Application of Machine Learning in Multimorbidity Research: Protocol for a Scoping Review

Area of Science:

• Health Informatics
• Artificial Intelligence in Medicine
• Public Health

Background:

• Multimorbidity, the presence of multiple chronic conditions, significantly burdens global healthcare systems, increasing mortality, reducing quality of life, and escalating costs.
• The complexity of multimorbidity necessitates advanced analytical tools for effective management and intervention.
• Machine learning (ML) offers powerful capabilities for disease prediction, treatment development, and clinical strategy optimization in managing complex health conditions.

Purpose of the Study:

• To conduct a scoping review identifying and exploring the existing literature on machine learning applications for patients with multimorbidity.
• To recognize diverse ML models, patient cohorts, input data types, and algorithm maturity in multimorbidity research.
• To map the outcomes and performance metrics of ML models applied to multimorbidity.

Main Methods:

• Adherence to PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) guidelines.
• Comprehensive literature search across five major databases: PubMed, Embase, IEEE, Web of Science, and Scopus.
• Independent screening of titles, abstracts, and full texts by two reviewers using Covidence, focusing on studies involving multiple chronic diseases or at-risk individuals.

Main Results:

• The review will present a PRISMA-ScR flow diagram detailing the study selection process.
• Findings will be synthesized through a narrative approach, supplemented by charts and tables for extracted data.
• Results will be published in a peer-reviewed journal, providing a comprehensive overview of the field.

Conclusions:

• This scoping review is potentially the first to systematically investigate ML in multimorbidity research.
• It aims to summarize current approaches, identify diverse ML applications, and highlight research gaps in managing multiple chronic conditions.
• The findings will inform future research directions, contributing to improved patient outcomes through advanced ML techniques.