Doctorate Projects

This thesis contributes to our understanding of the strategies and features necessary to build and maintain trust in digital health. Myriad experts foresee digital health technologies as a solution to alleviate the demands on contemporary health systems. Yet, these technologies introduce many ethical challenges, one of which is trust. Besides altering the status quo in physician-patient relationships, plenty of evidence suggests that certain digital health technologies may be unreliable or even promote detrimental behaviors. Moreover, these technologies remain increasingly susceptible to malicious actors; while a number of developers and implementers continue to act negligently. As a result, the perceived risks of digital health continue to grow and thus deteriorate trust. Since trust is imperative to the proper functioning of health systems, it is necessary to determine how to assess the trustworthiness of digital health technologies.

The thesis employs both empirical and theoretical methods. The first chapter offers an introductory analysis of digital health and trust in relation to this rapidly expanding form of medical technology. Chapter 2 discusses the major ethical issues posed by digital health in a global as well as in the Swiss context.Through a scoping review of the literature, Chapter 3 identifies the facilitators and inhibitors to trust in digital health as well as the main stakeholders. Meanwhile, since precision medicine relies on diverse datasets that may be obtained from digital health technologies, Chapter 4, stipulates some requirements for trust in precision medicine. Through focus group discussions, Chapter 5 identifies trust-related expectations with respect to mobile health apps, and translates such findings into a trustworthiness checklist. In Chapter 6, the checklist is validated by relevant stakeholders via an online survey. Finally, chapter 7 discusses the overall findings of this thesis, compares it to existing research, and determines its implications.

Rapid developments in big data and artificial intelligence (AI) give rise to many ethical challenges and concerns. In the domain of healthcare, AI-based applications have the potential to transform medicine and to become indispensable tools of the field. Set at the intersection of medicine and digital technologies, the area of digital medicine is, therefore, rapidly emerging and gaining increasing levels of attention. While many implications of such technological change are not yet known, a growing community of researchers from various fields has recognised specific ethical challenges and is actively working on addressing them. These include the fairness, transparency, accountability, clinical validity, and security of new forms of health data processing. New empirical methods are needed to increase our understanding of such novel ethical issues and to promote the emergence of ethically robust solutions. Therefore, my research sets out to enrich the toolbox of empirical bioethics with new methods, adapted to the specificities of digital medicine, and its ever more pressing challenges.

This project seeks to develop digital methods to investigate bioethical questions. A recent example of such a method is the application of sentiment analysis to study the public discourse on CRISPR on Twitter. More on the method and results from that study can be found in our preprint: “external pageCombining Crowdsourcing and Deep Learning to Assess Public Opinion on CRISPR-Cas9call_made” (2019).

Digitalisation is revolutionising every aspect of life, including biomedicine. The deployment of digital technology in the healthcare sector has enabled the use of big data; however, despite the fact that this holds important promise for healthcare, new challenges arise from the collection and analysis of large amounts of personal data. Such issues concern privacy, equality, fairness, identity, security, consent and risk-benefit. The lack of norms, and the existing ethical tools that regulate large biomedical data (e.g. the human subject research regulation; the medical ethics codes; the public health frameworks), seem inadequate to address these new challenges.

The goal of the project is twofold. First, it aims at mapping the ethics of big data in healthcare and biomedical research, providing a comprehensive ethical framework to guide common practices in governing the creation, collection, storage, processing, sharing, analysis, dissemination, and re-use of data in health. This analysis contributes to the development of useful tools to assess big data research projects, and to accommodate new risks and vulnerabilities typical of big data analytics. Second, it contributes policy recommendations to close existing regulatory gaps for research practices that involve big data.

Ultimately, this project investigates which ethical and normative standards for collection, analysis and protection of big data for health that should be included in a general framework.

Data is the lifeline to health research and precision medicine, yet it brings to the fore ethical concerns around privacy, security, interoperability, and accountability. Leaders strive to address such issues and manage the evolving conglomeration of actors, needs, and data processing practices via ethical guidelines for data governance. Written for a heterogeneous array of audiences and published online often in lengthy, verbose, and text-dominant PDF formats, these documents mobilise values and sensitise researchers, review committees, and the public to the challenges accompanying data usage in health research. However, despite being vital tools that make health data governance ethics transparent, stable and legible to multiple users, the authors generally disregard our present content-overloaded and digital-media milieu where accessibility and comprehensibility are as important as the content. At the same time, a growing number of visualisation approaches and digital tools could help communicate complex health-data ethics topics to diverse audiences and would allow for formats that engage today's users.