Does the individualized reference outperform a simple ultrasound-based reference applied to birth weight in predicting child neurodevelopment?

OBJECTIVES

Being small-for-gestational age (SGA) is associated with an increased risk of morbidity, but questions remain about how best to diagnose SGA, and thus, predict poor health consequences. The authors sought to compare an individualized reference for defining SGA with simple birth weight-based and ultrasound-based references applied to birth weight in predicting poor cognitive development at age five.

METHODS

The authors used data from the Successive SGA Births Study, a prospective study including 699 Alabaman and 618 Scandinavian women recruited from 1986 to 1988, and whose children had cognitive development scores measured at age five using the Wechsler Preschool and Primary Scale of Intelligence-Revised Intelligence Quotient. Sensitivity, specificity and positive predictive value (PPV) were estimated for each reference applied to birth weight using adverse cognitive development (score < 10(th) percentile) as the outcome. Relative risk of poor neurodevelopment was calculated, comparing infants classified as SGA by either the individualized or the simple ultrasound-based reference with infants not classified as SGA.

RESULTS

The individualized reference had higher specificity and PPV in predicting poor neurodevelopment. Neonates defined as SGA by the individualized reference alone had a higher risk (RR=2.20, 95% CI: 1.20, 4.00) of poor cognitive outcome, while those identified by the ultrasound-based reference alone did not (RR=0.95, 95% CI: 0.45, 2.01). None of the references could predict poor neurodevelopment well at age five.

CONCLUSIONS

The individualized birth weight reference modestly outperforms the simple ultrasound-based reference in identifying SGA infants with poor child neurodevelopment. However, neither reference can predict child neurodevelopment well.