Heart rate variability (HRV) increases several-fold in newborn infants who recover from severe cardiorespiratory failure. To test further the hypothesis that neonatal HRV is an index of illness severity, we measured both HRV and an illness severity score at 81 times in 45 infants in the NICU. Heart rate and HRV were measured at 32 beat intervals over a 2 hour period: HRV is represented by coefficient of variation (CV; S.D./mean) of a moving window of 2048 beats. Clinical illness severity was determined using the Score for Acute Neonatal Physiology and Neonatal Therapeutic Intervention Scoring System. The plots relate heart rate and HRV to the sum of the two clinical scores. The top plot shows no correlation of heart rate with illness severity (p=NS; Pearson product moment correlation). The bottom plot, on the other hand, shows strong inverse correlation of HRV and illness severity (r=0.7; p< 10-6); as the illness severity increased, HRV decreased. We conclude that HRV correlates well with the severity of neonatal illness, while heart rate itself does not.Figure

figure 1

Figure 1