The announcements came just days after the American Medical Association said the Department of Health and Human Services should create a national registry of clinical trials that could ensure that negative results become public.
But the benefits could extend far deeper, said David Handelsman, lead strategist for clinical research and development at SAS Institute Inc., whose modeling and statistical packages are used by nearly all drug companies to gauge drugs safety and efficacy. Depending on their quality, the data could help design clinical trials that cost less, require fewer patients and take less time.
Confirming its support for the results registry, Merck & Co Inc. suggested that a good starting point would be the current listing of ongoing clinical trials maintained by the National Institutes of Health.
GlaxoSmithKline, which has been accused of fraud after failing to publish unfavorable results of its antidepressant Paxil, announced Friday that it would make results from clinical trials of its marketed medicines publicly available online, perhaps in six months.
Handelsman predicted that industrywide information could be available in as little as two years. "The process of capturing clinical trials information exists in the marketplace. Its not being fully utilized yet, but the basic premise is there," he said.
Alan Goldhammer, vice president of regulatory affairs at the Pharmaceutical Research and Manufacturers Association of America, told The New York Times that his trade group thinks a registry might force companies to disclose proprietary information and that patients and physicians might lack the sophistication to properly interpret results.
In addition, government and academic researchers as well as drug manufacturers should be required to register trials, he said. Another possible, albeit unstated, objection is that such postings could allow easier comparisons of drugs for similar problems, which might complicate drug companies marketing strategies.
SAS Handelsman said he thinks more available data will help design better clinical trials, potentially bringing down drug costs and making medicine more effective. "The fact that Merck has jumped on board signifies that it thinks theres a business model and value to doing this."
SAS often helps drug companies use their own past results "to streamline their own operations," but being able to use data from other companies "would take that to another level," he claimed. Also, as genomics and other data are incorporated, pooling data could help to better understand the disease state and suggest new ways to design drugs.
In addition, Handelsman said he imagines providing services that might help patients and doctors pick the right drug, even if they havent mastered statistics or built a model. "We could package our technology so that it becomes a point-and-click interface," he said.
"They dont have to understand the architecture behind the scenes as long as we give them an interface thats useful for them." Subscription or advertising fees could fund such a service, he said, noting that such a project is a long way off.
Of course, the potential for using the data relies on the amount, quality and format of data that are made available. Right now, its unclear whether posted results would consist of a written abstract, statistical tables or something else.
If GlaxoSmithKline posts its results in the next six months, Handelsman said, that might help the industry realize which formats and information would be most useful.