Predicting cognitive outcomes in low‐risk infants at 3 months: A step too far?

The Hammersmith Infant Neurological Examination (HINE) is now used worldwide to assess the neurological status of infants both for diagnostic and prognostic purposes and it is a recommended tool for the early identification of infants at high risk of cerebral palsy. Given its widespread use a question is can it also help to identify infants at risk of cognitive delay. Brabant et al.1 addressed this issue using the HINE at 3.5 months corrected age in a moderately large cohort of low-risk infants born mid to late preterm, none of whom developed cerebral palsy. They excluded infants with major congenital or chromosomal anomalies, severe brain injury or serious illness, though the infants born late preterm that they included were only those admitted to the neonatal unit and may therefore have been at higher risk than those not admitted. The only cranial ultrasound scan finding reported was intraventricular haemorrhage, presumably lower grades, in 5%, though it is not clear if all infants were scanned. When comparing the total HINE score to Bayley-4 motor, cognitive, or language outcomes at 2 years using a cut-off of 1SD below the mean, or with the MacArthur–Bates Communicative Development Inventories, they did not find a relationship between HINE scores and outcomes. This is consistent with the 2021 findings of Romeo et al., who evaluated a large group of infants born preterm at 25 to 36 weeks of gestation. For infants with typical or mildly delayed outcomes (Bayley MDI scores 70–84, without cerebral palsy), their HINE scores at 3, 6, 9, and 12 months were similar to control infants.2 Brabant et al. found a median HINE score of 62/78, within the range expected for low-risk infants born preterm with favourable outcomes,3, 4 with no infant scoring below 50. It would have been interesting to know whether they would have had similar findings to those of Romeo et al. who found lower early HINE scores in infants with significant cognitive delay (Bayley cognitive score < 70).2 This distinction between severity of outcome is central to interpreting the present findings. In the Brabant et al. study only 3% of infants had Bayley-4 cognitive scores <70, three with very low scores but no comment on whether there were reasons which might aid understanding such exceptions. Further analysis of the distribution of asymmetries would have been informative, the authors arguing that as no more than four asymmetries were noted and no child developed cerebral palsy, this was not needed. More knowledge about patterns of asymmetry in children who do not develop cerebral palsy is of interest, and especially whether mild asymmetries persist or resolve on serial examinations. Interpretation should also consider that children with worrying findings at 3 months, or whose parents expressed developmental concerns, were referred for early intervention services as part of standard care. Such referrals may have improved the 2-year outcomes, although their potential effect cannot be determined from this study. The authors did not investigate the predictive value of individual HINE items, such as the quality of visual responses, pattern of central tone and movements with regard to cognitive or language function. This study supports earlier data and helps clarify the limitations of a single early HINE total score for predicting mild developmental difficulties in a low-risk preterm population. It is important to remember that the HINE is primarily a neurological examination, not designed to detect mild delays in cognition or language. Future studies could assess whether repeated HINE examinations, combined with developmental measures and parental concerns and even parental social, economic, and education status, are more informative than a single score at 3.5 months.3 An apparently reassuring total HINE score should not, on its own, determine whether an infant born preterm receives developmental follow-up. Not required.