Abstract: SA-PO1080
Validation of Revisions to the NxStage Dosing Calculator
Session Information
- Home Hemodialysis
November 09, 2019 | Location: Exhibit Hall, Walter E. Washington Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Dialysis
- 702 Dialysis: Home Hemodialysis
Authors
- Kubisiak, Kristine, Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Weinhandl, Eric D., Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Ofsthun, Norma J., Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Dalrymple, Lorien S., Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Kraus, Michael A., Fresenius Medical Care North America, Waltham, Massachusetts, United States
- Maddux, Franklin W., Fresenius Medical Care North America, Waltham, Massachusetts, United States
Background
The NxStage Dosing Calculator (DC) is an online tool that can be used to identify hemodialysis prescriptions that achieve a specified standardized Kt/V on the NxStage System One (NSO) platform. The DC has recently been revised to align with formulae in Kidney Disease Outcomes Quality Initiative guidelines and to permit incorporation of residual renal function (RRF). We used home hemodialysis (HHD) patient data from a large dialysis organization to validate DC revisions.
Methods
We identified adult patients who initiated HHD with NxStage equipment in Fresenius Kidney Care (FKC) clinics between March 1, 2016, and December 1, 2018. Patients were followed from completion of HHD training to discontinuation of HHD with NxStage equipment at FKC. We collected patient-days with pre- and post-dialysis blood urea nitrogen (BUN) measurements and retained those with exact adherence to prescribed treatment frequency during the preceding 7 days. We derived model-predicted standardized (std) Kt/V from the HHD prescription, and observed std Kt/V from BUN measurements, before and after DC revisions.
Results
The cohort included 3427 patients and 23,408 patient-days with BUN measurements. Over 91% of patient-days were accompanied by 4 or 5 weekly treatments, and RRF was measured in 19%. Model-predicted and observed std Kt/V, given DC revisions, are displayed in the figure, with a smoothed trend in orange. DC revisions increased the proportion of variation in observed std Kt/V that was explained by model-predicted std Kt/V from 26% to 58%.
Conclusion
Revisions to the NxStage DC have improved concordance of predicted and observed std Kt/V in HHD patients, especially near guideline targets. RRF is influential on forecasted Kt/V, so continued reliance on RRF in HHD patients requires accurate, timely data about RRF.