Objective

To compare the diagnostic effectiveness of selective versus universal ultrasonography as a screening test for large for gestational age (LGA) infants, and to determine whether previously described ultrasonic markers of excessive fetal growth could identify which suspected LGA fetuses were at increased risk of neonatal morbidity.

Methods

We analysed data from a prospective cohort study of nulliparous women, the Pregnancy Outcome Prediction study. All women had clinically indicated scans as per routine care. Additionally, all women had blinded ultrasonic estimated fetal weight (EFW) performed at around 36 weeks of gestational age (wkGA). Screen positive for LGA was defined as an EFW >90th percentile ≥34wkGA.

Results

The current analysis included 3,866 eligible women. Of these, 177 (5%) infants had a birth weight >90th percentile. 1,354 (35%) women had a clinically indicated ultrasonography ≥34wkGA. The sensitivity of selective ultrasonography was 27% and the sensitivity of universal ultrasonography was 38%. The specificity of both approaches was high (99% and 97%, respectively). Using universal ultrasonography, neonatal outcome differed (P for interaction) by abdominal circumference growth velocity (ACGV) for both any neonatal morbidity (P = 0.08) and severe adverse neonatal outcome (P = 0.03). LGA fetuses with increased ACGV had a relative risk (95% CI, P) of any neonatal morbidity of 2.0 (1.1-3.6, P = 0.04) and severe adverse neonatal outcome of 6.5 (2.0-21.1, P = 0.01), whereas LGA fetuses with normal ACGV were not at increased risk.

Conclusion

Screening using universal ultrasonographic fetal biometry increases the detection of LGA infants and combined with ACGV identifies infants at increased risk of adverse neonatal outcome.