ABSTRACT
One hundred twenty-one patients underwent an ultrasound examination within 48 hours
of delivery to assess prospectively the reliability of the diagnosis of intrauterine
growth retardation. Sonographic parameters examined included the abdominal circumference,
sonographic estimate of fetal weight, the head to abdominal circumference ratio, and
the femur length to abdominal circumference ratio. The best obstetric estimate of
gestational age was used. The diagnosis of growth retardation was based on the postnatal
ponderal index, and or the birthweight and crown-heel length percentiles. Seventeen
infants were growth retarded. Fifteen infants had a birthweight less than the 10th
percentile, but only nine (60%) were either asymmetrically growth retarded (by their
ponderal index) or symmetrically growth retarded (by virtue of a birthweight and length
less than the 10th percentile). All sonographic parameters were better able to predict
a birthweight below the 10th percentile for gestational age than growth retardation.
An abdominal circumference less than the 2.5 percentile for gestational age had the
highest sensitivity for growth retardation (88.0%) of the parameters studied. Only
the abdominal circumference centile identified all infants with either symmetric growth
retardation or asymmetric growth retardation associated with a birthweight below the
10th percentile. A sonographic estimate of fetal weight below the 10th percentile
had the highest positive predictive value for growth retardation-38%. In contrast
to the overall poor positive predictive values, the negative predictive values for
all parameters studied exceeded 90%. Combining the abdominal circumference percentile
with one of the three remaining techniques did not significantly improve diagnostic
accuracy. We conclude that the abdominal circumference percentile is the best screening
parameter for growth retardation, but that none of the studied parameters are specific
enough to rely on for the diagnosis of mild degrees of growth retardation.
