Heart rate variability in the early resuscitation of septic shock

Introduction

The assessment of heart rate variability (HRV) has provided valuable insight in sepsis. Impaired HRV has been shown to be diagnostic of sepsis, heralding its onset, and prognostic of its impact, correlating with the development of sepsis-induced organ failure and death.

Objective

To analyze the ability of HRV to act as a prognostic aid in the early resuscitation of septic shock, and to study the effect that early fluid resuscitation will have on the direction and magnitude of subsequent measurements of HRV in patients with septic shock.

Methods

Subjects were prospectively identified within the emergency department of an urban-based tertiary-care medical center during their initial evaluation and treatment for septic shock, defined as a systemic infection with a systolic blood pressure <90 mmHg after intravenous fluid or a serum lactate >4.0 mmol/l. Continuous cardiac telemetry was obtained for the assessment of HRV using a standardized set of multiple parameters including variables of frequency and time domain analyses. Fluid administration was recorded during the initial resuscitation. A composite endpoint of increasing organ failure and in-hospital mortality was measured.

Results

Prospective analysis of 15 patients with septic shock was made. The in-hospital mortality rate was 67% (10/15). Non-survivors had a significantly impaired HRV compared with survivors when measured through multiple parameters (P < 0.01), as seen in Figure 1. There was no difference between fluid administration between survivors and nonsurvivors (4,475 vs. 5,220 ml, P = NS). There was no relationship seen between intravenous fluid administration and the change in 2-hour HRV seen in the early resuscitation of septic shock.

Figure 1

Heart rate variability parameters measured during the initial resuscitation of patients with septic shock. Nonsurvivors showed significantly impaired variability as measured by LF and the HF/LF ratio. Data presented as logarithmic transformation to control for scale and allow visual comparison. Differences between survivors and nonsurvivors were calculated using raw averages without logarithmic conversion. Time domain factors: SDNN = standard deviation of the normal R-R interval; RMSSD = root mean square of the difference of successive R-R intervals. Frequency domain factors of HRV spectrum: LF = low frequency (0.04 to 0.15 Hz); HF = high frequency (0.18 to 0.4 Hz).

Conclusion

In the early resuscitation of septic shock, HRV assessment can differentiate survivors from nonsurvivors, independent of organ severity measurement. While the use of fluid administration in the early resuscitation of shock did not correlate with changes in HRV measured at 2 hours, this relationship may exist when measured at longer time points allowing for the physiologic response to manifest in the HRV trend. The assessment of HRV in patients with infection can identify those at high risk for clinical deterioration and can potentially serve as an endpoint of resuscitation in patients with septic shock.