In the aftermath of now discredited genomic research involving
a former Duke cancer researcher, Duke University is taking several steps to
safeguard data transparency and accountability and the chain of data evidence,
a top Duke research official told the Academic Council Thursday.
It's impossible to develop a system that will completely eliminate
academic fraud if a researcher is intent on misconduct, said Sally Kornbluth,
vice dean for the basic sciences in the School of Medicine. "But this case highlighted that we can
take a hard look at the infrastructure and the culture around research to
reduce it," she said. "And we
can provide safeguards for people who are trying to do the right things, but
make errors or are guilty of sloppiness. In research, errors are more likely to
be made through sloppiness than through fraud."
Among the changes will be a revised information technology system
for research data to ensure the data that is originally entered in a study is
the same data used in the end analysis and to document whenever data is altered.
Former Duke faculty member Anil Potti is currently being
investigated for research misconduct. A
key element of the investigation is that the original data developed in several
Potti research studies was not the data used in its analysis, Kornbluth said.
Kornbluth said the planned information technology will
create a "data lockbox."
"We want to see when changes are made and who makes the
changes," she said. "Had that
been in place in this case, the discrepancy in data would have been revealed
earlier."
But the Potti case also revealed the difficulty of
supervising modern medical research where it is unlikely that any one researcher
or even a single lab would have a wide set of necessary skills, including running
complex biostatistical calculations.
Kornbluth said there was "a dire need" in many
research labs for quantitative expertise to review data. As a result, Duke has taken steps to embed
biostatisticians in clinical research groups.
Already this change has attracted attention from other research
institutions looking to reduce errors in data analysis.
"We also want to ensure a culture of loyal
dissent," Kornbluth told faculty members.
"We want people to feel free to raise concerns when they see
research problems. This is through a
combination of creating places to raise concerns anonymously and to have
leadership throughout the institution where people feel comfortable to raise
their hands and speak out."
The university made errors in its follow-up to the concerns
expressed by the National Cancer Institute (NCI) in the Potti case, Kornbluth
said. The NCI concerns were prompted by
earlier investigations made by two MD Anderson Cancer Center researchers in
Texas that cast doubt on the Potti data.
After hearing of the NCI concerns, Duke brought in external
reviewers to examine Potti's data and determine whether the analysis was
correct. The team confirmed the
research, but Kornbluth said what was missed at the time was that the original
dataset was not used in the analysis.
"You often walk a fine line between trusting your
faculty, particularly those with a stellar track record," she said. "We thought this was an arcane
statistical review. They were asked to
validate the statistical methodology. We
gave the investigative team the data and they cranked through it and confirmed
the results. What we didn't appreciate
was that the raw data was different from the data used in the analysis.
"There was no intent to hide anything. We had a very vigorous response from the [Potti
lab] that the data should speak for itself, and we let that happen. But it was a fundamental mistake not to give
the investigative team the Texas concerns about the original data. They didn't know about those concerns, and I
wish they had."