Sharing Of Scientific Data Posed As Way To Diminish Fraud

WASHINGTON-Sharing notebooks and other data with someone outside their laboratory is an idea that is anathema to many scientists. But they may need to get used to it as part of the price of performing science with public funds. The search for a better system to record, retain, and share data is emerging as a key issue in the ongoing debate over how to curb scientific misconduct. It promises to remain a significant issue long after scientists have finished arguing about whether Rep. John Dingell (D-Mich.) was justified in his grilling of Nobel laureate David Baltimore last month in connection with allegations of errors in a 1986 Cell paper that the MIT biologist coauthored. At the same time, the two days of hearings before Dingell's oversight and investigations panel of the House Energy and Commerce Committee (see excerpts on page 11) have highlighted some of the problems with how researchers keep their data. In particular, many scientists say that they are troubled by the admission of another Cell coauthor, Tufts University immunologist Thereza Imanishi-Kari, that she occasionally recorded data from experiments months after they had been performed. Imanishi-Kari also said that her notebooks were often in disarray, a condition that several scientists at the hearing admitted also applied to their own work. "I can't imagine any disadvantages in having your notes properly organized," says Arthur Ruben-stein, chairman of medicine at the University of Chicago and of an Institute of Medicine committee examining the responsible conduct of research. "If you do an experiment one day, you should write it down the same day. I certainly wouldn't want to defend her approach to recording data." Rubenstein's comments may reflect the fact that the era in which scientists could wear their legendary disorder as a badge of creative genius is coming to an end. Its demise is hastened by the growing concern about scientific misconduct, whether real or imagined, which has caused the government, universities, and professional societies to seek ways to make scientists more accountable for their use of public funds. Since scientists as a rule report progress by publishing their findings, the manner in which that research is presented has come under particular scrutiny. That is the key element in the allegations of immunologist Margot O'Toole against the Cell paper, and the reason behind Dingell's ordering of a United States Secret Service analysis of the notebooks, computer tapes, and photographs on which the paper was based. But the current debate on misconduct goes far beyond the integrity of the scientific literature. Although that question arises each time a scientist is found to have published a fraudulent paper, the new wave of concern touches on whether scientists are adequate custodians for a public trust, that is, the work that they perform as a result of receiving government money. And beyond that question lies an even more troubling one: Are scientific data being kept in a way that can be understood by an independent party, should an outside investigation become necessary? Any hint that Congress or the federal government may be contemplating standards for the retention of data sends shivers down the spines of scientists worried about how such rules could impinge on their working conditions. "One of the reasons that it's such a contentious issue is that we have no idea of the possibility of abuse," says Robert Weinberg, a biologist who works under Baltimore at the Whitehead Institute of Biomedical Research and who was a member of the Institute of Medicine committee. "Who would make such requests? Would there be any preconditions imposed, such as previous experience in the field?" Such fears are both groundless and beside the point, believes Rachelle Hollander, who as program officer for the ethics and values program at NSF has watched the debate over scientific misconduct blossom into a questioning of bow science is carried out. "At one time, scientists might have said, 'Don't bother me about safety in the lab. That would only get in the way of productive science.' But we no longer think that's an acceptable response," says Hollander. "Likewise, recording your data in a disorganized and messy fashion may result in poor science. I think that scientists would do well to clean up the way they keep their data before someone else tries to do it for them." Weinberg agrees that it is certainly possible for scientists to do better. "One of the constructive outcomes of the Dingell hearing is that it has illustrated the need for lab notebooks that can be read by a third party," he says. "That's not an unreasonable request." However, Weinberg and other scientists recoil in horror from any move toward a mandatory system of recordkeeping and access. And they believe that they must remain vigilant. Weinberg says that even the Institute of Medicine's seemingly mild recommendation that institutions should adopt guidelines on professional standards for research scientists could lead to undue restrictions. "There are some people who are afraid that it could be the beginning of a science police that will ultimately lead to a strangling of creative science in the United States," says Weinberg. "It sounds nice to propose that universities adopt such standards, but what if NIH isn't happy with what they submit and wants to decide on its own what's best for everybody?" That scenario would be a radical break from its past. Present NIH rules barely touch on the subjects of retention, ownership, and sharing of data. "The current policy in the guide [for grants and contracts] stipulates that data must be kept for three years beyond the end of the grant," explains Janet Newburgh of NIH's institutional liaison office. "To this point that has been interpreted to mean financial data, although I suppose that it could be extended to cover all data relevant to the grant." Federal rules are equally vague on the other issues, she admits. "Technically, the data belongs to the grantee's institution, and when the grantee leaves, the data should be retained by that institution," she says. "In practice, however, grantees have treated the data as their own. I suspect that most people feel that the data should be kept by the person who can make the best use of them, and that usually means the principal investigator." As for the sharing of data, Newburgh says that NIH "has no policy as of now." Although the government is silent on the issue, a growing number of journals have adopted policies that address the sharing of data. For several years, says Cell editor Benjamin Lewin, his journal has extracted a promise from all prospective authors that they will share reagents and other relevant materials from the published experiment with any scientist who requests them. Any scientist who fails to agree with such terms, Lewin notes, doesn't get into print. "Our policy is based on the premise that, if you aren't willing to distribute unique material, then your work is not reproducible," he explains. "And if it's not reproducible, then we don't want it to appear in Cell." Lewin says that the journal occasionally gets a call from someone who complains that an author is refusing to send something that has been requested, but he adds that Cell has never intervened directly. At the same time, Lewin draws a distinction between sharing reagents and sharing notebooks. "Lab notebooks are one's private records. I don't think they are the samething as reagents." Even so, he notes that many questions about conflicting results can be solved without resorting to the notebooks themselves. "A lot of the problems stem from the use of different procedures, and they can be resolved by sitting down with the author and going over what he or she actually did," he says. Imperfect or not, the ability of scientists to deal openly with criticism of their actions in the lab was openly questioned by Dingell at last month's hearings. And there are some observers who believe that many researchers are much too defensive about their freedom to practice science. "I think that David Baltimore picked a very poor analogy when he compared an outside agency's coming in to look at a scientist's notebooks to an IRS audit of your tax records," says one knowledgeable House aide not involved in the Dingell investigation. "The reason that people fear an IRS audit is not because their files are messy but because they often cheat on their taxes and don't want to be caught. I certainly don't think he [Baltimore] wanted to imply that scientists are afraid of an independent audit of their notebooks because they have something to hide."

Sharing Of Scientific Data Posed As Way To Diminish Fraud

Interested in reading more?

Become a Member of