Abstract: TH-OR100
Association of Baseline eGFR and AKI-D with Early Mortality after ESRD
Session Information
- Outcomes and Trends in Dialysis
October 25, 2018 | Location: 2, San Diego Convention Center
Abstract Time: 06:18 PM - 06:30 PM
Category: Dialysis
- 701 Dialysis: Hemodialysis and Frequent Dialysis
Authors
- Hsu, Raymond K., UCSF & SF VA Medical Center, San Francisco, California, United States
- Rubinsky, Anna, UCSF & SF VA Medical Center, San Francisco, California, United States
- Shlipak, Michael, UCSF & SF VA Medical Center, San Francisco, California, United States
- Peralta, Carmen A., UCSF & SF VA Medical Center, San Francisco, California, United States
- Lee, Benjamin J., UCSF & SF VA Medical Center, San Francisco, California, United States
- Hsu, Chi-yuan, UCSF & SF VA Medical Center, San Francisco, California, United States
Background
Abrupt decline in kidney function and non-recovery of dialysis-requiring AKI (AKI-D) have both been shown to be associated with adverse outcomes in patients with incident ESRD.We hypothesized that distinct patterns of transition to ESRD based on both the mean 1-year baseline eGFR and presence of AKI-D preceding ESRD would help prognosticate 90-day mortality after hemodialysis (HD) initiation.
Methods
We identified a national cohort of Veterans who initiated maintenance HD in 2009-2013 and had outpatient creatine (Cr) measured within 1 year prior to ESRD. Patients were categorized based on A) mean outpatient eGFR within 1 year prior to incident ESRD (≥ or < 30); and B) occurrence of in-hospital AKI-D leading to ESRD. AKI-D was defined using KDIGO criteria of [≥0.3mg/dL rise and/or ≥50% rise in Cr from mean 1-year baseline], and inpatient dialysis. The association between patterns of transition and all-cause mortality within 90 days after HD initiation were examined using multivariable Cox models.
Results
Patients with incident ESRD (N=22,815) were 69+11 years old, 98% male, 27% black, and 68% diabetic. 4114 (18%) had mean 1-year outpatient eGFR>30; 2644 (12%) experienced AKI-D with peak inpatient Cr rise was ≥50% from baseline. Relative to refence group of mean outpatient eGFR<30 and no AKI-D, patients with mean outpatient eGFR≥30 had ~2-fold adjusted risk of 90-day mortality, regardless of presence/absence of AKI-D (Table1). Among patients with mean eGFR<30, those with AKI-D and ≥50% rise in Cr had a small increase in mortality, whereas those with AKI-D and smaller peak Cr (<50% but ≥0.3mg/dl rise) had lower 90-day adjusted mortality.
Conclusion
Nearly 1 in 5 incident ESRD patients had a mean outpatient eGFR≥30 within the 1 year prior to HD initiation. This pattern of abrupt transition to ESRD was strongly associated with early mortality regardless of whether or not AKI-D directly preceded ESRD.
Pattern of Transition | [Death in first 90 days after ESRD] / [Total N] = crude death rate | Unadjusted Hazard Ratio | Fully Adjusted Hazard Ratio** |
eGFR<30*, no AKI-D | 1229 / 15,484 = 7.9% | Reference | Reference |
eGFR<30*, + AKI-D with peak Cr <50% from baseline | 39 / 1076 = 3.6% | 0.45 (0.32-0.61) | 0.62 (0.45-0.86) |
eGFR<30*, + AKI-D with peak Cr ≥50% from baseline | 170 / 2141 = 7.9% | 1.00 (0.85-1.17) | 1.31 (1.11-1.55) |
eGFR≥30*, no AKI-D | 655 / 3611 = 18.1% | 2.41 (2.19-2.65) | 1.94 (1.75-2.15) |
eGFR≥30*, + AKI-D with peak Cr ≥50% from baseline | 82 / 503 = 16.3% | 2.14 (1.71-2.68) | 2.09 (1.66-2.62) |
*refers to mean outpatient eGFR within 1 year prior to ESRD. **adjusted for demographics, co-morbidities, and pre-ESRD nephrology care.
Funding
- NIDDK Support