Prediction of birth weight discordance from fetal weight estimations at 21-24 weeks' scans in monochorionic and dichorionic twins.

OBJECTIVE

To evaluate if inter-twin estimated fetal weight (EFW) differences at 21-24 weeks' scans predict birth weight discordance in monochorionic and dichorionic twins born at three gestational age periods.

METHODS

We counted the number of pairs with discordant EFWs (EFWs greater than the mean + 1SD; i.e. a difference above 12% in dichorionic and 21% in monochorionic twins) derived during a 21-24 weeks' scan as compared to actual discordant birth weight (>25%) in monochorionic and dichorionic pairs born at ≤32, 33-36 and >36 weeks.

RESULTS

We studied 416 dichorionic and 244 monochorionic twins. The frequency of EFWs discordance at 21-24 weeks was the same for dichorionic and monochorionic twins but the total number of birth weight discordant pairs was significantly smaller among dichorionic twins (OR 0.2, 95% CI 0.1, 0.6). Generally, the positive predictive values for birth weight discordance by using the EFWs difference were quite poor, but always better in monochorionic than in dichorionic twins. Whereas the negative predictive value and hence the sensitivity for monochorionic twins was 100% irrespective of gestational age at birth, it was much lower in the dichorionic pairs.

CONCLUSION

Concordant EFWs at 21-24 weeks exclude discordant birth weight in monochorionic twins whereas discordant EFWs are poor predictors of birth weight discordance especially in dichorionic twins irrespective when the twins were born.