The empirical turn in bioethics and the deliberative turn in democracy theory occurred at around the same time, one at the intersection of bioethics and social science,1 2 the other at the intersection of political philosophy and political science.3–5 Empirical bioethics and deliberative democratic approaches both engage with immediate problems in policy and practice with normative intent, so it was perhaps inevitable that they would eventually find one another,6–8 and that deliberative research would become more common in bioethics.9 This commentary (...) responds to a paper by Ford et al, who ran a single jury of 18 people in Brighton in the UK in June 2018.10 The jury were asked to consider whether secondary research use of non-structured, free-text data from medical records should be permitted. I will consider three issues: the effort required in deliberative work, the challenge of sampling and selection of jurors and the basis for drawing normative conclusions from deliberative research. First: this report makes clear the effort involved in running a jury. These researchers contracted an external agency experienced in the method, established an oversight committee of policy-makers with power to act, developed expert witness briefs, selected witnesses and coordinated their testimony, and created a questionnaire to measure participants’ views, activities to support their understanding, and a scenario for consideration. …. (shrink)

The articles in this issue explore a number of difficult choices in medical care and research. They investigate ethical complexity in a range of decisions faced by policymakers and clinicians, and offer new evidence or normative approaches for navigating this complexity. In this issue’s feature article, Ford and colleagues engage with an ethical challenge faced by policymakers in relation to health research: should free text data contained in medical records be shared for research purposes?1 While some types of data from (...) electronic medical records are used widely in health research, privacy concerns have limited the use of free text data specifically. Ford and colleagues highlight the richness of the clinical information in free text and its centrality to some areas of healthcare such as mental health and primary care. In their view, while text analytic technology is developing rapidly, ‘policy makers routinely judge that the risk of re-identifying patients from text data is too high for these data uses’. Their citizens’ jury study in the UK found that participants were largely in favour of medical free text data being shared outside the NHS for research using an opt-out model: ‘the majority of jurors believed the benefits of sharing data for research outweighed the privacy risks’. In their commentary on this research, Largent and Morain raise some important issues about policymakers’ use of empirical findings …. (shrink)