Study: Academic medical centers do poor job reporting results of medical trials

Leading academic medical centers across the U.S. have a poor record of reporting results from clinical trials, leaving patients and doctors with a potentially skewed view of the safety and benefits of treatments, a new study finds.

Nationwide, only about two-thirds of results from more than 4,300 studies completed between 2007 and 2010 were published or reported – and only about one-third were released to the public within two years of completion.

The analysis was published last week in the journal BMJ.

“Once you realize that human experiments are being done and the results are not being reported, you realize that the entire scientific enterprise is being threatened,” said Dr. Harlan Krumholz, the Yale School of Medicine cardiologist and health care researcher who led the study.

Without accurate and timely reporting of clinical trial results, doctors and patients can’t make evidence-based decisions about conditions from cancer and heart disease to mental health disorders, Krumholz said.

“It’s biasing the medical literature,” he said.

The reasons researchers don’t report study results can range from the nefarious – they failed to get the results they wanted – to the mundane – they were too busy to input the data, experts said.

Krumholz and his colleagues analyzed results from more than 4,300 interventional clinical trials completed at more than 50 of the leading academic medical centers in the U.S. between October 2007 and September 2010. They looked at whether the institutions had published academic papers detailing their results or posted the results on ClinicalTrials.gov as of July 2014.

That’s the Web-based registry authorized by Congress in 2000 to allow public reporting of clinical trial results. In 2007, a new law expanded the site’s mandate, requiring reporting of basic results for public access. The site is one way patients with a range of rare and common diseases can research trials that may offer new treatment options.

There are nearly 209,000 studies in the U.S. and 192 countries now posted, but many completed trials include no results.

In the BMJ study, only 13 percent of academic medical centers nationwide posted study results on the federal website.

“This paper brings to light that the institution has a responsibility,” said Dr. Deborah Zarin, director of ClinicalTrials.gov. “They used to leave it to individual investigators.”

Part of the trouble is that there have been few incentives to encourage reporting or consequences for failure to do so. Under the federal Food and Drug Administration Amendments Act of 2007, results of certain trials are required by law to be reported within a year of completion and researchers can be fined up to $10,000 a day per trial for lapses. But, so far, no fine has been levied.

New rules are expected soon aimed at strengthening reporting of clinical trials funded by the National Institutes of Health and regulated by the Food and Drug Administration.

The BMJ study isn’t the first to call out academic institutions for poor reporting rates by name, but it’s part of a move to change the system to benefit doctors, scientists and patients, Zarin said.

“Nobody wants to be at the bottom of a list like that,” she said.