Predictors of Hyperkalemia among Patients on Maintenance Hemodialysis Transported to the Emergency Department by Ambulance

Abstract

Background: Hyperkalemia is common among maintenance hemodialysis (HD) patients and is associated with mortality. We hypothesized that clinical characteristics available at time of paramedic assessment prior to emergency department (ED) transport (ambulance-ED) would associate with severe hyperkalemia (K≥6.0). Rapid identification of patients who are at risk for hyperkalemia and thereby, hyperkalemia associated complications, may allow paramedics to intervene in a timely fashion, including directing emergency transport to dialysis-capable facilities. Methods: Maintenance HD patients from a single paramedic provider region, who had at least one ambulance-ED and subsequent ED potassium from 2014-2018, were examined using multivariable logistic regression to create risk prediction models inclusive of prehospital vital signs, days from last dialysis, and the presence of prehospital electrocardiogram (ECG) features of hyperkalemia. We used bootstrapping with replacement to internally validate each model and performance was assessed by discrimination and calibration. Results: Amongst 704 ambulance-ED visits, severe hyperkalemia occurred in 75 (11%); 26 patients with ED hyperkalemia did not have a prehospital ECG. Younger age at transport, longer HD vintage, more days from last hemodialysis session (OR 49.84, 95% CI 7.72-321.77 for 3+ days versus HD the same day (before) ED transport), and prehospital ECG changes (OR 6.64, 95% CI 2.31-19.12) were independently associated with severe ED hyperkalemia. A model incorporating these factors had good discrimination (c-statistic 0.82, 95% CI 0.76-0.89) and using a cut-off of 25% probability, correctly classified patients 88.9% of the time. Conclusions: Characteristics available at the time of ambulance-ED were associated with severe ED hyperkalemia. An awareness of these associations may allow healthcare providers to define novel care pathways to ensure timely diagnosis and management of hyperkalemia.

Predicting Urgent Dialysis at Ambulance Transport to the Emergency Department Using Machine Learning Methods

“…In this study, a risk prediction model incorporating presenting complaint, vital signs prior to transport, and time from last dialysis had very good discrimination (C-statistic 0.81, 95% CI: 0.76-0.86) for the primary outcome [1]. In subsequent work, Vinson et al [5] used the clinical characteristics of hemodialysis patients to predict hyperkalemia using multivariable logistic regression. In a study of 704 patients, following ambulance transportation to the ED, 75 patients (11%) had severe hyperkalemia (≥6 mmol/L), and a risk prediction model that included prehospital vital signs, days from last hemodialysis and a prehospital electrocardiogram with features of hyperkalemia had an AUC of 0.82 for severe hyperkalemia.…”

A Nomogram to Identify Hyperkalemia Risk in Patients with Advanced CKD

“…The clinical risk and potential economic value highlight the importance of predicting the risk of developing hyperkalemia in CKD patients with a high risk 16 . Traditional predictive factors associated with hyperkalemia in CKD include high CKD stages, dehydration, consuming excessive dietary potassium, heart failure (HF), diabetes, metabolic acidosis, and using potassium-increasing medications such as RAASis, potassium-sparing diuretics, and et al 9,[17][18][19][20][21] However, convenient tools for diagnosis or prediction the risk of hyperkalemia in CKD are few 2 .…”