Joseph O'Brien, Mayher J Patel, Jon W Schrock
Higher initial serum potassium concentrations were associated with greater short-term clinical risk in ED patients. Borderline potassium elevation was also associated with increased overall clinical risk, although not with higher isolated ventricular arrhythmia rates. These findings support serum potassium as a risk-stratification marker but do not establish a potassium-specific causal or dose-response effect.
BACKGROUND: Hyperkalemia is a common and potentially life-threatening electrolyte disturbance encountered in the emergency department (ED), but the graded association between serum potassium and short-term clinical outcomes in non-dialysis patients remains uncertain.
METHODS: We conducted a multicenter retrospective cohort study using the TriNetX Research Network to examine short-term outcomes among adults (≥18 years) presenting to U.S. EDs between 2010 and 2024. Patients were categorized by initial serum potassium concentration, including a borderline-elevation group (5.01-5.49 mmol/L) and prespecified hyperkalemia strata of 5.5-5.99, 6.0-6.49, 6.5-6.99, and ≥ 7.0 mmol/L, with each group independently compared with propensity score-matched normokalemic controls (3.5-5.0 mmol/L). Propensity models included demographics, comorbidities, laboratory values, and presenting vital signs. The primary outcome was a composite of cardiac arrest, ventricular tachycardia, or ventricular fibrillation within 48 h. Secondary outcomes included ICU admission, mechanical ventilation, hypoglycemia, and death within 48 h.
RESULTS: Among 141,938 eligible patients in the four prespecified hyperkalemia strata, 140,288 were retained after matching. Primary-outcome risk increased from 3.05% at potassium 5.5-5.99 mmol/L to 6.68% at ≥7.0 mmol/L, compared with 0.77-0.90% among matched controls (RR 3.97-8.08). Mortality increased from 2.04% to 5.79%, with similarly higher risks of ICU admission and mechanical ventilation. In the exploratory borderline group, 1,249,213 patients were matched to an equal number of controls; the primary outcome occurred in 1.52% versus 0.85% (RR 1.79, 95% CI 1.75-1.83), although isolated ventricular tachycardia and ventricular fibrillation were not increased.
CONCLUSION: Higher initial serum potassium concentrations were associated with greater short-term clinical risk in ED patients. Borderline potassium elevation was also associated with increased overall clinical risk, although not with higher isolated ventricular arrhythmia rates. These findings support serum potassium as a risk-stratification marker but do not establish a potassium-specific causal or dose-response effect.