by Eric Brand

Evidence-based medicine (EBM) is a major topic of discussion in the modern medical world. While Chinese medicine offers a rich and well-preserved historical record of medical interventions, its “evidence base” is fraught with confounding variables that hamper its acceptance into the mainstream medical paradigm. Chief among these variables are issues of inter-rater reliability (i.e., different doctors diagnose the same patient differently), the use of polypharmacy (multiple medicinal substances are combined together and are often varied for each patient at a given point in time), and the differences between modern research protocols and historical records of treatment (no controls, only treatment groups in ancient times).

Many practitioners have witnessed tremendous results from Chinese medicine, and the prevailing viewpoint in the field suggests the best results are obtained by applying appropriate therapies based on traditional medical theory. While the theory of Chinese medicine itself will never be “proven” by modern science, the clinical results of Chinese medicine are suitable to analysis by EBM. Even in the absence of a known mechanism of action, proper study design can demonstrate significant clinical effects from the utilization of Chinese medicine. Such studies strengthen the evidence base of Chinese medicine and are vital to the mainstream scientific acceptance of Chinese medicine’s potential.

The first two lines of the Dao De Jing (Tao Te Ching) state: “the dao that can be “dao-ed” is not the [real] dao; the name that can be named is not the [real] name.” Similarly, the theory that underlies Chinese medicine is not reality; it is a metaphor for reality that can be consistently used to replicate clinical effects if one understands CM theory well enough to employ its therapies. The description of the human body offered by biomedicine is also not a be-all, end-all description of reality; it is simply a cohesive collection of hypotheses with supporting evidence. The body and its state of disharmony can be viewed from multiple different paradigms; while no single paradigm will ultimately describe the totality of what the body truly “is,” different paradigms can be used to amass evidence that aids in the treatment of human suffering.

So how can we study the clinical effects of Chinese medicine while still preserving the traditional paradigm that it is based upon?

Follow up:

This is a very hot topic in the global world of Chinese medicine. Essentially the conundrum is that the clinical effects of “authentic” Chinese medicine can only be studied scientifically if the concept of pattern identification is preserved when selecting treatment. But how can we preserve the concept of pattern identification and pattern-specific therapy while maintaining a rigorous study design?

The solution in China has been the creation of a numeric rating system that allows patients to be grouped into cohorts based on their TCM pattern, with each pattern cohort receiving an herbal formula (or acumoxa treatment) appropriate for their pattern. While this is still imperfect in that each individual patient cannot receive fully customized therapy, it does allow for the preservation of pattern identification and pattern-based treatment, which many argue to be a fundamental aspect of TCM that must be preserved in order to properly assess its clinical potential.

The concept of a numeric rating system for signs and symptoms was created to maximize inter-rater reliability in TCM studies. Essentially, the problem is that different practitioners tend to diagnose the same patient differently, so scientific studies are easily limited when a given diagnostician is biased towards certain patterns or is generally unskilled in their diagnostic ability. Given the large number of studies done in China in a variety of different clinical settings, a national standard is required for research purposes.

While China does not yet have a single standard, many hospitals and universities have proposed numeric rating systems and approaches for the creation of diagnostic standards for research studies. A Beijing-based national committee on TCM diagnostic standards has been created, but they are currently still in the stage of gathering input from the various universities and hospitals so that a future national standard can be established. (Readers should note that there are no efforts in any way to standardize the treatment of patients in China, all Chinese medical doctors are free to use whatever theory and treatment methods that they feel are appropriate; national diagnostic standards are strictly limited to applications in research studies for the purpose of building Chinese medicine’s scientific evidence base. I mention this because many Western practitioners have the erroneous idea that Chinese medicine in China somehow draws from a limited theoretical base (“TCM” in the narrow sense), and I feel the need to point out that there are no “national standards” that restrict the theory and treatments in use there.)

Anyway, essentially these rating systems assign numeric values to specific signs and symptoms. These numeric values can be correlated to patterns, allowing the patients in the study to be split into different treatment cohorts based on their prevailing TCM pattern. For example, if a patient complains of fatigue, sore lower back and legs, fear of cold, frequent nocturia, and impotence, these signs and symptoms when taken together as a whole create a picture of kidney yang vacuity. On their own, any of these individual signs constitutes relatively weak evidence (for example, fatigue also commonly occurs in simple spleen vacuity as well as blood vacuity); however, taken together as a whole the picture overwhelming points to kidney yang vacuity. In the numeric rating system, signs that are relatively specific and offer strong evidence are weighed more heavily than non-specific but supportive signs (such as fatigue), and as a whole the numeric system can be used to group the patients into cohorts while reducing the variable of diagnostic bias that one encounters with any given practitioner.

Many Western practitioners know that individual symptoms and signs have different levels of diagnostic significance; this is basic TCM and it is exemplified in lectures such as Bob Flaws’ recent presentation on “Hallelujah Symptoms” at the 2008 Pacific Symposium. However, many Westerners do not know that there is a major trend in China, Hong Kong, and Taiwan to utilize common rating systems to subject pattern-based TCM therapies to the EBM approach. Two main hurdles remain when it comes to increasing inter-rater reliability for TCM studies: the traditional importance of pulse and tongue.
Pulse and tongue diagnosis both are tremendous confounding variables. Both are regarded as essential by traditionalists, but the information they reveal is interpreted differently by different practitioners and it is not as easily evaluated as a checklist of signs and symptoms. Consequently, tremendous effort and financial resources are currently pouring into technological solutions to improve the inter-rater reliability of TCM doctors when it comes to tongue and pulse diagnosis.

The tongue is far easier to interpret than the pulse. We acquire the skill to utilize tongue diagnosis effectively long before our pulse-taking abilities are mature, and the fact that the tongue is assessed visually instead of palpably makes suitable diagnostic machines for the tongue much easier to design than a similar machine for the pulse.

One of my teachers in Taiwan, Dr. Chang Hen-Hong, is the Vice Superintendent of Chang Gung Memorial Hospital and the head of Chinese medicine at Chang Gung University. Dr. Chang and his research assistants have spent many years researching (inventing!) machines that can digitally assess the tongue. Apparently, one of the greatest challenges is actually lighting, since the lighting used in the tongue photograph can easily cause the true color to come out differently in the image. From there, it is a question of pixels and their significance. Computers can actually capture tremendous detail, even greater than the naked human eye. Nonetheless, it takes a consensus of skilled doctors to tell the computer how to evaluate whether the tongue fur is thick or thin, how red or purple it is, etc.

At present, this research is already very advanced. Dr. Chang’s group was very far along several years ago during my stay at Chang Gung Hospital, and he recently told me that the completed tongue photographs will eventually be integrated into the patients’ digital files at Chang Gung. Currently, Chang Gung is a paperless hospital. Patients insert smart cards into a computer, and their personal file comes up on the doctor’s computer screen. Their past medical history is there, along with their drug and herbal treatments, TCM and biomedical diagnoses, etc. Soon, they will implement an image of the patient’s tongue so that the tongue changes can be tracked for each patient. Once thousands of tongue images (along with the corresponding diagnosis and herbal therapy for that patient) are in the system, Chang Gung will be able to analyze the data and tongue changes. In this way, they will be able to truly create an evidence base that allows them to see, for example, whether or not thick tongue fur is consistently associated with a clinical presentation of dampness, and whether or not it responds to damp-resolving therapies. Pretty stunning, really, given that Chang Gung can easily analyze the data of tens of thousands of cases based on a staggering variety of parameters.

Pulse machines are a bit slower in their development. I saw a presentation in Taiwan of a pulse machine that was being developed, and it was already strikingly advanced despite its limitations. For example, it could detect the prominence of the pulse at 27 different depths, and it could use extremely precise measurements to determine elements such as pulse width, turbulence of the blood in the vessels, rate, and forcefulness or lack thereof. Apparently it could even reveal global qualities such as the wiry pulse, but it lacked the sophistication to differentiate all of the traditional pulse qualities (though of course there is lack of consensus on the subjective sensation of some of the traditional pulse qualities, such as the rough/choppy pulse).

Overall, it will be a number of years before there is a complete global system that integrates all of the elements necessary for EBM-style studies that preserve traditional pattern diagnosis. It will take even longer before enough studies are done with rigorous methodology to “prove” the reliability of things like pattern identification and pattern-based TCM treatment, but clearly the field is moving in this direction. We live in fascinating times, and it is exciting to be in a field that has such a diverse spectrum of ideas and participants. While many of us will undoubtedly always choose to focus on patients close to home with customized treatments based on traditional medical theory, we cannot ignore the fact that the larger field of Chinese medicine is becoming deeply connected with EBM. In fact, EBM is the ideal paradigm to allow modern science to embrace and discover the miraculous effects of TCM that we see on a regular basis.