Blue NR, et al. - The ability of the Hadlock standard to predict neonatal morbidity and small for gestational age at birth was compared with that of The National Institute of Child Health and Human Development (NICHD) race/ethnicity-specific standard. These two were also compared regarding their performance among the Native American population, which is not accounted for in the NICHD standard. Findings suggest that for the prediction of both neonatal morbidity and small for gestational age (SGA), the Hadlock standard is superior to the NICHD race/ethnicity-specific standard despite its publication more than 25 years ago.

Methods

• Researchers performed this retrospective study of diagnostic accuracy via reviewing deliveries at the University of New Mexico Hospital from January 1, 2013 to March 31, 2017.
• Mothers with singleton, well-dated pregnancies and non-anomalous fetuses with an estimated fetal weight (EFW) within 30 days of delivery were included.
• On the NICHD EFW-percentile tables, they performed cubic spline interpolation to calculate percentiles specific to the gestational day.
• They then calculated EFW percentiles using both the Hadlock and NICHD race/ethnicity-specific standards according to maternal self-identified race/ethnicity.
• To predict composite and severe composite neonatal morbidity and SGA at birth (birth weight <10thpercentile), the receiver-operator area under the curve (AUC) of each method was calculated.
• The mean ultrasound-birth weight percentile discrepancy was calculated as an additional measure of method accuracy.
• For Native Americans, using the Hadlock and NICHD race/ethnicity standards percentiles were calculated (white, black, Hispanic, Asian) and test characteristics for each were calculated to predict neonatal morbidity and SGA.

Results

• Inclusion of 1514 women was performed; the mean ultrasonography-to-delivery interval was 14.4 days (± 8.8) and the SGA rate was 13.6% (n=206).
• The Hadlock method displayed superior performance for the prediction of both composite and severe composite neonatal morbidity, with higher AUCs than the NICHD method (p < 0.001 for both); however, neither of the two displayed good discriminatory value (all AUCs < 0.8).
• The Hadlock standard had higher sensitivity (61.1%) for the prediction of SGA at birth than the NICHD standard, both when using the interpolated NICHD method (36.2%, p < 0.01) and the NICHD whole-week 10th percentile cutoff (46.7%, p <0.01).
• For predicting SGA, also the Hadlock method showed a higher AUC than the NICHD interpolated method (0.89 vs 0.88, p < 0.01).
• Ultrasound-birth weight percentile discrepancy was lower with the Hadlock method than the NICHD method (6.1 vs 16.5 percentile points, p < 0.01).
• Significantly higher composite morbidity was identified among fetuses classified as growth restricted by Hadlock but not NICHD than normally grown fetuses.
• The Hadlock method had the highest AUC to predict composite and severe composite morbidity among Native American Women; while, to predict SGA, the Hadlock and all NICHD race/ethnicity-specific methods performed comparably.