As the COVID-19 pandemic rages on, this June 2021 issue of the JME contains several articles addressing pandemic-related ethical issues, including, discrimination against persons with disabilities,1 collective moral resilience,2 and stress in medical students due to COVID-19.3 It also contains a critical appraisal of the most recent WHO guidance document on the management of ethical issues during an infectious disease outbreak.4 This June issue of JME also addresses several important clinical ethics issues: covert administration of medication in food,5 educational pelvic (...) exams under anesthesia,6 consent to cancer screening,7 care of critically ill newborns when the birth mother is unwell,8–10 and ethical considerations related to recruiting migrant workers for clinical trials.11 Perhaps what is most unique about this issue is its Feature …. (shrink)

This perspective article explores the use of digital twins (DTs) in medicine, highlighting its capacity to simulate risks and personalize treatments while examining the emerging bioethical concerns. Central concerns include power dynamics, exclusion, and misrepresentation. We propose adopting a relational bioethical approach that advocates for a comprehensive assessment of DTs in medicine, extending beyond individual interactions to consider broader structural relations and varying levels of access to power. This can be achieved through two key relational recommendations: acknowledging the impact of (...) uneven relational structures on access to medical care and promoting social justice by evaluating resource allocation. While DTs in medicine offer promising advancements, a relational bioethical lens may provide a nuanced understanding, fostering equitable, inclusive and responsible integration of DTs into medical practice. (shrink)

In his contribution, ‘Represent me: please! Towards an ethics of digital twins in medicine’1, Braun shows that there is a fundamental ambivalence inherent in digital twins: they can either open up new freedoms for the simulated persons, or, conversely, endanger and restrict their freedom. To prevent digital twins from restricting people’s freedom, Braun suggests a strong focus on control. Braun’s focus on control is insufficient, I will argue, because his idea of control works only retroactively: it can only react to (...) restrictions of freedom that have already occurred, but it cannot prevent them, nor can it restore any lost freedom. Control is therefore insufficient to defend people’s freedom. To prove my claim, I will first use an everyday example to illustrate the fundamental ambivalence inherent in digital twins. Second, I will show that digital twins endanger people’s freedom in a hermeneutic way. Finally, I will demonstrate why controllability cannot prevent this hermeneutic endangerment. I understand digital twins, like Braun, as simulation of a person, as their dynamic and precise effigy. Simulated twins use machine-learning and artificial intelligence capabilities to make predictive analyses about the simulated person and the future development of her health. Imagine: You are eating dinner. Meanwhile, your partner lectures you about the calorie density and nutrient composition of your food, warns you that your portion size exceeds your daily calorie requirement by x calories. It calculates how your dinner will negatively impact your body mass index and physical fitness, and predicts what the consequences will be for your overall health. Based on these data and predictions, it creates an individualised diet plan for you—and at the same time shows you by how much you could improve your vital values by sticking to that plan. This example helps to illustrate the fundamental …. (shrink)

Digital twinning in medicine refers to the idea of simulating a person’s organs, muscles or perhaps their entire body, in order to arrive more effectively at accurate diagnoses, to make treatment recommendations that reflect chances of success and possible side-effects, and to better understand the long-term trajectory of an individual’s overall condition. Digital twins, in these ways, build on the recent movement toward personalised medicine,1 and they undoubtedly present us with exciting opportunities to advance our health. Of course, the opportunities (...) are accompanied by newfound challenges—a common refrain in discussions surrounding emerging technologies. In a recent article, Matthias Braun surveys numerous problems, including the precision of simulations, ownership and consent to the use of digital models, and issues of justice in assuring equitable access to novel medical systems.2 In particular, he places special emphasis on the notion of representation. According to Braun, proper representation naturally calls for correspondence—namely a precise model offering dynamic, real-time feedback as well as transparency, both for the patient and the practitioner, in terms of the data being employed. While technical challenges remain, these goals are quite clear and relatively uncontroversial. Where a host of questions arise, however, is in the discussion on the interaction between the represented persons and their simulations. Alongside questions regarding ownership of one’s digital twin, Braun raises the concern that ‘giving …. (shrink)

The construction of digital simulacra in healthcare and medical research purportedly strives to virtually recreate some aspect of reality, whether that be a piece of human tissue, an entire organ,...

The concept of a digital twin comes from engineering.1 It refers to a digital model of an artefact in the real world, which takes data about the artefact itself, data about other such artefacts, among other things, as inputs. The idea is that the maintenance of artefacts—such as jet engines—can be vastly improved if we work with digital twins that simulate actual objects. Similarly, personalised medicine might benefit from the digital modelling of body parts or even whole human bodies. A (...) medical digital twin could use data about the patient, more general population data, and other inputs to generate predictions about the patient. This could lead to highly personalised interventions and nuanced judgments about the patient’s health. Matthias Braun2 discusses this intriguing prospect, asking how we should think about the way in which a digital twin could represent a patient. I will respond to Braun’s striking suggestion that we can regard a digital twin as an extension of the patient’s body. Notably, Braun does not compare his just-mentioned idea with the extended mind thesis popularised by Andy Clark and David Chalmers.3 But I am sure many readers will be reminded of the extended mind thesis. Accordingly, I will consider this comparison. I cannot discuss this comparison in detail, nor fully evaluate Braun’s suggestion. But I can say something about how we might approach this comparison, and provide some …. (shrink)

In ‘Represent me: please! Towards an ethics of digital twins in medicine’,1 Braun analyses the potential for simulations of human organs and bodies, or ‘digital twins,’ to faithfully a person. Drawing from several French philosophers, he introduces ‘first conditions for digital twins to take on an ethically justifiable form of representation’.1 The analysis predominantly focuses on challenges that arise in terms of representation, embodiment, control by the patients after which the twin is modelled. Many challenges are posited on the existence (...) of digital twins as autonomous technologies capable of acting on behalf of the patient. One may be reasonably sceptical about the potential to achieve ‘digital twins’ in the near future; the author, to his credit, acknowledges this fact in the section ‘Simulating a precise efficacy’ while still providing a faithful overview of a nascent field of development. Given the state of development described in the paper, the risk of autonomous digital twins in the near future seems remote. There are two clear ethical risks of digital twin which, to my reading, deserve more attention than they receive in Braun’s analysis. Let us refer to these as the ‘near-term’ challenges of digital twins, in the sense that their significance does not turn on the degree of agency or embodiment of digital twins, but rather their mere existence and potential …. (shrink)

Machine intelligence already helps medical staff with a number of tasks. Ethical decision-making, however, has not been handed over to computers. In this proof-of-concept study, we show how an algorithm based on Beauchamp and Childress’ prima-facie principles could be employed to advise on a range of moral dilemma situations that occur in medical institutions. We explain why we chose fuzzy cognitive maps to set up the advisory system and how we utilized machine learning to train it. We report on the (...) difficult task of operationalizing the principles of beneficence, non-maleficence and patient autonomy, and describe how we selected suitable input parameters that we extracted from a training dataset of clinical cases. The first performance results are promising, but an algorithmic approach to ethics also comes with several weaknesses and limitations. Should one really entrust the sensitive domain of clinical ethics to machine intelligence? (shrink)

BackgroundHealth data-driven activities have become central in diverse fields (research, AI development, wearables, etc.), and new ethical challenges have arisen with regards to privacy, integrity, and appropriateness of use. To ensure the protection of individuals’ fundamental rights and freedoms in a changing environment, including their right to the protection of personal data, we aim to identify the ethical approaches adopted by scientists during intensive data exploitation when collecting, using, or sharing peoples’ health data.MethodsTwelve scientists who were collecting, using, or sharing (...) health data in different contexts in Sweden, were interviewed. We used systematic expert interviews to access these scientists’ specialist knowledge, and analysed the interviews with thematic analysis. Phrases, sentences, or paragraphs through which ethical values and norms were expressed, were identified and coded. Codes that reflected similar concepts were grouped, subcategories were formulated, and categories were connected to traditional ethical approaches.ResultsThrough several examples, the respondents expressed four different ethical approaches, which formed the main conceptual categories: consideration of consequences, respect for rights, procedural compliance, and being professional.ConclusionsTo a large extent, the scientists’ ethical approaches were consistent with ethical and legal principles. Data sharing was considered important and worth pursuing, even though it is difficult. An awareness of the complex issues involved in data sharing was reflected from different perspectives, and the respondents commonly perceived a general lack of practical procedures that would by default ensure ethical and legally compliant data collection and sharing. We suggest that it is an opportune time to move on from policy discussions to practical technological ethics-by-design solutions that integrate these principles into practice. (shrink)

In his Feature Article ‘Represent me: please: Towards an ethics of digital twins in medicine’1, Mattias Braun considers several important bioethical issues in relation to the use of digital twin simulations in health and medical contexts. He focuses on the ways these simulations are used or are proposed to be deployed in these domains, including to what extent they are a ‘true’ or ‘real’ representation of human bodies. In this response, I want to take a step back and delve into (...) the metaphor of the ‘digital twin’. I consider the implications of using this metaphor in the context of medical care and how this choice of language participates in what science and technology scholars term the ‘sociomaterial imaginaries’ that are used to support and promote novel technologies.2 Language is important. It matters. The choice of language is both symbolic and persuasive. It intervenes in political and policy debates and research and healthcare funding decisions.2 It has material effects and affects: to some of which Braun has gestured in his piece. The …. (shrink)

Matthias Braun’s1 concise reflections on the ethical challenges posed by digital twins in medicine are briefly supplemented below by a thought that, in my view, seems to run through the text like a thread, but whose ethical implications are not explicitly stated. Braun states in the final paragraph that ‘digital twins do not fundamentally threaten the embodied person’, at least not as long as the person in question has ‘control over her simulated representation’. I agree with this, but would like (...) to point out that it is not just the embodied person who has control. Regardless of whether the person embodied by the digital twin has permanent control over it, which enables them to revoke their agreed representation at any time, the company that developed or uses the digital twin also has control over it. The control of the company or companies and other involved parties lies in the fact that data and information are …. (shrink)

Braun explores the use of digital twin technology in medicine with a particular emphasis on the question of how such simulations can represent a person.1 In defining some first conditions for ethically justifiable forms of representation of digital twins, he argues that digital twins do not threaten an embodied person, as long as that person retains control over their simulated representation via dynamic consent, and ideally with the option to choose both form and usage of the simulation. His thoughtful elaboration (...) provides insight into the challenges inherent in interactions between a person and their digital twin, emphasising the modes of control required to respect personal modes of freedom. This individual-centric approach to the question of representation leaves out the important ethical consideration of ensuring that all, not merely some, can be represented. Braun describes dynamic consent as a necessary mode of control, arguing that where a person is unable to give consent, ‘such simulations threaten to become illegitimate representations. They would then shift the kind of interaction from representation to illegitimate forms of prediction or surveillance and thereby could lead to infringements to individual modes of freedom.’ Consequently, the implementation of digital twin technology in healthcare would seem to exclude the most vulnerable members of society from participation in and benefiting from medical innovation, …. (shrink)

When making substituted judgments for incapacitated patients, surrogates often struggle to guess what the patient would want if they had capacity. Surrogates may also agonize over having the (sole) responsibility of making such a determination. To address such concerns, a Patient Preference Predictor (PPP) has been proposed that would use an algorithm to infer the treatment preferences of individual patients from population-level data about the known preferences of people with similar demographic characteristics. However, critics have suggested that even if such (...) a PPP were more accurate, on average, than human surrogates in identifying patient preferences, the proposed algorithm would nevertheless fail to respect the patient’s (former) autonomy since it draws on the ‘wrong’ kind of data: namely, data that are not specific to the individual patient and which therefore may not reflect their actual values, or their reasons for having the preferences they do. Taking such criticisms on board, we here propose a new approach: the Personalized Patient Preference Predictor (P4). The P4 is based on recent advances in machine learning, which allow technologies including large language models to be more cheaply and efficiently ‘fine-tuned’ on person-specific data. The P4, unlike the PPP, would be able to infer an individual patient’s preferences from material (e.g., prior treatment decisions) that is in fact specific to them. Thus, we argue, in addition to being potentially more accurate at the individual level than the previously proposed PPP, the predictions of a P4 would also more directly reflect each patient’s own reasons and values. In this article, we review recent discoveries in artificial intelligence research that suggest a P4 is technically feasible, and argue that, if it is developed and appropriately deployed, it should assuage some of the main autonomy-based concerns of critics of the original PPP. We then consider various objections to our proposal and offer some tentative replies. (shrink)