Artificial Intelligence in the National Health System: Assessment of Patient Satisfaction and Trust in Primary Care Services

Abstract

In recent years, artificial intelligence has gained relevance in multiple domains, and in the specific domain, AI agents for patient care have consolidated. However, patients have lower levels of satisfaction with these agents mainly due to their lack of emotional responsiveness, perceived credibility, and even perceived practicality. This research aims to analyze whether the type of agent (AI vs. Human) exerts a significant influence on the level of perceived wellbeing of users in primary care services in the SNS 24 application. To materialize the objective, a between-subjects experiment was used, with 201 participants submitted to single-factor scenarios (AI Agent vs. Human). The results demonstrate that the type of agent exerts a significant influence on the perceived well-being of users, with the lowest level of satisfaction identified with the AI agent. The research also showed that perceived credibility is a mediator for the effect of the type of agent on the perceived well-being of users, as well as on perceived practicality. Finally, this research, theoretically based on empirical evidence, validates and extends the service encounter 2.0 and the AI Experiential Perspective, and in practice supports the implementation of automatic systems in the SNS 24 APP.