Researchers are increasingly reluctant to share the background details of their studies with other scientists according to new results from a survey of authors who published papers in the Annals of Internal Medicine in the last five years. This downward trend in researchers’ willingness to disclose such information is, unfortunately, at odds with the current surge in efforts to facilitate access to the types of study specifics that are vital to reproducing results.
Increasing transparency in research—by sharing the nitty-gritty details of studies that don’t make it into the published accounts, such as preliminary qualifying test results for clinical trial participants—is a hot topic. A report released on 29 March by the Institute of Medicine (IOM) based in Washington, DC, entitled “Sharing Clinical Research Data: A Workshop,” concluded that giving other scientists access to information from studies was increasingly important for the research community. But the report did acknowledge that researchers sometimes have fears that the data they share, for example clinical results, might be misused or misinterpreted if not enough attention is given to how the data were originally collected.
“The biomedical industry lags behind the rest of the world in how we share information,” says Sharon Terry, chair of the IOM workshop committee and president of the Genetic Alliance, a Washington, DC-based health advocacy group that focuses on issues related to gene testing. “We need to catch up with the other industries that have figured out ways to share data and still protect it.”
In the new survey, a majority of researchers said that they would be willing to share study materials with their colleagues, according to the results presented by Christine Laine, the editor-in-chief of the Annals of Internal Medicine, at the International Congress on Peer Review and Biomedical Publication held in Chicago today.
The findings came from theoretical questions answered by 389 respondents who published papers in the Annals of Internal Medicine between 2008 and 2012. During that period, 71% said they would share their study protocols beyond what was in the methods, and 72% were willing to share the full statistical methods used to analyze the data, including the computer algorithms employed. However, only 54% were willing to share all the data collected during the study, including information that didn’t end up being in the final report. Most of the researchers who answered the survey questions also added extra provisos under which they would share these types of information; for example some would only do so in response to a personal request from an interested party (rather than depositing the information in, for example, an online data bank).
What concerns the surveyors most is that over the five-year survey period the responses showed a noticeable decline in scientists’ willingness to share details about their study protocols. Based on the replies to the theoretical questions, around 80% of the survey respondents in 2008 said they would share additional details about their study protocols, beyond what is described in the methods section of the paper, but by 2012 only 60% were willing to provide colleagues with that information, a significant difference. There was a similar—but not statistically significant—slip in researchers’ willingness to share data. When the survey started, about 60% of researchers said they’d share raw data, but five years later that number had dropped to 45%.
“It was surprising to find that people seem less willing to share in more recent times,” says Laine. However, she pointed out that the culture of research is to work hard on your studies, collect data, and then try to get as much information as possible out of your own data, first. She adds that surveys may be a “baby step” toward getting researchers to view sharing information as a favorable thing to do.
Although Laine says there aren’t any plans to make information sharing a condition of publishing papers in the Annals of Internal Medicine, other journals are experimenting with ways to embed more detail into the research they publish. In July, the New England Journal of Medicine became the first journal to hyperlink additional study information through TrialShare, an instant-access platform developed with funding from the US National Institute of Allergy and Infectious Diseases, in Bethesda, Maryland, by the Immune Tolerance Network, based in Seattle.
But not everyone sees the community of researchers as reluctant to share. Terry says her experience at the IOM workshop was that people had already seen the handwriting on the wall. “It wasn’t a question of why we should share; the concern was how to share. Figuring out the nuts and bolts will be easier to do.”