Analysis Finds Clinical Trials Often Small, of Poor Quality
TUESDAY May 1, 2012 -- A new analysis of registered U.S. clinical trials -- the research that seeks to determine if medical treatments and prevention strategies work -- finds that many are small and of poor quality.
Studies of cancer treatments, in particular, often fail to follow the highest standards of medical research, the analysis found. Meanwhile, a full 7 percent of the studies didn't bother to mention their purpose, while others failed to provide other important details.
"We think expert groups need to start scrutinizing the database more and really examine the quality of the trials one by one," said report author Dr. Robert Califf, vice chancellor for clinical and translational research at Duke University. "For the first time, we have a chance to look at the entire universe [of clinical research] and we can see that we can do better."
Califf and his colleagues examined the Clinicaltrials.gov database, which includes medical research studies that enroll people. By law in the United States, many kinds of studies must be registered in this database.
As the law has been strengthened, the number of registered trials has grown. A total of 40,970 trials were registered between October 2007 and September 2010, compared to 28,881 trials registered over the previous three years.
"In the past, there would be a lot of studies done, but many would never be studied or reported anywhere," Califf explained. "Sometimes you'd only see the ones that were positive. Now, we have much more accurate information."
However, the analysis found that most clinical research in the database isn't very extensive, raising questions about reliability.
The new report says 62 percent of the trials from 2007-2010 were small, with 100 or fewer participants. Only 4 percent had more than 1,000 participants.
"There are 330 new clinical trials being registered every week, and a number of them are very small and probably not as high quality as they could be," Califf said.
Ideally, medical studies compare randomly chosen groups of people to each other, with one group getting a treatment and the other getting another treatment or an inactive placebo. However, 65 percent of cancer studies didn't randomize their participants, compared to 26 percent of cardiovascular studies.
The finding raises questions about whether cancer studies should be done differently, Califf said.
One expert agreed that the quality of medical research in the United States is lacking.
It's clear that "a lot of studies are of poor quality," said Kay Dickersin, director of the Center for Clinical Trials at Johns Hopkins Bloomberg School of Public Health in Baltimore and co-author of an editorial accompanying the report.
That's a problem because the medical world relies on research not only to determine whether new medications work but also to figure out guidelines for common medical procedures, such as whether people should get colonoscopies, Dickersin said.
Dickersin noted that the federal database includes incomplete information about some studies. For example, 7 percent of studies registered from 2007-2010 didn't mention their purpose, as is required; 4 percent didn't say how many volunteers would take part.
The lack of correct data slows down researchers who are trying to analyze a swath of research for studies known as meta-analyses that help doctors figure out guidelines, Dickersin explained.
The report appears in the May 2 issue of the Journal of the American Medical Association.
The U.S. National Library of Medicine has a fact sheet about clinicaltrials.gov.
Posted: May 2012