PAY their wages but don鈥檛 tell researchers what to write. That is the message
to drugs companies from the medical establishment this week.
Several of the world鈥檚 leading medical journals, including The
New England Journal of Medicine and The Lancet, announced that
they are taking a tougher stance on publishing the results of clinical trials
funded by the pharmaceuticals industry. Their action has been prompted by
concerns that companies sponsoring the research are trying to influence the
design and interpretation of trials, and that unpalatable results are being
swept under the carpet.
However much the pharmaceuticals industry may protest, these concerns are
well founded. Surveys show that company-funded trials are more likely to yield
positive results than independent studies. The results of clinical trials that
prove negative are only one-third as likely to appear in a journal as the
results of successful trials.
Advertisement
From now on, many journals will publish papers only if the main authors, not
the sponsor, make the final decision on publication. To achieve this, the
journals will ask authors of papers to sign a statement declaring that they had
access to all the data, and accept full responsibility for the conduct of the
trial and its findings.
The measures should help put a stop to the sort of abuses that prompted this
move. These include the case of Knoll Pharmaceuticals, which for seven years
blocked publication of findings suggesting that generic thyroid drugs worked
just as well as one of its brand-name products. We only know about such cases
because a few researchers have had the courage to stand up to their
sponsors.
The unpalatable truth is that some scientists still remain silent or collude
in the distortion or suppression of trial results. Indeed, the pressure to
obtain and maintain funding can be so great that a few even resort to
fabricating data to keep their paymasters happy.
Getting authors to sign statements isn鈥檛 likely to put a stop to this kind of
corruption. But there are other measures that companies and academic
institutions could be taking. For instance, several organisations, including
drugs giant GlaxoSmithKline, have already set up online registers of clinical
trials. Companies enter basic information about trials into the register when
they begin and as they progress. This makes it possible to find out if trials
have been abandoned or haven鈥檛 been published. They also make it harder for
facts to be twisted at a later date.
At the moment, such registers are entirely voluntary. But regulators could
refuse to give studies ethical approval unless they are registered. They could
also ask the researchers running trials to declare any interests they have in
the outcome.
The more information there is in the public domain, the harder it will be for
companies or researchers to distort their results. And this is in everybody鈥檚
best interests. For in the long run, bad science is bad business.
