JMIR Publications

ABSTRACT

Background:

Health professionals face significant psychological burdens including burnout, anxiety, and depression, which can negatively impact their well-being and patient care. Traditional psychological health interventions often encounter limitations such as a lack of accessibility and privacy. Artificial intelligence (AI) chatbots are emerging as promising solutions to these challenges by providing accessible and immediate support. Therefore, it is necessary to systematically evaluate the characteristics and effectiveness of AI chatbots designed specifically for health professionals.

Objective:

This scoping review aims to evaluate the existing literature on the use of AI chatbots for psychological health support among health professionals.

Methods:

Following Arksey and O’Malley’s framework, a comprehensive literature search was conducted across eight databases, covering articles published from 2018 to 2024, including backward and forward citation tracking and manual searching from the included articles. Articles were screened for relevance based on inclusion and exclusion criteria, among 2,465 articles retrieved, 10 met the criteria for review.

Results:

Among the 10 articles, three employed a development study design. Six chatbots were delivered via mobile platforms, and four via web-based platforms, all enabling one-on-one interactions. Natural language processing (NLP) algorithms were commonly used and cognitive behavioral therapy (CBT) techniques were applied to psychological health in four articles. Usability was evaluated in six articles, with generally high satisfaction rates; however, dropout rates remained high, particularly in emotionally intense situations. Improvements in anxiety, depression, and burnout were observed in four studies, although one reported an increase in depressive symptoms.

Conclusions:

AI chatbots show promise as tools to support the psychological health of health professionals by offering personalized and accessible interventions. Nonetheless, further research is required to establish standardized protocols and validate the long-term effectiveness of these interventions. Future studies should focus on refining chatbot designs and assessing their impact on diverse health professionals.