Mean DBP, short-term scaling exponents from DBP, SBP, MAP predict IVH in preterm infants
Early prediction of intraventricular haemorrhage (IVH) in preterm infants may be facilitated by mean diastolic blood pressure (DBP) and short-term scaling exponent from beat-to-beat DBP, systolic BP (SBP) and mean arterial pressure (MAP), reports a new study.
The study involved physiological data collected from 27 infants within 1 to 3 hours after birth. Only those without severe congenital abnormalities, with birth weight <1,500 g and with gestational age <30 weeks were included.
Each infant underwent raw air flow waves and intra-arterial blood pressures measurement, and cranial ultrasound. From these, the beat-to-beat SBP, DBP MAP and pulse interval (PI) were derived and used to construct models to predict IVH.
Of the 27 infants, eight subsequently developed IVH. While there were mostly no significant differences in baseline characteristics between the groups, those that developed IVH had significantly lower detrended DBP (p=0.019) compared to those who did not.
Univariate logistic regression showed that mean DBP (area under the curve [AUC], 0.807; 95 percent CI, 0.62 to 0.99; p=0.022) and short-term fractal exponents from MAP (AUC, 0.779; 0.60 to 0.96; p=0.359), SBP (AUC, 0.771; 0.58 to 0.96; p=0.382) and DBP (AUC, 0.807; 0.64 to 0.97; p=0.278) were good predictors of IVH. Respiratory factors were not significant predictors of IVH.
Multivariable analysis showed that combinations of predictors yielded higher AUC values than single-variable models. Long-term scaling exponents derived from PI combined with mean DBP yielded the highest AUC values (AUC, 0.921; 0.82 to 1.02; p=0.035).
This was followed by mean DBP combined with short-term scaling exponents from MAP (AUC, 0,871; 0.74 to 1.01; p=0.027), mean MAP and long-term scaling exponents from PI (AUC, 0.864; 0.72 to 1.02; p=0.068), and mean DBP and short-term scaling exponents from DBP (AUC, 0.864; 0.72 to 1.00; p=0.022).