A novel advantage of the use of machine learning (ML) systems in medicine is their potential to continue learning from new data after implementation in clinical practice. To date, considerations of the ethical questions raised by the design and use of adaptive machine learning systems in medicine have, for the most part, been confined to discussion of the so-called “update problem,” which concerns how regulators should approach systems whose performance and parameters continue to change even after they have received regulatory (...) approval. In this paper, we draw attention to a prior ethical question: whether the continuous learning that will occur in such systems after their initial deployment should be classified, and regulated, as medical research? We argue that there is a strong prima facie case that the use of continuous learning in medical ML systems should be categorized, and regulated, as research and that individuals whose treatment involves such systems should be treated as research subjects. (shrink)

The application of artificial intelligence and machine learning technologies in healthcare have immense potential to improve the care of patients. While there are some emerging practices surro...

Institutional review boards distinguish health care quality improvement and health care quality improvement research based primarily on the rigor of the methods used and the purported generalizability of the knowledge gained. Neither of these criteria holds up upon scrutiny. Rather, this apparently false dichotomy may foster under-protection of participants in QI projects and over-protection of participants within QIR.

The Institute of Medicine has called on health care leaders to transform their health systems into “learning health care systems,” capable of studying and continuously improving their practices. Learning health care systems commit to carrying out numerous kinds of investigations, ranging from clinical effectiveness studies to quality improvement research and implementation science. There has been progress in realizing the IOM's vision, but also many challenges. One of these challenges has been lingering uncertainty about whether the data collection and monitoring central (...) to learning health care systems is actually research and if so, what kind of ethical oversight it should have. This is not a new question. Yet so far, there has been no foundational analysis of the fit between the existing human subjects protection framework in use in the United States and the new kinds of data collection activities that are being, and increasingly will be, undertaken by learning health care systems. Two companion feature articles in this volume, by a team at Johns Hopkins, fill this void. (shrink)

Quality improvement is an important function of learning health systems, and public policy should promote QI activities. Use of systematic methodologies in QI has prompted substantial confusion regarding when QI is human subjects research under the Common Rule, and this confusion persists with the revised Rule. Difficulty distinguishing research from QI imposes costs on the quality improvement process. I offer guidance to IRBs to mitigate these costs and suggest a new regulatory exclusion for minimal risk quality improvement activities.

A number of unique problems plague human research protection efforts at United States (U.S.) State and Territorial Departments of Health (DOHs) problems which might be ameliorated through a consolidated national or regional, independent, not-for-profit Institutional Review Board (IRB).

A number of unique problems plague human research protection efforts at United States State and Territorial Departments of Health. The first problem is related to the number of Institutional Review Boards operated by and Federalwide Assurances held by DOHs. The lack of these two essential regulatory human research protection program mechanisms points to a possible inadequacy of infrastructure at DOHs for protecting human subjects. The second and third problems are related to the use and interpretation of research protection laws and (...) regulations at DOHs. Parsing public health activities into research or practice can be vexing and likely produces variability in how individual DOHs make the determination between research and practice and the outcomes of such decisions for the same or similar projects. A related problem is that DOH or Institutional Review Board officials might misclassify public health practice activities as research, which can have costly, adverse consequences. (shrink)

Using mobile health research as an extended example, this article provides an overview of when the Common Rule “applies” to a variety of activities, what might be meant when one says that the Common Rule does or does not “apply,” the extent to which these different meanings of “apply” matter, and, when the Common Rule does apply, how it applies.

Char and colleagues provide a useful conceptual framework for the proactive identification of ethical issues arising throughout the lifecycle of machine learning applications in healthcare. Th...

Calls are increasing for American health care to be organized as a learning health care system, defined by the Institute of Medicine as a health care system “in which knowledge generation is so embedded into the core of the practice of medicine that it is a natural outgrowth and product of the healthcare delivery process and leads to continual improvement in care.” We applaud this conception, and in this paper, we put forward a new ethics framework for it. No such (...) framework has previously been articulated. The goals of our framework are twofold: to support the transformation to a learning health care system and to help ensure that learning activities carried out within such a system are conducted in an ethically acceptable fashion. (shrink)

In this paper, we argue that there are important ethical questions about healthcare improvement which are underexplored. We start by drawing on two existing literatures: first, the prevailing, primarily governance-oriented, application of ethics to healthcare ‘quality improvement’ (QI), and second, the application of QI to healthcare ethics. We show that these are insufficient for ethical analysis of healthcare improvement. In pursuit of a broader agenda for an ethics of healthcare improvement, we note that QI and ethics can, in some respects, (...) be treated as closely related concerns and not simply as externally related agendas. To support our argument, we explore the gap between ‘quality’ and ‘ethics’ discourses and ask about the possible differences between ‘good quality healthcare’ and ‘good healthcare’. We suggest that the word ‘quality’ both adds to and subtracts from the idea of ‘good healthcare’, and in particular that the technicist inflection of quality discourses needs to be set in the context of broader conceptualisations of healthcare improvement. We introduce the distinction between quality as a measurable property and quality as an evaluative judgement, suggesting that a core, but neglected, question for an ethics of healthcare improvement is striking the balance between these two conceptions of quality. (shrink)