Abstract: FR-PO783
Predictors of Hyperkalemia-Related Emergency Department Encounters Among Patients Receiving Hemodialysis Care
Session Information
- Dialysis: Hospitalization and Mortality
October 26, 2018 | Location: Exhibit Hall, San Diego Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Dialysis
- 701 Dialysis: Hemodialysis and Frequent Dialysis
Authors
- Ronksley, Paul E., University of Calgary, Calgary, Alberta, Canada
- Wick, James, University of Calgary, Calgary, Alberta, Canada
- Elliott, Meghan J., University of Calgary, Calgary, Alberta, Canada
- Weaver, Robert G., University of Calgary, Calgary, Alberta, Canada
- MacRae, Jennifer M., University of Calgary, Calgary, Alberta, Canada
Background
Almost 10% of emergency department (ED) visits among dialysis patients are for conditions that could potentially be managed in an outpatient setting such as hyperkalemia. We used population-based data to identify factors that place hemodialysis patients at increased risk of hyperkalemia-related ED events.
Methods
We identified all chronic hemodialysis patients age ≥18 years from March 2009–March 2015 within southern Alberta, Canada. We used a nested case-control design to identify differences between patients with and without hyperkalemia-related ED events (defined by ICD-10 related codes and/or serum K+ ≥6mmol/L). Cases were matched to controls based on dialysis site type (satellite or in-centre) and time period. Clinical and dialysis-specific variables were measured 2-4 weeks prior to outcome dates. We assigned a random date within each control’s period on hemodialysis to serve as a proxy for an outcome date. Potential predictors included demographic/clinical characteristics, prior health system use, and dialysis run sheet variables. Conditional logistic regression models were used to identify significant predictors of hyperkalemia-related ED events.
Results
Of 2012 patients on chronic hemodialysis, 129 had 180 hyperkalemia-related ED events (cases) within the study timeframe. Controls were matched to cases at a ratio of ~3-to-1 (508 controls). In bivariate analysis, cases were younger, had higher levels of comorbidity, higher acute care use in the prior 6 months, and received more intensive dialysis treatment (e.g. higher ultrafiltration (UF) volume, cumulative duration of dialysis) in the weeks prior to an ED event. Multivariate modeling identified the following predictors of hyperkalemia-related ED events: ED use in prior 6 months (OR: 2.70; 95% CI: 1.69-4.33), dialysate potassium concentration ≤3.0mmol/L (2.95; 1.28-6.77), average UF volume>2.5L per dialysis session (2.73; 1.59-4.71), >15 hours of cumulative dialysis time in the prior week (5.84; 2.32-14.70), dialysis access via fistula (1.79; 1.20-2.66).
Conclusion
We identified a number of predictors that place patients at greater risk of presenting to the ED with hyperkalemia. Identification of such patients may allow for targeted strategies for preventive care, thus avoiding unnecessary acute care use and cost while improving patient quality of life.