BMJ  2006;333:391-394 (19 August), doi:10.1136/bmj.333.7564.391

Analysis and comment

Research priorities in traditional Chinese medicine

¹ Hong Kong Cochrane Branch, Chinese University of Hong Kong, Hong Kong, China

Is the current Western model of research—trying out unknown treatments in animals—suitable for studying treatments that have long been used in humans?

Introduction

Evidence based medicine re-emphasises applied clinical research in human subjects.¹² However, research in traditional Chinese medicine has had a mechanism centred approach and has been dominated by studies of basic and intermediate mechanisms. Though tremendous efforts have been made, and despite occasional successes, such as in acupuncture,³ most questions—for example, the nature of disease in traditional Chinese medicine—have not been satisfactorily answered.⁴

According to Liang, "Since the early 1990s, the search for the nature of disease has descended into a downward spiral. All the breakthroughs once cheerily anticipated seemed to have become an illusion. The entire traditional Chinese medicine research is currently in a state of disarray. Basic research had come to a standstill. What has gone wrong? Where should we go from here?"⁴ In this article, I argue that research priorities in traditional Chinese medicine need to be reviewed, and I propose an efficacy driven strategy.

Two possible approaches to research

The mechanism centred approach is primarily concerned with the search for the molecular, cellular, and pharmacological bases of traditional medicines. It seeks to identify the active substances of herbal treatments and investigate the mechanism of action. This strategy is shaped by a belief that traditional Chinese medicine need not be evaluated,⁵ by the general prosperity of basic biomedical sciences, and by drug development models in conventional medicine. As many research activities in traditional Chinese medicine aim to develop new drugs, the model used by conventional medicine to design and develop new drugs, such as drugs for cancer,⁶ seems directly relevant.⁷ It has three important conventions (fig 1, left). Firstly, the potential drug must be a single chemical entity, or a combination of known substances. Secondly, for ethical reasons, evaluation of the drug's safety and efficacy must begin in vitro and in animals. Thirdly, the drug's pharmacology and mechanism of action must be well understood before it is evaluated fully in humans. This model has been successful in modern conventional medicine and is readily available for traditional Chinese medicine.

View larger version (53K): [in this window] [in a new window]   Fig 1 Comparison of approaches to research on novel and traditional treatments

 

Fig 2 Qinghao, from which qinghaosu (artemisinin), a powerful new anti-malaria drug, is extracted Credit: SUE SNELL/WELLCOME PHOTO LIBRARY

 
However, as traditional Chinese medicines are already in use, it would be better when studying them to start with showing efficacy in humans by randomised controlled trials (fig 1, right). Efficacy refers to the capacity of a drug to bring about more good than harm in human subjects and is what matters most for any medical intervention. In the efficacy driven approach, studies on mechanisms and active substances are still important but should be undertaken after efficacy has been confirmed.

Why an efficacy driven approach?

Studies of efficacy will have direct, immediate applications for patient care whether the results are positive or negative: if a treatment is proved to be efficacious, promoting its use will benefit patients; if a treatment lacks efficacy, stopping using it will save resources. Such studies would also avoid unnecessary basic research on inefficacious interventions.

Even if a treatment is efficacious, it may not be possible to determine the mechanism of action and the active substances immediately, particularly in herbal treatments that contain many compounds. In scientific inquiries, success is often hard to predict.⁸ However, we need not wait for such successes. Lack of knowledge about mechanisms and active substances need not prevent the use of efficacious treatments. Many powerful interventions, such as penicillin and smallpox vaccination, were accepted well before their mechanisms were understood.⁹ Inefficacious interventions will eventually be discarded whether or not the mechanism has been explained. Bloodletting is a classic example, and vitamin E for preventing cancer provides a modern equivalent.^{10 11}

Mechanisms are really only theories and change over time as new knowledge becomes available. How does cowpox vaccine prevent smallpox? The answer to this question today differs from the explanation given 200 years ago. Some would even argue against heedless searching for deep explanatory models: any good empiricist can attest that sound evidence is hard won and that the human mind can concoct a theory to support any set of notions and observations.¹²

The primary value of medical theories lies in successfully guiding medical practice and generating efficacious interventions. Thus, demonstrating efficacy would be the best empirical test of the theories behind traditional Chinese medicine. Based on therapies of confirmed efficacy, newly developed and improved theories of traditional Chinese medicine will be on a more solid ground.

A long history of use, traditions, faith, popularity, and anecdotes are widely taken as evidence for the efficacy of traditional Chinese medicines. Some traditional therapies are undoubtedly effective (Qinghao, for example; see fig 2),^{13 14} but this does not mean that all are. Randomised controlled trials are the most rigorous method for evaluating the efficacy of any interventions.

Is it ethically acceptable to study traditional Chinese medicines in humans first? These medicines have been used for thousands of years. Whether tested or not, they will continue to be used in places where traditional Chinese medicine is officially recognised. New treatments will continue to be invented and given to patients by individual doctors without any systematic evaluation. A randomised controlled trial is simply a systematic use of the medicine, with an explicit research objective. In determining what and how much basic research should be required before randomised controlled trials of traditional medicines are allowed to start, it is important to note that the delay in starting a trial is proportionally related to the amount of basic research required, and also that it is unethical not to conduct or to delay trials of treatments that are widely used but have uncertain efficacy.

Difficulties in conducting clinical trials of traditional Chinese medicine

A small number of randomised controlled trials in traditional Chinese medicine have been done; most are of poor methodological quality.¹⁵ The situation may partly be due to methodological difficulties, such as design and implementation of placebo blinded trials of individualised treatents. Other crucial matters also need to be addressed.

Incommensurability
Both traditional and conventional medicine originated from different world views: the former from ancient Chinese philosophy and the latter from ancient Greek and Roman medicine.^{16 17} Consider the old Buddhist tale, where a few blind men try to find out what an elephant is like. The man who touched the leg thought that the elephant was like a post, while the one who grabbed the tail believed that it was like a rope. Conventional medicine may see only the "leg" and traditional Chinese medicine only the "tail." The same disease presents different problems in the two paradigms of medicine. For example, in traditional Chinese medicine, hypertension would be several different syndromes (for example, Gan Yang Shang Kang—predominance of the yang of the "liver"), and blood pressure need not be referred to in either making the diagnosis or in judging whether the syndromes have improved or deteriorated.

Would this incommensurability mean that traditional Chinese medicine can be evaluated only within its own paradigm but not by the standards of conventional medicine (particularly those for diagnosis and outcomes), and that conventional medicine could not understand and accept it? To continue with the tale: if the elephant is gone, all the men would agree that there was no longer an elephant. The very existence of the elephant is essential to any deduction from the evidence. If the two medicine systems see and deal with the same underlying disease in different ways and the disease is "cured" by either form of medicine, the disease will be "gone" regardless of methods. This suggests that traditional Chinese medicine could be evaluated by using outcomes defined in both medical systems, especially when a disease can be cured.

To resolve the problem of incommensurability in diagnosis, patients for trials of traditional Chinese medicines can be recruited in two ways. Firstly, patients with the same traditional Chinese medicine syndrome can be recruited from those with a particular disease in conventional medicine. In such a trial, the same traditional treatment can be evaluated, and generalisation about the therapy would be valid, but the eligible patients available for study could be small in number and difficult to recruit. Secondly, patients with the same disease in conventional medicine could all be recruited, regardless of their diagnosis in traditional Chinese medicine. The eligible patients would be many, but they must be given different traditional treatments. Many current traditional Chinese medicine trials fall into the latter category. They may show that it is efficacious, but generalisation about each treatment is difficult unless the trial is sufficiently large to allow subgroup analyses.

Individualised treatments
It has been argued that because in traditional Chinese medicine treatment is individualised, randomised controlled trials cannot be used as such trials require similar patients who need similar treatments. This misunderstanding is not new. Sir Austin Bradford Hill noted 40 years ago: "The most frequent and the most foolish criticism of the statistical approach in medicine is that human beings are too variable to allow of the contrasts inherent in a controlled trial of a remedy." He challenged his critics: "If each patient is unique, how can a basis for treatment be found in the past observations of other patients?"¹⁸

Can similar patients who need similar treatments be found in traditional Chinese medicine and studied in clinical trials? It is true that no patients are exactly the same. What matters, however, is whether the dissimilarity is relevant to the efficacy of the treatment. Like conventional medicine, traditional Chinese medicine has a limited number of "syndromes" available (the state of a disease in a person at a particular time, which determines the choice of treatment) and probably only a few hundred are common. In traditional Chinese medicine, it is legitimate to give patients with the same syndrome similar treatments.

For example, classic herbal formulas and proprietary drugs, such as Liu Wei Di Huang Wan for deficiency in the yin of the "kidney," were developed for specific syndromes. Good evidence for further tailoring beyond the standard formulas is often lacking. In addition, the same traditional medicine is often prescribed to patients with the same disease, regardless of the syndrome.⁴ It would be unreasonable to deny the value of randomised controlled trials in traditional Chinese medicine as that would be to refuse the scientific tenets of induction and causation.

Despite these arguments, it is possible to design randomised, blinded trials of individualised treatments in traditional Chinese medicine. In such a trial, every patient can consult the doctor and be prescribed the herbs best tailored to their needs. Then the patient will go to see the trial administrator, who will randomise him or her to receive either the herbs as prescribed by the doctor or placebo herbs. Placebo herbs should be chosen by experienced traditional Chinese medicine doctors to make sure they are highly unlikely to have relevant effects. Then the prescribed and placebo herbs can be prepared in an identical ready to use form (such as solutions) or made into small pieces and packed in gauze bags which can be simmered in water by patients at home without being opened.

Summary points Research in traditional Chinese medicine has been dominated by the search for its biological basis, identifying active substances, and investigating mechanisms of action Searching for active substances and mechanisms of action is meaningless if a treatment has no clinical efficacy As traditional medicines are already in use, research should adopt an efficacy driven approach, and start by showing efficacy in humans through randomised controlled trials Cultivating a culture around uncertainty and evaluation is needed more than resolving the methodological difficulties in conducting randomised controlled trials of traditional medicines

However, as the efficacy of individualised treatments is determined by two factors: the treatment itself and the competence of the treating doctor, trials of individualised treatments could have problems for the explanation and application of their results: the efficacy shown in a trial may not be reproducible in other doctors and the absence of efficacy could be explained as resulting from poor skills of the doctor rather than from the treatment. Cluster trials that randomise patients to practitioners of traditional Chinese medicine and conventional medicine and allow practitioners to treat in an individualised manner have the same problems.

Attitude towards evaluation
In social research, "One of the greatest methodological fallacies of the last century...is the belief that science is a particular set of techniques; it is, rather, a state of mind, or attitude, and the organisational conditions which allow that attitude to be expressed."¹⁹ Is this also true of research in traditional Chinese medicine? The attitude towards the need for evaluation, rather than methodological difficulties, has been the greatest hindrance in evaluating traditional Chinese medicine. For advocates, every treatment works, so evaluation is unnecessary. For sceptics, traditional Chinese medicine is quackery, so evaluation is pointless. The truth probably lies somewhere in between: some treatments may be efficacious, and some are probably not. This stark division between proponents and opponents shows the collective professional uncertainty about traditional Chinese medicine and represents a typical clinical equipoise that calls for a randomised controlled trial.^{20 21} Methodological difficulties can be resolved only when the need for evaluation is widely endorsed.

Many interventions in the current practice of conventional medicine are of doubtful value and have not been proved by scientific research, and are thus also candidates for the efficacy driven approach to research. Research on basic mechanisms that is crucial for future advancement should not be downplayed in the new approach—but clinical research should precede basic research for interventions already in use.

---

Contributors and sources: This article benefited hugely from discussions with and comments from colleagues, notably Ke-Ji Chen, Ping-Chung Leung, Jean Woo, Tze-Wai Wong, Chris Saligy, Mike Clarke, Guo-Ming Qi, Joseph Lee, Jim Dickinson, Thomas Chan, Joe Liu, Hong-Wei Zhang, and those at the World Economic Forum 2002 China Business Summit, which invited a speech on this topic.

Funding: JLT received grants for related research from the Chinese University of Hong Kong (6900511, 2041041, 2040972).

Competing interests: None declared.

References

1. Complementary medicine: time for critical engagement [editorial]. Lancet 2000;356: 2023.[CrossRef][ISI][Medline]
2. World Health Organization. World report on knowledge for better health. 2004. www.who.int/rpc/meetings/pub1/en/ (accessed 12 Aug 2005).
3. Ulett GA, Han S, Han JS. Electroacupuncture: mechanisms and clinical application. Biol Psychiatry 1998;44: 129-38.[CrossRef][ISI][Medline]
4. Liang MX. [The predicaments and future of the search for the nature of disease in traditional Chinese medicine.] Beijing: People's Medical Press, 1998: 1-48. (In Chinese.)
5. Yuan Z. [Development of traditional Chinese medicine is an important part of the advancement of scientific medicine.] In: Chen KJ, ed. [Rational thinking in the development of traditional Chinese medicine.] Beijing: People's Medical Press, 1997: 36-9. (In Chinese.)
6. Steele VE, Boone CW, Lubet RA, Crowell JA, Holmes CA, Sigman CC, et al. Preclinical drug development paradigms for chemopreventives. Hematol Oncol Clin North Am 1998;12: 943-61.[CrossRef][ISI][Medline]
7. Hayvey AL. Sense, nonsense and antisense: different approaches to drug discovery. In: Harvey AL, ed. Advances in drug discovery techniques. 2nd ed. Chichester: Wiley, 1998: 1-12.
8. Kuhn TS. The structure of scientific revolutions. Chicago: University of Chicago Press, 1970: 77-91.
9. Wynder EL. Studies in mechanism and prevention: striking a proper balance. Am J Epidemiol 1994;139: 547-9.[Free Full Text]
10. Warren KS. All effective treatments could be free. Ann NY Acad Sci 1993;703: 1-4.[ISI][Medline]
11. Rapola JM, Virtamo J, Haukka JK, Heinonen OP, Albanes D, Taylor PR, et al. Effect of vitamin E and beta carotene on the incidence of angina pectoris. A randomized, double-blind, controlled trial. JAMA 1996;275: 693-8.[Abstract]
12. Haynes RB. A warning to complementary medicine practitioners: get empirical or else. BMJ 1999;319: 1632.[ISI][Medline]
13. Sheehan MP, Rustin MH, Atherton DJ, Buckley C, Harris DW, Brostoff J, et al. Efficacy of traditional Chinese herbal therapy in adult atopic dermatitis. Lancet 1992;340: 13-7.[CrossRef][ISI][Medline]
14. Tran TH, Day NP, Nguyen HP, Nguyen TH. Tran TH. Pham PL, et al. A controlled trial of artemether or quinine in Vietnamese adults with severe falciparum malaria. N Engl J Med 1996;335: 76-83.[Abstract/Free Full Text]
15. Tang JL, Zhan SY, Ernst E. Review of randomised controlled trials of traditional Chinese medicine. BMJ 1999;319: 160-1.[Free Full Text]
16. She JB. The beginning of the Chinese and Western medicines. In: He YM, ed. [Differences, perplexity, and choices—a comparison of traditional Chinese medicine with western medicine.] Shen Yang: Shen Yang Press, 1990. (In Chinese.)
17. Tiao KY. [The philosophical considerations of pathology of traditional Chinese medicine.] Shanghai: Shanghai Science Popularisation Press, 1997. (In Chinese.)
18. Hill AB. The clinical trial. New Engl J Med 1952;247: 113-9.[ISI][Medline]
19. Dingwall R. "Don't mind him—he's from Barcelona": qualitative methods in health studies. In: Daly J, MacDonold I, Willis E, eds. Researching health care: designs, dilemmas, disciplines. London: Tavistock/Routledge, 1992: 161-75.
20. Freedman B. Equipoise and the ethics of clinical research. N Engl J Med 1987;317: 141-5.[Abstract]
21. Weijer C, Shapiro SH, Glass KG, Enkin MW. Clinical equipoise and not the uncertainty principle is the moral underpinning of the randomised controlled trial. BMJ 2000;321: 756-8.[Free Full Text]

Received for publication 14 February 2006.
CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    Technorati    What's this?

Related Articles

For and against: Clinical equipoise and not the uncertainty principle is the moral underpinning of the randomised controlled trial • FOR • AGAINST

Charles Weijer, Stanley H Shapiro, Kathleen Cranley Glass, and Murray W Enkin
BMJ 2000 321: 756-758. [Extract] [Full Text] [PDF]

Review of randomised controlled trials of traditional Chinese medicine

Jin-Ling Tang, Si-Yan Zhan, and Edzard Ernst
BMJ 1999 319: 160-161. [Full Text] [PDF]

Rapid Responses:

Read all Rapid Responses

Clinical Evaluation Already Embedded in Traditional Chinese Medicine Practice

Rey C. Tiquia
bmj.com, 25 Aug 2006 [Full text]

Time to recognize African Traditional Medine & suitable research approaches

Sam W. Mhlongo, et al.
bmj.com, 26 Aug 2006 [Full text]

Clinically Evaluating SARS

Dr. Rey C. Tiquia Ph.D.
bmj.com, 8 Nov 2006 [Full text]