Impact of Neuroimaging Modalities and Grading Systems on Prognostic Accuracy for Neurodevelopmental Outcomes in Preterm Infants with Intraventricular Hemorrhage

Background Intraventricular hemorrhage (IVH) is a frequent complication following preterm birth and is linked to adverse neurodevelopmental outcomes (NDO). Objective This study assesses the impact of neuroimaging modalities and grading systems on the prognostic accuracy of NDO. Materials and Methods We conducted a retrospective single-center observational cohort study involving 507 consecutively born very preterm infants (2009–2018). We analyzed clinical parameters, neuroimaging modalities (cerebral MRI (cMRI)/cranial ultrasound (cUS)), and grading systems (Papile/Volpe) in infants with and without IVH regarding NDO (Bayley Scales of Infant Development) using univariable and multivariable regression analyses. Results IVH was detected in 85 of 507 infants (GA: 26.49 ± 2.57 weeks, male: 48.2%), occurring more frequently in less mature infants and following vaginal delivery (p = 0.011). Significant clinical risk factors for IVH included catecholamine treatment, low Apgar score, delivery mode, and bronchopulmonary dysplasia (all p < 0.032); non-invasive ventilation also contributed to mild IVH. IVH was associated with worse cognitive (CO, p = 0.004) and motor outcomes (MO, p = 0.001), and worse outcomes in IVH II° vs. IVH I° (CO: p = 0.035; MO: p = 0.016). CMRI identified IVH more frequently than cUS, revealing significant classification differences (p = 0.025). In the presence of intraventricular blood, prognostic accuracy differed significantly between the Papile and Volpe classification, with different NDO (CO: 94.72 ± 1.11 vs 80.82 ± 6.38, p = 0.047; MO: 89.91 ± 1.12 vs 73.53 ± 5.89, p = 0.015). Conclusion The choice of neuroimaging modality and grading system impacts the accuracy for neurodevelopmental outcome prediction in preterm infants with IVH. Our results underscore the importance of detecting intraventricular blood for outcome assessment due to its association with adverse NDO.