Business your connection to The Boston Globe

Drug-trial registry data grows, but report faults some firms

Drug companies are making public more information about medical studies they are conducting, but some still withhold key details, a new analysis of a federal registry finds.

Merck & Co., stung by allegations that it hid information on Vioxx's dangers, gets somewhat better marks in the new analysis than it did in an earlier one. However, Pfizer Inc., GlaxoSmithKline PLC, and Novartis AG are lagging, according to the report in yesterday's New England Journal of Medicine.

In May, the journal's editor-in-chief accused Merck, Pfizer, and Glaxo of making a mockery of efforts to increase the transparency of such experiments, called clinical trials.

The new report shows some progress, said its chief author, Dr. Deborah Zarin of the National Library of Medicine, which runs the registry.

The registry,, was created in 2000 as part of an overhaul of Food and Drug Administration monitoring. It requires certain types of studies to be listed, such as late-stage experiments involving life-threatening illnesses like cancer.

But it didn't get wide participation from the drug industry or many voluntary listings until September 2004, when editors of leading medical journals said they would no longer publish results of any studies that were not first listed in a public registry.

The idea was to make it easier for scientists, regulators, and the public to cross-check what studies were being done on a drug and get the big picture of risks and benefits. The registry includes studies by universities, governments, and industry, but concerns about openness center on the industry.

The analysis covers May 20 to Oct. 11. Entries rose to 22,714 from 13,153 in that time, spiking around Sept. 13 when the medical journals' new policy took effect.

Zarin focused on whether company listings revealed two things: the name of the drug or device being tested and the main outcome being measured, such as death or cancer recurrence within five years.

Pfizer was worst on giving names; 14 of its 224 new listings lacked this information. Glaxo named the drug in all but one of its 128 new listings but has the worst overall record: 21 percent of its total registry entries lack specific drug names.

Merck used to be the worst but amended its entries and now has the best record, listing the drug 99 percent of the time. However, it did poorly on the second measure in the analysis, listing what outcome was being measured only 20 percent of the time.

Today (free)
Yesterday (free)
Past 30 days
Last 12 months
 Advanced search / Historic Archives