THE JOURNAL OF ALTERNATIVE AND COMPLEMENTARY MEDICINE
Volume 12, Number 3, 2006, pp. 303–310
© Mary Ann Liebert, Inc.
A Review and Analysis of Placebo Treatments, Placebo Effects,
and Placebo Controls in Trials of Medical Procedures When
Sham Is Not Inert
STEPHEN BIRCH, Ph.D., Lic.Ac. (US), M.B.AcC. (UK)
Researchers examining the efficacy of medical procedures make assumptions about the nature of placebo.
From these assumptions they select the sham interventions to be used in their trials. However, placebo is not
well defined. A number of definitions are contradictory and sometimes misleading. This leads to problems in
sham-controlled studies of medical procedures and difficulties interpreting their results. The author explores
some of the contradictory definitions of placebo and assumptions and consequences of these. Principal among
these is the assumption that the placebo is inert when it is not, which introduces bias against the tested med-
ical procedures and devices. To illustrate the problem, the author examines the use of sham procedures in clin-
ical trials of the medical procedures surgery and acupuncture in which the sham was assumed to be inert but
was not. Trials of surgery and acupuncture should be re-examined in light of this.
placebo effects are small4and large.5Some claim placebo
does not exist.6Placebo effects vary with different treat-
ments;5,7,8effects for drugs are different than for devices;9
different-colored placebo pills change the placebo ef-
fect;5,8,10how treatment and placebo are explained in the
study changes the placebo effect5,7,11and can trigger its op-
posite, the nocebo effect.12Placebo effects interact with
other treatment effects,7,13making it difficult if not impos-
sible to control for them.5Some argue that placebo con-
trolled trials may not be possible in complex interventions
because they cannot separate placebo from other treatment
effects.14–16How should this bewildering array of opinions,
claims, counterclaims, and contradictory findings be under-
stood? When something supposedly so ubiquitous as the
placebo effect is not really understood, what does it mean
for clinical research?
Placebo is a very difficult concept to define.8,17–22Al-
though commonly understood as, “A pharmacologically in-
he literature on placebo is confusing evidenced by re-
cent books,1–3and conferences.1,3Authors claim
active agent given to a patient as a substitute for an active
agent and where the patient is not informed whether he is
receiving the active or inactive agent,”23recent descriptions
highlight some of the inherent ambiguities.
One assumption is that placebo is by definition inert. “The
one thing of which we can be absolutely sure is that place-
bos do not cause placebo effects. Placebos are inert and don’t
cause anything.”8The inert placebo triggers complex inter-
nal mechanisms in a patient that lead to observable placebo
effects, a process especially related to how the patient at-
tributes meaning.8,22,24However, contrary definitions of
placebo treatment include noninert therapies, so long as they
are not thought to be specific for the condition under treat-
The Shapiros described placebo as any therapy that has
not been proved effective in double-blind controlled stud-
ies25and that “The placebo may be an inert sugar pill, an
active drug, or any treatment no matter how potentially spe-
cific or by whom administered.”25The assumption that an
unproven treatment can be considered a valid placebo leads
to the paradox in placebo trials that a placebo is being com-
pared to a placebo. Further, what if it is discovered later that
Stichting (Foundation) for the Study of Traditional East Asian Medicine (STEAM), Amsterdam, The Netherlands.
what was called a “placebo” according to this definition was
later shown to be an active specific treatment?26–28Because
the definition will have contradicted itself, is not the defin-
ition now logically false?
Some have discussed the distinction between “perceived
placebo effects” and “true placebo effects.”17,23,29,30Per-
ceived placebo effects lump the “true placebo effect” to-
gether with a number of “nonspecific effects,” including re-
gression to the mean, natural course of the disease.30
However, this model makes no clear statement about the na-
ture of the placebo treatment itself, and leaves it up to the
imagination of the researcher to define. The same confusion
can be introduced as was seen in the last definition.
Finally, a recent approach is based on an “operational”
definition: “For the purpose of this article, placebo is an in-
tervention used in a clinical trial that is administered with
the intention of mimicking some other intervention so that
an unbiased comparison can be made.”31This definition ex-
plicitly assumes that any attempt at mimicking a therapy is
a placebo intervention and is thus similar to the definition
by Shapiro and susceptible to the same errors and contra-
dictions as that definition. It includes in the definition all
mistakes made by researchers in defining, constructing, and
implementing their mimic therapy.
Although there is a lot of agreement that a number of dif-
ferent factors such as treatment context, patient expectations,
and enthusiasm of the practitioner can contribute to placebo
effects,1–3,5,8,23there is considerable confusion about what
constitutes a placebo therapy. Is the placebo to be an inert
therapy or is it not inert, and include other “non-specific ef-
fects” caused by trying to mimic a therapy? This paper ex-
plores some of the consequences of this confusion.
Why the placebo control?
Why are randomized placebo-controlled clinical trials, or
“efficacy”32,33trials conducted? The simplest answer is that
“ . . . tests of efficacy always presume an underlying mech-
anism, and so a test of efficacy is necessarily a test of an
assumed mechanism.”32Because a therapy has specific ef-
fects (mechanisms or active ingredients) associated with it
and a variety of other effects not specific to that therapy,
then placebo-controlled studies are conducted to examine
the size of the specific effects by controlling for (subtract-
ing) the nonspecific effects.13,33–35These nonspecific effects
include effects caused by placebo responses, and effects
such as regression to the mean, natural course of the dis-
ease.29,30Randomization is thought to distribute the latter
effects equally between treatment groups.6,30,33,36Placebo
effects are controlled for if the test or active treatment is
compared to a “sham” treatment. In drug trials this is not
difficult to do provided double blinding (blinding of thera-
pist and patient) is used and the active and sham treatments
appear the same.33The sham treatments involve using a
chemically inert similar-looking pill with no active chemi-
cals in it and blinding those administering the treatment so
that neither they nor the patient know what is being given.
In such trials, the sham intervention is inert. Randomization
is also usually thought to distribute the amount of placebo
responses equally between the treatment groups. Under these
basic conditions efficacy or randomized placebo-controlled
clinical trials are conducted. This type of study is also known
as an “explanatory” trial33because it is designed not only
to examine if the treatment is more effective than its com-
parison treatment, but also to explain the effectiveness in re-
lation to the active components of the test treatment (e.g.,
However, what about therapeutic techniques such as
surgery and acupuncture? How can one guarantee that the
sham treatment is inert like a placebo pill? How can one im-
itate surgery while not initiating other nonplacebo-related
treatment effects resulting from trauma, etc.? Is it possible
to make inert sham treatments in ACT (alternative and com-
plementary therapies) like acupuncture?
If it is not possible to use inert sham treatments, how does
one deal with the additional beyond-placebo “nonspecific”
effects of the technique or device? When sham treatments
are used that are not inert, what impact does this have on
the understanding and interpretation of sham-controlled
clinical trials? Many researchers have assumed that the sham
treatments in medical technique trials are inert when they
are not. This opens the door to difficult questions about how
to interpret sham trials of, for example, acupuncture and
The inertness of “sham”: implications when wrong
What are the implications of a faulty assumption that the
sham procedure is inert? It introduces bias against finding
the tested therapy to be effective. As de Craen and col-
leagues demonstrated, “if the placebo has a large deviation
from inertness it is obvious that the results of any trial will
be biased, but even low placebo activity could bias the com-
parison when the absolute treatment effect is small.”26Be-
sides the threat to internal validity that a noninert placebo
can create,26three other compounding problems occur.
Researchers risk overestimating the size of the placebo
effects because they have inadvertently included other ef-
fects that are not placebo related,6,36making it more diffi-
cult to show that the real therapy is significantly more ef-
fective than sham.23This problem is compounded by the
Sample sizes often are based on an estimate of the ex-
pected size of the test and control treatment effects. If the
sham intervention is assumed to be inert, no greater than
placebo, calculations often are based on general ideas about
the size of the placebo effect. Usually this is taken to be
around 30% based on Beecher’s original calculations (how-
ever, these have since been found to be erroneous).17,19To
demonstrate a significant difference between the 30% ef-
fectiveness of the sham (placebo) and the expected effec-
tiveness of the intervention itself (say 60% to 70%), rela-
tively small sample sizes are necessary.37However, if the
sham intervention turns out to be greater than the estimated
30% effectiveness because the sham is not inert (e.g., if the
sham is 50% effective), then the smaller sample size based
on a comparison of 30% versus 70% will be inadequate to
demonstrate the effectiveness of the therapy.23,37The ma-
jority of acupuncture trials have suffered from inadequate
sample size,37–39and a number of the surgery trials have
had small sample sizes.40,41Trials that assume their sham
intervention is inert when it is not are at greater risk of mak-
ing false-negative judgments. De Craen and colleagues
showed that if the effect of the noninert placebo is small, it
can bias a study if the treatment effects are small. This bias
increases if the effects of the noninert placebo are larger and
especially if they are specific to the condition under study.26
Finally, whenever a trial concludes that “therapy X is not
more effective than placebo,” it is either implied by the re-
searchers and/or understood by readers that the placebo was
inert.11The training of most people reading published trials
is not adequate to understand the complexity of the placebo
concept and people revert to the original idea from the 1950s
that the sham procedure was inert, which usually damns that
therapy as ineffective.19These various factors compound
each other to produce bias against the test therapy.
Biologic and therapeutic effects of surgical
In sham surgical trials, both groups of patients receive the
surgical incision or other invasive procedure to expose the rel-
evant organs or anatomical structures. The real surgical group
receives the surgical procedure, whereas the sham group has
the incision closed. Both groups receive the same postopera-
tive care. In the 1950s two trials of mammary artery ligation
surgery were performed for the treatment of angina.40,41The
mock or sham surgery included making an incision and ex-
posing the relevant blood vessels, and then surgical closure of
the wound. Although these trials described the procedure as a
placebo procedure, one implied that the additional postsurgi-
cal care may have had an effect41the other that “spontaneous
improvement in collateral circulation cannot be ruled out.”40
Thus these trials seem to imply that their “placebo” may not
have been inert. However, they have been discussed in the lit-
erature as though they were inert placebos.24,42In the 1990s
a trial of the surgical implantation of fetal tissue in the treat-
ment of Parkinson disease used a sham procedure in which a
hole was drilled in the skull and then the wound was closed
without transplantation of tissue.43Although this trial did not
refer to the sham surgery as a placebo procedure, an earlier
discussion of the trial44and later researchers45–47described it
It is generally held that surgery is associated with strong
placebo effects. Various reasons are offered: the absence of
regulatory demands of proof of efficacy,25the degree of
physical discomfort from the procedure,17and the meaning
attributed to the procedure and shedding of blood.24The last
two reasons touch on a key and usually ignored aspect of
any surgical intervention, including the sham. The human
body has protective and reparative mechanisms that are trig-
gered by injury. These are not placebo effects, they are di-
rect and expected by-products of the actual trauma of the
surgical incision. All surgical wounds provoke a cascade of
physiologic responses,48which not only repair the wound,
but also trigger a range of biologic effects that could affect
other (patho)physiologic processes in the body. Local and
systemic reparative mechanisms are triggered by surgical in-
cisions, including: vasoconstriction followed by clot forma-
tion, vasodilation, migration of white blood cells into the
damaged tissues, inflammatory reactions, fibroplasia, ep-
ithelialization, and wound contraction.48,49Additionally, the
body responds to traumatic painful injury with analgesic
mechanisms.50It cannot be ruled out that these effects may
affect problems near the site of the injury as well as other
parts of the body. If the biologic activities resulting from
surgical incisions have not been adequately investigated for
their potential therapeutic effects, they cannot be assumed
to be therapeutically inert, a necessary feature if sham stud-
ies are to be considered placebo controlled.
Additionally, one can speculate that other effects might
result from the loss or letting of blood that accompanies the
procedure. Bloodletting has been used as a medical proce-
dure over more than two thousand years in multiple cul-
tures.51–53It is used in modern medicine for treatment of
conditions such as hemochromatosis.54It also has been in-
dicated for a number of other problems,55–58and often is in-
dicated because it helps improve blood circulation.59–61It is
not known whether the loss of blood from the surgical in-
cision on the chest of angina patients40,41can be therapeu-
tic for angina, but bloodletting on the torso has been indi-
cated for the treatment of some cardiac diseases.61Blood
loss triggers sympathetic responses that cause decreased
blood flow in most parts of the body, but it can cause slight
vasodilation of the coronary arteries,62which could in prin-
ciple help trigger improved coronary artery blood flow. One
report indicates that bloodletting was specifically helpful for
angina pectoris.63Likewise, it is not known if the letting of
blood from incisions on the skull of patients with Parkinson
disease undergoing surgical implantation of fetal tissue is
therapeutic for Parkinson’s syndrome,43but bloodletting of
the head, neck, or upper torso has been indicated for a wide
range of problems, including some neurologic disorders,60
with one report indicating successful use for improvement
of Parkinsonian symptoms.64
The same problem is found when other sham surgical tri-
als are scrutinized. The sham and real surgical procedure for
osteoarthritis of the knee had similar effects,65whereas the
use of leeches for bleeding on the knee was more effective
than the usual treatment.66This suggests that the sham
PLACEBOS WHEN SHAM IS NOT INERT
surgery may have been an active therapeutic method rather
than a placebo treatment because it involved making inci-
sions that cause bleeding, and bleeding of the knee was
demonstrated effective for osteoarthritis of the knee. Re-
searchers conducting sham surgical trials have not consid-
ered the possibility of these effects; thus, they have not been
subjected to relevant investigations and have not been con-
However one interprets placebo responses, they will most
likely be inseparable from the biologic effects of the surgi-
cal incision (both in the sham and real procedures). When
placebo effects interact with or are inseparable from other
effects, placebo-controlled trials become difficult if not im-
possible,5,15as they cannot test the purported mechanism
(specific effects) of the therapy.13,32In other words, placebo-
controlled trials of surgical interventions may not be possi-
ble because there is no way to factor out or control for the
actual placebo effects.
General biologic effects of acupuncture
Acupuncture is an ancient therapy with a variety of dif-
ferent explanatory models in its clinical practice.67–70How-
ever, there are a few important principles that are generally
agreed upon and common to virtually all these models.
Treatment involves applying selected techniques to selected
points. Traditional and modern theories and diagnoses guide
and dictate the selection of which techniques at which
points.68,71Clinical trials testing the specific claims of
acupuncture generally have tried to focus on testing the ef-
ficacy of applying specific techniques and/or specific
points.72–74However, insertion of needles into the body can
stimulate other effects not dependent on the locations of
stimulation and are thus nonspecific. Regardless of where
the needles are placed, a range of nonspecific mechanisms
can be activated.37,74–80These include heterosegmental
analgesic mechanisms,50,81,82homosegmental analgesic ef-
fects,83microcirculatory effects,84and relaxation response
effects.78,85One can speculate a cascade of normal physio-
logic effects involved in prevention of infection and heal-
ing of damaged tissues after the insertion of any needles
anywhere on the body.49These effects occur in addition to
placebo effects. An international group of researchers dis-
cussing these effects recently concluded,
we cannot be satisfied that a truly inert intervention is
possible as the control treatment in acupuncture stud-
ies. Therefore the sham intervention cannot be con-
sidered equivalent to placebo in the same way that a
placebo pill is considered to be an inert intervention in
a placebo controlled pharmaceutical trial74
This is a point raised by others.11,79,80However, it has been
common for researchers to not separate these different
effects and lump them together with any placebo effects.74,75
It is also possible that the different treatment effects of
acupuncture may interact,74evidenced by the finding of the
endorphin involvement in specific needling of acupoints,86
heterosegmental analgesic nonspecific effects,81
placebo effects.87Such interactions probably make placebo-
controlled trials of acupuncture difficult if not impossi-
ble.5,15Further, it has been postulated that the principal pur-
pose of the traditionally based diagnoses and treatments in
acupuncture is to target a specific improvement in the in-
nate healing abilities of each patient, which may employ the
same mechanisms of action as those harnessed by placebo.74
There is growing evidence for the claim that placebo har-
nesses self-healing mechanisms.22Because the traditionally
based acupuncture treatment targets improvement of these
mechanisms, controlling for this in placebo-controlled trials
of acupuncture may thus require attempting to control for
the specific mechanisms and effects of the therapy being in-
vestigated, which contradicts the reasons for conducting
placebo-controlled trials.32,74Some of the nonspecific ef-
fects that have been documented with needling are involved
in the natural healing mechanisms of the body, which also
may employ some of the same pathways that placebo acti-
vates. These potential problems with sham acupuncture
treatments also need to be investigated further.
Different sham methods have been employed in sham tri-
als of acupuncture, including invasive and noninvasive sham
methods.74,75,88Although some researchers consider that in-
vasive sham techniques produce only placebo effects,45,72
there is general agreement that any invasive sham acupunc-
ture cannot be inert.74,79,88It is not clear how active the non-
invasive sham methods are. The question of whether they are
truly inert has been raised.74,75,79Putting aside questions about
the practical value of the noninvasive sham for research,75
any noninvasive sham acupuncture method at least must be
investigated to establish if it produces unintended physiologic
effects from touch, pressure, and so on, so that they can be
controlled for, or whether it is truly inert.79The recently de-
veloped noninvasive sham needle of Streitberger and Klein-
henz is considered to be a “placebo needle” by its inventors,89
but reports of its use clearly indicate otherwise. One re-
searcher in Germany caused bleeding in one patient as a re-
sult of the required mechanical stimulation of the noninva-
sive sham (Birgit Seybold, personal communication, 2003).
Recently various authors have suggested that acupunc-
ture is a complex intervention,15,16because for example, the
interview and discussion with patients are a necessary part
of treatment, and a number of different methods may be used
in treatment. Thus, it may not be possible to conduct
placebo-controlled studies of acupuncture because they run
the risk of mixing other specific and nonspecific effects with
any placebo effects in the sham group and thus generate
false-negative results.15Other research models need to be
investigated and developed for complex interventions.14–16
This argument parallels the preceding discussions and com-
pounds the problems identified.
Validity of controls in sham intervention trials
It has been argued that in sham studies in which the ther-
apist cannot be blinded, it is essential to assess the credi-
bility of the treatment and success of blinding,73,75,79,90
something that is curiously missing in sham surgical trials.
Methods have been developed to assess the blinding and
credibility of the treatment,79,90but these are not sufficient
to allow control of the physiologic nonspecific effects. Fur-
ther methods have been developed to attempt these addi-
tional controls,78,91but so far only a few trials have at-
tempted to make distinctions between placebo and control
for the physiologic “nonspecific” effects and placebo “non-
specific” effects.85,91,92Although the methodology74,75,78,93
has not yet been generally accepted, it is the only published
method developed to try to address these nonplacebo non-
specific physiologic effects. Trials that cannot or have made
no attempt to separate placebo from these other effects be-
come difficult to interpret. Reviewers may need to reclas-
sify many trials before they attempt to interpret their results.
Some reviews have attempted to analyze efficacy of
acupuncture compared to placebo,94–96which for the rea-
sons discussed here cannot be accurate. Such reviews need
to restate the comparison as acupuncture versus sham. Some
reviews have made the comparison to sham acupunc-
ture,39,97,98which is a more accurate comparison. However,
in both cases it needs to be explicitly stated that the sham
is an active treatment of unknown effectiveness. If in at-
tempting to investigate the efficacy of purported mecha-
nisms of acupuncture it is not possible or feasible to control
for all the nonspecific effects of needling, trials that inves-
tigate the effectiveness of acupuncture might do better to
use pragmatic models33together with laboratory methods to
investigate purported mechanisms.32
Attributing meaning: the need for change and
In sham trials of acupuncture researchers project their de-
finition of placebo effects onto patients and assume that ob-
served effects in the control group match the projected
placebo effects. Sham surgical trials can be seen to make
similar conceptual mistakes. Surgeons do not really believe
that the physiologic effects of making incisions and causing
blood loss can be therapeutic; rather, they believe that the
specific surgical procedure produces the therapeutic effects.
These assumptions lead them to conclude that their sham is
inert. This introduces bias against surgery and acupuncture
in these studies.
Clinical trials of medical procedures need to be re-ex-
amined and redesigned to remove this bias. “Studies in-
volving placebo effects must be designed to separate actual
placebo effects from various artefacts. These artefacts in-
clude investigator, observer, and patient bias, specific bio-
logical effects attributable to physical or chemical proper-
ties of the placebo” (emphasis added).99The biologic effects
of any intervention need to be understood in addition to any
projected therapeutic effects of that intervention. When these
unintentionally produce therapeutic effects in addition to the
projected therapeutic effects, trials must take this into ac-
count. The practice of calling noninert sham treatments
placebo in publications should stop.11
Early trials of cholesterol-lowering agents in heart dis-
ease used capsules with olive oil or corn oil as the placebo
treatment. However, later research showed that these two
oils are both capable of reducing low-density lipoprotein and
thus were inappropriate as placebo treatments.26It is now
clearly known that surgical and acupuncture sham proce-
dures are not inert; this must be dealt with in future trials
and interpretation of previous trials.
The assumption that a sham medical procedure is inert
when it is not introduces bias against the medical procedure
for which the sham is used as a control. It makes it difficult
to find significant differences between the sham and real
procedure. The common conclusion from trials of medical
procedures that the procedure was no better than placebo is
misleading to the reader and is inaccurate. In these studies
the conclusion is either wrong, or it is not possible to show
that it is correct, thus making many of these trials difficult
if not impossible to interpret and reviews of these trials open
to question. Trials of medical procedures should stop stat-
ing that they are “placebo-controlled” or that their sham pro-
cedure is a placebo procedure unless they can present evi-
dence that their sham is in fact inert and have validated it
as a placebo treatment. The practice of making a priori as-
sumptions about what are acceptable placebo effects in a
trial needs to be re-examined, because this has been re-
sponsible for introducing faulty assumptions about the na-
ture of placebo and bias against the therapy being investi-
gated. Often these a priori assumptions are based on
ignoring published literature about treatment effects and bi-
ologic mechanisms that are known to arise in response to
the proposed treatment methods.
Placebo-controlled surgical trials appear to be impossi-
ble. Placebo-controlled acupuncture trials cannot use inva-
sive sham procedures. If a noninvasive sham procedure is
used in an acupuncture trial, the sham must be investigated
for physiologic effects to establish that it is in fact inert, and
more complex and complete control procedures and meth-
ods must be used to separately control for other nonspecific
physiologic effects. Additionally, the conceptual and theo-
retical basis of acupuncture practice needs to be better con-
sidered, to ensure that the natural healing processes that are
the target of therapy are not confused with placebo, and are
“controlled” for in sham trials. This is an extremely com-
plex and difficult area in which few studies have attempted
PLACEBOS WHEN SHAM IS NOT INERT
comprehensive controls and it remains to be seen how prac-
tical, acceptable, and feasible proposed methods are for do-
Thanks to Jackie C. Wootton, M.Ed., for her helpful com-
ments on the manuscript.
1. Guess HA, Kleinman A, Kusek JW, Engel LW, eds. The Sci-
ence of the Placebo, London: BMJ Books, 2002.
2. Harrington A, ed. The Placebo Effect. Cambridge, MA: Har-
vard University Press, 1997.
3. Peters D, ed. Understanding the Placebo Effect in Comple-
mentary Medicine. Edinburgh: Churchill Livingstone, 2001.
4. Hrobjartsson A, Gotzsche P. Is the placebo powerless? An
analysis of clinical trials comparing placebo with no treatment.
5. Kaptchuk TJ, Edwards RA, Eisenberg DM. Complementary
medicine: Efficacy beyond the placebo effect. In: Ernst E, ed.
Complemetary Medicine: An Objective Appraisal. Oxford,
UK: Butterworth Heinmann, 1996:42–70.
6. Kienle GS, Kiene H. Placebo effect and placebo concept: A
critical methodological and conceptual analysis of reports on
the magnitude of the placebo effect. Altern Ther Health Med
7. Kleijnen J, de Craen AJM. The importance of the placebo ef-
fect: A proposal for further research. In: Ernst E, ed. Com-
plementary Medicine: An objective appraisal. Oxford, UK:
Butterworth Heinmann, 1996:31–41.
8. Moerman DE, Jonas WB. Deconstructing the placebo effect
and finding the meaning response. Ann Int Med 2002;136:
9. Kaptchuk TJ, Goldman P, Stone DA, Stason WB. Do medical
devices have enhanced placebo effects? J Clin Epidemiol
10. De Craen AJM, Lampe-Schoenmaeckers AJEM, Kleijnen J.
Non-specific factors in randomized clinical trials: Some
methodological considerations. In: D. Peters, ed. Understand-
ing the Placebo Effect in Complementary Medicine. Edin-
burgh: Churchill Livingstone, 2001:179–187.
11. Linde K, Dincer F. How informed is consent in sham-con-
trolled trials of acupuncture? J Altern Complement Med
12. Flaten MA, Simonsen T, Olsen H. Drug-related information
generates placebo and nocebo responses that modify the drug
response. Psychosom Med 1999;61(2):250–255.
13. Kleijnen J, de Craen AJM, van Everdingen J, Krol L. Placebo
effect in double blind clinical trials: A review of interactions
with medications. Lancet 1994;ii:1347–1349.
14. Medical Research Council. A framework for development and
evaluation of RCTs for complex interventions to improve
health. Online document at: http://www.mrc.ac.uk/pru/pdf-
mrc_cpr.pdf Posted in 2000. Accessed October 5, 2005.
15. Paterson C, Dieppe P. Characteristic and incidental (placebo)
effects in complex interventions such as acupuncture. BMJ
16. Verhoef MJ, Lewith G, Ritenbaugh C, et al. Complementary
and alternative medicine whole systems research: Beyond
identification of inadequacies of the RCT. Complement Ther
17. Ernst E. Towards a scientific understanding of placebo effects.
In: D. Peters, ed. Understanding the Placebo Effect in Com-
plementary Medicine. Edinburgh: Churchill Livingstone,
18. Gotzsche P. Is there logic in the placebo? Lancet 1994;344:
19. Harrington A. Seeing the placebo effect: Historical legacies
and present opportunities. In: Guess HA, Kleinman A, Kusek
JW, Engel LW, eds. The Science of the Placebo. Towards an
Interdisciplinary Research Agenda. London: BMJ Books,
20. Richardson PH. Placebo effects in pain management. Pain Rev
21. Roberts AH. The powerful placebo revisited: Magnitude of
nonspecific effects. Mind/Body Med 1995;1(1):35–43.
22. Walach H, Jonas WB. (2004). Placebo research: The evidence
base for harnessing self-healing capabilities. J Alt Compl Med
23. Meinert CL. Clinical Trials: Design, Conduct and Analysis.
Oxford, UK: Oxford University Press, 1986.
24. Moerman DE. Explanatory mechanisms for placebo effects:
Cultural influences and the meaning response. In: Guess HA,
Kleinman A, Kusek JW, Engel LW, eds. The Science of the
Placebo. Towards an Interdisciplinary Research Agenda. Lon-
don: BMJ Books, 2002:77–107.
25. Shapiro AK, Shapiro E. The placebo: Much ado about noth-
ing? In: Harrington A, ed. The Placebo Effect. Cambridge MA:
Harvard University Press, 1997:12–36.
26. De Craen AJM, Tijssen JGM, Kleijnen J. Is there a need to
control the placebo in placebo controlled trials? Heart
27. Wyon Y, Lindgren R, Lundeberg T, Hammar M. Effects of
acupuncture on climacteric vasomotor symptoms, quality of
life, and urinary excretion of neuropeptides among post-
menopausal women. Menopause 1995;2:3–12.
28. Birch S. Issues to consider in determining an adequate treat-
ment in a clinical trial of acupuncture. Complement Ther Med
29. Ernst E, Resch KL. Concept of true and perceived placebo ef-
fects. BMJ 1995;311:551–553.
30. Resch KL, Ernst E. Research methodologies in complemen-
tary medicine: Making sure it works. In: Ernst E, ed. Com-
plemetary Medicine: An Objective Appraisal. Oxford UK:
Butterworth Heinmann, 1996:18–30.
31. Vickers AJ. Placebo controls in randomized trials of acupunc-
ture. Eval Health Profess 2002;25(4):421–435.
32. Hyland ME. Methodology for the scientific evaluation of com-
plementary and alternative medicine. Complem Ther Med
33. Thomas K, Fitter M. Possible research strategies for evaluat-
ing CAM interventions. In: G Lewith, WB Jonas, H Walach,
eds. Clinical Research in Complementary Therapies. Edin-
burgh: Churchill Livingstone, 2002:59–91.
34. Linde K, Vickers A, Hondras M, et al. Systematic reviews of
complementary therapies—an annotated bibliography. Part 1:
Acupuncture. BMC Complem Alt Med 2001;1:3.
35. Walach H. The efficacy paradox in randomized controlled tri-
als of CAM and elsewhere: Beware of the placebo trap. J Alt
Compl Med 2001;7(3):213–218.
36. Davis CE. Regression to the mean or placebo effect? In: Guess
HA, Kleinman A, Kusek JW, Engel LW, eds. The Science of
the Placebo. Towards an Interdisciplinary Research Agenda.
London: BMJ Books, 2002:158–166.
37. Lewith GT, Machin D. On the evaluation of the clinical ef-
fects of acupuncture. Pain 1983;16:111–127.
38. Lytle CD. An overview of acupuncture: U.S. Department of
Health and Human Services, Public Health Service, Food and
Drug Administration, Center for Devices and Radiological
39. Melchart D, Linde K, Fischer P, et al. Acupuncture for recur-
rent headaches: A systematic review of randomized controlled
trials. Cephalalgia 1999;19:779–786.
40. Cobb LA, Thomas GI, Dillard DH, et al. An evaluation of in-
ternal-mammary-artery ligation by a double-blind technique.
41. Dimond GE, Kittle CF, Crockett JE. Comparison of internal
mammary artery ligation and sham operation for angina pec-
toris. Amer J Cardiol 1960; April:483–486.
42. Benson H, McCallie DP. Angina pectoris and the placebo ef-
fect. NEJM, 1979;300(25):1424–1429.
43. Freed CR, Greene PE, Breeze RE, et al. Transplantation of
embryonic neurons for severe Parkinson’s disease. NEJM.
44. Freed CR, Breeze RE, Fahn S. Placebo surgery in trials of ther-
apy for Parkinson’s disease. NEJM, 2000;342:353–355.
45. Kaptchuk TJ. The placebo effect in alternative medicine: Can
the performance of a healing ritual have clinical significance?
Ann Intern Med 2002;136:817–825.
46. Tenery R, Rakatansky H, Riddick FA, et al. Surgical “placebo”
controls. Ann Surg 2002:235(2):303–307.
47. Verhaar JA. Placebo surgery in randomized research: Possi-
ble and ethically justified in some cases. Ned Tijdschr Ge-
48. Bryant WM. Wound healing: Clinical symposia. CIBA 1977;
49. Oschman JL. Energy Medicine in Therapeutics and Human
Performance. Edinburgh: Churchill Livingstone, 2003.
50. Le Bars D, Willer JC, de Broucker T, Villanueva L.
Neurophysiological mechanisms involved in the pain-re-
lieving effects of counter-irritation and related techniques
including acupuncture. In: Pomeranz B, Stux G, eds. Sci-
entific Bases of Acupuncture. Berlin: Springer Verlag,
51. Epler DC. Bloodletting in early Chinese medicine and its re-
lationship to the origin of acupuncture. Bull Hist Med
52. Unschuld PU. Huang Di Nei Jing Su Wen: Nature, Knowl-
edge, Imagery in an ancient Chinese Medical text. Berkeley:
University of California Press, 2003.
53. Watts S. Disease and Medicine in World History. New York:
54. Silbernagl S, Lang F. Color Atlas of Pathophysiology.
Stuttgart: Thieme, 2000.
55. Girelli CM, Mirata C, Casiraghi A. Effect of blood letting on
serum aminotransferase levels of patients with chronic he-
patitis C and iron overload. Rec Prog Med 1998;89:241–244.
56. Kumar H. Repeated blood donation effective in treating hy-
perlipidemia. J Assoc Physicians India 1994;42:468–469.
57. Rombos Y, Tzanetea R, Kalotychou V, et al. Amelerioration
of painful crises in sickle cell disease by venesections. Blood
Cells Mol Dis 2002;28:283–287.
58. Sadahiro T, Endoh H. Continuous blood-letting for congestion
in replantation of the amputated finger. J Hand Surg 1984;
59. Hartig GK, Connor NP, Heisey DM, Conforti ML. Compar-
ing a mechanical device with medicinal leeches for treating
venous congestion. Otolaryngol Head Neck Surg 2003;129(5):
60. Kudo K. Shirakuho: Micropunction, 3rd ed. Yokosuka, Japan:
Ido no Nippon Publishing Company, 1963.
61. Shiroda F (ed), Asakawa K (translator), Wang FY, Ren HZ.
(authors). Kyugyoku Ryoho: Cupping Therapy. Ichikawa,
Japan: Toyo Gakujutsu Publishing Company, 1985.
62. Guyton AC. Textbook of Medical Physiology, 5th ed. Philadel-
phia: WB Saunders, 1976.
63. Piccirillo G, Fimognari FL, Valdivia JL, Marigliano V. Effects
of phlebotomy on a patient with secondary polycythemia and
angina pectoris. Int J Cardiol. 1994;44:175–177.
64. Birch S. A case of postpartum complications? In: MacPher-
son H, Kaptchuk TJ, eds. Acupuncture in Practice. Edinburgh:
Churchill Livingstone, 1997:125–138.
65. Moseley JB, Wray NP, Kuykendall D, et al. Arthroscopic treat-
ment of osteoarthritis of the knee: A prospective randomised,
placebo-controlled trial. Results of a pilot study. Am J Sports
66. Michaelsen A, Klotz S, Ludtke R, et al. Effectiveness of leech
therapy in osteoarthritis of the knee: A randomised controlled
trial. Ann Int Med 2003;139:724–730.
67. Birch S. Diversity and acupuncture: Acupuncture is not a co-
herent or historically stable tradition. In: Vickers AJ, ed. Ex-
amining Complementary Medicine: The Sceptical Holist.
Cheltenham, UK Stanley Thomas, 1998:45–63.
68. Birch S, Felt RO. Understanding Acupuncture. Edinburgh:
Churchill Livingstone, 1999.
69. Birch S, Kaptchuk T. The history, nature and current practice
of acupuncture: An East Asian perspective. In: Ernst E, White
A, eds. Acupuncture: A Scientific Appraisal. Oxford UK: But-
70. Macpherson H, Kaptchuk TJ. Acupuncture in Practice. New
York: Churchill Livingstone, 1997.
71. Birch S. Testing the claims of traditionally based acupuncture.
Complem Ther Med 1997;5:147–151.
72. Mayer DJ. Acupuncture: An evidence-based review of the clin-
ical trial literature. Annu Rev Med 2000;51:49–63.
73. Vincent C, Lewith G. Placebo controls for acupuncture stud-
ies. J Roy Soc Med 1995;88:199–202.
74. Birch S, Hammerschlag R, Trinh K, Zaslawski C. The non-
specific effects of acupuncture treatment: When and how to
control for them. Clin Acup Orien Med 2002;3:20–25.
75. Birch S. Controlling for non-specific effects of acupuncture in
clinical trials. Clin Acup Orien Med 2003;4:59–70.
76. Lao L, Ezzo J, Berman BM, Hammerschlag R. Assessing clin-
ical efficacy of acupuncture: Considerations for designing fu-
PLACEBOS WHEN SHAM IS NOT INERT
ture acupuncture trials. In: Stux G, Hammerschlag R, eds. Clin- Download full-text
ical Acupuncture: Scientific Basis. Berlin: Springer Verlag,
77. Linde K, Streng A, Jurgens S, et al. Acupuncture for patients
with migraine: A randomized controlled trial. JAMA 2005;
78. Margolin A, Avants SK, Kleber HD. Rationale and design of
the cocaine alternative treatments study (CATS): A random-
ized, controlled trial of acupuncture. J Alt Complem Med
79. White AR. Acupuncture research methodology. In: G Lewith,
WB Jonas, H Walach, eds. Clinical Research in Complemen-
tary Therapies. Edinburgh: Churchill Livingstone, 2002:307–
80. White P, Lewith G, Berman B, Birch S. Reviews of acupunc-
ture for chronic neck pain: Pitfalls in conducting systematic
reviews. Rheumatology 2002;41:1224–1231.
81. Bing Z, Cesselin F, Bourgoin S, et al. Acupuncture-like stim-
ulation induces a heterosegmental release of Met-enkephalin-
like material in the rat spinal cord. Pain 1991;47:71–77.
82. Murase K, Kawakita K. Diffuse noxious inhibitory controls in
anti-nociception produced by acupuncture and moxibustion on
trigeminal caudalis neurons in rats. Jpn J Physiol 2000;50:
83. Melzack R. Acupuncture and related forms of folk medicine.
In: Melzack R, Wall PD, eds. Textbook of pain. London:
Churchill Livingstone, 1984:691–700.
84. Itaya K, Manaka Y, Ohkubo C, Asano M. Effects of acupunc-
ture needle application upon cutaneous microcirculation of
rabbit ear lobe. Acup Electrother Res Int J 1987;12:45–51.
85. Avants SK, Margolin A, Holford TR, Kosten TR . A ran-
domized controlled trial of auricular acupuncture for cocaine
dependence. Arch Int Med 2000;160:2305–2312.
86. Pomeranz B. Scientific research into acupuncture for the re-
lief of pain. J Alt Complem Med 1996;2:53–60.
87. Price D, Soerensen LV. Endogenous opioid and non-opioid
pathways as mediators of placebo analgesia. In: Guess HA,
Kleinman A, Kusek JW, Engel LW, eds. The Science of the
Placebo. Towards an Interdisciplinary Research Agenda. Lon-
don: BMJ Books, 2002:183–206.
88. Hammerschlag R. Methodological and ethical issues in clini-
cal trials of acupuncture. J Alt Complem Med 1998;4:159–171.
89. Streitberger K, Kleinhenz J. Introducing a placebo needle into
acupuncture research. Lancet 1998;352:364–365.
90. Vincent CA. Credibility assessment in trials of acupuncture.
Complem Med Res 1990;4:8–11.
91. Birch S, Jamison RN. A controlled trial of Japanese acupuncture
for chronic myofascial neck pain: Assessment of specific and
nonspecific effects of treatment. Clin J Pain 1998;14:248–255.
92. Margolin A, Kleber HD, Avants SK, et al. Acupuncture for
the treatment of cocaine addiction. A randomized controlled
trial. JAMA 2002;287:55–63.
93. Birch S. Clinical research of acupuncture: Part two. Controlled
clinical trials an overview of their methods. J Alt Complem
94. Ezzo J, Berman B, Hadhazy VA, et al. Is acupuncture effec-
tive for the treatment of chronic pain: A systematic review.
95. Linde K, Worku F, Stor W, et al. Randomized clinical trials
of acupuncture for asthma—a systematic review. Forsch
96. van Tulder MW, Cherkin DC, Berman B, et al. The effec-
tiveness of acupuncture in the management of acute and
chronic low back pain: A systematic review within the frame-
work of the Cochrane Collaboration Back Review Group.
97. Ernst E, White AR. Acupuncture as a treatment for temporo-
mandibular joint dysfunction: A systematic review of ran-
domized trials. Arch Otolar Head Neck Surg 1999;125:
98. White AR, Resch KL, Ernst E. A meta-analysis of acupunc-
ture techniques for smoking cessation. Tobacco Control
99. Wilentz JS, Engel LW. The research and ethical agenda. In:
Guess HA, Kleinman A, Kusek JW, Engel LW eds. The Sci-
ence of the Placebo. Towards an Interdisciplinary Research
Agenda. London: BMJ Books, 2002:283–285.
Send reprint address requests to:
Stephen Birch, Ph.D., Lic.Ac. (US) M.B.Ac.C. (UK)
Stichting (Foundation) for the Study of Traditional East
Asian Medicine (STEAM)
W.G. Plein 330
1054 SG Amsterdam