New York Times article on Peter Gotzsche
CANCER: SCIENTIST AT WORK--PETER GOTZSCHE; A Career That Bristles With
CANCER: SCIENTIST AT WORK -- PETER GOTZSCHE
By DONALD G. McNEIL Jr. Published: April 9, 2002, New York Times
COPENHAGEN— Seated at an eerily neat desk in a modern
office on the grounds of Denmark's state hospital, Dr. Peter C. Gotzsche does
not seem a man whose bite is much worse than his very gentle bark. But this
tall, gray-haired statistician is
sending some eminent doctors into fits of apoplexy as he quietly implies that
they've wasted their lives defending old wives' tales, maltreated their
patients or assisted in frauds that perhaps ought to land them in jail.
Most recently, he has made
headlines for a study published in The Lancet in October concluding that the
major studies of mammography do not show that mammograms can reduce death from
Earlier last year, in The New
England Journal of Medicine, he purported to pop the old shibboleth known as
the placebo effect. Many doctors who have for years repeated what they had
learned in medical school -- that about 35 percent of patients improve if you
give them only a sugar pill -- learned that Dr. Gotzsche felt they had fallen
for a theory as grounded in hard evidence as the one that swamp gas causes
In 1999, in The Journal of the
American Medical Association, he argued that the research behind the Food and
Drug Administration approval of fluconazole, an expensive and highly touted
anti-fungal drug, was based on cooked books.
Before he was director of the
Nordic Cochrane Center here and paid to make trouble like this, he was in
private research. There, he asserted that a multinational drug company was
deliberately overdosing AIDS patients with AZT because it was profitable to do
His critics say he is too
harsh, rejecting important studies with small but acceptable data-collection
flaws. Dr. Gotzsche replies that he does
not go around looking for trouble, but it seems to come looking for him. ''We
had no preconceptions about mammograms
when we were asked by a government agency to look into this,'' he said in an
interview in his blue and white office, where a thick stack of mammogram data
fills drawer space. ''It just wasn't one of my interests.'' ''Now I've received
more than 2,000'' -- he
stops, and checks his computer -- ''I've received 2,276 e-mails about this.
It's incredible what we've unleashed.''
Cancer researchers who still
endorse mammograms have loudly attacked the work he did with his colleague Dr.
Ole Olsen. The seven landmark studies
that back the use of mammograms for early cancer detection were begun between
1962 and 1982 in New York, Canada and Sweden. Dr. Gotzsche's specialty,
meta-analysis, blends multiple study results. But he and Dr. Olsen found five too
flawed to use; the
remaining two showed no benefit.
Dr. Anthony B. Miller at the
German Cancer Research Center said some of the five were precipitously thrown
out because the authors couldn't or wouldn't answer all of Dr. Gotzsche's
questions about their data.
Attacks on him have come from
three groups, Dr. Gotzsche said: disinterested experts, oncologists who spent
careers recommending mammograms, and those who earn money in the $1 billion
screening industry. The first group, he
said, are ''academic and reasonable,'' the rest have vested interests ''and
want to shoot the messenger.''
Dr. Gotzsche was born 53 years
ago in Naestved, near Copenhagen, where his mother taught mathematics and home
economics. Her father, a doctor, was
his role model. ''He was very bright,'' Dr. Gotzsche said. ''I never succeeded
in beating him in chess.'' His own
grandfather never once let him win? ''No,''
he answered, looking perplexed. ''Fairness and honesty were important values in
the family. I would have found it unfair if he lost to me on purpose.'' His
first job was a paper route, and he
planned to be a doctor, too, ''until I realized when I was 17 that many of my
grandfather's patients really weren't ill.''
They were affluent women,'' he said. ''Seeing the doctor was a social
event, so I decided it was a waste of time. I wanted to be a biologist and walk
in rain forests.''
Instead, as a biologist, he
went to work doing clinical trials for Astra Pharmaceuticals. But,
disillusioned by its glowing pitches for the arthritis drug he tested, he
started medical school in 1978. By 1990, he had a medical degree and a
doctorate. His thesis: ''Bias in Double-Blind Trials.'' ''It was provocative,''
he said quietly. ''In
1990, the pharmaceutical industry had a good image.''
The thesis was read by Dr. Iain
Chalmers, an Oxford professor who founded the Cochrane Collaboration, and
invited Dr. Gotzsche to direct its Nordic branch. The self-supporting collective,
worldwide chapters in 50 medical specialties, was named for Archie Cochrane,
author of a history of fatal flaws in medical practice, like the refusal by
19th century doctors to wash their hands between dissecting corpses and
examining patients. Its mission is to
question medical tenets by analyzing all the trials of any treatment. It has,
for example, compared 1,000 acupuncture studies and concluded there is no
evidence it works, even for pain or smoking.
His placebo study, with Dr. Asbjorn Hrobjartsson, was done by hunting
down trials that experts thought hardly existed: those with three patient
groups, the treatment, a placebo and no treatment. They hoped to find a dozen,
but by getting the United States National Library of Medicine to rewrite its
software so ''Treatment: none'' could be searched, they found 130, in
everything from herpes to nausea to schizophrenia. Over all, the untreated patients
-- those still on the waiting list,
for example -- got better at the same rate as those receiving placebos. ''We
started the project believing there were important placebo effects,'' Dr.
Gotzsche said. ''But when we finished, there wasn't much left.''
A small effect may exist in subjective
symptoms, like pain, he conceded, but not in doctor-observable ones like
Dr. James M. Robins, a Harvard
epidemiologist who follows Dr. Gotzsche's work, said he would probably have
reached different conclusions. Dr. Gotzsche excluded acceptable mammogram
studies and averaged pain studies in a way that made the placebo effect seem
not to exist, rather than to exist in some studies for unknown reasons. He called
Dr. Gotzsche's harshness toward
other people's randomized trial data ''unscientific and even bizarre.'' ''He
seems to have a very black-and-white
mind,'' Dr. Robins said. ''When he writes, he seems to be saying: 'You've done
such terrible things against the pristine world of randomization that you're
not going to be allowed in the covenant.' ''
Aware of his critics, Dr.
Gotzsche said he was ''not cooking up difficulties or being nihilistic.'' ''There
are many things in health care that,
if you look at them closely, you are disappointed,'' he said. His 1999 fluconazole
study, done with Dr.
Helle Krogh Johansen, his wife, attacked the research backing a drug that sells
in the United States for $25 a pill for yeast infections and toenail fungus but
also kills funguses that can be fatal to AIDS patients. He found that studies
fluconazole's patent holder, compared oral doses to mixed results for
amphotericin B and nystatin. But amphotericin B should be given intravenously
and nystatin doesn't work at all, Dr. Gotzsche said, so the studies were, in
In a later letter to the
medical journal, Pfizer conceded that the comparison drugs did not work and
that it would change how it reported results of trials. A Pfizer spokesman,
Andrew McCormick, says the company stands by that response.
In 1987, Dr. Gotzsche said, he
was the coordinator for all Nordic AIDS drug trials, and found that AZT was
just as effective, with fewer side effects, at 400 milligrams as at 1,200
milligrams, the dose recommended by its maker, Burroughs Wellcome. The company,
which sold it in 100-milligram pills, did not lower its recommendation, he
Patricia R. Seif, a spokeswoman
for GlaxoSmithKline, which now owns the Burroughs Wellcome company, said 1987
was ''ancient history'' but recalled that the company's recommended AZT dosage
was reduced from 1,200 to 600 milligrams in mid-1990, based on studies
sponsored by the National Institute of Allergic and Infectious Diseases between
1986 and 1989. About the price-gouging accusation, she said the company cut its
price for AZT by 20 percent in 1987 and another 20 percent in 1989. No one she
could find recalled Dr. Gotzsche's
Nordic study, she said.
Asked if he had ever confronted
Burroughs Wellcome by holding a news conference or getting a forceful group
like Act Up into the fray, Dr. Gotzsche said he had not, then added: ''Perhaps
we should have -- but that was not our style here in the Nordic countries.''