(1047) Multimodal Treatment Effect on Heart Rate Variability Among Vasoactive Medication Use in Sepsis

Introduction: Heart rate variability (HRV) as a surrogate marker, of autonomic dysregulation has been proposed as a useful early detection for patients with sepsis; however, to date, no investigations have rigorously evaluated how guideline-based medication treatments affect HRV physiomarkers. In this analysis of treatment effects, the objective was to characterize the complex dynamics between HRV physiomarkers and multi-modal vasopressor treatment in the context of disease severity and other sepsis treatments.

Methods: This retrospective cohort study included intensive care unit patients meeting Sepsis-3 criteria admitted to Emory University from January 2016 to December 2018. Two cohorts were compared (norepinephrine vs. no vasopressor) and were propensity score matched based on age, volume of fluid resuscitation, and Sequential Organ Failure Assessment (SOFA) score at 24 hours. The norepinephrine cohort was further divided and compared according to single adjunct and multiple vasopressors treatment. Continuous electrocardiograms (ECG) obtained from bedside monitors were used to derive HRV features in the time, frequency, and nonlinear domains. Treatment effects on HRV for single and combination vasopressors were evaluated by two-sample T test and multiscale entropy.

Results: A total of 881 patients were included (norepinephrine, n=303; no vasopressor, n=578). The average age was 55.2 years, and SOFA score was 3.8 and the average fluid volume administered in the first 24 hours was 8.91mL/kg for the norepinephrine cohort (437.6 mL) and 13.97ml/kg for the no-vasopressor cohort (1097 mL). The following HRV markers had significant differences (p < 0.001) between cohorts: beat to beat interquartile range, power in the low frequency range, ratio of low to high frequency power, sample entropy, acceleration capacity, deceleration capacity, SD2, ratio of SD1SD2 measures of Poincare plot and multiscale entropy for scale1 to scale19. A similar statistical significance was observed for the administration of multiple vasopressors.

Conclusions: The results of this investigation suggest that medications potentially complicate interpretation of HRV markers, and further evaluation of HRV in context of disease progression and ongoing treatments is necessary for bedside translation.