Background: Serum potassium levels are recognized for their prognostic significance in patients presenting with acute myocardial infarction (AMI). However, the correlation between potassium level fluctuations and mortality rates among AMI patients remains unclear.
Methods: A retrospective cohort study was conducted using the MIMIC-IV database, including deidentified data from patients admitted to the Beth Israel Deaconess Medical Center from 2008 to 2022. Potassium fluctuation was assessed using parameters including mean postadmission serum potassium levels (K+[mean]), first measurable value upon admission (K+[admission]), minimum (K+[min]) and maximum (K+[max]) measurable values, and coefficient of variation (K+[CV]). The primary outcome was all-cause in-hospital mortality; secondary outcomes included ventricular tachycardia or fibrillation (VT/VF) and cardiac arrest. Restricted cubic spline models and logistic regression models were used to assess the associations between potassium fluctuation and clinical outcomes.
Results: A J-shaped correlation between serum potassium levels and the risk of in-hospital mortality was identified. Both high and low potassium levels were significantly associated with increased mortality. Specifically, K+[mean] levels below 3.5 mmol/L and above 4.5 mmol/L were associated with higher mortality. Elevated K+[CV] values were also associated with higher in-hospital mortality in both univariate and multivariate analyses. Increased potassium variability was correlated with elevated risks of both VT/VF and cardiac arrest.
Conclusions: Serum potassium fluctuation is an independent predictive factor for in-hospital mortality among AMI patients. These findings underscore the importance of maintaining potassium homeostasis in the management of AMI, suggesting that monitoring and stabilizing potassium levels are crucial for reducing in-hospital mortality.
Keywords: Acute myocardial infarction; In-hospital mortality; MIMIC-IV database; Potassium fluctuation; Serum potassium; Ventricular arrhythmias.
© 2025. Springer-Verlag GmbH Germany, part of Springer Nature.