More than four years in, officials are still working on draft regulations to further define the ClinicalTrials.gov submission requirements laid out in the FDA Amendments Act of 2007.

When the rules will come out is anyone’s guess. Nonetheless, the mandates specified in the law remain in effect, and the office at the National Institutes of Health that runs the database says it is making a consistent effort to ensure study sponsors are educated on their new responsibilities.

The ClinicalTrials.gov staff, housed within the National Library of Medicine, conducts several conference calls a day with study sponsors struggling to submit information required by law, Deborah Zarin, director of ClinicalTrials.gov, said in an interview. They also receive hundreds of emails per week with questions.

The rate of inquiry has remained roughly constant since the new federal trial disclosure mandates were enacted into law in 2007 and Zarin says she doesn’t expect a change any time soon.

“There are always sponsors who are entering results for the first time,” Zarin said. “So no matter how many people we train ... there are always many new ones waiting in line behind them.”

Mandates Laid Out, But Gaps Remain

The FDA Amendments Act of 2007 significantly expanded the government’s online database of clinical trials, which originally launched in 2000, by requiring registration and results reports on a large proportion of studies evaluating FDA-regulated devices and drugs.

All device studies other than early stage feasibility trials need to be registered soon after they start, although NIH is not supposed to publicly post the information on device trials until after the product is approved.

Results need to be posted on the site, irrespective of the publication status of the data, for all trials that require registration once a product is cleared or approved, or within a year after a trial is complete, if the product was already approved.

The statute is relatively detailed in describing the trial information and data that need to be entered into the database, as well as establishing basic deadlines and financial penalties for noncompliance. But it raises a number of specific questions for the Department of Health and Human Services to address in regulations.

The reg will go beyond the specific issues requested by Congress, defining the requirements for all ClinicalTrials.gov provisions in the FDA Amendments Act of 2007.

For instance, the statutory language asks officials to determine whether study data for products that are never cleared or approved by FDA need eventually to be reported on the database, and whether narrative summaries of clinical trials and results can be included in the databank without being misleading or promotional.

The planned regulation, however, will not be limited to addressing the specific issues requested by Congress. Rather, it is intended to be the implementing regulation for all ClinicalTrials.gov provisions in FDAAA, including mandates for trial registration, results reporting and adverse event disclosures. Particular emphasis will be placed on issues that have raised confusion among users so far.

“We have learned a lot and learned about areas of ambiguity in the legal language that require clarification through rulemaking,” Zarin said. “There are things that we could not have realized ahead of time could be ambiguous, until you see them play out in real life.”

However, Zarin declined to disclose details on the specific issues that will be addressed in the proposed rulemaking.

One area the device industry has in the past pointed to with concern relates to reporting of adverse events. Companies have noted that the law’s mandates do not align with FDA rules about what adverse events must be reported from a device trial. (See Also see "Adverse Event Reports: Posting On ClinicalTrials.gov Begins Soon" - Medtech Insight, 27 April, 2009..)

Education Is The Focus

Regulations aside, Zarin says her group is focused on engaging data submitters. In addition to conducting routine conference calls and replying to emails, the effort includes posting resources online, including webinars, documents defining basic submission elements and papers published in peer-reviewed journals.

The process of registering trials and submitting data to the site is not trivial, Zarin stressed.

“There is a learning curve,” she noted. “Sponsors need to allocate resources to this task, and ... learn the ins and outs of doing it. In general ... we find that more experienced sponsors are more successful at entering results that meet our criteria.”

Another important factor in successful submissions is making sure the right people are involved, she said.

“It usually takes the participation of, if not the principal investigator, a lead scientist in the study, as well as people that understand the ClinicalTrials.gov system,” Zarin explained. “And when sponsors try to do it with non-scientific employees, it does not tend to go well, because it is a scientific exercise to submit results.”

The NIH group also is taking an approach that puts a little bit more pressure on companies and other submitters, but the activity falls short of what officials would label as “enforcement,” the agency says.

The ClinicalTrials.gov staff does not have processes or resources in place to monitor what qualifying trials have not been posted to the database, but it does track for missing required data elements in trial registrations and, in particular, for completed trials for which results have not been submitted, when the timing suggests they should have been.

“We may say, ‘Alert! Pay attention to this trial, we think you might be missing something,’” ClinicalTrials.gov Director Deborah Zarin says.

NIH works with FDA to notify individual sponsors of potential problems in so-called “FDAAA issues reports” that can be viewed when a submitter logs on to their ClinicalTrials.gov account. In addition, the government emails issues reports to sponsors two times per year, Zarin said.

“We can’t do it with total specificity because there are things that have to be worked out in rulemaking,” she explained. “But we may say, ‘Alert! Pay attention to this trial, we think you might be missing something.’ And then the sponsor, we would hope, would look at it and make their own decisions about whether they need to report something that they haven’t reported yet.”

Compliance Rates Hard To Pin Down

NIH lacks specific data on the current level of compliance with ClinicalTrials.gov requirements across the industry. While outside academics have started to look at the issue, the focus of recent analyses has been on drugs and not devices.

A study published in the December 2011 issue of the journal Health Affairs found that a significant proportion of drug trials are registered late (past the deadline of 21 days following the start of patient enrollment) and that only 12% of completed drug studies registered in the database from September 1999 to May 2010 reported results within one year.

A study posted Jan. 3 in the British Medical Journal found that only 22% of qualifying trials for approved drugs had a results report on ClinicalTrials.gov within one year of trial completion.

It is not straightforward to perform a parallel analysis to these evaluations for device studies. There is also the added complication that the law includes an exception for device studies, where sponsors are expected to register trials with NIH according to the normal timelines, but the government is not supposed to make the registration public until after the device is cleared or approved.

Searches of the database performed by “The Gray Sheet” show that the number of registrations of potentially qualifying (phase II-IV) industry-funded trials that included a device intervention spiked significantly in 2005 (to 296 from eight in 2004) when editors of major medical journals began requiring early registration of trials as a condition of data publication.

There was another, smaller, spike in 2008 (to 357 from 272 in 2007) after the FDAAA requirements broadly took effect. After that, the number of trials registered per year leveled off through 2010 and then dropped back closer to 2007 levels in 2011 (281 trials).

In terms of results reporting, as of Jan. 19, about 16% of “closed” industry-funded phase II-IV device trials that were first registered on the site in 2007 now include results entries, and 20% of such trials first registered in 2008 include results. As would be expected for more recent trials, the percentage of entries that include results declined for trials that were registered in later years.

However, this measure is highly imperfect for assessing compliance, since the “closed” designation includes trials that have stopped enrolling but are still collecting data, as well as trials that have been prematurely terminated, in addition to completed trials. Further, one cannot search or sort results by completion date on the database.

Zarin says the database is receiving about eight result entries per month for device-related trials out of the 120 result entries per month received overall.

“So, a fairly small percentage of the results postings are device trials,” she said. “Whether or not that is the right number, I can’t say.”

[Editor’s note: Read more about this topic in “The Gray Sheet,” your source for in-depth coverage of the medical devise and diagnostics industry.]

ClinicalTrials.gov: NIH Busy Educating Sponsors On Mandates While Regs Remain In The Works

More from Clinical Trials

More from R&D