Artificial Intelligence in Multimorbidity Care: A Systematic Review of Opportunities, Challenges and Limitations

Aim: This study identified AI’s role in supporting patients with multimorbidity in primary healthcare and explored associated opportunities, challenges, and limitations. Methodology: A systematic literature search was conducted in Scopus, Web of Science, and PubMed using terms related to AI and multimorbidity in PHC. After screening, 53 articles met inclusion criteria and were analysed per PRISMA guidelines. Results: AI can support multimorbidity patients through personalized treatment, clinical support, virtual assistance and coordinated care via monitoring and interventions. However, adoption remains limited by poor data quality, disease interactions, low model interpretability and underrepresentation of multimorbid populations in training data. Consequently, AI’s role in supporting multimorbidity patients remains limited, with potential benefits unrealized. Implications and recommendations: Realizing AI’s potential for multimorbidity patients requires addressing technical, ethical, clinical and systemic challenges. Research must prioritize developing representative datasets and interpretable models that capture multimorbidity’s complexity to help deliver coordinated, patient-centred care. Originality/value: This analysis shows limited AI integration in multimorbidity care, highlighting the need for quality data and interpretable models. It identifies key challenges that must be overcome to transform AI from conceptual promise to practical support for individuals with multimorbidity.