Abstract
BACKGROUND: Heart rate variability (HRV) is a parameter of autonomic nervous system function. A decrease of HRV has been associated with disease severity, risk of complications and prognosis in several conditions. OBJECTIVE: We aim to investigate the feasibility and the reliability of in-ambulance HRV registration during emergency interventions, and to evaluate the association between prehospital HRV parameters, patient characteristics, vital parameters and short-term outcome. METHODS: We conducted a prospective study using a non-invasive 2-lead ECG registration device in 55 patients transported by the paramedic intervention team of the Universitair Ziekenhuis Brussel. HRV assessment included time domain parameters, frequency domain parameters, nonlinear analysis, and time-frequency analysis. The correlation between HRV parameters and patient and outcome characteristics was analyzed and compared to controls. RESULTS: Artifact and ectopic detection rates were higher in patients during ambulance transportation compared to controls in resting conditions, yet technical reasons precluding in-ambulance HRV analysis occurred in only 9.6% of cases. HRV acquisition was possible without safety issues or interference with routine emergency care. Reliability of the results was considered sufficient for Sample entropy (SampEn), good for the ratio of low frequency and high frequency components (LF/HF ratio) in the frequency and the time frequency domain, and excellent for the triangular interpolation of the NN interval histogram (TINN), and for the short-term scaling exponent of the detrended fluctuation analysis (DFA α1). HRV indices were significantly reduced inpatients with unfavorable outcome compared to patients with favorable outcome and controls. Multivariate analysis identified lower DFA α1 as an independent predictor of unfavorable outcome (OR, 0.155; 95% CI 0.024-0.966; p = 0.049). CONCLUSION: In-ambulance HRV registration is technically and operationally feasible and produces reliable results for parameters in the time, frequency, nonlinear and time frequency domain. Especially non-linear HRV analysis during emergency ambulance transportation may be a promising approach to predict patient outcome.