The performance of second trimester long bone ratios for Down syndrome screening is influenced by gestational age

To determine if gestational age (GA) at the time of ultrasound impacts the positive predictive value of shortened femur and humerus lengths (FL, HL) for trisomy 21 (T21).

Sonograms from 14 to 21 and 6/7 weeks' gestation were collected over a 28 month period. Multiple gestations or fetuses with major structural anomalies were excluded. Biometric data and GA were obtained; the expected HL (or FL): observed HL (or FL) ratios were calculated using two regression formulas (Benacerraf and Nyberg). A HL ratio <0.90 and a FL ratio <0.91 were considered shortened. T21 fetuses were identified through database and chart review. Positive predictive values (PPV) for T21 of the shortened bone ratios were determined, then stratified by GA.

Of the 2606 ultrasounds, 8.9% and 18.9% of fetuses had shortened HL and FL ratios, respectively, using the Benacerraf formula. Shortened ratios were noted significantly less commonly (2.3 and 4.4%, respectively, P < 0.001 for each) using the Nyberg formula. With either formula, abnormal bone ratios were more frequently documented with a GA less than 17 weeks (P < 0.001). There were 17 T21 pregnancies.

GA and formula selection influence the performance of long bone ratios as soft markers for T21 in the second trimester.