Objective: To determine the most appropriate ultrasonic measurement for the prediction of a small for gestational age (SGA) infant.
Data sources: A literature search of all English language journals over the last 15 years was undertaken. One hundred seventeen articles on the use of morphometric and Doppler ultrasonic measurements in the diagnosis of SGA were reviewed.
Methods of study selection: Studies were included if antenatal and postnatal criteria for diagnosis were clearly defined and data for SGA and normal fetuses were reported, allowing the construction of a 2 x 2 table.
Data extraction and synthesis: Studies with the same criteria were grouped according to whether the population was high or low risk. The sensitivity, odds ratio (OR), and false-positive rate were calculated for each study; summary statistics were then calculated for each ultrasonic measurement provided the individual sensitivities, ORs, and false-positive rates were not significantly different (P > .01).
Conclusions: In high-risk subjects, abdominal circumference below the tenth percentile had the highest common sensitivity, and estimated fetal weight below the tenth percentile had the highest common OR. In studies in which morphometric and Doppler ultrasonic measurements were compared in the same subjects, the Doppler ORs tended to be lower than the ORs for morphometric measurements. In low-risk subjects, much lower ORs were found for all ultrasonic measurements. The heterogeneous nature of the studies reviewed may have contributed to the different results within each group.