Barriers and Facilitators to the Implementation of Artificial Intelligence Enabled Diabetes Interventions in Lower-Middle-Income Countries: A Systematic Review Protocol
Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background
Diabetes represents an emerging global health crisis, with lower-middle-income countries experiencing a fast growth in prevalence. Diabetes care in these regions often faces significant challenges, including inadequate healthcare infrastructure, limited financial resources, a shortage of trained healthcare personnel, and a dual burden of communicable and non-communicable diseases. Artificial intelligence (AI) offers tailored and scalable solutions for addressing these systemic barriers by enabling early diagnosis and risk prediction, integrating diabetes care delivery across all levels of healthcare. However, the successful implementation of AI interventions requires an understanding of the unique infrastructural, technological, socio-political, and cultural factors influencing diabetes care in these regions.
Objective
This protocol outlines a systematic review to synthesize evidence on the barriers and facilitators to implementing AI-enabled diabetes interventions in lower-middle-income countries, and to examine the specific AI applications being deployed in these settings.
Methods
A comprehensive literature search will be performed across five databases: Medline, Web of Science, CINAHL, IEEE Xplore, and ACM Digital Library, encompassing peer-reviewed publications from 2015 to 2025. The review will include studies which assess AI-enabled interventions implemented in healthcare settings for diabetes prevention, diagnosis, and management in lower-middle-income countries. Studies that assess the efficacy of artificial intelligence tools without direct evaluation of these tools in clinical decision-making or patient care processes will be excluded. Two independent reviewers will assess studies for inclusion using a predefined search strategy. The reporting of results will adhere to the PRISMA 2020 checklist (1). The risk of bias in individual studies will be evaluated using Hawker’s tool for disparate study designs.
Conclusion
The findings of this systematic review will identify key considerations for implementing AI technologies in diabetes care and provide evidence to support policymakers, healthcare providers, and technology developers in designing context-appropriate interventions that improve care delivery and health outcomes in lower-middle-income countries.
