Optimizing Care Trajectories with Artificial Intelligence: A Systematic Review and Meta-Analysis

Ben Khizzou, Nouhaila; Aarabe, Mourad; Alla, Lhoussaine; Benjelloun, Ahmed

doi:10.31557/APJCP.2026.27.5.1589

Optimizing Care Trajectories with Artificial Intelligence: A Systematic Review and Meta-Analysis

Document Type : Systematic Review and Meta-analysis

Authors

¹ National School of Business and Management, Sidi Mohamed Ben Abdellah University, Fez, Morocco.

² National School of Applied Sciences, Sidi Mohamed Ben Abdellah University, Fez 31000, Morocco.

10.31557/APJCP.2026.27.5.1589

Abstract

Objective: This study aims to provide a critical and comprehensive summary of existing research on the effectiveness of AI-assisted care pathways by analyzing their impact on hospital stays and readmissions. Method: This review was conducted in accordance with the PRISMA 2020 guidelines, which specify the selection and analysis steps from the outset. The search strategy targeted three international databases: Scopus, Web of Science (WoS), and PubMed (Medline). The inclusion criteria focused on randomized controlled trials (RCTs), before-and-after studies, quasi-experimental studies, longitudinal studies, and literature reviews involving hospitalized patients without age or pathology restrictions, provided that their care was based on the “clinical pathway” method. Studies that did not meet these criteria were excluded. A PRISMA diagram illustrates the systematic selection process, which resulted in the selection of 31 studies. Data extraction was performed using a structured methodology to ensure the validity and comparability of the results. The description of random sequence generation processes was adequate in 24 studies, partially addressed in five studies, and insufficient in two studies. This limits the ability to assess the risk of bias. Results: With regard to economic outcomes, such as length of stay and readmission, most studies [reported] positive effects associated with AI-assisted pathways. The Covidence tool (version 101) was used for the selection and extraction of data on the 18 studies analyzing pathways, AI, length of stay, and readmission. Group analyses indicated that the implementation of clinical protocols incorporating AI optimized care without increasing readmission rates or length of stay. Conclusion: However, the limited number and heterogeneity of studies on the application of artificial intelligence currently prevent the establishment of a universal framework for the implementation of AI-assisted clinical protocols.

Keywords

Main Subjects