Model predicted Down syndrome detection rates for nuchal translucency screening in twin pregnancies

To estimate the Down syndrome detection rate for nuchal translucency (NT) screening in twins when fetus-specific risk allows for between-fetus NT correlation.

The between-fetus correlation coefficient of log NT, in multiples of the median (MoM), was estimated from a series of 977 unaffected twins scanned at a single centre. Results were expressed in multiples of the normal median using a curve derived from 515 unaffected singleton pregnancies at the same centre. A screening result was classified as positive if the risk for at least one fetus exceeded the cut-off. Detection rates were estimated for a fixed 1-5% false-positive rate, at different gestational weeks, separately for risk calculation using an algorithm which takes account of between-fetus NT correlation or not.

The correlation coefficient in unaffected pregnancies was 0.43 (P < 0.0001) and estimated to be 0.23 and 0.11 in discordant and concordant twins. At 12 weeks of gestation, the model predicted detection rate for a 3% false-positive rate was 68% when between-fetus correlation is not taken into account, increasing to 73% when it is applied. Similarly, for other false-positive rates and gestational weeks there was a predicted 4-6% increase in detection.

Using a fetus-specific Down syndrome risk algorithm leads to a worthwhile increase in detection.