TL;DR
This study delves into the attitudes of five key stakeholder groups (patients, healthcare professionals, researchers, political stakeholders, and the general public) towards AI-based contactless sensors (AI-CS) in healthcare. It reveals a surprisingly high level of openness, with experience and knowledge correlating to reduced emotional reactions and greater acceptance. The findings highlight critical tensions around touch, privacy, and unobtrusiveness, offering crucial guidance for the responsible development and governance of AI-CS.
What Problem Does This Paper Address?
The rapid advancement of AI-based contactless sensors (AI-CS) promises to revolutionize health monitoring and assessment. However, there's a significant gap in empirical research specifically examining attitudes towards these technologies in healthcare. Existing studies often focus narrowly on medical or patient perspectives, neglecting the crucial viewpoints of other stakeholders like researchers, policymakers, and the general public. This oversight hinders a comprehensive understanding of the ethical, social, and practical implications of AI-CS, potentially leading to misaligned development and implementation. This paper aims to bridge this gap by exploring the multifaceted attitudes of diverse groups towards AI-CS in health.
Methods
The researchers employed a qualitative, cross-sectional study design, utilizing a semistructured interview guide with 104 participants. The interviews were analyzed using qualitative content analysis with ATLAS.ti software, following a three-component model of feelings, thoughts, and behavioral aspects. This approach allowed for an in-depth exploration of participants' attitudes, capturing not just their opinions but also the underlying emotions and behavioral tendencies related to AI-CS. The study also incorporated a sociodemographic questionnaire and the Hospital Anxiety and Depression Scale (HADS) to mitigate sampling bias and rule out the influence of anxiety or depression on attitudes.
Datasets
Not discussed in the paper.
Key Insights and Ideas
- High Openness, Nuanced Views: Despite the novelty of AI-CS, all stakeholder groups showed a significant level of openness. However, this openness was not uncritical, with participants articulating complex considerations around privacy, control, and the human element in healthcare.
- Experience Breeds Acceptance: A strong correlation emerged between increased experience and knowledge of technology and a reduced tendency for emotional reactions. Those with more exposure to technology, particularly positive experiences, were generally more accepting of AI-CS.
- The Paradox of Touch: The concept of "contactlessness" sparked a significant tension. While it was seen as beneficial for avoiding unwanted intimate contact, it also raised concerns about the loss of human touch, which is deeply valued in healthcare for its diagnostic and relational roles.
- Protection vs. Surveillance: A critical dichotomy was identified between the potential of AI-CS for protection (e.g., constant monitoring in palliative care) and the risk of surveillance, particularly when used outside of controlled institutional settings. The blurring lines between these contexts due to portable devices are a major concern.
- Unobtrusiveness: A Double-Edged Sword: The unobtrusive nature of AI-CS was recognized as a technical advantage for minimizing measurement distortions. However, it also raised concerns about a potential loss of visibility, which could impact control and oversight.
Key Results
The study found a high level of openness to AI-CS across all five stakeholder groups. Two key trends emerged:
- Experience and Knowledge: Greater experience and knowledge correlated with a reduced tendency to react emotionally. Researchers, for instance, were less likely to express strong emotions, tending towards a more sober view.
- Positive Experiences: Participants with positive past experiences with technology were generally more open and positive towards contactless sensors.
The analysis also highlighted three key tensions:
- Contact(lessness) and Touch: The importance and ambivalence of touch were central, with participants valuing its human element while also appreciating the avoidance of unwanted intimate contact.
- Protection and Surveillance: The path- and context-dependency of AI-CS use created a tension between protection and surveillance, especially concerning privacy.
- Unobtrusiveness: The benefits of unobtrusiveness were weighed against challenges related to control and governance.
Limitations and Future Work
The authors acknowledge that the study was unable to recruit an equal number of political stakeholders, which may limit the full representation of this group. Additionally, a lack of clarity surrounding the definition and applications of AI-CS among participants led to varied interpretations and associations, highlighting a need for better public information. Future work could focus on developing targeted information strategies and hands-on experiences to mitigate rejection and foster acceptance. The study also suggests that regulatory frameworks need to be robust to address the ethical complexities and ensure the responsible implementation of AI-CS, particularly concerning consent, data management, and the balance between individual and societal benefits.
Summary updated