Biomedical research is shifting to become more open and transparent by providing increasing amounts of information about funding, conflicts of interest and data sharing in their publications, according to a survey of recent papers.
John Ioannidis at Stanford University in California and colleagues examined 149 biomedical research papers published between 2015 and 2017 to see how many included information on key indicators of transparency, such as who funded the work, potential conflicts of interest on the part of the researchers, and the availability of the raw data and complete research protocols.
They found that the majority of papers contained statements on funding and conflicts of interest (69% and 65%, respectively), and almost one in five mentioned publicly available data — although only one paper included a link to a full study protocol. The work was published in PLoS Biology on 20 November1.
This was a big improvement on the results of a previous study by some of the same researchers, which found that, in a sample of 441 articles published between 2000 and 2014, most contained little, if any, information on funding, conflicts or data sharing2 (see 'Opening up').
“This is a reason to be optimistic,” says co-author Joshua Wallach, who studies research practices at the Yale School of Public Health in New Haven, Connecticut.
Statements on funding and conflicts are important because they are sources of potential bias, which can affect how studies are designed and conducted, he says. And easy access to data and experimental protocols is vital for replicating scientific findings. “The fact we are now seeing statements on data sharing shows the culture is shifting,” says Wallach.
Around 5% of the papers studied were a full or partial replication of previous work — an encouraging sign, but still a lower proportion than Wallach would like. “We need greater adoption of replication,” he says. “It shouldn’t be treated as second-class science.”
Although the results indicate that the biomedical community is showing greater acceptance of the importance of open science, Wallach credits some of the progress to changes at journals: many now require statements on funding and conflict of interest, and have polices on data sharing.
But Wallach says that journals could do more to promote protocol sharing. And public repositories such as PubMed should make it easier to find information about funding, conflicts and data for people without access to the full paper, he says.
Jim Woodgett, director of research at the Lunenfeld-Tanenbaum Research Institute in Toronto, Canada, says that papers like this are changing how scientists work by revealing how little transparency there has been in the past. But this kind of culture change takes a lot of effort, and universities and journals need to take the lead. “Nobody says no to a journal requirement,” he says.