The benefits of measuring heart rate variability in sepsis*

Abstract

Early diagnosis and treatment is an important key to improve the prognosis of sepsis. The fluctuations in heart rate with time named as heart rate variability (HRV) are modified by sepsis. HRV can be measured by a panel of mathematical methods using the usual descriptive statistics of time series, frequency domain analysis and complex methods issued from the chaos theory, the information theory, the fractal geometry or using different models. Most of these methods can currently be used for real-time monitoring at the bedside. Specific profiles of HRV have been documented in sepsis. Some changes of HRV have been identified early in the course of sepsis and some others have been correlated with the prognosis. This is well-documented in neonatal intensive care units where HRV monitoring seems to reduce the mortality after late onset sepsis in very low birth weight infants. In this review, we evaluate the value and the limits of HRV in the diagnosis and prognosis of sepsis. We think that HRV is a real-time, non-invasive, non-expensive, physiological marker that could help to draw new strategies to deal with patients presenting with high-risk of sepsis.