Objective: To assess the predictive accuracy of serial measurements of cervical length (CL) for preterm birth in asymptomatic women with triplet pregnancy.
Methods: A retrospective study of women with triplets who underwent serial sonographic measurements of CL until 28-32 weeks of gestation. The predictive accuracy of CL for preterm birth was determined at 4 periods along gestation: 18-20 weeks (period 1), 21-24 weeks (period 2), 25-27 weeks (period 3) and 28-32 weeks (period 4).
Results: A total of 431 measurements of CL from were analyzed. CL decreased in a linear manner across gestation: 40.8 ± 7.1 mm, 36.5 ± 8.4 mm, 29.9 ± 11.4 mm and 25.0 ± 11.8 mm in periods 1, 2, 3 and 4, respectively. The difference in CL between women who did and did not deliver prematurely was small before 25 weeks (periods 1&2) but became more pronounced later in pregnancy (periods 3&4), mainly due to a rapid cervical shortening between periods 2 and 3 (shortening rate -29.0 ± 20.0% vs. -12.6 ± 20.5%, respectively, p = .01). The best predictors of preterm birth were either a single measurement of CL during period 3 or the degree of cervical shortening between periods 2 and 3.
Conclusions: Care providers should be aware of the limited predictive value of cervical length before 25 + 0 weeks in triplet pregnancies.
Keywords: Cervical length; change; prediction; preterm birth; serial; triplets.