Abstract: PO1136
Predicting Dialysis Facilities at Risk of Low ICH-CAHPS Quality of Center Scores
Session Information
- Hemodialysis and Frequent Dialysis - 3
October 22, 2020 | Location: On-Demand
Abstract Time: 10:00 AM - 12:00 PM
Category: Dialysis
- 701 Dialysis: Hemodialysis and Frequent Dialysis
Authors
- Lofty, Karen, Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Jiao, Yue, Fresenius Medical Care, Global Medical Office, Waltham, Massachusetts, United States
- Weber, Stephen, Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Larkin, John W., Fresenius Medical Care, Global Medical Office, Waltham, Massachusetts, United States
- Usvyat, Len A., Fresenius Medical Care, Global Medical Office, Waltham, Massachusetts, United States
- McClure, Angela, Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Hymes, Jeffrey L., Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Maddux, Franklin W., Fresenius Medical Care AG & Co. KGaA, Global Medical Office, Bad Homburg, Germany
Background
Medicare administers biannual ICH-CAHPS surveys to capture patients’ perceived experience of outpatient hemodialysis (HD) care. Recent operations efforts at a national dialysis provider aimed to develop prediction models to identify HD facilities at risk of low ICH-CAHPS rating in the subsequent survey period.
Methods
We used retrospective data from HD facilities at a national dialysis provider during 2018-2019. Two models were built to predict HD facilities that continued to have (Model 1) or decreased below (Model 2) a <60% top box ICH-CAHPS rating on the dialysis center staff, care, and operations subdomain in the spring 2019 survey period. Facility variables in 2018 included were: fall and spring ICH-CAHPS ratings; patient/employee net promoter (NPS) scores; employee retention rate; center quality Five Star rating; years of certification; facility size; composite clinical quality score; and % of HD non-adherence. Predictor variable importance was evaluated, and the performance of various modeling methods was assessed using several machine learning algorithms. We randomly selected derivation (70%) and validation (30%) datasets.
Results
We found the highest performance using GLM and GAM methods for both Models (Figure 1). The assessment of performance via the area under curve (AUC) showed use of GLM modeling correctly predicted true/false positives in 73% of facilities that continued to have (Model 1) and 70% of facilities that decreased below (Model 2) a <60% top box ICH-CAHPS rating.
Conclusion
The developed prediction models may be used as a tool in identifying HD facilities at risk of low patient ICH-CAHPS ratings. Prospective use in quality improvement efforts appears warranted.
Funding
- Commercial Support –