Persistent normality on serial cranial ultrasound as a marker of early developmental trajectory in preterm infants: a single-centre retrospective cohort study

The first years of life are critical for defining developmental trajectories after preterm birth and planning early support. Serial cranial ultrasound (CUS) is routinely used for neonatal brain surveillance, but the reassurance provided by a normal early scan, and by persistent normality thereafter, is not well quantified. We assessed whether persistent normality on serial CUS identifies a low-risk early developmental trajectory. Single-centre retrospective cohort of live-born infants < 32 weeks’ gestation and/or birth weight ≤ 1500 g from a prospectively maintained neonatal and neurodevelopmental follow-up registry (1991–2020). CUS was performed in four predefined windows, recording intraventricular haemorrhage, ventriculomegaly and parenchymal lesion pattern at each scan. Gestational-age (GA) and epoch-adjusted odds ratios (aORs) were estimated for selected developmental outcomes. Of 3081 infants, 2841 had complete first-scan data; 1881/2841 (66.2%) had a normal first CUS and 1833/1881 (97.4%) had subsequent CUS follow-up. Among these, 1296/1833 (70.7%) maintained a completely normal serial CUS trajectory, increasing to 1248/1544 (80.8%) after two consecutive normal scans and 1199/1308 (91.7%) after three. This probability was GA-dependent, ranging from 42.3% at 24–25 weeks to 82.2% at ≥ 30 weeks after a normal first CUS. Persistent normality was associated with lower frequencies of cerebral palsy (CP) at 2 years (0.7% vs. 11.9%; aOR 0.07), late or absent independent walking (4.0% vs. 8.6%; aOR 0.50), sustained early-childhood developmental/educational support (10.0% vs. 21.9%; aOR 0.44) and cognitive sequelae around the preschool-to-school transition (6.8% vs. 15.2%; aOR 0.48). In the broader cohort, infants with later abnormality after a normal first CUS had outcome frequencies close to those with abnormal first CUS. Persistent normality did not clearly discriminate documented or suspected autism spectrum disorder, preschool hyperactive/inattentive behavioural concerns, learning difficulties after adjustment, or sensory impairment. Persistent normality across the serial CUS pathway identifies a low-risk, but not risk-free, early developmental trajectory and may guide risk-stratified follow-up and support planning. It supports reassurance regarding major motor outcome, sustained early-childhood support needs and cognitive sequelae, while reinforcing balanced counselling: normal ultrasound findings should not be used to exclude later behavioural, social-communication, language, educational or broader neurodevelopmental vulnerabilities. Ongoing developmental surveillance remains necessary.