Audit finds only one U drug trial death but questions lax reporting

Updated 5 p.m. | Posted 10:55 a.m.

A state review of industry-sponsored drug trials by the University of Minnesota found no deaths beyond one previously reported — but criticized the U for failing to properly report medical problems that arose during the experiments.

Legislative Auditor Jim Nobles said lack of proper documentation means he can't say for sure how many test subjects have suffered serious medical problems — such as hospitalizations and suicide attempts — during experiments carried out by the Department of Psychiatry between 2004 and 2014.

In a review (.pdf) released Thursday, Nobles found that the U met federal reporting guidelines. But he said that's a low bar, and that some university researchers haven't promptly or thoroughly reported problems to their research oversight panel.

"It's a very big deal," Nobles said, "because that's the way you achieve accountability, and adequate oversight and protection [of] human subjects."

Brian Herman, the university's vice president for research, called Nobles' review "a relatively well-balanced, fair and accurate report," and said the U is already carrying out some of the suggestions.

U of M bioethicist Leigh Turner, one of the university's main critics, called the U's lack of documentation a "cause for concern. It compromises the ability to conduct an independent investigation" of problems that occur in drug trials.

He and other critics have been calling for a review of university records on its clinical drug trials for many months.

Their concerns were prompted by the 2004 death of Dan Markingson, a mentally ill participant who killed himself during a corporate-sponsored drug trial at the U.

They wondered how many other deaths or serious injuries have occurred over the years.

Their concerns were heightened in February. An outside panel criticized the U for inadequately protecting patients and allowing a "culture of fear" in the psychiatry department to hinder attempts to improve protections.

After Nobles issued a scathing report in March over how the university handled the Markingson, he said he would review the university's files.

Nobles' report on Thursday criticized university's recordkeeping. It said researchers have often given the U's oversight panel little documentation of problems, and that those reports have often left out important details.

He said some researchers waited too long to report problems. For example, reviewers found eight cases in which researchers waited at least six months to report hospitalizations to their oversight panel. And they said 17 percent of records never said when a problem occurred.

Nobles said such delays could keep oversight officials from quickly stopping a potentially dangerous experiment.

"They may need to and want to and should intervene — and not wait 10 days, 30 days or 60 days," he said. "They maybe need to do something immediately."

Because of the lack of proper paperwork, Nobles said he was unable to determine how well researchers met university reporting requirements.

He also said some people connected to drug trials who told investigators of problems for which he could not find documentation.

"I think there are events that have occurred that have not been reported that should have been," he said. He declined to say what those were.

The list of problems that must be reported immediately appears to have shrunk over the years as the federal government has relaxed guidelines, the report notes, and the U has followed suit.

As a result, Nobles said, researchers have had too much latitude in deciding what should be reported and when.

The report recommends a tighter accounting of what happens during experiments. The university should require its researchers to report more types of problems, it says, and report them more promptly and in greater detail.

It also recommends more frequent reviews of drug trials that appear to pose a greater risk to patients.

"The reporting ought to err on the side of reporting [serious events]," he said, "and not just leave it totally to the discretion and judgment of the researcher to decide" whether a problem was related to the experiment.

Nobles said the university continues to believe that compliance with federal guidelines is the key.

But he said, "If you really want to be — as President Eric Kaler says — 'beyond reproach," you've got to move further than just complying with federal regulations."

Herman said the U is beginning to carry out dozens of research reforms recommended by an outside panel in February.

Those already cover many of the concerns Nobles has raised, he said, as do changes the U has made in previous months. The U will address others soon, he said.

"We are going to be more active in making sure there's appropriate oversight for all of our studies," he said. "We will increase the information we ask the investigators to provide us, and we'll increase the kinds of events that will have to be reported, and the frequency of the monitoring process."

But he cautioned that the U should not change reporting requirements for studies that are externally monitored, either by a federal agency such as the Food and Drug Administration, or by a corporate sponsor that must answer to a federal agency.

Many such drug trials are run at multiple universities, he said, all of which follow federal guidelines that provide "adequate protection."

"If every [university] had a different reporting criteria," Herman said, "it would be very difficult to coordinate between these multi-center trials.

Although Nobles questioned that objection, he said overall the U of M is "moving in the right direction."

He said he plans on monitoring the university to ensure it's making the necessary changes.