Withdrawal of clinical trials policy by Canadian research institute is a “lost opportunity for increased transparency”

Canadian researchers and academics are puzzled by the Canadian Institute for Health Research’s decision to withdraw its policy on clinical trial registration and results just three months after posting it on its website.

The institute has declined to comment on the decision. A statement on its website says that its policy has been “superseded” by a more general guidance document on the ethics of research involving humans that was prepared for, and approved by, Canada’s three major public funding agencies.

“The CIHR [Canadian Institute for Health Research] policy certainly was leading the drive towards increasing transparency,” said An-Wen Chan, a scientist with the Women’s College Research Institute in Toronto and co-author of the Ottawa Statement on Principles and Implementation of Clinical Trial Registration and Results Reporting (BMJ 2005;330:956-8; doi:10.1136/bmj.330.7497.956).

If the institute’s policy is permanently rescinded, the result would be “a lost opportunity for a federal funding agency to make a statement that increased transparency is important for . . . ensuring that publicly funded research has maximal impact,” said Dr Chan.

The “policy on registration and results disclosure of controlled and uncontrolled trials funded by the CIHR,” posted on the agency’s website on 20 December 2010 was removed sometime during the second last week of March 2011. An apparently hastily written note on the website now states that the Tri-Council policy statement, Ethical Conduct for Research Involving Humans (TCPS-2) “supersedes our previously policy” (sic).

The term “superseded,” however, has raised eyebrows because TCPS-2 was submitted to the institute in August 2010, before it published its own policy.

Trudo Lemmens, of the University of Toronto law school, said the decision to remove the policy “sends a bad message.” The TCPS-2 requirements are more general and vague than the CIHR policy, and it is not clear who will implement and enforce TCPS-2, he said.

The advantage of the CIHR policy was that “if you wanted to obtain funding you explicitly had to comply and report to CHIR directly. Now the obligation is indirect,” said Professor Lemmens, a member of the World Health Organization Expert Advisory Panel on Clinical Practice Guidelines and Research Methods and Ethics.

The TCPS-2 encourages researchers to “make reasonable efforts” to publish clinical trial results in a “timely manner,” whereas the CIHR policy had required public disclosure within 12 months after the end of the trial and reserved the right to disclose the final report itself within 18 months. The CIHR policy also required that serious adverse events be included in the publication of trial results, whereas TCPS-2 makes reference only to incidental findings, new information and unanticipated issues that should be reported during the trial.

The TCPS-2 places responsibility for many of its provisions on the research ethics boards

at institutions, yet there is a “worrisome lack of a coherent regulatory framework” surrounding research ethics boards, said Professor Lemmens.

Unlike TCPS-2, the CIHR policy required that when a trial was registered, a “reference to the systematic review that justified the need for the proposed trial” would have to be included. Also, trial information had to be updated at least once a year, and major protocol changes and early termination reported to CIHR and the trial registry within 30 days. All trial data had to be retained for 25 years unless deposited in a freely accessible data repository.”

The CIHR has recently been “in discussions” with Rx&D, the trade association for Canada’s brand name drug companies, on renewing the CIHR/Rx&D collaborative research programme (which funds awards, grants, and clinical trials) and finding ways “to improve the clinical trial environment,” Rx&D president Russell Williams stated in an email.

Professor Lemmens, who has published widely on transparency and clinical trials, said that at an international level, the pharmaceutical industry has been critical of trial registration and requirements about having to give details of results. He speculated that “maybe it is not pure accident” that the CIHR policy, widely regarded as pushing for more transparency in clinical trial reporting, has been withdrawn “when there is a clear push (in CIHR) to promote collaborative research with industry.”

The final TCPS-2 document, which had been several years in development, was formally approved in the second week in December, said Susan Zimmerman, executive director of the federal Interagency Secretariat on Research Ethics.

But there’s no reason that the CIHR could not continue to have its own more detailed policy, Dr Chan suggested. “It is one thing to adopt another agency’s policy. It is somewhat stronger, if only in a symbolic sense, to create your own policy to reinforce the importance of the issue, to reiterate the key principles [in the guidance document] and then to provide mechanisms to enforce those principles,” he said.

In a form letter sent to Canadian researchers who expressed concern about the disappearance of the CIHR policy, Ian Graham, CIHR’s vice president Knowledge Translation and Public Outreach, stated the CIHR policy was removed “as the overlap [with TCPS-2] will cause confusion and inconsistent application of the requirements.”