Hvordan påvirker kunstig intelligens sygeplejerskernes rolle og autonomi i et aktør-netværk perspektiv? Illustreres via case: ”Tidlig opsporing af sygdom med kunstig intelligens i den kommunale hjemme- og sygepleje”

Abstract

This assignment considers the introduction of artificial intelligence (AI) in municipal home care, specifically in the early detection of diseases among the elderly.
The purpose of the assignment is to examine how AI affects the role of nurses and the interaction between the actors in the network, with a focus on nurses' autonomy. This is analyzed by using actor-network theory combined with professional theory. Empirically, the analysis is based on semi-structured interviews, mainly with nurses, as well as ongoing data collection through observations and experiences gained as a manager of AI project in home care.
The main finding is that AI can be seen as a new actor in the network, creating a series of shifts in the roles of the actors and changing the dynamics. For nurses, it means moving towards a more hybrid role where they need to use and combine new logics (engineering, management, etc.) to exercise their nursing expertise in the transformed network.
AI distributes new healthcare knowledge, which was previously exclusive to nurses, are now visible to all actors in the network. For example, knowledge about hospital admissions and medical conditions. This makes it more challenging for nurses to maintain control over their professional domain “occupational control”. Similarly, AI brings engineering knowledge into the network, such as the risk of hospitalization, which also affects the network dynamics.
In the transition to the new network, nurses receive and take a range of new and changed roles, including being the central interpreter of AI's knowledge. This happens as nurses position themselves as obligatory passage points, where they interpret and filter the information from AI before redistributing it in the network, thus attempting to maintain “occupational control”. However, this strategy is not entirely effective as the visibility of knowledge makes it difficult to maintain control. Therefore, nurses engage in ongoing dialogue to "negotiate" specific healthcare interventions with healthcare assistants. Nurses address this by supplementing "occupational control" with more "reflexive control," focusing on controlling the meaning, which is exercised through new roles such as facilitator and interpreter of AI.
Likewise, there is a shift in nurses' autonomy. Nurses cannot make independent decisions to the same extent as before because multiple actors are involved. Nurses must argue and negotiate their viewpoints more extensively, but this does not necessarily result in a decrease in autonomy if they are able to establish a meaningful connection between healthcare, AI, clients, organization for both themselves and the healthcare assistants. In this way, nurses can gain autonomy through increased influence over the actions of the assistants. On the other hand, nurses may lose autonomy by losing decision-making authority since more tasks can be performed by the assistants.
Another sub-conclusion is that the organization surrounding AI is not considered entirely stable. This is because AI potentially creates internal fragmentation among nurses and between nurses and social and healthcare assistants, which can challenge stability in the long run. Fragmentation among nurses can occur because some nurses seemingly experience an "overload of information," while the majority of nurses perceive AI as providing useful knowledge, that can be translated into healthcare actions for the clients.
The results of the assignment provides new insights into how AI can affect municipal nursing care. It can be used in former management to understand the dynamics and thus exploit the benefits of AI in practice while mitigating its drawbacks.