ABSTRACT
OBJECTIVES: To evaluate the usefulness of first-trimester crown-rump length (CRL) discordance in predicting adverse outcome in twin pregnancies. METHODS: This retrospective study included a large cohort retrieved from local ultrasound databases at 14 obstetric departments in Denmark, comprising all twin pregnancies with two live fetuses scanned between 11 and 14 weeks' gestation during the period 1 January 2004 to 31 December 2006. The association between CRL discordance ≥ 10 % and adverse outcome was evaluated. RESULTS: Among 1993 twin pregnancies, 1733 were dichorionic (156 (9%) discordant; 1577 (91%) concordant) and 260 were monochorionic (32 (12%) discordant; 228 (88%) concordant). In dichorionic twin pregnancies we found an association between CRL discordance ≥ 10% and preterm delivery before 34 weeks' gestation (P=0.007), birth weight discordance (P=0.001) and mean birth weight (P=0.033). In monochorionic twin pregnancies we found an association between CRL discordance ≥ 10% and birth weight discordance (P=0.02) and mean birth weight (P=0.03). To evaluate CRL discordance as a predictor of fetal loss and preterm delivery before 34 weeks' gestation, receiver-operating characteristics curves were created for each outcome. For CRL discordance ≥ 10% as a predictor of fetal loss and preterm delivery in dichorionic twin pregnancies, sensitivity was 0.17 (95% CI, 0.06-0.28) and 0.14 (95% CI, 0.10-0.18), respectively, and in monochorionic twin pregnancies it was 0.10 (95% CI, 0.03-0.22) and 0.16 (95% CI, 0.06-0.26), respectively. CONCLUSIONS: CRL discordance in twin pregnancies is associated with, but is not a strong predictor of, adverse outcome.
