Anderson, Christopher. "Growing Up In Public," Nature 355 (2 January 1992), p. 6.
Here, in its entirety, is this brief statement:
If the modern era of scientific misconduct wasborn two and a half years ago when the National
Institutes of Health (NIH) created the Office of Scientific Integrity (OSI), 1991 was is awkward
adolescence. The year opened with a nation watching the investigations of AIDS pioneer Robert
Gallo and the immunologist Thereza Imanishi-Kari (and by extension, her co-author David
Baltimore), as well as open warfare over the operation of the OSI. And unfortunately, it closed
just the same way.
In the intervening 12 months, Suzanne Hadley resigned as the deputy director of OSI, and
Representative John Dingell (Democrat, Michigan) strongly criticized NIH for its bungled
handling of the whole issue. Other than that, not much changed. Investigation of scientific
misconduct was a mess last year, and it is a mess today.
However, 1992 may be the year in which misconduct grows up. For one thing, the investigations
of Imanishi-Kari and Gallo - OSI's flagship cases - seem to be winding down, though slowly ...
And although those cases have been long, ugly affairs, they have opened up the misconduct
system as never before.
Through congressional hearings, a phenomenal amount of news coverage, and the attention of
virtually every element of the scientific community, the pitfallsof misconduct investigating are
now a matter of public record. Leaks are one problem. So are inconsistent procedures (for
instance, prominent researchers got special review committees, although others did not). In both
the Imanishi-Kari and the Gallo case, NIH investigators were often reduced to a role of following
up allegations in the press, which made nearly everyone but Dingell uncomfortable. In an
important debate over'due process' in OSI investigations has, intentionally or not, essentially
halted several cases.
Even OSI admits that some of its most prominent investigations were badly handled. But it has
also learned some tricks on the job: to avoid leaks, sensitive drafts reports now go only to
principal parties, and OSI is increasingly employing forensic and statistical analysis to add some
quantitative rigour to what has often been a disquietingly subjective process. Investigations now
focus on whether misconduct occurred, and no longer stumble on the question of a researcher's
intent. As OSI discovered, claims of "unintentional" misconduct have flummoxed many
university investigations, even when they turned out to be red herrings that obscured clear
abrogation of scientific responsibilities.
Other changes at OSI are coming from outside. After losing a lawsuit that challenged the way it
developed its procedures, OSI published a set of proposed new rules last year. Public comments
were generally scathing, mostly focused on the proposed definition of misconduct, which
included, together with the usual "fabrication, falsification and plagiarism," the category of
"other practices that seriously deviate from those that are commonly accepted from the scientific
community." An NIH advisory committee has recommended that the catchall phrase be changed
to"other fraudulent activities in proposing, conducting, reporting or reviewing research," a
definition that OSI says it can live with.
The committee also proposed - and NIH agreed - that OSI's staff be increased from 19 to 28,
including, for the first time, three lawyers (OSI investigators havetraditionally been scientists).
And the committee recommended open hearings, in which accused and accusercan face each
other. OSI director Jules Hallum opposesthat move, arguing that face-to-face confrontations
"would destroy the willingness of whistle-blowers to come forward."
Even as it reconsiders it role, however, OSI languishes in a sort of bureaucratic limbo. Both
Congress and some Administration officials are contemplating taking OSI away from NIH and
placing it instead under the wing of the Department of Health and Human Services, NIH's parent
agency. When Dingell held a hearing last summer accusing Bernadine Healy, the NIH director, of
a conflict of interest in an OSI investigation of a case at the Cleveland Clinic, Healy's former
Institution, it only reinforced the concern that OSI might be more independent if it operated like
any other government investigative office - firmly entrenched in the bureaucracy. If the
administration does not propose the move itself, congressional legislation to that effect may
appear this year.
But the worst may be behind the misconduct controversy, if not OSI itself. Perhaps the most
encouraging sign is the improving quality of university investigations. Whereas academic panels
in the past often erred on the side of finding no misconduct, Hallum says that recent university
investigations, such as two last year at the California Institute of Technology, have been more
thorough and fair. "If they keep it up, they may put us out of business," he says. Nevertheless,
until conspicuous mishandlings such as Imanishi-Kari's inquiry at the Massachusetts of
Technology (sic) and that of whistle-blower Erdem Cantekin at the University of Pittsburgh (...)
become a thing of the past, OSI wants to keep tight reigns onthe universities. The proposed new
procedures would allow OSI to intercede earlier in an academic investigations if things seem to
be going awry, and Hallum is hoping to have the rules clarified to give OSI explicit authorization
to investigate the universities themselves, to explore the possibilities of cover-ups.