Using Z-scores to compare biometry data obtained during prenatal ultrasound screening by midwives and physicians

To compare retrospectively the distribution of foetal biometry data as measured by midwives and physicians during second and third trimester screening of an unselected population of pregnant women.

Standard measurements of biparietal diameter (BPD), head circumference (HC), abdominal circumference (AC), and femur length (FL) were performed by four midwives and ten physicians at 20 to 24 weeks of gestation and at 30 to 34 weeks of gestation as part of routine ultrasound examinations over a 26-month period (Jan. 2005-Mar. 2007). All measurements were converted into Z-scores using different prediction equations. The reference chart best fitting our practice was determined for each fetal parameter (French College of Sonographers for BPD, Chitty et al. for HC and FL, Snidjers and Nicolaides for AC). The means and SDs of the Z-score distributions for data collected by midwives and physicians were compared using Student's t-test for means and the Fisher-Snedecor test for SDs.

We retrieved 1566 and 1631 measurements made by midwives and physicians respectively between 20 and 24 weeks of gestation, and 1710 and 1578 measurements made by midwives and physicians respectively between 30 and 34 weeks of gestation. Mean values recorded by midwives were significantly closer to 0 (p < 0.05) for many foetal parameters. SD values were also significantly lower and were below 1.

In this study, midwives have a greater tendency than physicians to normalize biometry data. Such normalization may hamper the sensitivity of routine ultrasound screening for abnormal foetal growth.