Related Experiment Video
Updated: Jul 5, 2026
Robotically Delivered fMRI-Guided Personalized Transcranial Magnetic Stimulation Therapy for Treatment-Resistant Depression
Published on: April 10, 2026
Alleviating Nurse Burnout With an Artificial Intelligence-Selected Mobile Cognitive Behavioral Therapy-Based
Yeongeun Kim1,2, Chiyoung Cha2,3, Gumhee Baek2
1Dongnam Institute of Radiological and Medical Sciences, Busan, Republic of Korea.
Background:
Nurse burnout is a pervasive global problem. Cognitive behavioral therapy (CBT) has been shown to reduce burnout; however, most digital CBT programs use standardized approaches that overlook individual differences in burnout profiles. With advances in artificial intelligence (AI), algorithm-based recommendation systems now enable personalized intervention delivery by matching specific CBT modules to users.
Objective:
This study aimed to test the effects of an AI-selected mobile CBT-based intervention on nurse burnout and to describe participants' experiences with the intervention. Specifically, it evaluated whether an AI-selected CBT-based intervention differentially reduced burnout subdomains compared with an information-only control group and explored how nurses perceived and engaged with the AI-selected program.
Methods:
This study adopted a mixed methods design, integrating a 2-group randomized controlled trial and qualitative content analysis exploring participants' experiences. For this randomized controlled trial, a total of 125 nurses were enrolled and randomly assigned to either the experimental group (n=62) or the control group (n=63) between October 2024 and December 2024. The experimental group received an AI-selected mobile CBT-based intervention, in which an AI algorithm assigned CBT modules based on participants' burnout profiles (client-related, personal, and work-related), job stress, and coping characteristics. The control group received information related to burnout management. Primary outcomes, client-related, personal, and work-related burnout, were assessed at baseline, 2 weeks, and 4 weeks. Secondary outcomes, including coping strategies, job stress, and stress response, were assessed at baseline and 4 weeks. Between-group differences in burnout over time were examined using repeated measures analysis of variance, with adjustment for job stress and stress response. Within-group changes and postintervention group differences were analyzed using t tests. Open-ended survey responses and follow-up interviews (n=5 in the experimental group) were analyzed using thematic content analysis.
Results:
Follow-up completion rates were 84.6% (137/162) at both 2 and 4 weeks. The experimental group showed a greater reduction in client-related (F1,121=7.548; P=.007), personal (F1,121=6.533; P=.01), and work-related burnout (F1,121=38.194; P<.001) than the control group, reflecting more pronounced within-group improvements over time. No significant between-group differences were observed for coping strategies, job stress, or stress response. Qualitative findings suggested that some participants were receptive to the AI-selected CBT-based intervention and reported increased self-awareness and reflective engagement with coping strategies that they might not have selected independently.
Conclusions:
The findings suggest that participants were receptive to AI-selected CBT-based interventions, suggesting the potential of such interventions as a supportive approach for alleviating nurse burnout. Future research should explore the sustainability of these effects and optimize the intervention duration to enhance engagement and impact.
