CI 0.40-1.1). However, the numbers in this subgroup were too smallto draw definitive conclusions.CONCLUSION: In women with a twin gestation, treatment with progesto-gens neither prolongs pregnancy nor improves neonatal outcome. Thevalue of screening for short cervix in twin gestations and subsequent pro-gestogen treatment should be explored in prospective studies.

473 The NIFTY study: a multi-centre randomiseddouble blind placebo controlled trial of nifedipinemaintenance tocolysis in fetal fibronectin positivewomen in threatened preterm labourEmma Parry1, Carolien Roos2, Peter Stone3, LynseyHayward4, Ben Willem Mol5, Lesley McCowan6

1Auckland District Health Board, Maternal-Fetal Medicine, Auckland, NewZealand, 2Dutch Consortium for Studies in Women’s Health andReproductivity, Department of Obstetrics and Gynaecology, Amsterdam,Netherlands, 3University of Auckland, Obstetrics and Gynaecology,Auckland, New Zealand, 4Counties Manukau District Health Board,Obstetrics and Gynaecology, Auckland, New Zealand, 5Dutch consortiumAMPHIA trial, Netherlands, 6University of Auckland, Departmentof Obstetrics and Gynaecology, Auckland, New ZealandOBJECTIVE: Maintenance tocolysis is frequently prescribed for womenin preterm labour, but evidence of effectiveness is lacking. Womenwho are Fetal Fibronectin positive have increased risk of progressionto preterm birth. We hypothesised that maintenance nifedipinewould reduce neonatal intensive care admission by � 4 days andprolong pregnancy by � 7 days.STUDY DESIGN: Between 2003 and 2008 women with a singleton preg-nancy in threatened preterm labour (24�0 weeks to 33�6 weeks)with a positive Fetal Fibronectin swab who had completed a course ofcorticosteroids and 48 hours of acute nifedipine tocolysis were invitedto participate. Exclusion criteria were: ruptured membranes, bleed-ing, fetal anomaly, suspected chorioamnionitis or contraindication toNifedipine. Nifedipine (same dose as prescribed in preceding 24hours) or placebo was continued until 37 completed weeks gestation.The primary endpoints were a reduction in NICU admission of � 4days and prolongation of pregnancy of � 7 days (sample sizes of 38and 100 respectively).RESULTS: Of the 60 participants, 29 received nifedipine and 31 pla-cebo. Baseline characteristics were comparable. Gestational age atrandomization was 29.8 and 29.5 weeks for nifedipine and placebo,respectively. Gestational age at delivery was 36.1 5.1 weeks for nifed-ipine and 36.8 3.6 weeks for placebo (p� 0.03). Prolongation of preg-nancy by � 7 days occurred in 22/29 (76%) in the nifedipine groupand 25/31 (81%) in the placebo group (RR 0.94 [0.72 to 1.2]). Eightneonates (28%) in the nifedipine group and 7 (23%) in the placebogroup were admitted to NICU (RR 1.2 [0.51 to 2.9]). Length of NICUadmission [median (IQR)] was 27 (24 - 41) days and 16 (8 - 37) daysin nifedipine and placebo groups, respectively (p� 0.17).CONCLUSION: In women with threatened preterm labour who are FetalFibronectin positive, maintenance tocolysis with nifedipine neitherprolongs pregnancy, nor reduces length of NICU admission.

474 Potential benefits of vaginal delivery onsignificant morbidity in preterm AGA infantsErika F. Werner1, Christina S. Han2, David Savitz3, AnnaSfakianaki2, Edmund F. Funai2, Heather Lipkind2

1Johns Hopkins University, Gynecology and Obstetrics, Baltimore, MD,2Yale University, Ob/Gyn & Reprod Sci., New Haven, CT, 3Brown University,DepartmentsofEpidemiologyandObstetricsandGynecology,Providence,RIOBJECTIVE: According to National Vital Statistics, the cesarean deliv-ery (CD) rate for very preterm infants (delivery at �34 weeks gesta-tion) is rapidly approaching 50 percent. The increase in CDs has oc-curred with relatively sparse data regarding what is best for the healthof the neonate. This study was undertaken to compare neonatal out-comes by method of delivery in preterm, appropriate for gestationalage (AGA) infants in a large diverse cohort.STUDY DESIGN: Birth data for 1995 to 2003 from New York City werelinked to hospital discharge data. Data were limited to singleton, liveborn, cephalic neonates delivered between 25 and 34 weeks. Exclusioncriteria included congenital anomalies, forceps or vacuum assistance,small for gestational age (less than the 10th percentile), and motherswith a history of prior CD. Any neonatal diagnosis of intraventricularhemorrhage (IVH), seizure, sepsis, subdural hemorrhage, respiratorydistress syndrome (RDS), or five minute Apgar � 7 was considered asignificant neonatal morbidity. Associations between method of de-livery and neonatal morbidities were estimated using logistic regres-sion.RESULTS: 18,370 neonates meeting the study criteria were identified;74% were delivered vaginally and 26% were delivered by CD. Aftercontrolling for maternal age, ethnicity, education, primary payor,pre-pregnancy weight, gestational age at delivery, diabetes and hyper-tension, CD was associated with increased odd of RDS (OR 1.9, 95%CI 1.7-2.0) and five minute Apgar �7 (OR 2.5, 95% CI 2.2-3.0) com-pared to vaginal delivery. Prior to adjusting for the aforementionedconfounding factors, CD was associated with increased odds of allneonatal morbidities compared to vaginal delivery.CONCLUSION: While preterm AGA neonates delivered by CD may havea higher baseline risk for some neonatal complications that can beexplained by preexisting maternal and fetal factors, those born by CDhad a higher risk of RDS and low Apgar score. In all variables studied,in no instance were outcomes more favorable among the CD group.

Poster Session III Doppler Assessment, Fetus, Prematurity www.AJOG.org

S216 American Journal of Obstetrics & Gynecology Supplement to JANUARY 2012