Statistics
The incidence of sIUFD in MC twin pregnancy was planned to be calculated using ONS data as the denominator. Maternal, fetal and neonatal outcomes were described and reported as percentages, mean and SDs or medians and IQRs as appropriate. Sub-group analyses were planned to determine if TTTS or sIUGR, were associated with any characteristics, or if they were associated with a poorer outcome for the surviving co-twin. We also planned to explore prognostic indicators for mother and infant associated with sIUFD in MC twin pregnancy, e.g. gestation of IUFD of the first twin, using logistic regression. No sample size calculation was performed, but prior to commencing the study, we had estimated, based on ONS data from 2012 and existing literature, that of 11,228 twin pregnancies, 30% would be MC, and 7.5% would result in a sIUFD, thus there would be 253 cases per annum. No imputations of missing data were performed, cases with missing outcomes were not included in certain analyses as stated below. The study is funded by the British Maternal Fetal Medicine Society and Twins Trust (formerly TAMBA).