[Infowarrior] - The danger of drugs … and data

[Richard Forno]

The danger of drugs … and data
Comments (88)

     * Ben Goldacre
     * The Guardian, Saturday 9 May 2009
     * Article history

http://www.guardian.co.uk/commentisfree/2009/may/09/bad-science-medical-journals-companies

A fascinating court case in Australia has been playing out around some  
people who had heart attacks after taking the Merck drug Vioxx. This  
medication turned out to increase the risk of heart attacks in people  
taking it, although that finding was arguably buried in their  
research, and Merck has paid out more than £2bn to 44,000 people in  
America – however, they deny any fault.

British users of the drug have had their application for legal aid  
rejected, incidentally: the health minister, Ivan Lewis, promised to  
help them, but documents obtained by the Guardian last week showed  
that within hours Merck launched an expensive lobbying effort that  
convinced the minister to back off.

This is a shame, because court cases can be tremendously revealing.

The first fun thing to emerge in the Australian case is email  
documentation showing staff at Merck made a "hit list" of doctors who  
were critical of the company, or of the drug. This list contained  
words such as "neutralise", "neutralised" and "discredit" next to the  
names of various doctors.

"We may need to seek them out and destroy them where they live," said  
one email, from a Merck employee. Staff are also alleged to have used  
other tactics, such as trying to interfere with academic appointments,  
and dropping hints about how funding to institutions might dry up.  
Institutions might think about whether they wish to receive money from  
a company like that in future. Worse still, is the revelation that  
Merck paid the publisher Elsevier to produce a publication.

The relationship between big pharma and publishers is perilous. Any  
industry with global revenues of $600bn can afford to buy quite a lot  
of adverts, and pharmaceutical companies also buy glossy expensive  
"reprints" of the trials it feels flattered by. As we noted in this  
column two months ago, there is evidence that all this money distorts  
editorial decisions.

This time Elsevier Australia went the whole hog, giving Merck an  
entire publication which resembled an academic journal, although in  
fact it only contained reprinted articles, or summaries, of other  
articles. In issue 2, for example, nine of the 29 articles concerned  
Vioxx, and a dozen of the remainder were about another Merck drug,  
Fosamax. All of these articles presented positive conclusions. Some  
were bizarre: such as a review article containing just two references.

In a statement to The Scientist magazine, Elsevier at first said the  
company "does not today consider a compilation of reprinted articles a  
'journal'". I would like to expand on this statement: It was a  
collection of academic journal articles, published by the academic  
journal publisher Elsevier, in an academic journal-shaped package.  
Perhaps if it wasn't an academic journal they could have made this  
clearer in the title which, I should have mentioned, was named: The  
Australasian Journal of Bone and Joint Medicine.

Things have deteriorated since. It turns out that Elsevier put out six  
such journals, sponsored by industry. The Elsevier chief executive,  
Michael Hansen, has now admitted that they were made to look like  
journals, and lacked proper disclosure. "This was an unacceptable  
practice and we regret that it took place," he said.

The pharmaceutical industry, and publishers, as we have repeatedly  
seen, have serious difficulties in living up to the high standards  
needed in this field, and bad information in the medical literature  
leads doctors to make irrational prescribing decisions, which  
ultimately can cost lives, and cause unnecessary suffering, not to  
mention the expense.

It has been estimated it would take 700 hours a month to read the  
thousands of academic articles relevant to a GP; doctors skim, they  
take shortcuts, they rely on summaries, or worse. We could perform  
better when giving them information, but for now, it will often be  
"actually, I think I've seen at least two studies on that, and in  
different journals".

The real tragedy is that the cost of distorted information, and  
irrational prescribing, is far greater than the cost of the research  
that could prevent it. Health systems pay for these drugs – state- 
funded in almost every single developed country – and they largely pay  
for the journals, too. In a sensible world, countries would band  
together and pay for comparative research themselves, and the free,  
open distribution of the results, to prevent all this nonsense.

We do not live in a sensible world.