Prediction of Intraventricular Hemorrhage Using Echocardiography and Near Infrared Spectroscopy

Moderate-severe intraventricular hemorrhage (msIVH, Grades II-IV) is a significant neurological complication among extremely low gestational age neonates (ELGANs, <=27+6 weeks) and is associated with long-term neuro-disabilities. In Canada, msIVH affects ~25-30% of the 1300 ELGANs born annually, with little change in incidence over last decade. Typically, it occurs between days 2-7 of age, providing a finite window of opportunity. Instituting therapies at the population level, however, exposes many low-risk infants to side effects, adversely affecting risk-benefit profile and requiring large sample sizes in trials. A targeted preventative approach, though ideal, is currently challenged by our inability to reliably identify at-risk ELGANs early after birth. Near-infrared spectroscopy (NIRS) has emerged as a promising non-invasive bedside neuromonitoring tool. Pilot studies using NIRS, including ours, found lower cerebral saturations (CrSO2) and greater periods of altered cerebral autoregulation in infants who later developed msIVH. However, a systematic planned investigation is needed to establish the predictive characteristics of NIRS-derived markers, using clinically translatable methods (cumulative burden over time-period vs. single time-point values) and identify their relative performance at different time-points during transition. Further, incorporating echocardiographic (ECHO) hemodynamic markers, known to be associated with msIVH, may allow for the establishment of robust multi-model prediction models and the gain of mechanistic hemodynamic insights to inform future management. Hence, our objective is to investigate the utility of multi-modal assessment using NIRS and ECHO for early identification of ELGANs at risk of msIVH, and generate clinically applicable predictive model(s).