Sepsis is associated with impairment in autonomic regulatory function. This work investigates the application of heart rate and photoplethysmogram (PPG) waveform variability analysis in differentiating two categories of sepsis, namely systemic inflammatory response syndrome (SIRS) and severe sepsis. Electrocardiogram-derived heart period (RRi) and PPG waveforms, measured from fingertips (Fin-PPG) and earlobes (Ear-PPG), of Emergency Department sepsis patients (n = 28) with different disease severity, were analysed by spectral technique, and were compared to control subjects (n = 10) in supine and 80° head-up tilted positions. Analysis of covariance (ANCOVA) was applied to adjust for the confounding factor of age. Low-frequency (LF, 0.04-0.15 Hz), mid-frequency (MF, 0.09-0.15 Hz) and high-frequency (HF, 0.15-0.60 Hz) powers were computed. The normalised MF power in Ear-PPG (MFnu(Ear)) was significantly reduced in severe sepsis patients with hyperlactataemia (lactate > 2 mmol/l), compared to SIRS patients (P < 0.05). Moreover, in a group of normal controls, MFnu(Ear) was not altered by head-up tilting (P > 0.05), suggesting that there may be a link between 0.1 Hz ear blood flow oscillation and tissue metabolic changes in sepsis, in addition to autonomic factors. The study highlighted the value of PPG spectral analysis in the non-invasive assessment of peripheral vascular regulation in sepsis patients, with potential implications in monitoring the progression of sepsis.

Sepsis has been defined as the systemic response to infection in critically ill patients, with severe sepsis and septic shock representing increasingly severe stages of the same disease. Based on the non-invasive cardiovascular spectrum analysis, this paper presents a pilot study on the potential use of the nonlinear support vector machine (SVM) in the classification of the sepsis continuum into severe sepsis and systemic inflammatory response syndrome (SIRS) groups. 28 consecutive eligible patients attending the emergency department with presumptive diagnoses of sepsis syndrome have participated in this study. Through principal component analysis (PCA), the first three principal components were used to construct the SVM feature space. The SVM classifier with a fourth-order polynomial kernel was found to have a better overall performance compared with the other SVM classifiers, showing the following classification results: sensitivity = 94.44%, specificity = 62.50%, positive predictive value = 85.00%, negative predictive value = 83.33% and accuracy = 84.62%. Our classification results suggested that the combinatory use of cardiovascular spectrum analysis and the proposed SVM classification of autonomic neural activity is a potentially useful clinical tool to classify the sepsis continuum into two distinct pathological groups of varying sepsis severity.

A rabbit model of endotoxaemia was employed to study the regional changes in photoplethysmogram (PPG) waveform and its low frequency fluctuations, and how they are related to the physiological events during the time course of endotoxic shock. Endotoxin (1 mg kg(-1) lipopolysaccharide) was injected into eight anaesthetized and mechanically ventilated rabbits. The 90 min monitoring period was divided into six stages, with the onset of hypotension separating the first three (EDTX1-3) and last three (HYPO1-3) stages. The most significant finding was a substantial but transient elevation in sympathetic-related toe PPG variability (PPGV) spectral power in EDTX3 and HYPO1 (P < 0.01 versus EDTX1). The group average response showed that the rapid rise started 15 min before and peaked at the onset of hypotension, which indicated a surge in sympathetic vasomotor activity that preceded the decompensatory blood pressure fall. Ear skin vasoconstriction was evident by a marked and sustained fall in ear PPG amplitude along with a decrease in ear skin temperature at the onset of hypotension, during which the sympathetic-related ear PPGV spectral power was elevated (P < 0.01, HYPO1 versus EDTX1). The results demonstrate the value of PPGV in characterizing regional vascular control and provide important insights into the physiological mechanisms of endotoxic shock.

Sepsis is a potentially lethal condition, and is one of the major causes of death in non-coronary intensive care units. Sepsis syndrome progresses through a number of increasingly severe stages, from systemic inflammatory response syndrome (SIRS) through sepsis, severe sepsis and septic shock. Each stage of sepsis is potentially characterized by differing autonomic nervous system responses. Spectral analysis of cardiovascular variability has been regarded as a possible non-invasive method to study this autonomic regulation, and in this study, the variabilities of heart period (RRi) and pulse transit time (PTT) derived from electrocardiogram and photoplethys-mogram were investigated in three different groups: normal subjects (n = 11), SIRS (n = 7) and severe sepsis patients (n = 16), by computing spectral and cross-spectral measures in the low-frequency (LF) and the high-frequency (HF) ranges. SIRS and severe sepsis patients were found to have lower RRi (p < 0.01), augmented LF power in PTT (p < 0.01) and a lower RRi-PTT ratio (alpha(PTT)) in the LF and HF bands (p < 0.01) as compared with the normal subjects, which might indicate a suppression of baroreflex-mediated autonomic control of heart rate and an increased sympathetic influence on ventricular contractility in sepsis. The results have highlighted the potential value of spectral analysis of RRi and PTT variabilities as a non-invasive tool for clinical evaluation of cardiac autonomic regulation in sepsis patients.