24 Use of outcomes data to determine fetal growth standards Jacob Larkin 1 , Hyagriv Simhan 1 1 University of Pittsburgh, Dept of OBGYN, Div of Maternal-Fetal Medicine, Pittsburgh, PA OBJECTIVE: Extremes of fetal growth are universally recognized as de- terminants of perinatal morbidity, and screening for aberrant fetal growth is a standard component of routine obstetric practice. None- theless, criteria for discerning normal from abnormal growth are crude and arbitrary. By convention, SGA or LGA thresholds are set at the 10th and 90th centile in a population distribution of birthweight, and are not determined by clinical risk thresholds or outcome data. Fractional polynomial regression permits estimates of risk along all values of continuous predictor variables, without assumptions of lin- earity or constant slope. We sought to examine changes in the risk of neonatal death along the continuum of birthweight centiles, and to use these findings to generate outcome-driven fetal growth standards. STUDY DESIGN: We determined birthweight centile for gestational age in 87,184 liveborn, non-anomalous, singleton neonates delivered at Magee-Womens Hospital. Using fractional polynomial logistic re- gression, we used gestational age at delivery and birthweight centile to predict neonatal death. RESULTS: The probability of neonatal death as determined by birth- weight centile and gestational age is shown in the Figure. The risk of neonatal death declines with decreasing gestational age, nadirs at the 55th centile of birthweight, and increases incrementally with excur- sions from the 55th centile. A fetalgrowth standard, listing weight thresholds which correspond to a 3-fold increase in risk of neonatal death over the gestational age-specific nadir (9.2% in birthweight dis- tribution for SGA, 98.1% for LGA), is shown in the Table. CONCLUSION: Standards of normal fetal growth can be determined by clinically relevant absolute or relative risk thresholds of important perinatal outcomes (e.g., neonatal death, stillbirth or NICU admis- sion). To our knowledge, this is the first instance in which thresholds of normal fetal growth have been determined by perinatal outcome data. 25 Does having an EFW less than the 10th centile really matter? Results of the National Multicenter Prospective PORTO trial Julia Unterscheider 1 , Sean Daly 2 , Michael Geary 3 , Mairead Kennelly 4 , Fionnuala McAuliffe 5 , Keelin O’Donoghue 6 , Alyson Hunter 7 , John Morrison 8 , Gerard Burke 9 , Patrick Dicker 10 , Elizabeth Tully 1 , Fergal Malone 1 1 Royal College of Surgeons in Ireland, Obstetrics & Gynecology, Dublin Ireland, 2 Coombe Women and Infants University Hospital, Obstetrics & Gynecology, Dublin, Ireland, 3 Rotunda Hospital, Obstetrics & Gynecology, Dublin, Ireland, 4 Coombe Women and Infants University Hospital, UCD Center for Human Reproduction, Dublin, Ireland, 5 National Maternity Hospital, UCD Obstetrics & Gynecology, School of Medicine and Medic Science, Dublin, Ireland, 6 University College Cork, Cork University Maternity Hospital, Obstetrics & Gynecology, Cork, Ireland, 7 Royal Jubilee Maternity Hospital, Obstetrics & Gynecology, Belfast, Ireland, 8 National University of Ireland, Obstetrics & Gynecology, Galway, Ireland, 9 Mid- Western Regional Maternity Hospital, Obstetrics & Gynecology, Limeri Ireland, 10 Royal College of Surgeons in Ireland, Epidemiology & Public Health, Dublin, Ireland OBJECTIVE: The PORTO Trial is a multicenter prospective trial con ducted at the seven largest obstetric centers in Ireland, with it being to evaluate optimal management of the IUGR fetus. For the purposes of the Trial, IUGR was defined as EFW less than the 1 centile. It is unclear however whether this definition is of clinic nificance. The objective of this analysis is to document the out of this population. STUDY DESIGN: A total of 1,056 ultrasound-dated singleton pregna cies with EFW ⬍10th centile were recruited between 24 0/7 an weeks‘ gestation between January 2010 and June 2012. Perina early neonatal outcomes were documented for all participants RESULTS: Of 1,056 pregnancies with EFW ⬍10th centile at recruit ment, 820 (78%) remained ⬍10th centile until delivery. 492 (4 had abnormal umbilical artery (UA) Dopplers and 82 (8%) deve UA AEDF or REDF. Table 1 summarizes the maternal and fetal acteristics. Mean gestational age (GA) at enrollment and delive 29.8 and 37.6 weeks, respectively. There were 8 aneuploidies congenital anomalies. The overall perinatal mortality rate in th hort was 14.2 per 1000 births. Among the normally formed inf with normal karyotype, there were 6 stillbirths (1:170) and 5 n deaths (1:200). CONCLUSION: Having an EFW less than the 10th centile is a transie finding in 22% of pregnancies. For the remaining 78% with per tently low EFW, constitutionally small size,rather than pathologic IUGR, is by far the most likely outcome. This calls into question utility of EFW less than the 10th centile as a definition for poss IUGR. A careful evaluation of possible underlying structural or typical abnormalities is warranted in these pregnancies. Risk of neonatal death predicted by birthweight centile and gestational age at delivery Fetal growth standard SGA threshold ⫽ 9.2%, LGA threshold ⫽ 98.1%. Oral Concurrent Session 2 Fetus/Ultrasound www.AJOG.org S16 American Journal of Obstetrics & Gynecology Supplement to JANUARY 2013