- Evaluate effectiveness of AI and machine learning interventions in psychiatric settings for improving continuity of patient care.
- Conduct a mixed methods systematic review (2016–2025 literature), using multiple databases, two independent reviewers, and JBI convergent segregated synthesis.
- Map clinical effectiveness, ethical constraints, and implementation factors to inform clinical guidelines, governance frameworks, and design of proactive learning mental health systems.
JMIR Res Protoc. 2026 Aug 5;15:e95931. doi: 10.2196/95931.
ABSTRACT
BACKGROUND: Continuity of care is essential in psychiatric services due to the chronic, relapsing nature of mental health conditions, yet care pathways remain heavily fragmented at critical transition points. Although advancements in AI and machine learning (ML) offer powerful capabilities to track longitudinal data and automate clinical decision-making, a structured appraisal of their efficacy in supporting continuity of psychiatric care is lacking. This protocol outlines a mixed methods systematic review to evaluate how AI-driven workflows can proactively enhance monitoring, optimize triage and care resource allocation, and address systemic coordination gaps.
OBJECTIVE: The primary objective of this systematic review is to evaluate the effectiveness of AI and ML interventions in psychiatric care settings in improving the continuity of patient care. Secondary objectives include stratifying the types of AI architectures used and identifying implementation barriers and facilitators.
METHODS: A systematic literature search of MEDLINE, Embase, CENTRAL, CINAHL, and APA PsycInfo will be conducted to identify peer-reviewed randomized controlled trials, nonrandomized interventional studies, and qualitative or mixed methods evaluations published between January 1, 2016, and December 31, 2025. Two independent reviewers will perform study screening, data extraction, and quality assessment. A mixed methods convergent synthesis using the Joanna Briggs Institute (JBI) convergent segregated approach will be carried out to synthesize quantitative evidence on effectiveness and qualitative data on implementation.
RESULTS: This review is self-funded and was officially registered with PROSPERO on January 24, 2026 (CRD420251245352). Comprehensive database searches have commenced, with full-text screening and transcript reviews projected to conclude by late August 2026, followed by data analysis and submission of the systematic review manuscript targeted for early spring 2027.
CONCLUSIONS: By systematically mapping interventions across patient, institutional, and health system levels, this review will clarify the clinical effectiveness, ethical boundaries, and logistical implementation factors of psychiatric AI tools. Ultimately, these consolidated insights will provide an evidence-based foundation to inform clinical guidelines; governance frameworks; and the design of proactive, learning mental health systems.
TRIAL REGISTRATION: PROSPERO CRD420251245352; https://www.crd.york.ac.uk/PROSPERO/view/CRD420251245352.
INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): PRR1-10.2196/95931.
PMID:42555933 | DOI:10.2196/95931
Share Evidence Blueprint
Save to Google Notes
Search Google Scholar
Save as PDF
⭐ My Revision List
