(OSA) in pregnancy Kathleen Antony, Alpna Agrawal, Melanie Arndt, Adrienne Murphy, Philip Alapat, Kay Guntupalli, Kjersti Aagaard Baylor College of Medicine, Obstetrics and Gynecology, Division of Maternal-Fetal Medicine, Houston, TX, Baylor College of Medicine, Department of Pulmonary-Critical Care, Houston, TX, University of Texas Health Science Center, Department of Medical Education, Houston, TX OBJECTIVE: While polysomnography (PSG) is the gold standard for diagnosisofOSA, it is cumbersomeandexpensive toperform.Severalquestionnaire-based scales for diagnosis have been developed and validated in non-pregnant subjects, namely the Berlin (BQ) and Epworth sleepiness scale (ESS). We sought to ascertain the validity of these scales in pregnancy in order to establish the true prevalence of OSA in pregnancy. STUDY DESIGN: Following sample sized estimation based on pilot data, the BQ and ESS were administered to low and high-risk gravidae at initial presentation. Screen positive subjects were referred to a certified sleep lab for PSG, and incentivization reimbursement was provided. All subjects were tracked and, if admitted for antepartum care, underwent a modified sleep study (ResMed ApneaLin (AL) recording continuous pulse oximetry and nasal airflow). RESULTS: 1507 subjects were enrolled (results summarized in Fig.). The prevalence of subjects who screened positive varied by method (15.1% BQ, 19.0% ESS, and 30.1% by either BQ or ESS); screening positive on the BQ was significantly associated with screening positive on the ESS (p 0.05). Neither the BQ nor ESS were reliable diagnositic measures for OSA by either AL or PSG (ROC 0.45 for BQ, ROC 0.55 for ESS). Despite incentivization, written and verbal contact, and flexbile scheduling, 86.8% of subjects did not undergo diagnostic testing. Among 13.2% received PSG or AL, neither the BQ nor ESS were predictive of OSA and a higher than expected AL-negative rate was observed (92.6%). Among those screen positive subjects undergoing PSG or AL testing, 14.5% ultimately met OSA diagnostic criteria for an estimated point prevalence of 4.35%. CONCLUSION: In the largest reported prospective trial on OSA to date, we have observed that while screening by either Berlin or Epworth scales may be validated measures in non-pregnant subjects, it is poorly predictive among gravidae. These findings raise concern for studies which report OSA-linked outcomes without corroborating PSG diagnostic criteria having been met. 672 General anesthesia at cesarean delivery portends worse maternal and neonatal outcomes Kathleen Brookfield, Sarah Osmundson, Mariam Naqvi, Alexander Butwick, Dierdre Lyell Stanford University, Obstetrics & Gynecology, Stanford, CA, Stanford University, Anesthesia, Stanford, CA OBJECTIVE: The cesarean delivery rate is more than 34% in the U.S., and is expected to continue to rise. Women undergo cesarean delivery under general anesthesia (GA) for several reasons, some of which may increase maternal and neonatal morbidities. We compared maternal and neonatal outcomes for women undergoing cesarean delivery under GA vs. other types of anesthesia (spinal, epidural, narcotics, and local) to assess the effect of anesthetic on outcome. STUDY DESIGN: A subset of women who underwent primary or repeat cesarean delivery were identified from the NICHHD MFMU Network registry, which collected data on pregnancy outcomes from patients at 19 centers from 1999-2002. We compared outcomes including time to delivery, birthweight, Apgars, need for prolonged ventilation, and maternal/neonatal death among patients receiving GA vs. other anesthesia. RESULTS: 54,632 patients underwent cesarean, of whom 4,172 underwent GA and 50,460 underwent another type of anesthesia. Women who received GA were more likely to have an abruption, chorioamnionitis, pulmonary edema, emergency cesarean, abnormal placentation, maternal ICU admission and death, stillbirth, neonatal hypotension, prolonged hypotonicity, and respiratory morbidity (p 0.01). Although mean incision to delivery time was faster for women undergoing GA (6.2 mins vs. 10.7 mins; p 0.001), mean neonatal umbilical cord pH was lower (7.21 vs. 7.24; p 0.001). When emergent cesareans were excluded (n 365), a multilogistic regression model demonstrated that exposure to GA vs. other types of anesthesia was an independent predictor of 5 minute Apgar score 5 (p 0.01), as was delivery at 32 weeks. In a second model including only women who received GA for non-emergent cesarean deliveries, prolonged incision-to-delivery time ( 18 mins) was an independent predictor of 5 minute Apgar scores of 5 (p 0.037). CONCLUSION: Multiple maternal and neonatal effects exist from the use of general anesthesia. Shortening the incision to delivery time may improve neonatal outcomes.
Read full abstract