An evidence-based approach to predicting low IQ in very preterm infants from the neurological examination: outcome data from the indomethacin Indomethacin Intraventricular Hemorrhage Prevention Trial.

We evaluated whether the degree of cerebral palsy (CP) at age 3 in very preterm children is predictive of full-scale intelligence quotient (FSIQ) <70 at age 8 by calculating likelihood ratios (LRs) for findings on the neurologic examination. Data from the follow-up phase of the Indomethacin Intraventricular Hemorrhage Prevention Trial, which includes periodic neurologic examination and neuropsychometric testing, were used. Information was available on 366 of 440 (83%) children with birth weight of 600 to 1250 g who survived. Neurologic examination at age 3 was grouped by presence and type of CP, and the Weschler Intelligence Scale for Children-Third Edition FSIQ at age 8 was grouped dichotomously (<70 or > or =70). CP was identified in 35 of 366 3-year-olds (9.5%). An FSIQ <70 was identified in 47 of 366 children at 8 years old (12.8%). FSIQ <70 occurred in 14 of 17 children with tri- or quadriplegia (82%), 8 of 18 children with di- or hemiplegia (44%), and 25 of 331 children without CP (7.5%). Useful LRs were calculated for tri- or quadriplegia (30), di- or hemiplegia (5.7), and children without CP (0.55). These LRs have greater impact on posttest odds for FSIQ <70 than those for birth weight <1000 g, history of bronchopulmonary dysplasia, and Stanford-Binet Intelligence Score <70 at age 3. We conclude that the neurologic examination at 3 years old predicts FSIQ <70 at age 8 with LRs that allow evidence-based parental counseling and intervention planning.