Cardiac variability in critically ill adults: Influence of sepsis

Objective: To evaluate, in critically ill adults, factors associated with i mpaired sympathovagal balance. Design: One-month inception cohort study. Setting: Twenty-six-bed medical intensive care unit of a teaching hospital. Patients: critically ill adults with an expected duration of intensive care unit stay of greater than or equal to 48 hrs were enrolled. Patients with permanent arrhythmia or cardiac pacing were not included. Interventions: None. Measurement and Main Results: Sympathovagal balance was assessed on the day after intensive care unit admission by the low-frequency/high-frequency ra tio obtained from spectral components of heart rate signal: overall variabi lity, low frequency, and high frequency. Results: Forty-one patients, 13 with sepsis and 28 without sepsis, were ass essed. Predictors of low-frequency/high-frequency ratio with the automatic interaction detection method were sepsis and age. Binary logit analysis adj usted for age showed that sepsis remained a strong and independent factor o f a low-frequency/high-frequency ratio of < 1.50, with an odds ratio of 3.6 3 (95% confidence interval, 1.47-9.01, p = .005). Use of mechanical ventila tion, catecholamines, or sedation did not add any information. The use of t he low-frequency/high-frequency ratio in diagnosing sepsis may be supported by a likelihood ratio for low frequency/high frequency < 1 at 6.47. Conclusions: This work suggests that impaired cardiac variability and notab ly sympathovagal balance (i.e., a low-frequency/high-frequency ratio < 1.0) may be a diagnostic test for sepsis.