JMIR Publications

ABSTRACT

Background:

As populations age globally, artificial intelligence (AI)-enabled digital health interventions are increasingly being adopted to support integrated, person-centered care for older adults. However, while evidence on the technical performance of AI technologies is expanding rapidly, their economic value and cost-effectiveness in elderly care settings remain poorly understood.

Objective:

This systematic review aims to synthesize and critically appraise the current evidence on the economic evaluation of AI technologies for older adults, mapping studies across the World Health Organization's Integrated Care for Older People (ICOPE) pathway to identify evidence gaps, assess economic outcomes, and examine factors influencing the cost-effectiveness of AI-assisted interventions in elderly healthcare.

Methods:

A systematic review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. PubMed, Embase, Web of Science and EconLit were systematically searched for studies published up to October 2025. Eligible studies were full economic evaluations examining artificial intelligence (AI)–based technologies used in elderly healthcare. Titles, abstracts, and full texts were screened independently by two reviewers. Reporting quality was assessed using the Checklist for Health Economic Quality Evaluations (CHEQUE) quality assessment tool. Included studies were synthesized using descriptive statistics and narrative synthesis. The analysis was structured according to the four-step ICOPE care pathway.

Results:

In total, 30 studies published between 2018 and 2025 were included, covering 14 countries. Most evaluations focused on screening and diagnostic applications of AI (n = 24), particularly in cancer and ophthalmologic conditions. Across studies, model-based approaches predominated, with decision trees, Markov models, and discrete-event simulations commonly used. AI-related costs varied widely across studies, with only a minority drawing on trial-based data. Many evaluations (63.3%, n=19) relied on assumptions, manufacturer quotes, or expert opinion, and cost components were often insufficiently detailed. Most AI interventions (90%, n=27) were found to be cost-effective, with half (n=15) demonstrating cost savings driven by improved diagnostic accuracy, earlier detection, and labor substitution. Key drivers of cost-effectiveness included AI performance parameters, AI-related costs, population characteristics, disease burden, and local healthcare system features.

Conclusions:

This systematic review provides a comprehensive synthesis of economic evidence on AI-assisted technologies in elderly healthcare following the ICOPE care pathway. Findings show that formal evaluation evidence is largely missing for personalized planning and monitoring interventions, with the majority of evaluations focused on screening and diagnostic technologies.