Objective: The purpose of this study was to determine whether the use of race-specific definitions of short femur and humerus lengths improves Down syndrome detection.
Methods: This was a retrospective cohort study over 16 years. For each self-reported maternal race (white, African American, Hispanic, and Asian), we evaluated the efficiency of Down syndrome detection using published race-specific formulas compared with a standard formula for short femur and humerus lengths (observed versus expected lengths < or =0.91 and < or =0.89, respectively). The sensitivity, specificity, and 95% confidence intervals for each parameter were compared. Screening performance was compared by areas under the receiver operating characteristic curves.
Results: Of 58,710 women, 209 (0.3%) had a diagnosis of a fetus with Down syndrome. Although the race-based formula increased sensitivity in each population, the increase was statistically significant only in the white population, whereas a decrease in specificity was statistically significant in all 4 populations, as denoted by nonoverlapping confidence intervals. The area under the receiver operating characteristic curve for the model using the race-specific definition of short femur length was 0.67 versus 0.65 compared with the standard definition, and for humerus length it was 0.70 versus 0.71.
Conclusions: The use of race-based formulas for the determination of short femur and humerus lengths did not significantly improve the detection rates for Down syndrome.