Service quality in AI versus human healthcare: a dual-path model driving patient outcomes

This study investigates how service quality shapes patient responses to AI versus human doctors by developing a dual-path model that integrates cognitive and emotional evaluation mechanisms. Using a 2 × 2 experimental design manipulating medical agent type (AI versus human doctor) and service quality, with trust as a measured moderator, data from 296 participants were analyzed via structural equation modeling. The results show that human doctors generate significantly higher patient satisfaction and continuance intention than AI-based services. This effect is fully mediated by three key service evaluation dimensions – perceived diagnostic accuracy, perceived decision transparency, and emotional companionship – capturing functional, informational, and relational aspects of service quality. Furthermore, service quality and trust significantly moderate the cognitive pathway, such that the performance gap between AI and human doctors narrows under high service quality or high trust conditions but widens when these factors are low, whereas the emotional pathway remains less sensitive to these contextual factors. By conceptualizing medical agent type as a source of service quality differentiation and revealing the dual cognitive–emotional mechanisms underlying patient evaluation, this study contributes to the service quality and healthcare management literature, offering new insights into how AI-enabled healthcare services can be designed and optimized to improve patient outcomes.