Altman, Lawrence T. "British Cast Spotlight on Misconduct in Scientific Research," New York

 

Altman, Lawrence T. "British Cast Spotlight on Misconduct in Scientific Research," New York
Times, 9 June 1998, p. F3.

LITTLE can rival the misery of a medical researcher who discovers that a co-author has
committed scientific misconduct. One who feels that most keenly is Dr. Cameron Bowie,
emeritus director of public health in Somerset, England.

Last week, Dr. Bowie retracted a published paper that had been influential in developing health
policy on community care in England and that was incorporated in national testing of public
health doctors. The paper, on the need for community-based care for the disabled and the cost
effectiveness of such care, had won acclaim for Dr. Bowie and his coauthor, Dr. Mark H.
Williams. Fellow researchers had called it a fundamental report.

Dr. Bowie had trumpeted the findings with gusto until British officials found that Dr. Williams
had committed scientific misconduct, including falsifying statistics in another research paper. Dr.
Williams, a trusted friend, had put Dr. Bowie's scientific credentials into question, and in an
interview last week, Dr. Bowie said his "reaction was one of disbelief that it could have
happened."

The British Medical Journal published Dr. Bowie's retraction in an issue focusing on various
forms of scientific misconduct like fabrication of data, plagiarism, redundant publication and
conflict of interest between reviewers and authors.

Citing a string of recent cases, the journal's editor, Dr. Richard Smith, said that scientific
misconduct was a bigger problem than scientists were willing to admit and called for a national
body with powers to investigate researchers without warning.

Though most scientists have contended that scientific misconduct is a trivial problem, it has been
the focus of Congressional inquiries in the United States, and after years of inactivity is drawing
increasing attention elsewhere. Medical leaders say that scientific misconduct flourishes, despite
their efforts to encourage research institutions to eliminate it. The magnitude of misconduct is
unknown and its detection is largely serendipitous. It is an international problem because
scientists from one country routinely publish in another country's journals.

This summer, a White House committee is to review a final draft of a plan to develop uniform
Governmentwide policies for research misconduct in the United States.

Dr. Bowie's retracted paper described the unmet needs for ways to allow 181 severely disabled
adults to live independently. The paper said the patients were interviewed and examined in 1989
and 1990. The study found that health professionals inadequately monitored such needs but that
assistance from a social worker in a coordinated approach could successfully improve care at
minimal cost. Dr. Bowie and Dr. Williams published the research in 1993.

But in 1996 Dr. Bowie was shocked to learn that Dr. Williams, who had moved to a nearby
university, faced disciplinary action for falsifying statistics in another research paper, forging
letters of recommendation and citing bogus degrees and qualifications to secure academic
positions.

Dr. Bowie, hoping to find evidence to support Dr. Williams and to forestall any challenge to their
joint paper, searched for the questionnaires and computerized records of their study but could not
find them. So, using Dr. Williams's list, he visited the patients to check they had been
interviewed. None said they had.

In February, acting on the other charges, British officials revoked Dr. Williams's medical license.
Dr. Bowie said he moved to retract the paper because he "could not assume the information was
bona fide." Dr. Bowie said he had confronted Dr. Williams, who at their most recent meeting last
week repeated denials that he had fabricated the data.

Efforts to reach Dr. Williams by telephone were unsuccessful.

Editors say co-authors must check each other's original data, not just the compilations, before
publishing a report. But such checking is impractical, particularly for large multi-center studies,
Dr. Bowie said.

Describing himself as naive for not checking Dr. Williams's data, Dr. Bowie said he now
believed that "when research findings are important, a co-author must be more suspicious than
you otherwise might be" and must scrutinize the raw data.

Researchers also need to store original records systematically so that years later any qualified
scientist could study them for any number of reasons beyond checking their validity, Dr. Bowie
said.

No one has thoroughly reviewed Dr. Williams's other publications to determine if they were
flawed, said Dr. Richard Smith, editor of The British Medical Journal.

Misconduct in research, if it is detected, is usually lethal to a scientist's career. Yet an unknown
number of scientists plagiarize, fabricate and falsify evidence, apparently believing they will not
get caught.

"Editors have decided that we can no longer ignore misconduct," Dr. Smith said, and recent
experience shows "that once editors begin to pay serious attention to misconduct, it is there
before their eyes."

Because taxpayers often finance medical care and because fraudulent research can put lives at
risk, protecting the public from misconduct in research is a public health measure similar to
protecting the water supply or safety of foods.

"Surely public concern about the entry of erroneous material in biomedical publication on health
and disease is at least of equal importance" as such standard health protection measures, Dr.
Michael J. Farthing, a leading gastroenterologist, wrote in The British Medical Journal. Dr.
Farthing is editor of Gut, a British medical journal.

Although the journal focused on scientific misconduct in England, the message holds elsewhere.

Dr. Drummond Rennie, a deputy editor of The Journal of the American Medical Association,
advised British officials in an article in The British Medical Journal to learn from experience in
the United States. In the 1980s, when Congressional committees looked into research
misconduct, responses from institutions were often "slow, bungled, idiosyncratic, and unfair to
almost everyone," Dr. Rennie said.

Recalling his experience as deputy editor of The New England Journal of Medicine 20 years ago
when it published a fraudulent research report, Dr. Rennie cited an editorial written by Dr.
Arnold S. Relman, then the journal's editor: "A request from an editor for primary data to
support the honesty of an author's findings in a manuscript under review would probably poison
the air and make civil discourse between authors and editors even more difficult than it is now."

In the British journal, Dr. Farthing wrote: "From an editor's point of view, doing one's duty is
simplified if fraudulent material is actually published. In this case, retraction of explanation is
required, the matter is in the public domain, and the offenders face public disgrace."

To aid in dealing with a more difficult problem - what to do with an unpublished suspect
manuscript on an editor's desk - Dr. Smith led the creation of a self-help group for editors.
Formation of the group grew out of an informal meeting in 1997 when a group of editors, mostly
in England, discussed the lack of clear guidelines on scientific misconduct.

At a second meeting last fall, a larger group of participants "confirmed our worst suspicions:
misconduct continues to thrive in all its various forms," Dr. Farthing wrote.

Participants said they were frustrated at having only limited powers to deal with questions of
research dishonesty that were raised in editing manuscripts. Editors contend they have neither
mandate nor money to investigate. Most said they usually rejected papers without informing
responsible officials at the medical centers where the authors worked. And no one tracks how
often the manuscripts suspected of being dishonest were published elsewhere.

Dr. Smith said that in its first year the group reviewed 25 cases without knowing the names of the
authors. The problems included plagiarism and conducting trials without receiving permission
from an ethics committee. When the group referred a case to the British licensing council
involving a doctor who said he had tested an unusual combination of drugs in many patients, Dr.
Smith said, they learned that the author had already lost his license.

Proposals have been made for journals to randomly audit papers as a quality control measure. But
editors have repeatedly rejected them.