OBJECTIVES: To determine whether single or serial estimates of both the amniotic fluid index (AFI) or pulsatility index (PI) of the umbilical artery Doppler waveform can usefully identify infants with anthropometric features of intrauterine growth restriction (IUGR). METHODS: A total of 274 women underwent serial antenatal ultrasound examinations at predetermined intervals. Four biophysical parameters were considered: AFI and PI prior to delivery and change in AFI and PI over a 28-day period in the third trimester. All values were expressed as standard deviation scores. IUGR was diagnosed if the neonate demonstrated skinfold thickness < 10th centile, ponderal index < 25th centile or mid-arm circumference to occipitofrontal circumference (MAC:OFC) < -1 SD. Receiver operator characteristic curves were used to determine an optimal cut-off point, and test performance of the biophysical parameters were expressed as likelihood ratios (LR). RESULTS: The test performances of all four ultrasound parameters for any of the three diagnostic criteria for IUGR was poor. The highest positive LR was only 2.5 (95% CI 1.5-4.1) for PI prior to delivery in the prediction of ponderal index < 25th centile. CONCLUSION: Despite positive associations between single and serial estimates of AFI and PI with abnormal neonatal morphometry, the likelihood ratios remained low. The results of this study do not support the use of single or serial estimates of AFI or umbilical artery PI in the prediction of IUGR.
OBJECTIVES: To determine whether single or serial estimates of both the amniotic fluid index (AFI) or pulsatility index (PI) of the umbilical artery Doppler waveform can usefully identify infants with anthropometric features of intrauterine growth restriction (IUGR). METHODS: A total of 274 women underwent serial antenatal ultrasound examinations at predetermined intervals. Four biophysical parameters were considered: AFI and PI prior to delivery and change in AFI and PI over a 28-day period in the third trimester. All values were expressed as standard deviation scores. IUGR was diagnosed if the neonate demonstrated skinfold thickness < 10th centile, ponderal index < 25th centile or mid-arm circumference to occipitofrontal circumference (MAC:OFC) < -1 SD. Receiver operator characteristic curves were used to determine an optimal cut-off point, and test performance of the biophysical parameters were expressed as likelihood ratios (LR). RESULTS: The test performances of all four ultrasound parameters for any of the three diagnostic criteria for IUGR was poor. The highest positive LR was only 2.5 (95% CI 1.5-4.1) for PI prior to delivery in the prediction of ponderal index < 25th centile. CONCLUSION: Despite positive associations between single and serial estimates of AFI and PI with abnormal neonatal morphometry, the likelihood ratios remained low. The results of this study do not support the use of single or serial estimates of AFI or umbilical artery PI in the prediction of IUGR.