13 Things That Don't Make Sense (26 page)

The other big problem was that Shang’s study involved pooling data from trials measuring different effects: different kinds
of homeopathic treatments for different kinds of ailments with different kinds of outcomes—pain reduction, resolution of an
infection, reduction of inflammation, and so on. That is OK if homeopathy really is nothing more than a placebo, because all
trials are, effectively, measuring one kind of response. Linde and Jonas’s 1997 study had pooled data on this assumption.
But since then several studies had found some effect above placebo in trials of homeopathy in specific cases. If those studies
have any truth to their results, Shang’s pooling of results invalidates the entire analysis; it skews the statistics, giving
a significant risk of producing a false negative.

Lastly, once Shang and colleagues had whittled down the trials they decided were worthy of attention, this meta-analysis ended
up studying only eight clinical trials of homeopathy. With such a small pool, the outcome “could easily be due to chance,”
Linde and Jonas said. And that means their suggestion that they had proved the clinical effects of homeopathy are placebo
is “a significant overstatement.”

In 1997 Linde and Jonas had concluded that the results of their own study made it impossible to claim that the effects of
homeopathy are completely due to placebo. It was hardly a ringing endorsement, but neither was it a nail in homeopathy’s coffin.
In fact, they lamented in their letter to the
Lancet
that their own study had been “misused by homeopaths as evidence that their therapy is proven.” The
Lancet
, they said, was misusing the new study in a similar way. “The
Lancet
should be embarrassed by the Editorial that accompanied the study,” Linde and Jonas said. “A subversive philosophy serves
neither science nor patients.”

Strong words, especially since they came from nonbelievers. But then Jonas had been encountering his own frustrations with
homeopathy. A couple of months later, in October 2005, he published a paper with Harald Walach in the
Journal of Alternative and Complementary Medicine
. It is a balanced review, admitting that there are “some hints” that diluted and succussed homeopathic substances are biologically
active—but that there is “no single clinical area where reported effects have been demonstrated unequivocally.” Meta-analysis,
they say, just doesn’t help; whether homeopathy comes out as a placebo or not depends on the way a study is done. Overall,
the problem with analyzing homeopathy’s credentials “is not in finding a stunning initial result . . . the real problem is
replicating the effect once it has been seen.” In other words, they too had failed to prove homeopathy’s inefficacy. Yet again.

This all seems implausible. Given more than two centuries, science has failed to show that homeopathy is bunkum. How is this
possible? How can we put this issue to rest? The answer may lie in the pages of the
Homeopathic Repertory,
the sprawling catalog of symptoms, remedies, and appropriate dilutions consulted by homeopaths before offering a prescription.
The clinical trials that have tested homeopathy’s efficacy pick out some homeopathic remedies and use them to treat ailments
such as the inflammation caused by rheumatoid arthritis. A six-month study carried out by the director of research at London’s
Royal Homeopathic Hospital gave a negative result for forty-two homeopathic remedies in this case. But what if some of the
remedies being used are in fact effective? Could it be that concentrating on just a few of the myriad available remedies would
cause the results of trials to come out as significantly better than placebo?

It would certainly make sense of the gulf between the generally unimpressive trial results and the anecdotal claims that sane
individuals make of homeopathy’s successes. Lionel Milgrom, a chemist at Imperial College London, for example, trained as
a homeopath because he was so impressed at how quickly and conclusively a homeopathic remedy cured his partner’s recurrent
pneumonia. Another acquaintance, an author of science books and a professional science communicator, once told me he watched
in amazement when
Apis mel
, a homeopathic remedy made from a honeybee drowned in alcohol, deflated his two-year-old daughter’s swollen tongue after
she had been stung by a bee.

It might be that all these miracle stories could be properly balanced by reports of incidents where homeopathy fails to have
an effect. It’s called publication bias in the pharmaceutical trade; people seldom bother to announce exclusively negative
results. But here’s the rub: Could it be that the haphazard nature of homeopathic diagnosis and prescription—and the flakey,
unproven (and yet unquestioned) status of some of its remedies—is occluding a truly impressive phenomenon?

Bob Lawrence certainly thinks so, and he’s on a mission to prove it. Lawrence is another convert; he had a fifteen-year skin
condition cleared up by a homeopathic remedy. Antibiotics also cleared it up, he says, but the side effects were too horrible
to live with. When a friend recommended a homeopathic treatment, he took it with a large dose of skepticism, but he hasn’t
looked back. He subsequently gave up a perfectly good job in engineering to train as a homeopath and is now a pharmacist at
one of Britain’s largest homeopathic dispensaries, the Helios Homeopathic Pharmacy in the English spa town of Tunbridge Wells.
Take a tour of this place with Lawrence, and you’ll encounter everything that’s wrong—and right—with homeopathy in the twenty-first
century.

I
was expecting something a little more disconcerting, something a little more
Romeo and Juliet
. Something like an apothecary’s lair. Instead there’s a brightly lit shop, a service counter, and behind that, a load of
very normal-looking people. They were bustling about in white coats, sliding boxes off shelves, then opening them up and pulling
out tiny vials from which they dripped liquids into other tiny vials.

There were three disconcerting things about the scene, though. The first was the strange names typewritten onto the box labels—one
of them said “Lava,” for instance. Also troubling was the occasional violent banging noise as a pharmacist succussed a vial’s
contents. The third oddity was the surface on which Lawrence was doing his own succussion. It was a huge, black, leather-bound
King James Bible.

Having done three raps on the Bible, his fist clenched around a vial containing a homeopathic remedy made from amethyst, Lawrence
looked up. His face said, “I wish you hadn’t seen that.” You don’t have to use a Bible, he assured me. What you need, apparently,
is a flexible but hard surface. It was Hahnemann, the founder of homeopathy, who suggested a leather-bound Bible might be
the most suitable tool.

Hahnemann is followed enthusiastically at Helios. The pharmacy is a kind of central clearinghouse for homeopathic remedies;
the staff are expert at the kind of dilution, succussion, and redilution that lies at the heart of homeopathy, and they get
sent substances to “potentize” in this way from all over the world. The pharmacy’s Bible has obviously seen a lot of action;
its covers are now held on with rubber bands.

Lawrence is no mystic, though; he is not a pathological believer who thinks the Bible will convey some special force to the
remedy. This was confirmed when he took me downstairs to show off the machines he built to do the most laborious succussions
and dilutions. Sometimes, for an ultrapotent remedy, you have to repeat the process thousands of times; Lawrence has used
his engineering skills to automate the process. He wants to put homeopathy on a more scientific footing, he says. Sometimes
people will send him a bit of bat to potentize. Or the wing of a cicada. He won’t turn it into a remedy until he knows exactly
what species it came from; he wants the Latin name. He would dearly love to come at the
Homeopathic Repertory
, the sprawling catalog of symptoms, remedies, and appropriate dilutions, with a scientific ax, thinning it down to include
only what’s provably effective.

As we struggled to talk above the repetitive knocking of the machines, I noticed some more boxes. I could feel Lawrence willing
me not to see the names on the labels, but his will is evidently not as powerful as that of a Chinese Qigong grand master.
The names that stood out were “F Sharp Minor,” “G Major Chord,” “Crop Circle,” and “Flapjack.” When I asked Lawrence about
them—how you capture F Sharp Minor in a bottle, for instance—he raised his eyebrows and rolled his eyes. Looking around again,
I spied a box marked “Gog and Magog, Oaks at Glastonbury.” Another said “Frog Spawn.” Down here, in the basement, we had bypassed
Romeo and Juliet
and gone straight to
Macbeth
.

It’s easy, in Helios’s basement, to see what’s wrong with homeopathy. It has, largely, become the preserve of people who want
to believe in the healing power of anything and everything “natural.” The range of homeopathic remedies is so vast, so all-encompassing,
as to make it virtually impossible to test homeopathy’s claims.

A homeopathic remedy is supposed to go through a system of checks known as a
proving
. The original substance is given to a group of volunteers, who note down any strangeness, any symptoms of anything that they
experience over the next few weeks. These symptoms are compiled and compared, and the ones that seem universal are then associated
with the substance. If a patient in a homeopathic consultation reports anything like those symptoms, the principle of homeopathy—literally,
“similar suffering”—means that a remedy made from the substance in question might make a useful treatment.

The trouble is, many of the medicines in the Helios pharmacy have not had anything like proper provings—and are obvious examples
of quackery. There are remedies on the shelves at Helios that have been made from condoms, pieces of lava, the blood from
an HIV-positive man, and even the whiff of antimatter.

What’s right about homeopathy, what it has going for it, is that someone like Lawrence is genuinely frustrated by this situation.
I could see the embarrassment in Lawrence’s eyes when I mentioned the musical remedies, and I feel a genuine empathy for his
predicament. He says he doesn’t have anything to do with these kinds of remedies, but he can’t stop others from “potentizing”
them. He believes homeopathy works, but he knows he doesn’t have a clue why, and the extraneous, strange stuff on his shelves
isn’t helping anyone to find out. He wants to keep an empirical, almost scientific approach to the claims for homeopathy,
while all around him flakes are making that nigh on impossible. Lawrence is actively trying to stem the tide of ridiculous
remedies, but there is only so much that one man can do. Lawrence is not alone, however. Forty miles north of the Helios pharmacy,
at the Natural History Museum in London, Vilma Bharatan is on a similar quest.

AS
well as holding down a job as one of the Natural History Museum’s botanists, Vilma Bharatan is a practicing homeopath. But
she is also a fierce critic of homeopathy. Its practitioners, she says, have been living off people’s reverence for the discipline
without applying any intellectual or scientific rigor. They have been lax in laying out their data. The plant names they use
are a mess, for instance, making it almost impossible to properly investigate relationships between known plant characteristics
and reports of homeopathic usefulness. It wasn’t always like this, she points out; there was a time when homeopathy was allied
with science.

The pages of Bharatan’s PhD thesis make interesting reading to anyone who wants to understand the problem with homeopathy.
First they tie together a group of homeopathic remedies—the flowering plants—with their proper biological names, the symptoms
they are meant to treat, and how effective homeopaths have found them to be. Then they sort these remedies using a computer
program that performs a
cladistics
analysis.

Biologists use cladistics to group plants or animals according to their physical characteristics or their genetic profiles.
Bharatan’s plan was to load the program with the homeopath’s idea of the therapeutic effects of the plants and see if there
was any correlation between the homeopathic groupings and any of the traditional biological groupings.

Her database is called a
matrix
; it’s a web of plant names, plus the claims for the various therapeutic effects each plant shows. Not that Bharatan included
every homeopathic claim in her matrix; she restricted the data set to those that had, at the very least, been “frequently
confirmed” in provings and their success confirmed in normal clinical use. In the end, the matrix comprised more than a quarter
of a billion therapeutic effects of plant remedies. When she ran the data through the computer program that analyzed and sorted
it, the museum’s server was creaking under the strain. This was the largest data set it had ever analyzed.

The output from a cladistics program is called a
cladogram.
It looks like a kind of family tree. The cladogram that shows how insects evolved into their various forms, for example, branches
off first for beetles. The other branch then splits into one branch for ants, bees, and wasps and another that branches into
two again: one branch is the butterflies and moths, the other branch is the flies. From this picture, we see how recently
two species descended from a common ancestor.

Bharatan’s cladogram showed very few “common ancestors” for the most part; in many cases, the program found no strong biological
relationships between the various homeopathic plant-based remedies. But on occasion it found very strong relationships. One
grouping, or
clade
, branching off and subdividing rather like the insect group on the tree of life, contained remedies whose curative properties
were associated with the cardiovascular system. Another group was plants used in treatments of female reproductive disorders.
Look at the raw data for a million years, and you would never see these groupings, Bharatan reckons. Because the plants are
used in such a wide variety of treatments, there is no way you would normally think of grouping them according to systems
of the human body. Nor do they belong to the same botanical family. Nonetheless, after running for thirty-two hours solid,
the computer decided they belonged together. The reason, it seems, is chemical.