1 (filmy), or 3 (dense). We then abstracted charts for delivery data, including the recorded skin incision to neonatal delivery times. We excluded cases where incision to delivery time was unavailable. We analyzed data using student t-test, chi square and survival analyses. RESULTS: A total of 145 women met study criteria and were analyzed, including 92 (63.5%) with adhesions and 53 (36.5%) without. More women with adhesion scores 3 (n=28) remained undelivered at 10, 20, 30 and 40 minutes after skin incision compared with women with scores of 3 (n=117) (Figure). By 30 minutes after incision, 17.9% of women with adhesion score 3 remained undelivered, vs. 5.1% of those with scores of 3 (p=0.04). The mean incision to delivery time in women with scores of 3 was 19.8 minutes, compared to 15.6 minutes in women with scores of 3 (p=0.04). CONCLUSIONS: After only one prior CD, adhesion severity predicted delayed delivery of the newborn. Identification of surgical techniques or materials to reduce adhesions at cesarean is needed, given that adhesions increase with each subsequent cesarean and national cesar- ean rates currently approach 32%. 679 Predicting delivery route in nulliparas with classification and regression tree analysis Michelle Kominiarek 1 , Paul VanVeldhuisen 2 , Kimberly Gregory 3 , Moshe Fridman 4 , Judith Hibbard 1 1 University of Illinois at Chicago, Chicago, IL, 2 EMMES Corporation, Rockville, MD, 3 Cedars-Sinai Medical Center, Los Angelos, CA, 4 AMF Consulting, Inc., Los Angeles, CA OBJECTIVE: To predict delivery route in nulliparas at labor admission. STUDY DESIGN: The Consortium on Safe Labor collected electronic data from 12 institutions on 228,668 deliveries from 2002-08. Nullip- aras with live, singleton gestations 37 weeks having a trial of labor with known body mass index (BMI,kg/m 2 ) at labor admission were analyzed. Classification and regression tree (CART) analysis was used to build a hierarchy of predictors for delivery route using race, age (30 or 30 years), height, admission BMI (25, 25-29.9, 30-34.9, 35-39.9, or 40), gestational age (37-39 or 40 weeks), diabetes, hypertension, admission cervical dilation (1cm, 1-3cm, 3.5-4.5cm, or 5cm), and induction. RESULTS: Cesareans occurred in 21% of the 53,905 included cases. Characteristics more common in cesarean than vaginal deliveries were greater age and BMI, African American or Hispanic race, diabe- tes, hypertension, induction, and less dilated cervices, P0.001 for all comparisons. Cervical dilation was the first branch of the CART. Ce- sareans occurred in 44%, 23%, 15%, and 10% at 1cm, 1-3cm, 3.5- 4.5cm, and 5cm respectively. BMI was the second branch for all dilation categories. BMI35 was associated with the most cesareans (60% for 1cm) and 19% of those with a BMI35 and 5cm still had a cesarean. The BMI influence was most evident in the 3.5-4.5cm category with 8% of BMI25 and 33% of BMI 40 having a cesarean. The third branch points included age, gestational age, and induction. The fewest cesareans (4.9%) occurred in those 5cm, BMI25, and 40 weeks. CONCLUSIONS: Cervical dilation followed by maternal BMI were im- portant predictors of delivery route in term laboring nulliparas. Given the rising cesarean rate and importance of decreasing this event in nulliparas, this provides data to encourage nulliparas and clinicians to await spontaneous labor, especially in obese patients, as the ultimate deterrent to cesarean delivery. Further research is needed to deter- mine if interventions in labor management can impact delivery out- come. 680 A randomized controlled trial of the effect of supplemental perioperative oxygen on the rate of surgical site infection Neena Duggal 1 , Vineela Poddatoori 1 , Sara Noroozkhani 1 , Carter Cherry 1 , Tim Park 1 , Aaron Caughey 2 1 Santa Clara Valley Medical Center, San Jose, CA, 2 Oregon Health & Science University, Portland, OR OBJECTIVE: To determine whether supplemental perioperative oxygen decreases surgical site wound infection and endometritis. STUDY DESIGN: This was a prospecttive randomized trial. Patients who were undergoing cesareans and had not received any antibiotics were recruited for the study. Each patient was randomized to 30% or 80% oxygen during the cesarean delivery and for one hour after surgery. The obstetrician and the patient were blinded to the concentration of oxygen used. The patient was evaluated for wound infection or endo- metritis during her hospital stay and at two weeks postpartum. Pri- mary end points were surgical site infection, as defined by the CDC, and endometritis. RESULTS: 827 patients were recruited for the study. 92 patients were deleted from the study and data analysis was done on 735 patients. Of these, 365 patients received 30% oxygen perioperatively and 370 pa- tients received 80% oxygen. The groups were well matched for age, parity, BMI, diabetes, number of previous cesareans, and scheduled versus unscheduled cesarean. There was no difference in wound in- fection in the two groups (6.8% versus 6.8%), and no significant dif- ference in endometritis in the two groups (9.4% in the patients who received 30% oxygen and 8.6% in the patients who received 80% oxygen). CONCLUSIONS: Patients who received supplemental oxygen periopera- tively did not have a lower rate of surgical site infection or endome- tritis as compared to patients who received the standard oxygen concentration. Table. Surgical Site Infection (SSI) / Endometritis 30% Oxygen 80% Oxygen Infection N365 N370 P value SSI Week 1 2.7% 3.5% 0.55 .......................................................................................................................................................................................... SSI Week 2 4.9% 4.1% 0.57 .......................................................................................................................................................................................... Total SSI 6.8% 6.8% 0.96 .......................................................................................................................................................................................... SSI and Endometritis 9.4% 8.6% 0.76 .......................................................................................................................................................................................... 681 Uterine rupture is it really so common and devastating obstetric event? Oz Gavish 1 , Rony Chen 1 , Assaf Bibas 1 , Yariv Yogev 1 1 Helen Schneider Hospital for Women, Petach Tikva OBJECTIVE: We aimed to assess the incidence of uterine rupture and to determine associated risk factors in a single tertiary center. STUDY DESIGN: A retrospective cohort study. Uterine rupture was de- fined according to ICD-9, code 6651. After collecting data from the computerized database, all patients’ charts were reviewed meticu- lously. In each case of suspected uterine rupture, two physicians as- sessed patients file for diagnosis verification. Primary study group was composed of women with complete uterine rupture. Secondary study Poster Session IV Epidemiology, Global Maternal-Fetal Public Health, Infectious Disease, Intrapartum Fetal Assessment, Operative Obstetrics www.AJOG.org S268 American Journal of Obstetrics & Gynecology Supplement to JANUARY 2011