Comprehensive Summary
This mixed-method study by Tutar et al. examined the anxiety levels, concerns, and metaphor perceptions of pediatric nurses regarding artificial intelligence (AI) technologies. 422 pediatric nurses in Turkey were interviewed face-to-face and filled out 3 forms: the Introductory Information Form, the Artificial Intelligence Anxiety (AIA) Scale, and a semi-structured interview form. To ensure an accurate, consistent, and reliable data collection processes, a single researcher was assigned to each hospital, therefore minimizing environmental variability. Text mining and sentiment analysis methods were used to categorize and theme open-ended questions into positive and negative results. On the AIA Scale, nurses reported a total score of 48.29 ± 13.53, slightly above the median, indicating heightened anxiety about using AI in clinical settings. There were positive correlation levels between learning and job replacement (0.645), misuse (0.557), and fear (0.544) as well as between job replacement and misuse (0.895), with fear (0.709), and between misuse and fear (0.717). In the open-ended analysis of interview responses, five main themes were found. First, the nurses believed AI could support clinical care through medication management. Second, they noted that AI can provide some psychosocial support but were concerned that emotional connection and empathy are irreplaceable and that AI is limited in this area. The third theme was that pediatric nurses have significant concerns about the accuracy and the reliability of these AI systems because small errors can cause large consequences for patients. The fourth theme revealed that pediatric nurses were concerned with their professional identity and job loss, noting no appropriate boundary recognized between the decision-making process of AI and clinicians’ use and potential exclusion from these decision-making processes. Lastly, pediatric nurses were significantly concerned about ethical boundaries, data security, and legal responsibilities, including patient privacy and HIPAA concerns. The discussion notes that nursing is fundamentally based upon a therapeutic touch, empathy, and human relationships and that it is important to continue these aspects within patient-centered care, especially since machines are unable to provide real human emotions to demonstrate human connection. Despite this, the pediatric nurses within this study believe that AI could improve the overall quality of healthcare using a more assistive approach to reduce nurses' workload so that they are able to focus directly on patient care.
Outcomes and Implications
These findings highlight the need to ensure effective and safe use of AI in pediatric nursing practices. There is a significant difference between emotional expression within an AI and an actual, real human. Hybrid models are being considered to support patient-nurse interaction, however, policies need to be put in place to protect the roles of these clinicians and patient-centered care. Ongoing research is needed to address whether AI should be offered as an assistive device rather than an autonomous device in pediatric nursing settings.