Third-Trimester Fetal Biometry and Neonatal Outcomes in Term and Preterm Deliveries.

OBJECTIVES

To determine whether specific biometric thresholds for head circumference, abdominal circumference, femur length, and estimated fetal weight can identify neonates at risk for adverse outcomes.

METHODS

We conducted a retrospective analysis of women with sonographic biometry after 26 weeks' gestational age (GA) followed by delivery of term and preterm neonates from 2007 through 2011. The head circumference, abdominal circumference, femur length, and estimated fetal weight were obtained. Sonographic data were merged with birth certificate and neonatal data. Biometry and estimated fetal weight were divided into percentile thresholds: 10th and above (reference), below 10th, below 5th, and below 3rd. Neonatal outcomes included neonatal intensive care unit admission, 5-minute Apgar score less than 7, and a composite of any morbidity/mortality (hypoxic-ischemic encephalopathy, periventricular leukomalacia, necrotizing enterocolitis, sepsis, renal failure, or death). Logistic regression yielded odds ratios and 95% confidence intervals for biometry and outcome, then adjusted for GA at delivery.

RESULTS

A total of 2237 patients delivered at term, and 455 delivered before term. Neonatal intensive care unit admission was not associated with any biometric threshold in the term and preterm groups. Five-minute Apgar score less than 7 was associated with head circumference below 10th, abdominal circumference below 3rd, and estimated fetal weight below 5th percentiles in the term group and head circumference below 10th, abdominal circumference below 10th, and femur length below 10th percentiles in the preterm group (P < .05). Composite morbidity/mortality was associated with abdominal circumference below 5th, femur length below 10th, and femur length below 3rd percentiles in the term group and head circumference below 5th, abdominal circumference below 10th, and femur length below 5th percentiles in the preterm group (P< .05). Adjustment for GA did not affect outcomes for term deliveries but did affect nearly all outcomes for preterm deliveries.

CONCLUSIONS

Irrespective of GA, no one biometric threshold can accurately predict adverse neonatal outcomes.