Shifting the consensus on sharing sensitive data

Researchers who work with human participants face competing mandates. Funding organizations increasingly require applicants to develop data management plans that discuss data sharing. Journals increasingly require that authors make accessible the data underlying articles. Institutional Review Boards (IRBs), in contrast, neither encourage scholars to discuss data sharing with study participants nor discourage the withholding of data. Because funders and publishers often defer to IRB mandates, data generated through interaction with human participants -- including data that could be unproblematically shared -- are often only available to the researcher who collected them. We report on two workshops for IRB personnel (a third is planned for May 2018), and a study we conducted of the guidance offered by 50 IRBs at R1 highest research activity universities. While workshop attendees have been open to the ethical sharing of research data, our initial analysis suggests that IRB guidelines rarely mention data sharing. We also propose guidance and informed consent language that anticipate data sharing, around which we seek to build consensus among IRBs.