After analysing the included literature, we identified groups of ethical, legal and social concerns raised by the authors. We have chosen trust as an umbrella term that encompasses all ethical, legal, social and technical challenges – from trust in the design and connectivity of a wearable device to confidence in its impact on oneself or social values. Given that trust “in terms of dispositions, beliefs or cognitions and feelings or affects and emotions [and is] essentially dyadic between two individuals, even if one of these `individuals` is a collective entity” (Barbalet, 2019), it is important that the ELS problems of digital phenotyping be addressed in terms of trust-building – or “living a reality that provides `good reasons`” (Möllering, 2001) if the full potential of digital phenotyping is to be exploited. Nevertheless, the fact that technologically sound effectiveness and efficiency considerations are the only determinants of eHealth planning falls victim to the technological imperative, as eHealth is associated with a number of potentially serious problems. These are not insurmountable. However, they should be addressed and resolved before they are implemented so that downstream difficulties do not jeopardize otherwise beneficial development. However, the challenges are not limited to the technical field. It would be disastrous for healthcare providers to forget that eHealth makes patients (and sometimes their loved ones) active participants in the delivery of care. This gives the question of responsibility a completely new form. First, the fact that eHealth must operate in an environment that includes different types of physical elements and protocols means that they can only be implemented if they can be seamlessly integrated into existing systems that are part of the established health structure. Interoperability is therefore a necessary material condition for their success, otherwise processing risks being affected by technology and the fiduciary duty of associated healthcare professionals and institutions can be compromised – with serious legal consequences.
This goes beyond interoperability within the institutional environment, which is hospital. This includes the problem of integrating eHealth into databases and operating systems used by doctors and other healthcare providers whose patients participate in eHealth services. Maher, N. A., Senders, J. T., Hulsbergen, A. F., Lamba, N., Parker, M., Onnela, J. P., & Broekman, M. L. (2019). Passive Data Collection and Use in Healthcare: A Systematic Review of Ethical Issues. International Journal of Medical Informatics, 129, 242-247.
doi.org/10.1016/j.ijmedinf.2019.06.015 Our keyword search string was limited to 11 keywords in all search databases. We included keywords based on our preliminary scope of the literature. However, many other terms from the lexical field of numerical phenotyping could have been included, such as: Lifelogging, which has been “displaced” by the term Quantified Self, as claimed by Deborah Lupton (2017, p.1). In addition, we decided to use the keyword crowdsensing in our search term (Mehdi, 2019; Pryss, 2019), which are commonly used in computer science and information science – although this buzzword has led to quite a large number of publications, they were not selected by our title and abstract selection, as these publications did not address ethical, legal or social issues in the field of digital phenotyping. eHealth promises to be a cost-effective and efficient way to provide affordable health care to patients who would otherwise be excluded or underserved. However, eHealth also poses a number of ethical and legal challenges that, if not resolved before implementation, could jeopardize its success. Among other things, changes in the nature of the relationship between health professionals and patients, in informed consent requirements and in the division of responsibilities are affected. There are also privacy concerns and the position of health informatics and service providers is also affected. Another complicating factor is outsourcing.
With e-health, patients are involved in the care process not only as test people, but also as participants. If eHealth is not fully automated through in-house telemetry, patients will have to report the relevant data – and they may make mistakes when measuring or declaring values. Even if the process is automatic, patients may accidentally interfere with these automated measurements or their transmission. The division of responsibility therefore takes on a new aspect, and when family members or partners are involved, the question of their shared responsibility also arises. Certainly, there are legal precedents in other areas of health care when patients contribute to negative outcomes. However, these are based on the traditional model of the relationship between health professionals and the patient, which is based on three premises: first, that care involves a direct encounter between patient and profession; secondly, that the patient`s contribution to the functioning of the techniques and technologies employed by the skilled person is essentially non-existent; and third, that its own expertise determines the availability and reliability of patient instrumental data and is independent of the patient`s abilities in the use or operation of these tools. eHealth is changing this image considerably. It is therefore doubtful whether eHealth can rely on the traditional approach of accountability. Even the standard consent model may no longer apply. To be included in the review, publications had to include some sort of ethical, social, or legal review of digital phenotyping and focus on the use of digital phenotyping in diagnosis and decision-making, both in the medical and anthropotechnical sense. Based on our preliminary literature review, we decided to expand the WHO (1946) definition of health. Consumers, researchers and others may have different views on health.
In addition, they can assimilate all numerical phenotype data with health-related data. We have therefore decided to adopt a broader sense of health, encompassing both well-being and fitness, and we refer to a holistic notion of health based on physical, mental, spiritual, emotional and social well-being, all aspects of a symptomatology of patients` daily lives that digital phenotyping promises to preserve. Ultimately, we should be thinking about the same ethical issues of telemedicine that we have always considered when caring for our patients. When we focus on maintaining a strong patient-physician relationship, protecting patient privacy, promoting equal access and treatment, and achieving the best possible outcomes, telemedicine can improve medical practice and patient care in ways we can all feel comfortable with. What recommendations can we draw from the literature to assess the ethical, legal and social implications of the use and implementation of digital phenotyping? Finally, although the most commonly used theoretical approach in our literature has been various theories of technology acceptance, another promising avenue in this direction could be explored through the supradisciplinary theory of social representation (Moscovici, 2001) and actor-network theory (Law, 2009).