds vs 367+/- 9 seconds). Several other variables (age, history of nosebleeds, blood pressure, and number of biopsies taken) were also associated with the BMBT, though the details were not reported.

BMBT (time for bleeding to stop from a standardized cut on the gums)- There was no difference between the groups.

TEG (a lab measure of clotting)- There was no difference between the groups.

Total blood loss- It is also not clear how this was measured. There was no difference between the groups.

Despite the failure to find an effect for most measures and the questionable clinical significance of the one difference seen (a difference from 46 to 88 seconds in the time for bleeding to stop), the authors naturally conclude that the study supports using YB routinely before this type of procedure.

Veterinary Evidence Summary
The table below lists all of the actual in vivo studies in veterinary species I have found. It also lists some key features of these studies that help evaluate how reliable their results are, including the number of animals studied, the use of key methods for preventing bias, such as randomization, blinding, and a control group, and the outcomes.

Table 1. Studies of Yunnan Baiyao in veterinary species.

Of these ten studies, 5 found no effect at all, and 2 others showed mixed results, with possible effects in some measures evaluated in the studies but not in others. Of the three fully positive studies, two did not report any of the major methods for controlling for possible bias and other sources of error.

Overall, this is a very unconvincing set of data. Even clearly ineffective methods can have some positive studies due to bias and error alone, so the lack of a clear, consistent pattern of expected effects is troubling. Not all of the studies used the same measures, so it is possible the product could have some clinical effects by some mechanism that doesn’t affect the laboratory measures of clotting usually used, but that is a big stretch.

It is also worth noting that the studies showing some effect didn’t look find any benefit in terms of clinically important outcomes, such as survival, need for transfusion, etc. Even under the most optimistic assessment of the evidence, it may be that Yunnan Baiyao speeds clotting in the case of small wounds by a few minutes, but this may not necessarily have any meaningful benefit for actual patients.

Bottom Line
The TCM rationale for using Yunnan Baiyao is part of an unscientific, quasi-religious belief system and cannot be accepted as a sufficient basis for using an otherwise unproven remedy on patients, especially when the ingredients in that remedy are not fully disclosed or regulated for quality, consistency, and safety. The more plausible scientific hypotheses for how Yunnan Baiyao might work remain unproven.

The clinical research evidence is mostly negative, and even positive studies have not shown any significant effects on clinically meaningful objective outcomes.  No clear evidence of harm has yet been found, though the limited nature of the evidence does not ensure that the product is truly safe.

References
Ogle CW, Soter D, Cho CH (1977) The haemostatic effects of orally administered yunnan bai yao in rats and rabbits. Comparative Medicine East and West 5:2, 155-160

Ogle CW, Dai S, Ma JC. The haemostatic effects of the Chinese herbal drug Yunnan bai yao: A pilot study. Am J Chin Med (Gard City N Y) 1976;4:147–152.

Graham L, Farnsworth K, Cary J (2002) The effect of yunnan baiyao on the template bleeding times and activated clotting times in healthy ponies under halothane anesthesia. Journal of Veterinary Emergency and Critical Care 12:4; 279; 2002 (abstract only)

Epp TS, McDonough P, Padilla DJ, et al. The effect of herbal supplementation on the severity of exercise-induced pulmonary haemorrhage. Equine and Comparative Exercise Physiology 2005;2:17-25.

Fan C, Song J, White CM. A comparison of the hemostatic effects of notoginseng and yun nan baiyao to placebo control. J Herb Pharmacother 2005;5:1–5

Murphy LA, Panek CM, Bianco D, Nakamura RK. Use of Yunnan Baiyao and epsilon aminocaproic acid in dogs with right atrial masses and pericardial effusion. J Vet Emerg Crit Care (San Antonio). 2016 Sep 26. doi: 10.1111/vec.12529. [Epub ahead of print]

Frederick J, Boysen S, Wagg C, Chalhoub S. The effects of oral administration of Yunnan Baiyao on blood coagulation in beagle dogs as measured by kaolin-activated thromboelastography and buccal mucosal bleeding times. Can J Vet Res. 2017 Jan;81(1):41-45.

Lee A. Boysen SR. Sanderson J. et al. Effects of Yunnan Baiyao on blood coagulation parameters in beagles measured using kaolin activated thromboelastography and more traditional methods. International Journal of Veterinary Science and Medicine. 2017;5(1):53–56.

MacRae R. Carr A. The Effect of Yunnan Baiyao on the Kinetics of Hemostasis in Healthy Dogs. ACVIM  Forum, National Harbor, MD, 2017.

Adelman L. Olin S. Egger CM. Stokes JE. Effect of Oral Yunnan Baiyao on Periprocedural Hemorrhage and Coagulation in Dogs Undergoing Nasal Biopsy. ACVIM Forum, National Harbor, MD, 2017.

Not everything that looks like science really is science. I’ve written about the faux science of some alternative medicine publications, such as the Integrative Veterinary Care Journal, which often publish articles that are held up as scientific evidence for alternative theories but which don’t qualify in any way as scientifically legitimate. And not all science is good science or reliable as evidence. This is, sadly, true of much of the veterinary literature, conventional or alternative. However, alternative medicine publications are particularly vulnerable to poor methodology and creating the appearance of reliable scientific evidence without the substance. A close reading of the literature cited by alternative practitioners to support their claims often finds that the evidence is unreliable or doesn’t actually say what they claim it does (e.g. The Evidence for Homeopathy: A Close Look).

So-called Traditional Chinese Medicine (TCM) (which is, arguably, not traditional at all), is a particularly clear example of this problem. There is an abundance of published research on TCM therapies, especially acupuncture. Most of this is published in China or in TCM-dedicated journals, and it almost always finds exactly what the investigators hope and expect to find. When almost every single research study reports positive results, and when most contain few effective controls for bias, the odds are good that the literature represents the strength of belief among TCM proponents more than the objective truth about the method.

A recent article from the American Journal of Traditional Chinese Veterinary Medicine (AJTCVM) illustrates this problem quite well:

Fowler MP. Deng-Shan S. Huisheng X. A Randomized Controlled Study Comparing Da Xiang Lian Wan to Metronidazole in the Treatment of Stress Colitis in Shelter/Rescued Dogs. AJTCVM. April, 2017.

The AJTCVM is usually not available except to members of the TCVM organization that publishes it, but this article was recently made public. This may be related to the ongoing efforts of the TCVM community to gain recognition for herbal medicine as a board-certified veterinary specialty, but this is just speculation.

This study compares a TCVM herbal remedy to a pharmaceutical for “stress colitis.” The structure of the experiment has many of the hallmarks of traditional scientific clinical research. Unfortunately, the work as a whole does not support the authors inevitable conclusion that the Chines herb is an effective treatment. I will summarize briefly the main problems with the study.

What is “Stress Colitis?”
Unfortunately, this is a commonly used term that does not have a meaningful definition based on established pathophysiology or research evidence. Dogs often develop acute large-bowel diarrhea, with the sort of symptoms described by the authors of this study. This is associated with many possible causes, from parasites to eating unfamiliar or unwholesome substances to changes in routine that are likely a source of psychological and physiological stress. Unfortunately, the term stress colitis is a vague one that can be applied to any acute large bowel diarrhea when no specific cause has been sought or found. In this study, the authors ruled out some common parasites as the cause of diarrhea, but no actual cause was identified. Therefore, the entity they are testing treatments for is ill-defined and could easily be a hodgepodge of different causes.

This wouldn’t be a fatal weakness in itself, so long as objective and consistent criteria were used to identify the problem. However, the criteria for including and excluding patients and for determining whether or not they responded to treatment were subjective and vague as well. This leaves it entirely to the judgement of the investigators whether the dogs had “stress colitis” or not, and whether they got better or didn’t get better with treatment. Such a source of bias is a serious flaw, especially given the lack of any blinding, as I will discuss shortly.

Metronidazole
The authors designed the study to compare the herbal product to metronidazole, a drug commonly used to treat all sorts of diarrhea, including whatever people choose to label “stress colitis.” Unfortunately, there is no scientific evidence showing that this drug is actually effective for acute large bowel diarrhea from most causes, including cases for which no specific cause can be identified. The authors claim the drug is recommended in a commonly used drug handbook, but this is not correct. All the scientific evidence for the use of metronidazole in treating colitis involves chronic, not acute cases, and mostly diseases with an established cause, such as ulcerative colitis, inflammatory bowel disease, and giardia. In other words, the authors cite evidence regarding chronic large bowel diarrhea due to a variety of different causes as evidence for using metronidazole as a treatment for acute large bowel diarrhea of unknown cause (presumed to be related to stress). This misrepresents the evidence.

Metronidazole as a treatment for acute idiopathic large bowel diarrhea is common, but that is an entirely anecdotal practice for which there is no specific scientific support. This means the owners are comparing an untested herb supported only by anecdotal evidence against an untested medication supported only by anecdotal evidence. Not a very robust study for determining the real safety and efficacy of either treatment.

The authors also seem quite ambivalent about metronidazole. On the one hand, they want to identify it as a treatment proven to be effective for stress colitis so they can claim their herbal remedy is effective if it seems to work as well. They state that metronidazole is “typically effective” and is “the drug of choice” for stress colitis. Since the diseases is not clearly defined or understood and the drug actually isn’t proven to be an effective treatment for it (though it is widely used in this way), the implication here is false.

On the other hand, the authors clearly want to emphasize the dangers of conventional therapy so they can make the common claim that their alternative therapy is safer. So only a few sentences after claiming metronidazole is typically effective, they refer to the “poor clinical response” and “ineffectiveness” in some dogs, while then providing a scary list of potential side effects with no discussion of how frequently these actually occur or their relationship to dose or other factors.

Metronidazole does, of course, have risks that have been established by scientific research, as does any treatment with any real effects at all. Nothing works without some risk of unintended effects. However, the risks are generally seen with high doses and long-term use. And while the authors identify no known side effects for their herbal remedy, this is simply because there is virtually no research on this product, and no previous clinical studies at all in dogs. It might be safer or mare risky than metronidazole, but without appropriate research we can’t know. In this study, at least, there were no apparent significant adverse effects. One dog in each group was reported to have vomiting, but whether this was related to treatment is unclear, and there was no difference between the groups and no other possible adverse reactions noted.

Study Design Problems
This study is a classic example of how to set up a study to show what you already believe and want to show. The first author specifically says in the paper that she wanted to run the study because she believes the herb is effective, and may work better than metronidazole, based on her clinical experience. This bias can freely influence the results because there is almost no control for bias in the study. The inclusion and exclusion criteria were subjective, the response to treatment was subjectively measured, and no one was blinded to the group assignment or treatment of any of the dogs. These are not pedantic details but the fundamental core of good, objective science that is absent from this study.

The results of the study seem, not surprisingly, to support the beliefs of the authors. There was no difference in the resolution of diarrhea between the two groups. More than 85% of dogs in both groups got better within 10 days, and the average time to resolution of diarrhea was nearly identical at about 3 ½ days. The authors claim this shows the herb to be just as effective as metronidazole. However, remember that metronidazole has never been actually tested for treating acute large bowel diarrhea, so we don’t actually know if it is effective. What we do know is that in almost every study of acute diarrhea in dogs, most of the dogs get better no matter what we do. Several other studies have shown that more than 80% of dogs with acute diarrhea get better within 3-5 days, just as in this case, with all sorts of other treatments besides metronidazole and the herb used in this study ( e.g. 1, 2). It is very likely that most of the dogs in this study got better regardless of treatment and would have with neither metronidazole nor the herb. The absence of a placebo or no-treatment control group is a serious flaw that makes it impossible for the authors to address this very significant potential explanation for their results.

TCVM Nonsense
It is also important to remember the concept of Tooth Fairy Science. You can design a scientific study to see whether the Tooth Fairy pays out more money depending on the type of tooth lost, the age or sex of the child, and many other variables. You can even do all sorts of fancy statistics to evaluate the results. However, none of this means anything if, as is likely, the Tooth Fairy does not exist.

TCVM is a great example of Tooth Fairy Science. The herb in this study is supposed to be appropriate for cases with “excess pattern of large intestine damp heat” because is can “clear damp heat,” “move Qi,” and “warm middle Jiao.” Since all of the folk mythology of TCVM theory is implausible, untestable, mystical nonsense, the results of controlled studies of it are meaningless. That is not to say the herbs used in TCVM might not have real, and useful biological effects. But these need to be determined by real science, evaluating the chemistry, biochemistry, and pharmacology of these herbs in patients with scientifically defined diseases and measured using appropriate scientific methods. None of this is part of this paper.

The authors suggest that the herb they test has been shown to have benefits for diarrhea, though it has not been clinically tested in dogs before. However, the literature they cite shows only a handful of lab animal studies, which don’t look at anything like acute idiopathic large bowel diarrhea, and a couple of weak studies in humans for, again, quite different problems such as chronic ulcerative colitis. The reality is that despite the attempt to recreate the trappings of science, the authors are basing this study on the pseudoscientific nonsense of TCVM theory, their own personal beliefs and experiences, and anecdotal evidence, not a plausible scientific rationale demonstrated through real research. Tooth Fairy science at its finest.

Bottom Line
This paper illustrates the deep problems in the research literature associated with much of alternative medicine, including TCVM. Implausible therapies are selected according to mystical, pseudoscientific theories and anecdote and then tested inappropriately with little or no serious attempt to control for bias and error, and then the results are overinterpreted to suggest equivalence to conventional treatment or even the superiority of the alternative methods. The body of sloppy and unreliable literature that results misleads not only the public but veterinarians and other scientists into believing there is real reason to take these therapies serious.

This, in turn, supports the deceptive Trojan Horse of integrative medicine, which seeks to blur the real and important distinction between science-based medicine and alternative medicine. Real scientific evaluation of many alternative therapies, especially herbal remedies, needs to be done. This ain’t it. Unfortunately, this is typical of what happens when alternative medicine proponents set out to sue science not to test the reality of their beliefs but to generate marketing tools to sell those beliefs to other clinicians and the public.