Abstract: TH-PO593
Artificial Intelligence Hospitalization Risk Model-Guided Selection for Enhanced Psychosocial Care
Session Information
- Health Maintenance, Nutrition, Metabolism - I
November 07, 2019 | Location: Exhibit Hall, Walter E. Washington Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Health Maintenance, Nutrition, and Metabolism
- 1300 Health Maintenance, Nutrition, and Metabolism
Authors
- Chaudhuri, Sheetal, Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Demaline, Jessica, Fresenius Health Partners, Austin, Texas, United States
- Johnstone, Stephanie, Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Weissman-Hunt, Amy R., Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Han, Hao, Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Jiao, Yue, Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Reviriego-Mendoza, Marta, Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Larkin, John W., Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Usvyat, Len A., Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Ketchersid, Terry L., Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Maddux, Dugan, Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Maddux, Franklin W., Fresenius Medical Care, Waltham, Massachusetts, United States
Background
A large dialysis organization (LDO) leverages an artificial intelligence (AI) risk model to help guide screening for a Social Work Intensive (SWI) program in value-based care (VBC) settings. SWI provides enhanced psychosocial care to improve quality of life, thereby reducing hemodialysis (HD) non-adherence/hospitalizations. In conventional models, patients are screened for SWI based on team identification of non-adherence or difficulty achieving outcome goals. We assessed if AI-directed SWI enrollment in VBC yielded comparable benefits to conventional clinician-based paradigms. The statements contained in this document are solely those of the authors and do not necessarily reflect the views or policies of the Centers for Medicare & Medicaid Services (CMS). The authors assume responsibility for the accuracy and completeness of the information contained in this document.
Methods
We used data from patients enrolled in SWI in 2017. In VBC, a 12-month hospital admission risk model guided SWI screening. In conventional care, SWI screening was based on clinician evaluation of risk. Patients screened positive for barriers in depression, stress, and sleep were enrolled into SWI and provided tailored weekly interventions for 8 weeks. We calculated admission and HD non-adherence 3 months before and after SWI enrollment in VBC and conventional settings.
Results
Among 6425 patients (conventional n=4646, VBC n=1779) enrolled in SWI program, hospital admission rates were similar before SWI enrollment between care settings, but HD non-adherence rate was lower in VBC settings. Admission and HD non-adherence rates were consistently lower 3 months after SWI enrollment in both a VBC and conventional setting (Figure 1A & B).
Conclusion
AI-directed SWI screening and enrollment in VBC appears to have consistent improvements in outcomes compared to clinician-based identification of risks in conventional settings. Use of this AI technology may help streamline efforts and allowing more time to focus on patient care.
Funding
- Commercial Support –