At the very least, editors say registries will keep scientists on track for reporting what they said they set out to prove.
Those initial trial plans will be traced by a registry number that authors receive when they first register the trial. "That
number allows the editor to look at the trial information in the registry and make sure that the submitted article agrees
with what the authors said the trial is going to do when they first registered," Sox says. "We want authors to report the
primary result they intended in the beginning of the trial rather than picking a result that happens to suit their purposes."
Sox says peer reviewers will also use registries to ensure that when authors describe other work that's been done, they describe
both published and unpublished studies, so readers get a balanced view.
The hope is clinical trial registries will make the peer review process better. "Instead of just sending a manuscript, with
clinical trial registries, you might have an opportunity to link to the spreadsheets or other individual case reports to offer
a more in-depth review," says Dougherty.
"If an article comes out of a registry so it is possible to trace the trial through its various stages, it would make the
review a lot easier," Lundberg says. "In theory, it would even make it faster and better. But we're a long way from having
There are still issues to sort out. "One problem is ambiguity over what type of trial should be registered," Sox says. "We
don't know enough about this middle ground of trials that fall between those clearly not affecting practice and those that
have a primary purpose to affect practice. There's no doubt about Phase III. It's the Phase II and small randomized Phase
I trials that are difficult to decide about. So we're going to get our feet wet and formulate what amounts to case law to
guide us so each journal will treat articles consistently."
Another issue of concern is how registries will affect intellectual property and innovation. "First-to-publish has been a
criterion for a long time in terms of who gets the credit for certain information," Bacon says. "The more you disclose during
the research process, the more you open up the opportunity for others to get their own results running earlier."
Either way, it's sure to be a long road ahead. "If trial registries function as they should, they would answer the problem
of transparency—but only for research that's funded by industry," Lundberg says. "Government also funds large numbers of trials,
and their transparency hasn't been all that great either. It's a tough nut to crack because there are so many organizations,
trials, and funding agencies."
The two customers of medical journals—readers and advertisers—have one thing in common: They tend to change their behavior
only after careful analysis of the benefits of what's "new."
Because of that, several medical journals have been slow to translate their businesses online. "Pharma is being dragged kicking
and screaming into the online arena," Weislogel says. "We who do electronic products think, 'Why wouldn't a pharma company
want to be in something that is going to physicians?' And then you look at the way decisions are made and the information
and metrics out there, which are very much in support of print products."
Most banner ads on journal Web sites are ineffective, says William Kelly, president of marketing consultancy Bioinformatics
and author of the report "Scientific and Medical Journals on the Web." Instead, journals are developing more formats and metrics
to get pharma executives over the hump in moving dollars online. "It is difficult to point to proven stickiness and proven
ROI," says David Marshall, senior publisher of The Neurologist. "So we are seeing interesting discussions going on over what works, draws eyeballs, and provides stickiness. Now we are
starting to see some interest in hot-linking, which allows someone to double-click on the name of a drug in a published article
and be taken to a site where they could get information about that particular drug."