provincial or regional, and 4 specialist).^15 The first trials were published in the
early 1980s. The number of trials had doubled every 2–3 years over the past
15 years. The number of RCTs published in all 100 journals by the end of
1996 was estimated to be around 7500. Comparison of trials searched for by
hand with trials of TCM found in electronic databases (which hold journals
of conventional medicine as well) shows that journals of conventional medi-
cine in China published about a quarter of the number of RCTs published in
journals of traditional Chinese medicine. Thus, almost 10 000 RCTs were
published in China before 1997. Over 90% of the trials in non-specialist
journals evaluated herbal treatments that were mostly proprietary Chinese
medicines. The 10 most common diseases in the trials were ischaemic heart
disease, stroke, chronic viral hepatitis, peptic ulcer, childhood diarrhoea,
hyperlipidaemia, primary hypertension, upper digestive tract bleeding,
diabetes mellitus and pneumonia. They accounted for a fifth of the trials.
Unfortunately much of the early research was considered to be inadmissible
because of problems associated with:^16
- poor translation of studies
- the quality and design of the research not being up to western
standards - the use of unvalidated methods
- methodological difficulties of establishing control groups and sham
procedures for the placebo arm of trials (e.g. it is impossible to ‘blind’
an acupuncturist) - variations in what is understood by different terms.
From an initial sample of 37 313 articles identified in the China National
Knowledge Infrastructure electronic database a study led by Wu of the
Chinese Cochrane Centre at Sichuan University, found 3137 apparent
randomized controlled trials on 20 common diseases published between
1994 and 2005.^17 Of these, 1452 were studies of conventional medicine
(published in 411 journals) and 1685 were studies of traditional Chinese
medicine (published in 352 journals). Interviews with the authors of 2235
of these reports revealed that only 207 studies adhered to accepted method-
ology for randomisation and could on those grounds be deemed authentic
randomised controlled. The reviewers considered that a randomisation
sequence generated from a random number table, calculator, or comput-
erised random number generator was authentic but that tossing a coin,
drawing straws, or allocating a participant according to date of birth or
hospital record number was not.
It is vital that correct plant species are used when researching traditional
herbal medicine and that tests are carried out on material prepared
according to ethnic methods.^18 The choice of test system might also be
Traditional Chinese medicine | 131