|
By Susan Levine
The U.S. Food and Drug Administration
has found a systemic breakdown in communication and operation
of the Johns Hopkins University review boards charged with
oversight of clinical
trials involving human subjects, according to documents
released September 7.
Agency inspectors said there was no
evidence that board members' questions about research applications
were ever sent to the researchers involved. It
also appeared that board members with conflicts of interest
in applications being reviewed did not always abstain from
voting on those proposals.
"There was a systemic problem in
the whole oversight scheme," said David Lepay, the FDA's
senior adviser for clinical science. "We had significant
concerns."
The FDA documents reflect in narrower
scope the malfunction that government regulators have identified
throughout the university's medical research program after
a healthy young volunteer's death during an asthma study in
June.
In late July, that triggered a brief suspension
of all of Hopkins's federally supported medical research with
human participants -- 2,800 clinical trials involving more
than $300 million.
The FDA and the Office for Human Research
Protections are conducting inquiries into 24-year-old Ellen
Roche's death. Both are part of the federal Department of
Health and Human Services.
The FDA's latest findings center on two
institutional review boards, including the one that approved
the protocol in which Roche enrolled at Hopkins's Bayview
Medical Center.
Because the boards' actions were documented
so poorly, the extent of the problem might not
be clear even in hindsight, Lepay said. Even with meeting
minutes and audiotape, inspectors could not reconstruct what
happened during those boards' deliberations and resolve discrepancies,
he said.
Much of the time, applications were reviewed
only by individual members or subcommittees, inspectors found.
They were not discussed and considered by the review boards
as a whole, and dozens of applications might have been were
approved by a one-time vote.
Unlike the Office for Human Research Protections,
which clamped harsh restrictions on Hopkins scientists even
after lifting its four-day suspension, the FDA has not taken
action against the university.
A final report by FDA inspectors will
not be sent to agency headquarters for several weeks. Though
punitive measures could be imposed, they seem unlikely given
Lepay's praise yesterday afternoon for the changes university
officials are implementing.
In contrast to Hopkins's early denial
of problems, "we do
feel that Hopkins is now taking this very seriously,"
he said.
In a statement yesterday evening, Chi
Dang, a vice dean at the university's medical school, said
the FDA's findings provide guidance "to further improve
our processes for conducting research involving human volunteers."
The issues documented are those "that Hopkins has been
reviewing and addressing throughout the summer."
Hopkins has already moved to double the
number of review boards at its medical centers; strengthened
training for board members and faculty in the regulations
governing clinical trials; and implement a meeting review
process to require full discussion and documentation by review
boards.
Roche died of acute respiratory distress
less than a month after inhaling hexamethonium. The chemical
was central to a clinical trial focused on why the airways
of healthy people such as Roche remain open after exposure
to allergens and other irritants.
After Roche's death, Hopkins was buffeted
by scathing criticism of research at its hospitals and affiliates,
as well as work by a faculty member as far afield as India.
In late August, the Maryland Court of
Appeals allowed lawsuits to proceed against the Kennedy Krieger
Institute over a lead paint study in low-income Baltimore
neighborhoods.
The judges said participants in the study,
which was overseen by a Hopkins board, were purposely misled
about the dangers involved. After the court ruling, the Office
for Human Research Protections opened a second Hopkins investigation.
Washington
Post September 8, 2001; Page B02
| 