This research theme focuses on the role of citizens and patients in the development and use of AI in healthcare. We do so in multiple ways. First, we study how AI impacts the role of patients. For example, AI can enable new forms of risk profiling and diagnostic practices that can reshape existing categories and perceptions of health and illness. This may also come with consequences for the self-management practices of patients. New roles of citizens and patients can also refer to new data practices in relation to reimbursement of healthcare interventions and treatments. For many innovative treatments and interventions, for instance, robust evidence is lacking and studies are done with a small patient population, in a limited controlled or uncontrolled setting or only the mechanism of action has been proven. Actors like the Dutch National Health Care Institute are therefore experimenting with new data practices: treatments are temporarily reimbursed on the condition that patients and health providers collect and share data that allows further evaluation of effectiveness and efficiency in practice.
Second, we study how patient experiences can be fed into AI and how they can be used to create experience-based, human-centred AI. Increasingly patients share their experiences and knowledge about their conditions through ego-documents like books, blogs and fora, many of which are collected in the patient science hub. While patients’ and citizens’ knowledge and experience are generally recognized to be of great value, due to their unstructured and idiosyncratic nature they often stay outside the reach of health care providers, who use more aggregated forms of knowledge. We develop various citizen science initiatives and explore different interdisciplinary methods that seek to combine the strengths of patient narratives, AI-techniques (e.g. natural language processing) and qualitative/interpretive research. Through such social science based experiments we reflect on the possibilities and impossibilities of using AI to create more person-centred services in a way that does justice to the lived experiences of patients.
Third, we focus on how patients and citizens perceive the legitimacy of AI and what this means for the use of AI in healthcare. Despite the great potential of AI applications for patient care, their use in practice has not gained traction. One of the key challenges that AI tools face is the ability to properly meet users’ needs, as failure to do so often results in misused or underutilized applications. To address these factors, more is required than technically robust AI models. If AI health solutions are to become embedded in clinical practice, they should be acceptable to the stakeholders involved in the healthcare service. We therefore investigate under what conditions patients and citizens find the use of AI desirable, reasonable and appropriate, the values and perceptions maintained by different groups, and the broader system of norms, values and beliefs that is considered to matter for different patient groups in relation to legitimate use of AI. What are according to them important conditions for the collection and use of data in order to be legitimate?
Links
Researchers
Dr. Hester van de BovenkampEmail address
Dr. Bert de GraaffEmail address
Dr. Rik WehrensEmail address
Dr. Iris WallenburgEmail address