Accuracy of fetal weight estimation by ultrasound in periviable deliveries.

OBJECTIVE

To evaluate the accuracy of ultrasound estimated fetal weight (EFW) near viability, and to determine the adequacy of use of EFW in place of birth weight (BWT) for predicting prognosis for infants born near the limit of viability.

METHODS

Retrospective chart review of women delivering between 22(0/7) and 25(6/7) weeks gestation (GA) with ultrasound performed within 7 days of delivery. Potentially relevant clinical factors were evaluated regarding their impact on accuracy of EFW. Estimated survival based on BWT and EFW, using an National Institute for Child Health and Human Development (NICHD) algorithm, were compared.

RESULTS

Study included 93 infants. Mean absolute percent difference (accuracy) of EFW for BWT was 9.4% (95%CI 7.4-11.3). There was no correlation between EFW accuracy and BWT, GA, maternal age, or BMI. There was a 3% overestimation of BWT per 100 g decrease in BWT (p = 0.001). Race, oligohydramnios, parity, smoking, or previous cesarean did not impact EFW accuracy. Mean predicted survival by the NICHD algorithm was 43.1% using BWT; 43.6% using EFW (p = 0.63). An overestimation of predicted survival (using EFW instead of BWT) greater than 20% was detected in only two cases.

CONCLUSION

Accuracy is similar to prior studies. Estimated newborn survival based on EFW is similar to that based on BWT.