Abstract
Background
Routine third-trimester ultrasound screening for suspected large-for-gestational-age (LGA) and macrosomic fetuses is increasingly used to guide obstetric management and reduce delivery-related morbidity. However, ultrasound estimation of fetal weight at term has limited diagnostic accuracy, and evidence supporting interventions to reduce morbidity in large fetuses remains limited. Together, these uncertainties raise questions regarding the benefit of routine ultrasound screening for fetal largeness at term.
Objective
To evaluate the diagnostic accuracy of routine third-trimester ultrasound for detecting large-for-gestational-age and macrosomic neonates at term and the impact of inaccurate prenatal diagnosis on obstetric management and perinatal outcomes.
Study Design
A retrospective cohort study of 21,743 singleton pregnancies undergoing routine 35 to 37 weeks of ultrasound at a tertiary referral center. Pregnancies with major fetal anomalies, genetic abnormalities, or missing outcome data were excluded. Estimated fetal weight ≥90th centile defined screen-positive status; birthweight ≥90th centile and >4000 g defined large-for-gestational-age and macrosomia at birth, respectively. Study outcomes included mode of delivery and composite adverse perinatal outcomes comprising maternal and neonatal outcomes. Composite adverse maternal outcomes included postpartum hemorrhage, high dependency or intensive care unit admission, obstetric anal sphincter injury, and intrapartum cesarean delivery, while composite adverse neonatal outcomes included neonatal intensive care unit admission >48 hours, continuous positive airway pressure support, mechanical ventilation, hypoxic-ischemic encephalopathy, intraventricular hemorrhage, respiratory distress syndrome, seizures, hypoglycemia, transient tachypnea of the newborn, meconium aspiration syndrome, and death. Adjusted relative risks with 95% confidence intervals were estimated using modified Poisson regression, adjusting for maternal and pregnancy characteristics. Diagnostic performance measures, including sensitivity, specificity, positive predictive value, negative predictive value, positive and negative likelihood ratios, and 95% confidence intervals, were calculated. The interval between the index test and reference standard was the time from ultrasound to delivery. Clinicians were aware of ultrasound findings, which informed counseling and management. Results are reported as adjusted relative risks with 95% confidence intervals.
Results
Ultrasound screening showed a sensitivity of 34.9% (33.2–36.6) for large-for-gestational-age and 35.6% (33.5–37.9) for macrosomia, with specificity of 97.4% (97.2–97.6) and 95.6% (95.3–95.9), respectively. Screen-positive pregnancies were significantly less likely to attempt labor (adjusted relative risk, 0.87; 95% confidence interval, 0.84–0.90; P<.001) and more likely to undergo intrapartum cesarean delivery (adjusted relative risk, 1.47; 95% confidence interval, 1.30–1.67; P<.001), with higher risks of composite adverse maternal (adjusted relative risk, 1.43; 95% confidence interval, 1.32–1.55; P<.001) and neonatal (adjusted relative risk, 2.37; 95% confidence interval, 1.85–3.05; P<.001) outcomes. Compared to true-positive cases, false-negative cases were associated with significantly lower rates of induction of labor (adjusted relative risk, 0.74; 95% confidence interval, 0.68–0.81; P<.001), intrapartum cesarean delivery (adjusted relative risk, 0.70; 95% confidence interval, 0.59–0.83; P<.001), and composite adverse maternal outcomes (adjusted relative risk, 0.87; 95% confidence interval, 0.77–0.98; P<.032), with no increased risk of adverse neonatal outcomes (adjusted relative risk, 0.77; 95% confidence interval, 0.56–1.06; P=.12). Compared to true-negative cases, false-positive cases experienced higher rates of operative vaginal delivery (adjusted relative risk, 1.29; 95% confidence interval, 1.07–1.57; P<.001), intrapartum cesarean delivery (adjusted relative risk, 1.25; 95% confidence interval, 1.0–1.55; P=.045), and composite adverse maternal outcomes (adjusted relative risk, 1.28; 95% confidence interval, 1.11–1.48; P<.001) with no difference in the risk of adverse neonatal outcomes (adjusted relative risk, 1.35; 95% confidence interval, 0.72–2.53; P=.347)
Conclusion
Routine third-trimester ultrasound screening for suspected large-for-gestational-age and macrosomia demonstrates limited diagnostic performance and appears to generate a labeling effect, whereby obstetric intervention and some adverse outcomes may be influenced more by suspected large-for-gestational-age status than by fetal size itself. These findings raise questions about the benefit of universal large-for-gestational-age screening at term and support consideration of more individualized, risk-based approaches to late-pregnancy assessment.