A prospective masked observational study of uterine contraction frequency in twins

This study was undertaken to compare uterine contraction frequency in twins versus singletons and to determine if contraction frequency can be an efficient predictor of spontaneous preterm birth in twin gestations. Fifty-nine twin and 306 singleton gestations were enrolled between 22 and 24 weeks at 11 centers. Contraction frequency was recorded with a home uterine activity monitor (HUAM) 2 or more times per day on 2 or more days per week until delivery or 36-6/7 weeks. Masked HUAM data were interpreted according to standard protocol. Repeated measures analyses were used to determine whether mean or maximum uterine contraction frequency per hour differed between singleton and twin gestations across gestational age, by time of day, and by delivery before 35 weeks or beyond. Uterine contraction frequency was also evaluated by logistic regression and receiver operator characteristic (ROC) curves as tests to predict spontaneous preterm birth. There were 34,908 hours of HUAM data recorded by the 306 singleton gestations and 5,427 hours by the 59 women with twins. Uterine contraction frequency was significantly greater in twins ( P = .002) compared with singletons, regardless of gestational age. Contraction frequency in twins increased significantly with gestational age and time of day (1600-0359 hours); but was not associated with spontaneous preterm birth. Maximum uterine contraction frequency was associated with preterm birth less than 35 weeks but only in the morning ( am) recording (0400-1559) and at the 29- to 30-week gestational age interval. This relationship was modest (odds ratio 1-2) and not consistent across gestational age or between the am and afternoon/evening ( pm) monitoring sessions. ROC analysis revealed no contraction frequency that efficiently identified twins who delivered prematurely at any 2-week gestational age interval. Mean uterine contraction frequency was significantly higher for twin gestations than singletons throughout the latter half of pregnancy and between 1600 and 0359 hours but was not higher among twins who delivered less than 35 weeks' gestation. Neither maximum am or pm contraction frequency predicted spontaneous preterm birth less than 35 weeks' gestation in twin pregnancies.