Understanding the Gap: A Cross-Sectional Survey of ELSI Scholars' Dissemination Practices and Translation Goals.

BACKGROUND: Researchers engaged in the study of the ethical, legal, and social implications (ELSI) of genetics and genomics are often publicly funded and intend their work to be in the public interest. These features of U.S. ELSI research create an imperative for these scholars to demonstrate the public utility of their work and the expectation that they engage in research that has potential to inform policy or practice outcomes. In support of the fulfillment of this "translational mandate," the Center for ELSI Resources and Analysis (CERA), funded by the National Human Genome Research Institute (NHGRI), aims to facilitate community-informed, ELSI research results synthesis and dissemination. However, little is known about how ELSI research scholars define the goals of translation and imagine the intended users of their research findings. METHODOLOGY: We distributed a Qualtrics survey to ELSI scholars that aimed to determine: (1) researchers' expectations for their research findings in relation to policy or practice outcomes, (2) the stakeholder groups researchers believe could benefit from their research findings, and (3) the methods researchers use to foster the uptake of their findings by those stakeholders. RESULTS: Most ELSI researchers surveyed thought there were stakeholders that could benefit from their research findings, including health care professionals, at-risk individuals, patients, and their family members, policy-makers, and researchers/scientists, and expected their research findings to inform the creation or revision of laws, policies, or practice guidelines. Most researchers planned to disseminate findings directly to relevant stakeholders, with fewer expecting dissemination support from research funders, universities, or other entities. CONCLUSION: The broad range of research topics, disciplines, and set of potential end users represented in ELSI reseach complicate the work of a knowledge broker. Nonetheless, the CERA can play an important role in disseminating ELSI results to relevant stakeholders. Further research should explore outreach mechanisms.

Moral Engagement and Disengagement in Health Care AI Development.

BACKGROUND: Machine learning (ML) is utilized increasingly in health care, and can pose harms to patients, clinicians, health systems, and the public. In response, regulators have proposed an approach that would shift more responsibility to ML developers for mitigating potential harms. To be effective, this approach requires ML developers to recognize, accept, and act on responsibility for mitigating harms. However, little is known regarding the perspectives of developers themselves regarding their obligations to mitigate harms. METHODS: We conducted 40 semi-structured interviews with developers of ML predictive analytics applications for health care in the United States. RESULTS: Participants varied widely in their perspectives on personal responsibility and included examples of both moral engagement and disengagement, albeit in a variety of forms. While most (70%) of participants made a statement indicative of moral engagement, most of these statements reflected an awareness of moral issues, while only a subset of these included additional elements of engagement such as recognizing responsibility, alignment with personal values, addressing conflicts of interests, and opportunities for action. Further, we identified eight distinct categories of moral disengagement reflecting efforts to minimize potential harms or deflect personal responsibility for preventing or mitigating harms. CONCLUSIONS: These findings suggest possible facilitators and barriers to the development of ethical ML that could act by encouraging moral engagement or discouraging moral disengagement. Regulatory approaches that depend on the ability of ML developers to recognize, accept, and act on responsibility for mitigating harms might have limited success without education and guidance for ML developers about the extent of their responsibilities and how to implement them.

The ELSI Virtual Forum, 30 Years of the Genome: Integrating and Applying ELSI Research.

This paper reports our analysis of the ELSI Virtual Forum: 30 Years of the Genome: Integrating and Applying ELSI Research, an online meeting of scholars focused on the ethical, legal, and social implications (ELSI) of genetics and genomics.

Developer Perspectives on Potential Harms of Machine Learning Predictive Analytics in Health Care: Qualitative Analysis.

BACKGROUND: Machine learning predictive analytics (MLPA) is increasingly used in health care to reduce costs and improve efficacy; it also has the potential to harm patients and trust in health care. Academic and regulatory leaders have proposed a variety of principles and guidelines to address the challenges of evaluating the safety of machine learning-based software in the health care context, but accepted practices do not yet exist. However, there appears to be a shift toward process-based regulatory paradigms that rely heavily on self-regulation. At the same time, little research has examined the perspectives about the harms of MLPA developers themselves, whose role will be essential in overcoming the "principles-to-practice" gap. OBJECTIVE: The objective of this study was to understand how MLPA developers of health care products perceived the potential harms of those products and their responses to recognized harms. METHODS: We interviewed 40 individuals who were developing MLPA tools for health care at 15 US-based organizations, including data scientists, software engineers, and those with mid- and high-level management roles. These 15 organizations were selected to represent a range of organizational types and sizes from the 106 that we previously identified. We asked developers about their perspectives on the potential harms of their work, factors that influence these harms, and their role in mitigation. We used standard qualitative analysis of transcribed interviews to identify themes in the data. RESULTS: We found that MLPA developers recognized a range of potential harms of MLPA to individuals, social groups, and the health care system, such as issues of privacy, bias, and system disruption. They also identified drivers of these harms related to the characteristics of machine learning and specific to the health care and commercial contexts in which the products are developed. MLPA developers also described strategies to respond to these drivers and potentially mitigate the harms. Opportunities included balancing algorithm performance goals with potential harms, emphasizing iterative integration of health care expertise, and fostering shared company values. However, their recognition of their own responsibility to address potential harms varied widely. CONCLUSIONS: Even though MLPA developers recognized that their products can harm patients, public, and even health systems, robust procedures to assess the potential for harms and the need for mitigation do not exist. Our findings suggest that, to the extent that new oversight paradigms rely on self-regulation, they will face serious challenges if harms are driven by features that developers consider inescapable in health care and business environments. Furthermore, effective self-regulation will require MLPA developers to accept responsibility for safety and efficacy and know how to act accordingly. Our results suggest that, at the very least, substantial education will be necessary to fill the "principles-to-practice" gap.

Stronger regulation of AI in biomedicine.

Regulatory agencies need to ensure the safety and equity of AI in biomedicine, and the time to do so is now.

To Understand Inequity, Bioethics Needs to Sort Things Out.

Bioethics is reexamining how to implement diversity, equity, inclusion, and justice concerns into scholarship. However, bioethicists should question the categories used to define diversity. The act of categorization is value laden, and classification systems confer power and benefits and generate harms. For example, what conditions count as disabilities? We should consider the equity implications of offering only "male" and "female" options for self-identification in health records. However, we should also interrogate all ideas about categorization, including how categories are formed, why they are formed, and who decides. Bioethicists cannot comprehend fully what equity and justice mean for underrepresented, underserved, or marginalized people until there is an understanding of how the boundaries of marginalization are created.

Trustworthiness matters: Building equitable and ethical science.

Trustworthy science requires research practices that center issues of ethics, equity, and inclusion. We announce the Leadership in the Equitable and Ethical Design (LEED) of Science, Technology, Mathematics, and Medicine (STEM) initiative to create best practices for integrating ethical expertise and fostering equitable collaboration.

Re-envisioning community genetics: community empowerment in preventive genomics.

As genomic technologies rapidly develop, polygenic scores (PGS) are entering into a growing conversation on how to improve precision in public health and prevent chronic disease. While the integration of PGS into public health and clinical services raises potential benefits, it also introduces potential harms. In particular, there is a high level of uncertainty about how to incorporate PGS into clinical settings in a manner that is equitable, just, and aligned with the long-term goals of many healthcare systems to support person-centered and value-based care. This paper argues that any conversation about whether and how to design and implement PGS clinical services requires dynamic engagement with local communities, patients, and families. These parties often face the consequences, both positive and negative, of such uncertainties and should therefore drive clinical translation. As a collaborative effort between hospital stakeholders, community partners, and researchers, this paper describes a community-empowered co-design process for addressing uncertainty and making programmatic decisions about the implementation of PGS into clinical services. We provide a framework for others interested in designing clinical programs that are responsive to, and inclusive and respectful of, local communities.

Epistemic Rights and Responsibilities of Digital Simulacra for Biomedicine.

Big data and AI have enabled digital simulation for prediction of future health states or behaviors of specific individuals, populations or humans in general. "Digital simulacra" use multimodal datasets to develop computational models that are virtual representations of people or groups, generating predictions of how systems evolve and react to interventions over time. These include digital twins and virtual patients for in silico clinical trials, both of which seek to transform research and health care by speeding innovation and bridging the epistemic gap between population-based research findings and their application to the individual. Nevertheless, digital simulacra mark a major milestone on a trajectory to embrace the epistemic culture of data science and a potential abandonment of medical epistemological concepts of causality and representation. In doing so, "data first" approaches potentially shift moral attention from actual patients and principles, such as equity, to simulated patients and patient data.

Not in my AI: Moral engagement and disengagement in health care AI development.

Machine learning predictive analytics (MLPA) are utilized increasingly in health care, but can pose harms to patients, clinicians, health systems, and the public. The dynamic nature of this technology creates unique challenges to evaluating safety and efficacy and minimizing harms. In response, regulators have proposed an approach that would shift more responsibility to MLPA developers for mitigating potential harms. To be effective, this approach requires MLPA developers to recognize, accept, and act on responsibility for mitigating harms. In interviews of 40 MLPA developers of health care applications in the United States, we found that a subset of ML developers made statements reflecting moral disengagement, representing several different potential rationales that could create distance between personal accountability and harms. However, we also found a different subset of ML developers who expressed recognition of their role in creating potential hazards, the moral weight of their design decisions, and a sense of responsibility for mitigating harms. We also found evidence of moral conflict and uncertainty about responsibility for averting harms as an individual developer working in a company. These findings suggest possible facilitators and barriers to the development of ethical ML that could act through encouragement of moral engagement or discouragement of moral disengagement. Regulatory approaches that depend on the ability of ML developers to recognize, accept, and act on responsibility for mitigating harms might have limited success without education and guidance for ML developers about the extent of their responsibilities and how to implement them.

Words matter: The language of difference in human genetics.

Diversity, equity, and inclusion efforts in academia are leading publishers and journals to re-examine their use of terminology for commonly used scientific variables. This reassessment of language is particularly important for human genetics, which is focused on identifying and explaining differences between individuals and populations. Recent guidance on the use of terms and symbols in clinical practice, research, and publications is beginning to acknowledge the ways that language and concepts of difference can be not only inaccurate but also harmful. To stop perpetuating historical wrongs, those of us who conduct and publish genetic research and provide genetic health care must understand the context of the terms we use and why some usages should be discontinued. In this article, we summarize critiques of terminology describing disability, sex, gender, race, ethnicity, and ancestry in research publications, laboratory reports, diagnostic codes, and pedigrees. We also highlight recommendations for alternative language that aims to make genetics more inclusive, rigorous, and ethically sound. Even though norms of acceptable language use are ever changing, it is the responsibility of genetics professionals to uncover biases ingrained in professional practice and training and to continually reassess the words we use to describe human difference because they cause harm to patients.

Latinx attitudes, barriers, and experiences with genetic counseling and testing: A systematic review.

As genetics is increasingly used across clinical settings, there is a need to understand the impact and experiences of diverse patients. This review systematically examined research literature on Latinx experiences with genetic counseling and genetic testing (GC/GT) in the United States, synthesizing key themes and knowledge gaps pertaining to both patient experience and hypothetical scenarios. Findings were based on a systematic search, inclusion, and thematic analysis of 81 empirical peer-reviewed articles published from January 1990 to July 2019 pertaining to Latinx populations and GC/GT. Studies most commonly addressed Latinas' perspectives on GC/GT in prenatal settings or for hereditary breast and ovarian cancer (HBOC). Costs, referrals, and communication were significant barriers to accessing genetic services for many Latinx patients, particularly those with low English proficiency (LEP). Studies highlighted difficulties accessing and communicating in healthcare settings, and how medical context and prior experience with healthcare workers and institutions influenced GC/GT decision-making. Providers' implicit biases about Latinx patients negatively impacted their care and impeded communication. Despite low awareness of cancer GT, Latinx patients often reported interest in learning more about GC/GT or unmet needs for GT discussion and provider involvement. This systematic review identified areas where providers can take action to improve Latinx experiences with GC/GT. Clinicians should elicit and respond to patient preferences about shared decision-making. For patients with low numeracy or LEP, providers should consider tailored educational and communication techniques. Most studies focused on HBOC and prenatal testing, and Latinx patients are heterogeneous, leaving many research questions about Latinx experience with GT/GC in other clinical areas.

Values and Practices to Strengthen Genetic Research Partnerships with Indigenous Communities.

Genetic datasets lack diversity and include very few data from Indigenous populations. Research models based on equitable partnership have the potential to increase Indigenous participation and have led to successful collaborations. We report here on a meeting of participants in four Indigenous community-university partnerships pursuing research on precision medicine. The goal of the meeting was to define values and practices that strengthen opportunities for genetic research. The group accorded the highest priority to developing trusting relationships, ensuring respect for Indigenous community authority, and pursuing research that has the potential to lead to community benefit. Supporting priorities included incorporation of Indigenous expertise in research planning, transparent communication, and development of community capacity, including capacity to participate in formulating research questions, informing research methodology, and leading research projects. Participants also noted the importance of attention to social determinants of health so that genetic contributors to health are evaluated in the appropriate context.

Data sharing and community-engaged research.

Data sharing must be accompanied by responsibility sharing.

A mixed-methods protocol to develop and validate a stewardship maturity matrix for human genomic data in the cloud.

This article describes a mixed-methods protocol to develop and test the implementation of a stewardship maturity matrix (SMM) for repositories which govern access to human genomic data in the cloud. It is anticipated that the cloud will host most human genomic and related health datasets generated as part of publicly funded research in the coming years. However, repository managers lack practical tools for identifying what stewardship outcomes matter most to key stakeholders as well as how to track progress on their stewardship goals over time. In this article we describe a protocol that combines Delphi survey methods with SMM modeling first introduced in the earth and planetary sciences to develop a stewardship impact assessment tool for repositories that manage access to human genomic data. We discuss the strengths and limitations of this mixed-methods design and offer points to consider for wrangling both quantitative and qualitative data to enhance rigor and representativeness. We conclude with how the empirical methods bridged in this protocol have potential to improve evaluation of data stewardship systems and better align them with diverse stakeholder values in genomic data science.

Three decades of ethical, legal, and social implications research: Looking back to chart a path forward.

More than thirty years ago in the United States, the National Center for Human Genome Research (NCHGR) at the National Institutes of Health (NIH) and its partner in the Human Genome Project (HGP), the Department of Energy (DOE), called for proposals from social scientists, ethicists, lawyers, and others to explore the ethical, legal, and social implications (ELSI) of mapping and sequencing the human genome. Today, nearly twenty years after the completion of the HGP, the ELSI Research Program of the National Human Genome Research Institute (NHGRI) continues this support. It has fostered the growth of ELSI research into a global field of study, uniquely positioned at the nexus of many academic disciplines and in proximity to basic and applied scientific research. We examine the formation of the first ELSI program and consider whether science policy in the public interest can exist within the confines of a set-aside from the NHGRI budget.