Autonomic dysfunction characterized by Heart Rate Variability among patients with Sepsis-related Acute Respiratory Failure

RationaleTo explore the association and implications of using Heart rate variability (HRV) derived from continuous bedside monitoring as a surrogate for detection of Acute Respiratory Failure (ARF) in critically ill sepsis patients. ObjectiveTo analyze HRV measures derived from continuous physiological data captured before ARF-onset to determine whether statistically significant markers can be characterized when compared to sepsis controls. MethodsRetrospective HRV analysis of sepsis patients admitted to Emory Healthcare ICUs was performed between ARF and age and gender-matched controls. HRV measures such as time domain, frequency domain, nonlinear, and complexity measures were analyzed up to 1 hour before the onset of ARF, and a random event time in the sepsis-controls. Statistical significance was computed by the Wilcoxon Rank Sum test. ResultsA total of 89 intensive care unit (ICU) patients with sepsis were included in this retrospective cohort study. Time-domain HRV measures including pNN50 (the fraction of consecutive NN intervals that differ by more than 50 ms), RMSSD (root-mean-square differences of successive NN intervals), standard deviation, interquartile range, variance, and approximate entropy for Beat-to-Beat intervals strongly distinguished ARF patients from the controls group. HRV measures for nonlinear and frequency domains were significantly altered (p<0.05) among sepsis patients with ARF compared to controls. Frequency measures such as low frequency (LF), very low frequency (VLF), high frequency (HF), and SD1/SD2 ratio nonlinear measure (SD1:SD2) also showed a significant (p<0.05) increase in the ARF group patients. Multiscale entropy complexity was lower for ARF patients compared to the control counterparts. Detrended fluctuation analysis (DFA) showed a decreasing trend in ARF patients. ConclusionsHRV was significantly impaired across sepsis patients who developed ARF when compared to sepsis controls, indicating a potential prognostic utility for earlier identification of the need for mechanical ventilation and management of patients suspected with sepsis.