Abstract: SA-PO398
eGFR Trajectory Pre-Dialysis Initiation Predicts 90-Day Hospitalization and Mortality Risk
Session Information
- Hemodialysis and Frequent Dialysis: CV and Risk Prediction
November 05, 2022 | Location: Exhibit Hall, Orange County Convention Center‚ West Building
Abstract Time: 10:00 AM - 12:00 PM
Category: Dialysis
- 701 Dialysis: Hemodialysis and Frequent Dialysis
Authors
- Kraus, Michael A., Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Wang, Yuedong, University of California System, Oakland, California, United States
- Wang, Catherine Y., Carnegie Mellon University, Pittsburgh, Pennsylvania, United States
- Hahn Contino, Carly, Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Usvyat, Len A., Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Kotanko, Peter, Renal Research Institute, New York, New York, United States
- Kossmann, Robert J., Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Kaufman, Harvey W., Quest Diagnostics Inc, Secaucus, New Jersey, United States
Background
Previous VA study1 showed 1-year hospitalization and mortality in ESRD patients was linked to eGFR rate decline in 2 yrs pre-dialysis. This study evaluated males and females and applied the CKD-EPI 2021 eGFR equation to define clusters that estimate risk of hospitalization and death within 90 days after dialysis initiation.
Methods
Patients initiated on dialysis at Fresenius Medical Care with matched Quest Diagnostics clinical testing, 1/1/2015 - 9/30/2020 were included and deidentified. Inclusion criteria: a) 2 yrs of lab data; b) >=10 serum creatinine (Scr) measurements; c) >= 1 Scr within 45 days; d) >= 2 Scr distributed over 2 quarters. Exclusion criteria were sudden eGFR increase > 20 mL/min/1.73 m2,or mean eGFR >30 mL/min/1.73 m2 within 45 days of dialysis. The revised CKD-EPI 2021 eGFR calculation was applied.
A cubic spline was fitted to eGFRs from each patient followed by functional data clustering methods to learn patterns of eGFR trajectories. Both K-mean and functional principal component analysis were used. One-way ANOVA and χ2 tests were used to compare the trajectory groups for continuous and categorical variables.
Results
2341 patients: 42% female, mean age 64.9 +/-12.3 years, 62.7% White, non-Hispanic, 15.7% Black non-Hispanic, 15.0% Hispanic. Patients grouped into 4 clusters of eGFR progression velocity: 1076 stable low cluster (slc); 920 slow decay cluster (sdc); 285 fast decay cluster(fdc), and 60 in the very fast decay cluster (vfdc). Comparing clinical laboratory test measures, faster decay groups tended to have lower levels of 25-Vit D, iPTH, and albumin; higher uACR, and slightly higher hgb A1C. Vfdc had higher WBC. 90-day mortality rates: 0.034 (slc), 0.043 (sdc), 0.084 (fdc), and 0.117 (vfdc). 90-day hospitalization rates: 0.219 (slc), 0.252 (sdc), 0.295 (fdc), and 0.500 (vfdc). Mortality HR compared to the slc was 2.98 fdc and 4.53 vfdc. Hospitalization odds ratio compared to slc 1.57 fdc and 3.84 vfdc. Cluster category was significantly associated with hospitalization and mortality, even after adjusting for demographic variables.
Conclusion
fdc and vfdc were associated with higher risk of hospitalization and mortality within 90-days after dialysis initiation.
1. O'Hare AM, et al. Am J Kidney Dis 2012 April: 59(4): 513-522
Funding
- Private Foundation Support