Hybrid theories, psychological plausibility, and the human/animal divide.

A hybrid theory is any moral theory according to which different classes of individuals ought to be treated according to different principles. We argue that some hybrid theories are able to meet standards of psychological plausibility, by which we mean that it's feasible for ordinary human beings to understand and act in accord with them. Insofar as psychological plausibility is a theoretical virtue, then, such hybrid theories deserve more serious consideration. To make the case for this view, we explain what psychological plausibility is and why we might value it, why the human/animal divide appears to be an entrenched feature of human psychology, and why Robert Nozick's hybrid theory doesn't go far enough. Finally, we make the case that a more promising psychologically plausible hybrid theory, with respect to humans and animals, will be (at least) at tribrid theory-that is, positing three domains rather than two.

Protecting health privacy even when privacy is lost.

The standard approach to protecting privacy in healthcare aims to control access to personal information. We cannot regain control of information after it has been shared, so we must restrict access from the start. This 'control' conception of privacy conflicts with data-intensive initiatives like precision medicine and learning health systems, as they require patients to give up significant control of their information. Without adequate alternatives to the control-based approach, such data-intensive programmes appear to require a loss of privacy. This paper argues that the control view of privacy is shortsighted and overlooks important ways to protect health information even when widely shared. To prepare for a world where we no longer control our data, we must pursue three alternative strategies: obfuscate health data, penalise the misuse of health data and improve transparency around who shares our data and for what purposes. Prioritising these strategies is necessary when health data are widely shared both within and outside of the health system.

Expanded FDA regulation of health and wellness apps.

This paper argues that the Food and Drug Administration's (FDA) policy for health and wellness apps is ethically problematic. Currently, the FDA does not regulate health and wellness apps that are not intended for medical use. As a result of this hands-off policy, preventing harm to consumers is left primarily to developers and app marketplaces. We argue that the FDA's duties to prevent harm and maintain accountability to the American public require that they play a much stronger role. We also discuss concerns about efficiency and fostering innovation, and argue that while they should help shape FDA regulation of health and wellness apps, they do not justify complete absence of FDA involvement.

Communicating Identifiability Risks to Biobank Donors.

Recent highly publicized privacy breaches in healthcare and genomics research have led many to question whether current standards of data protection are adequate. Improvements in de-identification techniques, combined with pervasive data sharing, have increased the likelihood that external parties can track individuals across multiple databases. This article focuses on the communication of identifiability risks in the process of obtaining consent for donation and research. Most ethical discussions of identifiability risks have focused on the severity of the risk and how it might be mitigated, and what precisely is at stake in pervasive data sharing. However, there has been little discussion of whether and how to communicate the risk to potential donors. We review the ethical arguments behind favoring different types of risk communication in the consent process, and outline how identifiability concerns can be incorporated into either a detailed or a simplified method of communicating risks during the consent process.

Biobank Participants' Attitudes Toward Data Sharing and Privacy: The Role of Trust in Reducing Perceived Risks.

Biobank participants are often unaware of possible uses of their genetic and health information, despite explicit descriptions of those uses in consent forms. To explore why this misunderstanding persists, we conducted semi-structured interviews and knowledge tests with 22 participants who had recently enrolled in a research biobank. Results indicated that participants lacked understanding of privacy and data-sharing topics but were mostly unconcerned about associated risks. Participants described their answers on the knowledge test as largely driven by their trust in the healthcare system, not by a close reading of the information presented to them. This finding may help explain the difficulties in increasing participant understanding of privacy-related topics, even when such information is clearly presented in biobank consent forms.

Collecting Biospecimens and Obtaining Biobank Consent From Patients in an Academic Health Care Setting: Practical and Ethical Considerations.

Academic health centers and health systems increasingly ask patients to enroll in research biobanks as part of standard care, raising important practical and ethical questions for integrating biobank consent processes into health care settings. This article aims to assist academic health centers and health systems considering implementing these integrated consent processes by outlining the 5 main issues-and the key practical and ethical considerations for each issue-that Indiana University Health and the Indiana Biobank faced when integrating biobank consent into their health system, as well as the key obstacles encountered. The 5 main issues to consider include the specimen to collect (leftover, new collection, or add-ons to clinical tests), whether to use opt-in or opt-out consent, where to approach patients, how to effectively use digital tools for consent, and how to appropriately simplify consent information.

Biobank Participants' Attitudes toward Requiring Understanding for Biobank Consent.

Biobank participants often do not understand the information they are provided during the informed consent process. Ethicists and other stakeholders have disagreed, however, on the appropriate response to these failures in understanding. This paper describes an attempt to address this issue by conducting knowledge tests with 22 recent biobank enrollees, followed by in-depth, semistructured interviews about the goal of understanding in biobank consent. The interviews revealed that while biobank enrollees thought the information on the knowledge test was important, they did not think that performance on the test should affect whether individuals are permitted to enroll in a biobank. Three main themes emerged from the interviews: helping others by contributing to research is more important than understanding consent forms, less understanding is required because biobank-based research is low risk, and only a small amount of information in the consent form is really essential. These perspectives should be considered in discussing the ethics and governance of biobank consent processes.

Conflicting roles for humans in learning health systems and AI-enabled healthcare.

The goals of learning health systems (LHS) and of AI in medicine overlap in many respects. Both require significant improvements in data sharing and IT infrastructure, aim to provide more personalized care for patients, and strive to break down traditional barriers between research and care. However, the defining features of LHS and AI diverge when it comes to the people involved in medicine, both patients and providers. LHS aim to enhance physician-patient relationships while developments in AI emphasize a physicianless experience. LHS also encourage better coordination of specialists across the health system, but AI aims to replace many specialists with technology and algorithms. This paper argues that these points of conflict may require a reconsideration of the role of humans in medical decision making. Although it is currently unclear to what extent machines will replace humans in healthcare, the parallel development of LHS and AI raises important questions about the exact role for humans within AI-enabled healthcare.

In Defence of Forgetting Evil: A Reply to Pilkington on Conscientious Objection.

In a recent article for this journal, Bryan Pilkington (2019) makes a number of critical observations about one of our arguments for non-traditional medical conscientious objectors' duty to refer. Non-traditional conscientious objectors are those professionals who object to indirectly performing actions-like, say, referring to a physician who will perform an abortion. In our response here, we discuss his central objection and clarify our position on the role of value conflicts in non-traditional conscientious objection.

Adolescent Assent and Reconsent for Biobanking: Recent Developments and Emerging Ethical Issues.

Research biobanks that enroll minors face important practical, ethical, and regulatory challenges in reconsenting participants when they reach the age of 18. Federal regulations governing research in the United States provide minimal guidance and allow for a range of practices, including waiving the requirement to obtain reconsent. Some commentators have argued that institutional review boards should indeed grant such waivers, given the low risks of biobank-based research and the impracticality of contacting all participants when they turn 18. There is also significant ethical debate about the age at which adolescents can make authentic, autonomous decisions regarding their research participation. This paper reviews these issues in detail, describes the current state of the ethical discussion, and outlines evidence-based policies for enrolling minors into research biobanks.

Incorporating Biobank Consent into a Healthcare Setting: Challenges for Patient Understanding.

BACKGROUND: Biobank participants often do not understand much of the information they are provided as part of the informed consent process, despite numerous attempts at simplifying consent forms and improving their readability. We report the first assessment of biobank enrollees' comprehension under an "integrated consent" process, where patients were asked to enroll in a research biobank as part of their normal healthcare experience. A number of healthcare systems have implemented similar integrated consent processes for biobanking, but it is unknown how much patients understand after enrolling under these conditions. Methods: We recruited patients who enrolled in a biobank while in a healthcare setting when receiving ordinary care. We assessed knowledge of consent materials using 11 true/false questions drawn from a well-known biobank knowledge test. After reviewing the results from 114 participants, we revised the consent form and repeated the knowledge assessment with 144 different participants. Results: Participants scored poorly on the knowledge test in both rounds, with no significant differences in overall scores or individual items between the rounds. In Phase 1, participants answered 53% of the questions correctly, 25% incorrectly, and 22% "I don't know." In Phase 2, participants answered 53% of questions correctly, 24% incorrectly, and 23% "I don't know." Participants scored particularly poorly on questions about data sharing and accessing medical records. Conclusions: Enrollees under an integrated consent model had significant misunderstandings that persisted despite an attempt to improve information specifically about those topics in a consent form. These results raise challenges for current approaches that attribute misunderstanding to overly complex consent forms. They also suggest that the pressures of the clinic may compound other problems with patient understanding of biobank consent. As health systems increasingly blend research and care, they may need to rethink their approach to educating patients about participation in a biobank.

Genetic Data Aren't So Special: Causes and Implications of Reidentification.

Genetic information is widely thought to pose unique risks of reidentifying individuals. Genetic data reveals a great deal about who we are and, the standard view holds, should consequently be treated differently from other types of data. Contrary to this view, we argue that the dangers of reidentification for genetic and nongenetic data-including health, financial, and consumer information-are more similar than has been recognized. Before different requirements are imposed around sharing genetic information, proponents of the standard view must show that they are in fact necessary. We further argue that the similarities between genetic and nongenetic information have important implications for communicating risks during consent for health care and research. While patients and research participants need to be more aware of pervasive data-sharing practices, consent forms are the wrong place to provide this education. Instead, health systems should engage with patients throughout patient care to educate them about data-sharing practices.

Forget Evil: Autonomy, the Physician-Patient Relationship, and the Duty to Refer.

Aulisio and Arora argue that the moral significance of value imposition explains the moral distinction between traditional conscientious objection and non-traditional conscientious objection. The former objects to directly performing actions, whereas the latter objects to indirectly assisting actions on the grounds that indirectly assisting makes the actor morally complicit. Examples of non-traditional conscientious objection include objections to the duty to refer. Typically, we expect physicians who object to a practice to refer, but the non-traditional conscientious objector physician refuses to refer. Aulisio and Arora argue that physicians have a duty to refer because refusing to do so violates the patient's values. While we agree with Aulisio and Arora's conclusions, we argue value imposition cannot adequately explain the moral difference between traditional conscientious objection and non-traditional conscientious objection. Treating autonomy as the freedom to live in accordance with one's values, as Aulisio and Arora do, is a departure from traditional liberal conceptions of autonomy and consequently fails to explain the moral difference between the two kinds of objection. We outline how a traditional liberal understanding of autonomy would help in this regard, and we make two additional arguments-one that maintains that non-traditional conscientious objection undermines society's autonomy, and another that maintains that it undermines the physician-patient relationship-to establish why physicians have a duty to refer.