Beat-to-beat analysis of the relation between RT and RR intervals in newborns.

Objective: To evaluate the dynamic RT (QRS apex-end of T wave) rate dependence in newborns.

Study Design: A Digital Holter ECG was acquired on day 15 in nine full-term and eight preterm infants. Ten-minute periods were recorded during wakefulness and sleep. The accuracy of fit with RT-RR pairs was individually assessed by 14 regression formulas (r coefficient, Akaike score, residual analysis). The medians of RT and Bazett's RT correction were calculated for each 10 milliseconds of RR.

Results: The mean RR and RT were 429+/-51 and 263+/-18 milliseconds. None of the prediction formulas were sufficiently accurate to describe RT over the whole range of RR (r<0.56). The Bazett correction produced differences of more than 50 milliseconds at different RR. Prematurity, sleep state and heart rate variability did not influence RT-RR relation.

Conclusion: None of the parametric formulas were found to be accurate in describing RT rate dependence in newborns.