Predicting neurodevelopmental impairment in preterm infants by standardized neurological assessments at 6 and 12 months corrected age

Acta Paediatr. 2010 Apr;99(4):526-30. doi: 10.1111/j.1651-2227.2009.01649.x. Epub 2010 Jan 5.

Abstract

Aim: Neurodevelopmental impairment in very preterm infants can be reasonably diagnosed by 18-24 months corrected age, whereas the predictive value of earlier assessments is debated. We hypothesized that neurological findings at 6 and 12 months indicative of subsequent cerebral palsy predict 18-24 months' neurodevelopmental impairment.

Methods: Neurodevelopmental examinations (Griffiths scales) at 20 months of age in 561 preterm infants (birth weight <1 500 g) were compared with results of standardized neurological examinations (Early Motor Pattern Profile; EMPP) and Griffiths scales at 6 (n = 451) and 12 months (n = 496) corrected age.

Results: Griffiths developmental quotients at 20 months were weakly but significantly related to EMPP scores at 6 (R(s) = 0.328) and 12 months (R(s) = 0.493). Areas under receiver operator characteristic curves for the EMPP to predict neurodevelopmental impairment (Griffiths scores <or=75) at 20 months were 0.772 (0.890) at 6 (12) months, compared to 0.915 (0.962) for Griffiths scores. By contrast, EMPP and Griffiths scores had equal power to predict unability to walk unaided at 2 years of age (EMPP 6/12 months: 0.946/0.983; Griffiths 6/12 months: 0.935/0.985).

Conclusion: Neurological examinations with the EMPP at 6 and 12 months corrected age are of limited value to predict neurodevelopmental impairment at 20 months.

MeSH terms

  • Cerebral Palsy / diagnosis*
  • Developmental Disabilities / diagnosis*
  • Early Diagnosis
  • Follow-Up Studies
  • Humans
  • Infant
  • Infant, Newborn
  • Infant, Premature*
  • Infant, Very Low Birth Weight
  • Motor Activity / physiology
  • Motor Skills Disorders / diagnosis*
  • Neurologic Examination*
  • Predictive Value of Tests
  • ROC Curve