JMIR Publications

ABSTRACT

Background:

Mental health disorders (MHDs) represent a growing global challenge and pose a significant risk to public health. Alongside developments in the field of large language models (LLMs), conversational mental health chatbots (CMHBs) have emerged and are increasingly being used by individuals in self-directed and independent ways to provide therapy and therapeutic support. While users’ perspectives on the use of CMHBs have been extensively examined and systematically synthesized, relatively little research has focused on how healthcare professionals (HCPs) perceive these tools. To develop a holistic understanding of the implications of CMHB use – including potential benefits, risks, and implementation barriers – it is essential to consider the perspectives of HCPs, who bring clinical expertise and psychological knowledge to the evaluation of mental health interventions. Accordingly, the objective of this review is to synthesize empirical evidence on HCPs’ perspectives regarding the use of CMHBs and to explore potential convergences and divergences between professional and user perspectives.

Objective:

This paper presents the protocol for a systematic review that aims to identify, synthesize, and critically appraise evidence on healthcare professionals’ perspectives regarding the use of CMHBs as tools for therapy or therapeutic support for individuals with MHDs, and to examine perceived benefits, barriers, and potential ethical concerns associated with their use.

Methods:

A systematic review of literature will be conducted in accordance with the PRISMA 2020 guidelines. Peer-reviewed qualitative, quantitative, and mixed-methods studies will be identified through searches of PubMed/MEDLINE, PsycINFO, Embase, CINAHL, Web of Science, and Scopus, with no restrictions on publication date. Study screening will be supported by AI-assisted active learning using ASReview, following the SAFE stopping procedure, with independent quality-assurance screening by a second reviewer. Data will be synthesized narratively, and methodological quality will be appraised using the Critical Appraisal Skills Programme (CASP) checklist.

Results:

The database search for this review was performed end of November 2025. The initial Title/abstract screening started in January 2026 and is currently underway. Data extraction is expected to be completed by April, and the final results are expected to be published by August 2026.

Conclusions:

In light of the rapid emergence of AI-driven chatbots in mental health care, this systematic review will synthesize current empirical evidence to address the urgent need to understand HCPs’ perspectives on the use of CMHBs. Specifically, it will examine how HCPs perceive CMHBs when used to simulate therapeutic interactions, as adjunctive support to conventional therapy, or as potential substitutes for specific therapeutic functions. By identifying perceived benefits, barriers, and ethical concerns, this review aims to contribute to a more comprehensive understanding of the implementation and broader implications of CMHBs in mental health care.