Poor predictive validity of the Bayley Scales of Infant Development for cognitive function of extremely low birth weight children at school age

Objective. The Bayley Scales of Infant Development, Second Edition (BSID II) are commonly used to assess outcomes of extremely low birth weight (ELBW) infants. We sought to assess the predictive validity of the BSID II Mental Developmental Index (MDI) for cognitive function at school age. Design/Methods. Of 330 ELBW infants admitted in 1992 - 1995, 238 (72%) survived to the age of 8 years, of whom 200 (84%) were tested at both 20 months' corrected age ( CA) and 8 years. Mean birth weight was 811 g, mean gestational age was 26.4 weeks, 41% were boys, and 60% were black. Measures included the BSID II at 20 months' CA and the Kaufman Assessment Battery for Children (KABC) Mental Processing Composite (MPC) at 8 years' postnatal age. BSID II MDI and MPC scores were compared and the predictive validity calculated for all 200 ELBW children and for the 154 ELBW neurosensory-intact subgroup. Predictors of stability or change in cognitive scores were examined via logistic regression adjusting for gender and sociodemographic status. Results. For all ELBW children, the mean MDI was 75.6 +/- 16 versus a mean KABC of 87.8 +/- 19. For the neurosensory-intact subgroup, the mean MDI was 79.3 +/- 16 and the mean KABC was 92.3 +/- 15. Rates of cognitive impairment, defined as an MDI or KABC of < 70, dropped from 39% at 20 months' CA to 16% at 8 years for the total ELBW population and from 29% to 7% for the neurosensory-intact subgroup. The positive predictive value of having an MPC of < 70 given an MDI of < 70 was 0.37 (95% confidence interval [CI]: 0.27, 0.49) for all ELBW infants, 0.20 (95% CI: 0.10, 0.35) for the neurosensory-intact subgroup, and 0.61 (95% CI: 0.42, 0.77) for the neurosensory-impaired subgroup. The negative predictive values were 0.98, 0.99, and 0.85 for the 3 groups, respectively. Neurosensory impairment at 20 months' CA predicted lack of improvement of cognitive function (odds ratio: 6.9; 95% CI: 2.4, 20.2). Children whose cognitive scores improved between 20 months and 8 years had significantly better school performance than those whose scores stayed at < 70, but they did less well than those whose scores were persistently > 70. Conclusions. The predictive validity of a subnormal MDI for cognitive function at school age is poor but better for ELBW children who have neurosensory impairments. We are concerned that decisions to provide intensive care for ELBW infants in the delivery room might be biased by reported high rates of cognitive impairments based on the use and presumptive validity of the BSID II MDI.