JMIR Publications

ABSTRACT

Background:

The rapid advancement of artificial intelligence (AI) has introduced new technological pathways for palliative care, particularly in data integration, predictive analytics, and real-time monitoring. However, for older adults characterized by multimorbidity, functional decline, and complex care needs, the context-specific applicability and implications of AI remain insufficiently understood.

Objective:

This review aimed to map the applications of AI in palliative care for older adults, examine its context-dependent applicability, and synthesize its advantages, limitations, and future directions.

Methods:

A scoping review was conducted following the framework of Arksey and O’Malley (2005). Literature searches were performed in PubMed, Web of Science, EBSCO, Embase, and Scopus. Eligible studies were screened, and data were synthesized using thematic analysis.

Results:

Artificial intelligence has the potential to support palliative care for older adults, but its performance varies depending on care contexts and patient needs. In older populations characterized by multimorbidity and complex care demands, its value should be recognized while its limitations are carefully considered. The key lies in effectively integrating AI into the specific care contexts of older adults. This study further clarifies the role of AI in this field, positioning it as a supportive tool that complements, rather than replaces, person-centered palliative care.

Conclusions:

A total of 11 studies were included. AI applications were embedded across clinical settings, provider workflows, and health system operations, supported by diverse modeling approaches and multi-source real-world data. From a trade-off perspective, AI demonstrates advantages in early detection of clinical deterioration, integration of multidimensional information, and improved care efficiency, but remains limited in addressing psychosocial needs, supporting individualized decision-making, and ensuring model interpretability and data representativeness. Future development should prioritize age-friendly design, integration of multidimensional patient-centered indicators, and enhanced data interoperability.