Neurological examination combined with brain MRI or cranial US improves prediction of neurological outcome in preterm infants.

BACKGROUND

The predictive value of the combination of neurological examination and brain magnetic resonance imaging (MRI) or cranial ultrasound (cUS) in preterm infants is not known.

AIMS

To study the prognostic value of the combination of neurological examination and brain MRI at term equivalent age (TEA) or serial neonatal cUS in very preterm infants for neurosensory outcome at 2 years of corrected age.

STUDY DESIGN

A prospective follow-up study.

SUBJECTS

A total of 216 very preterm infants (birth weight 1132 g [SD 331 g]) born in Turku University Hospital, from 2001 to 2006, were included.

OUTCOME MEASURES

The Dubowitz neurologic examination and brain MRI were done at TEA, and serial cUS examinations were performed until TEA. The Hammersmith Infant Neurological Examination (HINE) and neurosensory impairments (NSI) were assessed at 2 years of corrected age.

RESULTS

Of all infants, 163 (76%) had one or more deviant neurological items at TEA, and 32 (15%) had the HINE total score below the 10th percentile at 2 years of corrected age. A total of 17 (8%) infants had NSI. Neurological examination at TEA improved the negative and positive predictive values of brain MRI for NSI from 99% to 100%, and from 28% to 35%, respectively, and the negative and positive predictive values of cUS from 97% to 100%, and from 61% to 79%, respectively.

CONCLUSIONS

The combination of the Dubowitz neurologic examination and the brain MRI at TEA or serial neonatal cUS provides a valuable clinical tool for predicting long-term neurosensory outcome in preterm infants.