Depressed sympathovagal modulation indicates sepsis in patients with suspected infection

This study explored whether sympathovagal modulation assessed through frequency domains of heart rate variability (HRV) can indicate sepsis in patients with suspected infection. In total, 370 consecutive adult patients with suspected infection admitted to the emergency department were enrolled in this single-center cohort study. A continuous 10-minute electrocardiography for HRV analysis was recorded immediately for these patients after inclusion. Patients were stratified into non-sepsis and sepsis groups based on a sepsis-related organ failure assessment score of ≥2 that met the Third International Consensus Definitions for Sepsis. Seven frequency domains of HRV were compared between these 2 groups. Compared with the non-sepsis group (n = 98), the sepsis group (n = 272) had a significantly lower incidence of respiratory tract infection, higher total power, higher very-low-frequency component, higher high-frequency (HF) component, higher normalized HF component, lower normalized low-frequency (LF) component, and lower LF component/HF component ratio (LF/HF). Multiple logistic regression model identified HF component (odds ratio [OR] = 0.994; 95% confidence interval [CI], 0.990–0.999) and LF/HF (OR = 0.494; 95% CI, 0.423–0.578) as significant variables associated with sepsis. The area under receiver operating characteristic curves of HF component and LF/HF was 0.741 (95% CI, 0.685–0.797) and 0.930 (95% CI, 0.900–0.960), respectively, in identifying sepsis in patients with suspected infection. Tilted sympathovagal balance toward increased vagal activity and depressed sympathetic modulation, assessed by the HF component and LF/HF, may indicate sepsis in patients with suspected infection.