EP08.42: Pregnancy dating by Crown–rump length: searching a reference table suitable for our population

Fetal Crown–rump length (CRL), measured in the first trimester by ultrasound, along with an accurate adequate reference table are essential for the establishment of gestational age. The objective of the study was to assess the impact of different CRL reference tables on the gestational age confirmation, in our population. Standard CRL measurement was performed by various sonographers, on 3861 patients in the first trimester, as part of routine ultrasound exam. We included life singleton pregnancies, with certain last menstrual period (LMP), regular menstrual cycles, CRL between 15 and 84mm and an agreement between LMP and CRL dating ≤5 days. Cases with fetal death, congenital abnormalities by ultrasound or at birth, were excluded from the study. We calculated retrospectively the Z-scores, and the difference in days between LMP and CRL dating, using four reference curves, published by Robinson et al. (1975), Hadlock et al. (1992), Verburg et al. (2008) and INTERGROWTH-21 consortium (2014). The Z-score and number of days difference distributions were compared with the expected normal distribution by calculation of the mean and SD. Based on the analysis of the Z-score distribution, the best fit for our population was the Verburg reference curve (mean -0.161, SD 0.651), followed by INTERGROWTH-21 (mean -0.183, SD 0.717) and Hadlock (mean -0.286, SD 0.688) curves. Based on the analysis of the number of days difference distribution, the best fit for our population was the INTERGROWTH-21 reference curve (mean -0.659, SD 2.49), followed by Robinson (mean -0.905, SD 2.5), Verburg (mean 1.033, SD 2.5) and Hadlock (mean -1.093, SD 2.5) curves. In our population, all reference curves overestimated the gestational age. The bias is small, with an average error of almost one day. Modern reference tables have outperformed the older ones. However, however, for a CRL ≥15 mm, the venerable tables still have a good performance.