Cost-effectiveness analysis of prognostic-based depression monitoring. Issue 1 (2nd January 2019)

• Record Type: Journal Article
• Authors: Lin, Ying; Huang, Shuai; Simon, Gregory E.; Liu, Shan
• Is Part Of: IISE transactions on healthcare systems engineering. Volume 9:Issue 1(2019)
• Page Start: 41
• Page End: 54
• Publication Date: 2019-01-02
• Subjects: Prognostic-based monitoring -- cost-effectiveness analysis -- depression monitoring -- electronic health record -- mental health -- machine learning -- statistical learning -- Markov model -- chronic conditions -- health services -- Markov chains -- monitoring -- ROC curve -- clinical decision model; Biomedical engineering -- Periodicals; Medical informatics -- Periodicals; Medical care -- Periodicals; 610.28
• DOI: 10.1080/24725579.2019.1567627
• Languages: English
• ISSNs: 2472-5579

Abstract

Abstract: Chronic depression monitoring often relies on one-size-fits-all routine monitoring guidelines. Without considering heterogeneity in patients' disease progression, routine monitoring guidelines may lead to inadequate monitoring of sick individuals and unnecessary monitoring of healthy individuals. Prognostic-based monitoring that stratifies the individual's disease progression risk into different levels and adaptively allocates monitoring resources to high-risk individuals has the potential to improve patient health outcomes and cost-effectiveness of the monitoring service. However, challenges include how to best apply prognostic models to inform the design of monitoring strategies and identify the cost-effective strategies. To address these challenges, we develop a decision support framework that integrates individual prognostics, monitoring strategy design and cost-effectiveness analysis. We apply the proposed framework to simulate the adaptive monitoring of a depression treatment population from electronic health record data. Several prediction algorithms with increasing complexity, including natural history matching, logistic regression, rule-based method and Markov-based collaborative model, are simulated to monitor high-risk individuals for severe depression over time. We find six cost-effective monitoring strategies and demonstrate that two routine monitoring strategies are dominated by the prognostic-based monitoring strategies. Methods from this research show promise for implementing prognostic-based monitoring of chronic conditions in clinical practice.