Abstract: FR-PO306
Association of Potassium Level and Mortality in Massive Health Record Databases
Session Information
- Fluid and Electrolytes: Clinical
October 26, 2018 | Location: Exhibit Hall, San Diego Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Fluid and Electrolytes
- 902 Fluid and Electrolytes: Clinical
Authors
- Argyropoulos, Christos, University of New Mexico, Albuquerque, New Mexico, United States
- Bologa, Cristian George, University of New Mexico, Albuquerque, New Mexico, United States
- Pankratz, V. Shane, UNM Health Sciences Center, Albuquerque, New Mexico, United States
- Unruh, Mark L., University of New Mexico, Albuquerque, New Mexico, United States
- Roumelioti, Maria-Eleni, University of New Mexico, Albuquerque, New Mexico, United States
- Shah, Vallabh O., University of New Mexico Health Science Center, Albuquerque, New Mexico, United States
- Shaffi, Saeed Kamran, University of New Mexico, Albuquerque, New Mexico, United States
Background
Dyskalemias have been associated with mortality in studies of small to moderate samples. However the association of potassium (K) with mortality across the spectrum of renal function (est. glom. filtration rate, eGFR) remains poorly defined. Our aim was to characterize this relation in the Cerner Healthfacts database which abstracts data from 1/3 of US healthcare facilities over a period of 10 years.
Methods
Serum K levels, demographics, eGFR (CKD-Epi) and comorbidity (Charlson) collected within 24 hours of all adult patient encounters in Healthfacts were analyzed after excluding patients on dialysis. The most recent K level was associated with each death recorded in Healthfacts. Relative mortality risks (RR) were calculated by Poisson Generalized Estimating Equations, that accounted for the repeated measures of K levels in the same individual. Cubic splines were used to model the relation of K with death in fully adjusted models.
Results
We analyzed 20,697,035 K measurements in 15,376,693 individuals. Patients with CKD were more likely to be older, non-white and have higher K levels (Table). The RR of death with K level was U-shaped; the lowest RR was observed for K: 3.5-4.5 meq/l (Fig A). Hyperkalemia (K>5.5 meq/l) was associated with higher RR in those with low eGFRs (Fig B, p<0.001 for the interaction between eGFR and K).
Conclusion
The RR for hyperkalemia was higher relative to hypokalemia.The interaction between eGFR and K level reflects the variable causes of high K in patients with CKD & the inability of the kidneys to rapidly excrete the K load. Further studies are needed to understand this complex relationship.
| eGFR (ml/min/1.73m2) | |||||
| 15-30 | 30-45 | 45-60 | 60-90 | >90 | |
| N (measurements) | 825,283 | 1,579,185 | 2,371,562 | 7,341,935 | 8,579,070 |
| N (individuals) | 567,685 | 1,123,320 | 1,795,917 | 5,469,704 | 6,420,067 |
| Age (y) | 73 ± 14 | 75 ± 13 | 72 ± 13 | 63 ± 15 | 45 ± 15 |
| Whites (%) | 77% | 82% | 82% | 81% | 68% |
| Males (%) | 43% | 42% | 43% | 45% | 42% |
| Inpatient (%) | 46% | 36% | 30% | 24% | 20% |
| Charlson Score | 2.5 ± 2.8 | 1.8 ± 2.4 | 1.2 ± 1.9 | 0.8 ± 1.5 | 0.5 ± 1.3 |
| Serum K (meq/l) | 4.5 ± 0.8 | 4.3 ± 0.6 | 4.2 ± 0.6 | 4.1 ± 0.5 | 4.0 ± 0.5 |
All % refer to encounters
Funding
- Commercial Support –