[tt] CHE: Placebo Effect: Why we need less research on alternative medicine, not more

[Premise Checker]

Placebo Effect: Why we need less research on alternative medicine, not more
The Chronicle of Higher Education  The Chronicle Review
http://chronicle.com/weekly/v54/i27/27b01201.htm
From the issue dated March 14, 2008

By R. BARKER BAUSELL

In 1993 The New England Journal of Medicine published the results of a 
survey, conducted by David M. Eisenberg and his colleagues, estimating 
that tens of millions of Americans were frequenting an astonishing variety 
of bizarre, pseudomedical practitioners. Five years later, The Journal of 
the American Medical Association published a follow-up survey by the same 
group, concluding that visits to alternative practitioners were exceeding 
visits to primary-care physicians in the United States.

Both surveys were heavily cited in the medical literature, extensively 
covered in the popular media, and used to justify budget increases for the 
National Institutes of Health's nascent Office of Alternative Medicine and 
eventually for its successor, the National Center for Complementary and 
Alternative Medicine.

From a historical perspective, however, there was really nothing 
newsworthy about the public's infatuation with pseudomedical nonsense. In 
1784, for example, Benjamin Franklin, at the behest of the king of France, 
lent his considerable scientific expertise to a series of experiments that 
definitively debunked an animal-magnetism therapy then taking Paris by 
storm.

But while Franklin's experiments may have temporarily inconvenienced the 
therapy's inventor, Franz Anton Mesmer, they had little lasting effect on 
his reputation. His name entered our language as a verb — to mesmerize — 
with no pejorative implications, and he is currently recognized as the 
father of hypnosis, a therapy used in modern CAM, as complementary and 
alternative medicine is sometimes known.

Nothing has changed in recent years. In fact, the supplement and herbal 
industries had become so powerful by 1992 that their supporters were 
reputed to have sent an estimated two million letters to members of 
Congress, urging them to prevent the Food and Drug Administration from 
imposing restrictions on unsubstantiated health claims made on behalf of 
supplements and herbal products — restrictions that might have prevented 
the significant loss of life and thousands of adverse effects attributed 
to the herb ephedra, which was marketed as a weight-loss and 
athletic-performance supplement before it was finally banned a few years 
ago.

What is new is the role of reputable institutions in propagating such 
nonsense: The NIH now awards both research and educational grants in 
alternative medicine. According to the Web site of the National Center for 
Complementary and Alternative Medicine, "the immediate goal" of the latter 
program is "to encourage and support the incorporation of [complementary 
and alternative medicine] information into medical, dental, nursing, and 
allied health professional school curricula, into residency training 
programs, and into continuing education courses. An important longer-term 
goal [is] to accelerate the integration of CAM and conventional medicine." 
Many medical schools now work with complementary and alternative clinics, 
offer fellowships in alternative medicine, and both include material on 
alternative medicine in required courses and offer optional courses about 
it.

Fortunately, the National Center for Complementary and Alternative 
Medicine appears to be considerably more focused on scientific endeavors 
than its predecessor was, and — in addition to the curricular forays 
mentioned above — it has provided money for a number of high-quality 
clinical trials comparing selected alternative therapies to placebos. Not 
surprisingly, almost all those trials found no difference between the two, 
but that information has not come cheaply: The NIH has allocated 
approximately $860-million to the center since the 2000 fiscal year alone.

That leads naturally to a number of questions, including: Did we really 
need to spend hundreds of millions of dollars on full-fledged randomized 
and controlled trials to ascertain if thin needles inserted in the body 
can channel the flow of chi (an energy source as yet undetected by 
physicists) through a bodily system of meridians (as yet undetected by 
physiologists) to bring the patient's yin and yang into balance (as yet 
unmeasured) to ameliorate syndromes (which I doubt even the Chinese really 
understand) such as wind-damp-hot bi, and thus relieve patients' 
self-reported symptoms? Wouldn't it be wiser to allocate that money to 
other NIH agencies whose mission is to try to find an actual cure for a 
real disease, or to try to find better ways to manage patients' symptoms 
for those conditions that can't be cured? Or, if that is not politically 
feasible, in the spirit of transparency shouldn't the alternative-medicine 
center be renamed the National Center for the Study of Placebo Effects? 
And should we really keep this nonsense as part of our medical-school 
curricula?

Scientists, alas, are far less comfortable with multifaceted policy issues 
than they are with conducting ever-more clinical trials: one therapy for 
one medical condition at a time. But while that arduous and expensive 
approach may constitute a perpetual growth industry for both the National 
Center for Complementary and Alternative Medicine and medical schools, I 
suspect it would not move us any closer to any scientific conclusions.

And although I would not try to dictate what academics should and should 
not be allowed to teach, or what they might and might not study, I do 
believe that alternatives exist to compiling more and more evidence. In 
conventional scientific practice, the formulation of a plausible 
biological theory normally precedes its experimental evaluation. That is 
not possible with complementary and alternative medicine, however, because 
almost by definition such therapies do not have conventional physiological 
explanations. Instead they are experimentally evaluated first, and only 
when someone makes a positive finding — which is extremely rare in 
well-controlled clinical trials of the therapies — is the plausibility of 
the underlying theory examined to try to explain why such a result should 
have occurred.

Of course, a biologist would have little trouble passing judgment on the 
plausibility of any given complementary or alternative therapy's proposed 
mechanism of action. And lest anyone think that the acupuncture example 
above is an anomaly, consider homeopathy, the alternative therapy of 
choice in France that is also used by many Americans. The theory is that 
substances that elicit a particular symptom make excellent candidates for 
treating it — based on the principle of "like cures like" — as long as 
they are diluted to the extent that not a single molecule of the original 
substances (e.g., cuttlefish discharge or poison oak) is likely to remain 
in the solution. The substances, the therapy's advocates insist, retain or 
increase their effectiveness because the specialized homeopathic dilution 
process allows the water to "remember" them.

However, a biostatistician or research methodologist would have a much 
more difficult time synthesizing and evaluating the experimental evidence 
related to the efficacy of complementary and alternative medicine in 
general. That is because hundreds of individual therapies are in use, and 
thousands of trials — almost all methodologically suspect — have been 
conducted on them, often by advocates of alternative medicine.

Fortunately, impressive resources exist to make evaluation easier. First, 
we have a clear consensus on what does and does not constitute 
well-designed and well-executed evaluations of medical efficacy. Second, 
more than 3,000 high-quality systematic reviews have been completed by the 
Cochrane Collaboration, a research organization dedicated to locating, 
appraising, and synthesizing evidence regarding the effectiveness of 
specific medical interventions of all types, conventional and alternative.

A few years ago, I set out to test the very broad hypothesis that the 
therapeutic effects attributed to complementary and alternative medicine 
were basically nothing more than placebo and placebo-like effects. In the 
latter category, I included such experimental artifacts as natural 
history, like the fact that many symptoms wax and wane over time, or even 
resolve themselves completely without intervention; 
regression-to-the-mean, which can ensure that a symptom will be less 
severe after even an ineffectual intervention, if the intervention 
occurred when the symptom was at its worst; and experimental bias, such as 
failing to conceal from therapists, data collectors, and the participants 
themselves whether a participant was receiving a placebo or an 
intervention.

I used this logic: If there is a plausible biological explanation for why 
the placebo effect should occur, and credible experimental evidence that 
it does occur; but there is no plausible biological explanation for why 
any therapeutic effect from complementary or alternative medicine should 
exist, nor any credible experimental evidence that one does, which cannot 
also be attributed to the placebo effect; then a reasonable scientific 
conclusion would be that complementary and alternative therapies are 
nothing more than placebos.

Analyzing the evidence for the placebo effect, especially with respect to 
pain relief, was not difficult. We have known for decades, for example, 
that the mere expectation of relief can trigger the brain's release of 
endogenous opioids, which in turn cause actual relief. That has been 
demonstrated repeatedly, with both healthy subjects and patients 
recovering from surgery. The placebo effect is a most remarkable 
phenomenon, based on expectations resulting from past experiences and 
triggered by simple conditioning.

Although the process of evaluating the effectiveness of complementary and 
alternative therapies was much more arduous, I was able to employ a 
two-pronged approach that made the task more manageable. First, I looked 
only at high-quality trials that controlled for the placebo effect, and 
that had been published since January 2000 in the four American medical 
journals not limited to a specific disease that had the highest citation 
rates. Those journals employ more-stringent peer-review systems than other 
publications, and other criteria I used about sample sizes and attrition 
rates for subjects ensured the integrity of the results. And second, I 
examined all the systematic reviews from the Cochrane Collaboration that 
included two or more trials evaluating the effectiveness of complementary 
or alternative medicine.

The results were quite revealing. Only one of the trials published in the 
four selected journals in more than seven years demonstrated an effect of 
alternative medicine beyond the placebo effect — and that trial was 
sponsored and conducted, and its publication written, by employees of the 
company that marketed the product in question. The Cochrane reviews were 
almost as definitely negative.

Because there is little doubt that a placebo effect does exist, but little 
evidence to suggest that any complementary or alternative therapy is more 
effective than a placebo, I believe I have provided a logical answer — 
supported by empirical evidence — to the question of whether or not we 
really need to continue spending in excess of $100-million a year on 
alternative-medicine research to prove the obvious. My conclusion is that 
sometimes we do actually reach a point at which more research is not 
needed, and we have reached it with alternative and complementary 
therapies. Sometimes, in fact, certain types of research need never have 
been conducted in the first place.

I realize that if even Benjamin Franklin couldn't stem the ever-rising 
tide of alternative medicine, no one can. I don't even believe that the 
attempt to do so is a proper role for an educator. Ignorance, along with 
the Mesmers of the world, will be with us forever. I believe it is enough 
for us, as educators, to teach our students to recognize (and avoid) the 
impediments that prevent us from understanding the often complex links 
between cause and effect.

If nothing else, that topic would prove an appropriate complement to 
complementary and alternative medicine in our curricula. Perhaps it would 
even be an antidote.

R. Barker Bausell is a professor of research methodology in the School of 
Nursing at the University of Maryland at Baltimore, and director of 
evaluation in its Office of Research. He is author of Snake Oil Science: 
The Truth About Complementary and Alternative Medicine, published last 
fall by Oxford University Press.