Healthy Skepticism Library item: 15422

Warning: This library includes all items relevant to health product marketing that we are aware of regardless of quality. Often we do not agree with all or part of the contents.

Publication type: news

Goozner M.
Clinical Trial Process Due for Reform
Gooznews.com 2009 Apr 8
http://www.gooznews.com/archives/001378.html

Full text:

The Food and Drug Administration advisory committee that today will mull extending use of Seroquel XR (extended-release quetiapine) to general anxiety disorders and depression will consider evidence from clinical trials designed by clinicians working for AstraZeneca. Is it any better or worse than the numerous drugs already approved for those conditions? The trials didn’t test that proposition, and the FDA by law isn’t allowed to consider it.

But even if there wasn’t a single other drug available for GAD and depression, can the evidence submitted by AstraZeneca be trusted? At any FDA advisory committee meeting, the FDA presents its own analysis of the data, and it is always striking how at odds with the company presentation that can sometimes be.

Unfortunately, the same independent review does not exist for company-organized clinical trials that appear in the medical literature. There’s peer review, of course, and the funding source and conflicts of interest of the authors are put on public display. But as Bruce Psaty of the University of Washington points out in a commentary in today’s Journal of the American Medical Assciation:

It is not possible to look at a disclosure about industry funding of research, consulting, or speaking and know how to interpret the disclosure or its potential effect on a published clinical trial. Indeed, a funding disclosure does not necessarily mean that any bias is present. Even if the disclosures included exact dollar amounts [as would be required if the Physician Payments Sunshine Act passes], the interpretation would remain difficult.
How can bias be detected? Psaty suggests analyzing the key scientific components of the trial: Did the hypothesis have merit; did it address a pressing public health question; if conducted abroad, did the adverse events rate or outcomes in one part of the world differ from domestic data; did the authors analyze adverse events as rigorously as outcomes; was the discussion dispassionate and driven by data or did the discussion “have the look and feel of advocacy or promotion”? Psaty also asks clinicians — the consumers of the medical literature — to be on the lookout for editorials accompanying major clinical trials, and to look at whether the editorialist’s conclusions differ from the investigators.

That’s an awful lot to ask of readers, who rarely get beyond the abstract box printed on the first page of the study.

As long as most third stage proof-of-efficacy clinical trials remain controlled by industry — a circumstance “not likely” to change anytime soon as Psaty points out, readers who simply look at who funded a trial or the conflicts of interest of investigators will not really know if the results are biased.

The bias of conflict of interest is a behavioral phenomenon. Under the assumption of an accurate report, the design of the trial, the conduct of the study, and the interpretation of the results are perhaps the best measures that clinicians, researchers, and other readers have to assess the possibility of such a bias among their scientific colleagues and themselves.
I agree. Having spent the past five years professionally focused on ferreting out undisclosed conflicts of interest, and seeking to eliminate conflicted physicians for key posts in the regulatory process, I am well aware that conflict-of-interest disclosure is a poor signpost for identifying bias (for a good example of the misuse of COI analysis, see this Health Care Renewal blogpost by Roy Poses on the Lewin Group).

When it comes to published clinical trials, the sad truth is that lay readers and rank-and-file clinicians have little beyond conflict of interest analysis to judge bias, since journal editors do not use Psaty’s evaluative measuring sticks when determining whether to publish a study. Let’s not forget that they, too, have a conflict of interest — the unexpressed need to get advertising revenue and reprint sales from the same companies that are funding the clinical trials.

There is one alternative that Psaty dismisses too quickly with his comment that industry control of efficacy clinical trial testing isn’t going to end anytime soon. It’s an idea suggested by Marcia Angell, writing five years ago in her seminal The Truth About the Drug Companies. “Clinical testing could continue to be the industry’s responsibility, but it should be conducted at arm’s length, preferably through an Institute for Prescription Drug Trials,” she wrote. That way, independent physicians could design protocols that maximized the trial’s ability to generate information about comparative effectiveness, public health and safety.

Angell did not include this reform in her suggestions for top priorities for the new leadership at FDA that ran in the Boston Globe earlier this week. After all, this isn’t something that can be done by fiat by the agency’s new leaders. It will require action by Congress.

There once was a bill requiring that independent researchers conduct all science aimed at regulatory decision-making. It was introduced by Sen. Gaylord Nelson (D-WI) in the mid-1970s. Some enterprising staffer on the Hill should dredge it out of the Congressional Research Service’s archives and see if it isn’t worth another look.