LAUSR

Abstract Objective  A ventricle-to-brain index (VBI) >0.35 is associated with low scores on the Bayley Scales of Infant and Toddler Development, Third Edition (BSID-III) in preterm infants with birth weight <1,250 g. However, VBI obtained at the third ventricle has only moderate interobserver reliability. The objective of this study was to test (1) reliability of VBI measured at the foramen of Monro on the latest ultrasonogram (US) before discharge using the intraclass correlation coefficient (ICC) and (2) the relationship between VBI and BSID-III scores at ≥18 months corrected age. Study Design  The present study is a single-center retrospective cohort study. Results  The study included 270 preterm infants born at 23 0/7 to 28 6/7 weeks of gestational age. The ICC of VBI between independent measurements by two study radiologists on the first 50 patients was 0.934. Factors associated with the value of VBI included severe intraventricular hemorrhage, bronchopulmonary dysplasia, and systemic steroid administration for BPD but not postmenstrual age. In multivariate analysis, VBI was negatively and independently associated with cognitive ( p  = 0.002), language ( p  = 0.004), and motor ( p  < 0.001) BSID-III scores. The association between VBI and BSID-III scores was observed even in infants in whom the latest US was obtained before term equivalent age. The association between VBI and BSID-III scores was also observed after excluding those with severe intraventricular hemorrhage. Conclusion  In this very preterm cohort the measurement of VBI had excellent reliability. Moreover, VBI measurements were negatively associated with motor, language, and cognitive BSID-III scores. Key Points Mean values of VBI are stable with postmenstrual age. Values at the foramen of Monro are reliable and reproducible. VBI is negatively associated with Bayley scores. The association is observed even before term age.