Variability of cardiac frequency as an early diagnostic method for sepsis in neonates with risk factors

Abstract

Descriptive observational transversal research. Method: 100 newborns (NB) term and late pre-term at risk for early sepsis and other non-infectious pathologies were monitored to measure the variability of heart rate with a Polar V800 equipment, recording at least 1,000 consecutive r-r intervals, during the first 24 hours of life. Descriptive statistical measures, Mann-Whitney, Fisher, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), reason of likelihood (RL) and ROC (receiver operating characteristic) curves were used for statistical analysis. We found that the only statistically significant parameter with p< 0.05 (95% CI 0.043), which differed amongst the NB with and without risk of sepsis, was the average of the heart rate (HR) and it was the only parameter of heart rate variability (HRV) which had higher sensitivity (64%) and specificity (37%), the number of r-r intervals (m r) were inversely proportional to the average of the HR. Conclusions: The HRV by itself is not a reliable parameter for the diagnosis of early sepsis in RN with premature rupture of membranes (PPROM) over 18 hours, since it has a low sensitivity and specificity, therefore, should be evaluated together with the background; as well as clinical and laboratory tests.