Predicting neurodevelopmental outcomes in Australian First Nations infants: The transdiagnostic utility of early screening tools.

Aim: To determine the predictive relationship between evidence-based screening tools and neurodevelopmental outcomes in Australian First Nations infants.

Method: This prospective cohort study invited First Nations families to participate in a culturally adapted early developmental screening programme. A total of 156 infants (55.1% male, mean gestational age = 33.6 weeks, SD = 4.6) were screened using the Prechtl's General Movements Assessment, with optimality scoring using the Motor Optimality Score-Revised (MOS-R) at 3 to 5 months and the Hammersmith Infant Neurological Examination (HINE) at 4 to 9 months. Participants completed 'baby movement (BM) checks' at two time points (BM1, 3-5 months corrected age; BM2, 4-9 months corrected age), with final movement and learning checks at 12 months corrected age. At 12 months corrected age, standardized motor, cognitive, and communication assessments, neurodisability-specific symptomology, or a diagnosis made by a paediatrician classified infants as developing typically ('on track') or (1) with a high chance of or confirmed cerebral palsy (CP) or (2) non-CP neurodevelopmental delay (NDD), including autism and fetal alcohol spectrum disorder (FASD). Predictive relationships were investigated using logistic regression and diagnostic statistics.

Results: At 12 months, 127 of 147 (86%) eligible infants (n = 9 withdrawn or deceased) were classified as 'on track' (n = 55, 43%), NDD (n = 59, 47%), or CP (n = 13, 10%). MOS-R (≥ 14 weeks). The HINE distinguished infants as 'on track', CP, or NDD. Each 1-point decrease on both tools increased the odds of NDD (OR_MOS-R = 1.40, 95% confidence interval [CI] = 1.00-1.96; OR_HINE = 1.12, 95% CI = 1.05-1.21) and CP (OR_MOS-R = 1.47, 95% CI = 1.08-2.01; OR_HINE = 1.41, 95% CI = 1.21-1.65,). The MOS-R (cut-off of less than 23) and HINE (moderate to severely reduced) were best for identifying any NDD and CP (MOS-R: sensitivity = 84%, specificity = 38%; HINE: sensitivity = 64%, specificity = 63%). Combined trajectories across both tools were the strongest predictors of CP (sensitivity = 73%, specificity = 96%), autism (sensitivity = 59%, specificity = 95%), and FASD (sensitivity = 89%, specificity = 93%).

Interpretation: Evidence-based screening tools demonstrate promising transdiagnostic prediction of 'on-track' development and not only high chance of CP but also autism, FASD, and other NDDs.