The effect of fetal gender in predicting Down syndrome using long bone ultrasonographic measurements

Abstract

To determine if fetal gender affects the screening efficiency of short femur and humerus lengths in the prediction of trisomy 21. Retrospective cohort study of 62 111 patients presenting for ultrasound from 1990 to 2006. Short humerus and femur lengths were defined using (1) biparietal diameter (BPD) to femur/humerus length (FL/HL) ratios > 1.5 standard deviations above the mean, (2) the observed to expected (O/E) ratio of femur length ≤ 0.91 or humerus length ≤ 0.89, and (3) femur and humerus lengths < 5th percentile. The sensitivity, specificity, and likelihood ratios were calculated for the association of short FL/HL and trisomy 21 stratified by gender. Both BPD/long bone ratios as well as O/E ratios demonstrated a statistically significant higher specificity for the detection of trisomy 21 in female fetuses. This difference was most clinically significant when using the O/E ratio, which yielded a specificity of 82.6% in males and 90.6% in females for short femur, and 69.7% in males and 77.9% in females for short humerus, when these markers were evaluated as isolated findings. Gender-specific differences in the effectiveness of both short femur and humerus lengths for the prediction of trisomy 21 may exist, but their presence and magnitude are largely dependent on the formula used.