Do differences in diagnostic criteria for late fetal growth restriction matter?

Background: Criteria for diagnosis of fetal growth restriction (FGR) differ widely according to national and international guidelines and further heterogeneity arises from the use of different biometric and Doppler reference charts, making the diagnosis of FGR highly variable. Objective(s): To compare FGR definitions by Delphi consensus criteria and Society for Maternal Fetal Medicine (SMFM) definition, using different standards/charts for fetal biometry and different reference ranges for Doppler velocimetry parameters. Study design: From the TRUFFLE 2 feasibility study (856 women with singleton pregnancy 32+0 to 36+6 weeks of gestation at risk of FGR) we selected 564 women with available mid-pregnancy biometry. For the comparison we used standards/charts for estimated fetal weight (EFW) and abdominal circumference (AC) from Hadlock, Intergrowth, GROW and Chitty. Percentiles for umbilical artery pulsatility index (PI) and its ratios with middle cerebral artery PI were calculated using Arduini and Ebbing reference charts. Sensitivity and specificity for low birthweight and adverse perinatal outcome were evaluated. Results: Different combinations of definitions and reference charts identified substantially different proportions of fetuses within our population as FGR varying from 38% (with Delphi consensus definition, Intergrowth biometric standards and Arduini Doppler reference ranges) to 93% (with SMFM definition and Hadlock biometric standards). None of the different combinations tested appeared effective, with relative risk for birthweight below the 10th percentile between 1.4 to 2.1. Birthweight below the 10th percentile was observed most frequently when selection was made with the GROW/Chitty charts, slightly less using the Hadlock standard, and lowest with the Intergrowth standard. Using the Ebbing Doppler reference ranges resulted in a far higher proportion identified as FGR than with the Arduini Doppler reference ranges, while Delphi consensus definition with Ebbing Doppler reference ranges gave similar results to the SMFM definition. Delphi consensus definition using Arduini Doppler reference ranges were significantly associated with adverse perinatal outcome, with any biometric standards/charts. SMFM definition could not accurately detect adverse perinatal outcome irrespective of EFW standard/chart used. Conclusion(s): Different combinations of FGR definitions, biometry standards/charts and Doppler reference ranges select different proportions of fetuses categorized as FGR. The difference in adverse perinatal outcome may be modest, however these differences will have a significant impact in terms of rate of intervention.