9 Scientific Research
Waitlist:A control group formed by patients receiving no treat-
ment as a form of control. While a patient is waiting for his/her
treatment, all his/her data relevant to the treatment are collected
and later compared to those data obtained during treatment.
Since there is neither randomisation nor a placebo treatment, the
value of a waitlist group as a control group is very low.
Cross-over study:A particular design of controlled study. Each
patient will participate in two different courses of treatment (for
example treatment 1: medication; treatment 2: acupuncture). The
order in which the treatments are administered can differ from
patient to patient and be random (randomised cross-over study).
However, any long-term effect of the first form of treatment poses
a problem for the final result. Therefore so-called washout phases
during which patients receive no treatments at all are implemented
to eliminate these effects and prevent a distortion of the data to be
collected during the second phase. Since the effects of some
acupuncture treatments have been shown to last for up to a year,
cross-over studies are not very suitable for acupuncture research
and particularly non-significant results should be questioned.
The ‘n of 1’ study is another form of research: a patient will
receive various different treatments in order to find out which
form of treatment led to the best improvement for the relevant
disorder. However, the same limitations as discussed above will
affect this type of study.
Significance:In simple terms, the significance of a study indi-
cates if the difference of a particular parameter which has been
observed between two groups is caused by the treatment (and is
therefore a true difference) or whether the difference is the result
of random deviation. For example, if in a large number of treated
patients the observed data differ ‘significantly’ from a group of
untreated patients, it is highly probablethat this difference is
not due to a random deviation. This probability can be expressed
mathematically. If, for example, the significance is expressed as
p0.03 this means that, considering a random error of 3%, the
difference noted between the two groups is statistically signifi-
cant. The convention is to accept a random error of up to 5% as
acceptable. The significance of a study is based on the number
of patients treated and the extent of the difference in the treat-
ment result (for example by assuming the same random error,
the difference in treatment results has to be much larger in a
group of 10 patients (n10) than if the same study was carried
out with 100 patients (n100)). Only differences classified as
‘significant’ are accepted for research studies.
Power of the study:This refers to the ability of a study to be
statistically significant. The higher the number of participants,
the higher the power of the study. By international standards,80% is an acceptable level of power. Depending on the differ-
ence between the two study groups, the number of patients
required can therefore easily be in the 100s or 1000s.
Experimental studies:Through standardising and controlling
conditions of the experiment strictly, every attempt is made to
avoid random results, often leading to just a small number of
participants. Generally these tend to be healthy volunteers, who
are subjected to a clearly defined stimulus. In this case acupunc-
ture as a form of therapy will achieve a high level of validity.
More questionable are studies of healthy subjects without a
well-defined stimulus, since in that case acupuncture as a bal-
ancing form of treatment is not used according to its principles
(for example changes in peristalsis caused by acupuncture in
healthy subjects).
Blinding:The gold standard of clinical trials are the so-called
double blind studies in which both the patient and the therapist
remain ignorant as to which treatment (for example the tablet
with or without the active ingredient) is administered to the
patient. However, blinding of an acupuncturist is not possible.
Even in a single blind study (only the patient is ‘blinded’),
his/her behaviour will distort the result of the trial. This effect
can be counterbalanced to some extent by blinding the assessor
of the treatment results (for example changes in the range of
motion) so that he/she does not know which treatment (authen-
tic or sham acupuncture) was received by the patient.
Standardisation:In pharmaceutical trials the amount of the
active ingredient per tablet can easily be stated (for example
standardisation of a 10 mg tablet). However, acupuncture cannot
be standardised. The acupuncturist’s point selection will largely
be influenced by experience, prior knowledge, specialisation
and cultural background. Strictly speaking, acupuncture trials
therefore do not investigate acupuncture per se but acupuncture
as performed by an acupuncturist or a group of acupuncturists.
The fact that acupuncture is to a large extent centred on the ther-
apist has in the past led to a paradoxical situation: experienced
Chinese acupuncturists performing the ‘best’ acupuncture pro-
duced biomathematically ‘bad’ trials due to their lack of train-
ing, while ‘bad’ Western acupuncturists produced ‘good’ trials
from a biomathematical point of view. ‘Importing’ Chinese
acupuncturists into Western trials did not solve the problem
either, since the Chinese clinicians were now treating Europeans
with diseases not common in China and which reacted differ-
ently to acupuncture. Only as experience grows on both sides –
which would have to be evaluated individually for each trial – will
the value of acupuncture as a researched form of therapy increase
accordingly.