Predicting intradialytic hypotension using heart rate variability

Abstract

This study aimed to identify whether a new method using heart rate variability (HRV) could predict intradialytic hypotension (IDH) for one month in advance for patients undergoing prevalent hemodialysis. A total 71 patients were enrolled, and baseline clinical characteristics and laboratory results were collected when HRV was measured, then, the frequency of IDH was collected during the observation period. HRV parameters included heart rate, R-R interval, the standard deviation of N-N interval, the square root of the mean squared differences of successive NN intervals, very low frequency, low frequency, high frequency, total power, and low frequency/high frequency ratio. During the one-month observation period, 28 patients experienced 85 cases of IDH (10.0% of a total 852 dialysis sessions). Among the clinical and laboratory parameters, ultrafiltration rate, prior history of diabetes, coronary artery disease, or congestive heart failure, age, intact parathyroid hormone level, and history of antihypertensive drug use were integrated into the multivariate model, referred to as a basic model, which showed significant ability to predict IDH (the area-under-curve [AUC], 0.726; p = 0.002). In HRV parameters, changes between the early and middle phases of hemodialysis (referred to Δ) were identified as significant independent variables. New models were built from the combination of Δ values with the basic model. Among them, a model with the highest AUC value (AUC, 804; p < 0.001) was compared to the basic model and demonstrated improved performance when HRV parameters were used (p = 0.049). Based on our results, it is possible that future IDH might be predicted more accurately using HRV.