Daniel Steel

The use of artificial intelligence for diagnostic purposes is a significant topic of discussion in the present-day healthcare field. Developers working on this technology intend it to outperform human clinicians and therefore remove some of the burden from clinicians so that they can spend more time developing relationships of care with their patients. I focus on three ethical questions I believe should be answered prior to their widespread onboarding in healthcare systems. I first investigate what kind of artificial intelligence we ought to want for diagnostic systems in healthcare. I analyse three high-level categories of artificial intelligence: opaque black box systems, robustly transparent explainable systems, and semi-transparent grey box systems. I start by outlining the characteristics of these systems, culminating in the development of a novel definition for the black box problem. I then take this information and analyse each kind of high-level system. I defend the position that the best kind of systems for healthcare applications are grey boxes due to their customizability and semi-transparent nature. Second, I examine what obligations clinicians ought to have to their patients whenever they employ an artificial intelligence for diagnostics. I separate this chapter into three sets of obligations, one for each of the categories from chapter one, as a clinician may not always have the option to work exclusively with grey systems. By providing three lists, I ensure that clinicians have a minimum starting point regardless of what kind of system they employ.Finally, I address the implicitly articulated concerns surrounding whether the general push for more advanced AI in healthcare jeopardizes a patient’s ability to provide informed consent to a procedure or treatment plan. These concerns are raised nearly ubiquitously in the literature, yet to date no comprehensive analysis regarding the issue exists. I argue that the concerns over consent are actually due to a false dilemma in how we discuss the kinds of artificial intelligence in the literature. I then demonstrate how this false dilemma can be defeated through appeals to grey box systems and using human clinician secondary readers.

View record

This dissertation comprises three papers examining the historical, ethical, and social aspects of public health errors. My first paper explores how US health authorities responded to the discovery of the late health effects of radiation treatment. Based on the examination of multiple primary and secondary sources of evidence; archival research conducted at the National Archive in Washington, DC; and research conducted through media web-archives, I show how efforts by Michael Reese hospital in Chicago to locate and examine former patients (and the media attention these efforts attracted) led to a nationwide campaign by the National Cancer Institute (NCI) to warn those who underwent radiation treatment during childhood. My second paper investigates the ethics of evidence and post-market surveillance of pharmaceuticals in Canada. Drawing on philosophical discussions of inductive risk, the paper examines what evidence should have been sufficient for Health Canada (HC) to revise the misleading information that appeared in the product monograph for OxyContin. Given the stakes involved, I argue that a less strict standard of evidence would have been appropriate, yet HC in fact took the opposite course, insisting on a higher standard of evidence than it normally requires. The time it took for Health Canada to revise the monograph may have contributed to the prescription opioid epidemic in Canada. This paper also contributes to existing philosophical work by demonstrating that inductive risks in the post-approval stage are important and linked to pre-approval inductive risks. My third paper provides a new concept of public health errors—defined as acts of commission or omission, culpable or not, by public health officials, whose consequences for population health were clearly worse than those of an alternative that could have been chosen instead. This conception better corresponds to the task of public health, compared to policy failure literature, where achievement of political objectives is often used to measure success, and has practical and theoretical advantages. It also serves as a valuable analytical lens for understanding general mechanisms leading to public health errors, with utility for scholars who study policy errors as well as for public health actors interested in preventing them.

View record

Improving the quality of care and reducing preventable harm to patients/families are important obligations for health care professionals and institutions. Despite quality improvement’s (QI) significant contributions to health care, interventions that involve patients at the point of care raise ethical questions about heretofore unexamined wider impacts on patients. This qualitative empirical study challenged the assumption that QI is necessarily a benign and beneficial extension of routine medical care. The research questions focused on ethical and practical implications of patient/family experiences of a pilot QI intervention that gave patients a new opportunity to improve the quality and safety of their own care. Relational inquiry and ecological systems theory framed this ethical (values-based) analysis of moral agents in context. Using interpretive description methodology, I explored two hand hygiene QI interventions on two hospital units. Over 12 months, I observed patient care and staff meetings (~140 hours), interviewed/shadowed patients (n=25), families (n=8), providers/ volunteers (n=21), other staff (n=17), and analyzed documents. Findings about patient/family experiences of the QI interventions (what I term Work-as-Experienced) showed that a multifaceted set of perceptions and behaviours occurred at once. My thematic analyses centred on the value patients/families place on positive connections with their providers and health care settings. Contextual features of implementation showed disconnects between how quality leaders imagined the QI intervention’s success, how it was operationalized, and how patients/families experienced it. In interpreting these findings, I developed a Relational Safety Framework to explain how relational connections result from interactions with others and are potentially grounded in relational safety (feeling safe and valued). When people in vulnerable, uncertain circumstances have a sense of relational safety, they move towards connections with others, fostering trust. In this study I emphasize the importance of relational connections and shared meanings of quality/safety initiatives. Furthermore, evaluation is critical to assess the impact of QI on patients’/families’ lives. I recommend a stronger appreciation of the ethical implications of QI implementation, and strengthening relational safety through adopting a relational, multi-level perspective to foster ethical conduct of QI in health care. Relational safety is a new, ethically-based, values-added strategy to advance health care system improvement.

View record