Methods
This was a retrospective cohort study of all MC twins undergoing sIUD at 14-40 weeks from 2012 through 2018 at Shengjing Hospital of China Medical University. Inclusion criteria included delivery between January 2012 and April 2018 with complete information, including pregnancy outcomes. The gestational age of birth was assigned based on the first day of a woman’s last menstrual period. If this dating was not consistent with dating based on the earliest ultrasound (±7 days in the first trimester or ±10 days in the second trimester), the gestational age was reassigned. 9 The gestational age of sIUD was estimated according to the size of the dead fetus measured via ultrasound and previous ultrasound. Chorionicity was determined by ultrasound evaluation according to the presence of the lambda or T signs and confirmed after birth.10 If the patient did not undergo chorionic identification ultrasound during pregnancy, the placenta was examined after delivery to determine the chorionicity.11 sIUD was defined as spontaneous death and spontaneous death after fetoscopic laser occlusion after 14 weeks of gestation. Exclusion criteria included sIUD by radio frequency ablation and miscarriage occurring before 26 weeks and double fetal loss.
Information on each case was recorded to include the following: age (year), the gestational age of sIUD, parity, the gestational age at birth, history of fetoscopic laser occlusion, pregnancy method(including natural pregnancy and assisted reproductive conception), fetal complication, maternal complication, premature rupture of membranes, abnormality in amniotic fluid, and spontaneous labor or not.
According to the gestational age of birth, the cases were divided into PTB group (<37 weeks) and full term birth group (>37 weeks). We also divided these cases into four groups according to the gestational age of sIUD (14-19weeks, 20-24weeks, 25-30weeks, 31-36weeks) and compared the four groups using the chi-square test for trends.
Statistical analyses were performed with SPSS software (Version 17.0, SPSS). Chi-square analysis was performed to test the significance of the findings. Logistic regression analysis was conducted to determine risk factors for preterm delivery. P < 0.05 was considered statistically significant.