INTRODUCTION: Patients at end-stage renal disease (ESRD) undergoing hemodialysis (HD) are at high cardiovascular mortality risk due to potassium concentration ([K+]) changes out of normal ranges. The assessment of [K+] levels is limited as it can only be performed through blood tests in a hospital environment. Since [K+] levels are reflected on the T-wave in the electrocardiogram (ECG) signal, we hypothesized the T-wave morphology could be used to monitor [K+] changes in ESRD patients during HD. The aim of this study was to investigate whether dw, a time-warping-based ECG marker of T-wave morphology changes, with known association with ventricular repolarization dispersion, can monitor changes in [K+] levels. METHODS:We used 48-h long ECG recordings and [K+] measurements from a set of blood samples collected from 12 ESRD patients undergoing HD. An average T-wave (ATW) was obtained every hour of the ECG recordings and a reference T-wave was taken at the end of the HD session.The morphological changes between each ATW and the reference one were quantified with dw. The relative variations (Δ[K+]) in [K+] with respect to the reference values (end of treatment), during the HD session were computed. Spearman's correlation was computed between dw and Δ[K+]. RESULTS: The Figure shows the evolution of the distribution of Δ[K+] and dw during the HD session across all patients. The values of dw were significantly correlated with Δ[K+] (median [interquartile range] correlation coefficient of 0.95 [0.15]). CONCLUSIONS:T-wave morphology changes, quantified by dw, exhibit high correlation with [K+] changes. The results of this study support the use of dw to track changes in [K+] in ESRD patients undergoing HD.