Threatened delivery of a periviable neonate, defined as between 20 and 25 weeks of gestation, is a challenging clinical and ethical situation for the obstetrician and neonatologist with regard to various interventions to be offered at these vulnerable gestational ages.1 The American College of Obstetricians and Gynecologists and the Society for Maternal-Fetal Medicine recently released an Obstetric Care Consensus entitled “Periviable Birth” to assist health care providers in counseling affected patients.2 This document acknowledges that although interventions such as antenatal corticosteroid administration pose less risk to the mother, other interventions may result in both short- and long-term maternal morbidity.2
There is a scarcity of research, however, regarding the association of periviable birth with maternal outcomes (Rossi R, DeFranco E. Abstract 831: factors associated with cesarean delivery in previable births. Am J Obstet Gynecol 2017;216:S476–7.).3 Research in this area has been limited to the study of future reproductive outcomes and emphasizes maternal morbidity associated with cesarean delivery.4–8 Therefore, we sought to quantify the rate of maternal complications associated with a periviable birth (20–25 weeks of gestation) in a contemporary population.
MATERIALS AND METHODS
We performed a population-based retrospective cohort study of all live births in Ohio between 2006 and 2015 using U.S. live birth records from the state of Ohio. As of 2006, all birth records within Ohio used the newest version of the national birth certificate (2003).9 The protocol for this study was approved and a deidentified data set was provided by the human subjects institutional board review of the Ohio Department of Health. This study was exempt from review by the institutional review board at the University of Cincinnati, Cincinnati, Ohio. The primary outcome for this study was the frequency of maternal complications with periviable birth, defined as live birth between 20 and 25 weeks of completed gestation as reported on the U.S. birth certificate.9,10 The frequency of maternal complications among women who had a periviable delivery was compared with those who delivered preterm and at term and also stratified by 3-week epochs of gestational age at birth ranging from 20 to 40 weeks.
Gestational age was defined by the best obstetric estimate variable in the birth record, which takes into account a combination of last menstrual period, clinical, and ultrasound parameters.11 The exposure variable for this study, periviable birth, was defined as live birth between 20 and 25 weeks of completed gestation and acquired from the birth certificate based on data obtained from the obstetric medical record.9,10 The primary outcome, composite adverse maternal outcome, was also obtained from the birth certificate form and included the following complications: chorioamnionitis, blood product transfusion, unplanned hysterectomy, unplanned operative procedure, uterine rupture, and maternal admission to an intensive care unit (ICU).10
Cases with missing mode of delivery or gestational age at delivery (or both) were not included in this analysis. Body mass index was calculated using prepregnancy weight and height as recorded in the medical record and categorized by World Health Organization.12 All data variables included in analyses were obtained from the birth certificate as outlined in the National Vital Statistics System Guide for Completing the Facility Worksheets for the Certificate of Live Birth in the United States.10 All birth facilities in Ohio use this standard worksheet to collect information for birth certificates. This guide includes definitions, instructions for abstractors, sources from which to obtain data from the medical record, and keywords or abbreviations commonly used for each variable.10 All birth records included in this analysis used the most recent (2003) version of the U.S. birth certificate.9
Differences in baseline maternal demographic, obstetric, and neonatal characteristics among periviable (20–25 weeks of gestation), preterm (26–36 weeks of gestation), and term (greater than 36 weeks of gestation) groups were compared. Statistical comparisons were performed using analysis of variance and χ2 analyses for continuous and categorical data, respectively. Multivariate logistic regression was then used to estimate the adjusted relative risk of periviable birth and each preterm gestational age epoch on composite adverse maternal outcome as well as individual maternal complications using term birth as a comparison group. Women with multifetal gestations were excluded from the regression model of maternal complications. The final regression model was constructed using a backward selection process. Because outcome events were rare and there were significant demographic, socioeconomic, and obstetric differences between the periviable and term birth groups in univariate pretest comparisons, the number of variables in the regression model was limited to 10 events per variable according to Peduzzi et al to avoid overfitting the model and to ensure a reproducible, stable regression model.13,14 The variables selected for the adjusted model for composite adverse maternal outcome included race, socioeconomic status (Medicaid insurance), cigarette smoking, pregestational diabetes, chronic hypertension, parity, malpresentation, history of prior cesarean delivery, year of delivery, and cerclage placement during the pregnancy. Rates of individual maternal complications were also compared between 3-week gestational age epochs (20–22, 23–25, 26–28, 29–31, 32–34, 35–37, 38–40 weeks of gestation). Multivariate logistic regression was then used to estimate the adjusted relative risk of each delivery epoch on composite adverse maternal outcomes (chorioamnionitis, blood product transfusion, unplanned hysterectomy, unplanned operative procedure, uterine rupture, and ICU admission) compared with the “term” (38–40 weeks of gestation) referent group. Sensitivity analyses were performed including and then excluding women who delivered by cesarean, because this was likely an intermediate variable in the causal pathway of the relationship between periviable birth and maternal complications. As such, it would be statistically inappropriate to adjust for cesarean delivery in adjusted analyses.15 We performed a separate analysis, limiting the cohort to women likely representative of spontaneous preterm birth by excluding women who were induced or had complications such as preeclampsia or fetal growth restriction. Significant differences were defined as comparisons with P value of <.05 and 95% CI not inclusive of the null value of 1.0. Statistical analyses were performed using STATA 12.
Of 1,457,706 live births in Ohio during the 10-year study period, there were 6,085 periviable births between 20 and 25 weeks of gestation, 173,776 preterm births (26–36 weeks of gestation), and 1,267,020 term (greater than 36 weeks of gestation) births. The majority of periviable births occurred between 23 and 25 weeks of gestation. Periviable live births as a whole comprised 0.4% of all live births, whereas preterm and term births comprised 12.0% and 87.6%, respectively, during the study period. There were minimal missing data on gestational age at birth (n=7,838 [0.5%]) or mode of delivery (n=2,987 [0.2%]) during the study period. These births were excluded from the analysis.
Women who had a periviable birth were younger and more likely have other medical comorbidities such as obesity, pregestational diabetes, and hypertensive disorders in pregnancy compared with those who delivered at term (Table 1). Women who had a periviable delivery were more likely to be of non-Hispanic black race, smoke cigarettes, and have government-funded medical insurance (Medicaid) and were less likely to be married or to have completed a high school education (Table 1). As expected, the periviable group was also more likely to receive obstetric interventions such as antibiotics while in labor, tocolytic therapy, antenatal corticosteroids, and to have undergone placement of a cerclage compared with the other delivery groups (Table 2). Women were more likely to be transferred before delivery in the periviable group. Women in the preterm birth group had the highest cesarean delivery rates (44.0% vs 41.0% vs 28.9%, P<.001) compared with the periviable and term birth groups. The periviable birth group also had higher rates of malpresentation at delivery and multifetal pregnancies (Table 2).
The overall rate of composite adverse outcome (comprising chorioamnionitis, blood product transfusion, unplanned hysterectomy, unplanned operative procedure, uterine rupture, and ICU admission) for a periviable birth was 17.2% vs 5.1% and 2.7% for preterm and term births, respectively (P<.001). Periviable birth was also associated with higher rates of individual maternal complications compared with preterm and term birth including higher rates of chorioamnionitis (11.8% vs 1.7% vs 1.2%), blood product transfusion (1.5% vs 0.6% vs 0.2%), cesarean delivery (41.0% vs 44.0% vs 29.8%), unplanned hysterectomy (0.25% vs 0.14% vs 0.03%), unplanned operative procedure after delivery (4.6% vs 2.6% vs 1.3%), ruptured uterus (0.17% vs 0.06% vs 0.02%), and ICU admission (0.94% vs 0.45% vs 0.08%), but were less likely to receive epidural anesthesia (39.9% vs 50.0% vs 58.5%), respectively (all P values <.001; Table 2).
After exclusion of multifetal pregnancies and adjusting for maternal race, socioeconomic status (Medicaid insurance), cigarette smoking, pregestational diabetes, chronic hypertension, parity, malpresentation, history of prior cesarean delivery, year of delivery, and cerclage placement during pregnancy, multivariate logistic regression revealed an increased associated risk of composite adverse maternal outcome (adjusted relative risk [RR] 5.8, CI 5.4–6.2) in the periviable delivery group (Table 3). There was also significantly increased risk of individual maternal morbidities including chorioamnionitis, blood product transfusion, cesarean delivery, unplanned operative procedure after delivery, unplanned hysterectomy, ruptured uterus, and ICU admission during the periviable period compared with the term cohort (Table 3). Despite increased maternal complications associated with periviable birth, women were less likely to receive epidural anesthesia during delivery (Table 3).
The rate and risk of individual maternal complications were compared stratified by gestational age delivery group (20–22, 23–25, 26–28, 29–31, 32–34, 35–37, and 38–40 weeks of gestation). The composite adverse outcome rate was highest in the earliest gestational age groups (20–22 and 23–25 weeks of gestation) at 18.0% (95% CI 16.4–19.7) and 16.9% (95% CI 15.7–18.1), respectively (Table 4). The composite adverse outcome rate decreased with increasing gestational age (26–28 weeks of gestation; 11.6%, 95% CI 11.0–12.3]. After adjusting for confounding variables, the 20- to 22-week previable gestational age group had the highest association (adjusted RR 6.2, 95% CI 5.5–6.9) with the composite adverse outcome compared with the referent term group (Table 4). A sensitivity analysis was performed to minimize the influence of cesarean delivery on maternal morbidity because many of the complications evaluated were plausibly linked to operative risks (Table 3). After excluding women who underwent a cesarean delivery, we found statistically significant and persistent associations between maternal complications and periviable birth including chorioamnionitis (adjusted RR 10.4), transfusion (adjusted RR 6.1), unplanned hysterectomy (adjusted RR 8.4), unplanned operation after delivery (adjusted RR 10.0), and ICU admission (adjusted RR 6.7). This analysis was also stratified by 3-week epoch gestational age groups, showing an inverse relationship of maternal outcomes with advancing gestational age at birth (Table 4). In addition, we performed a separate analysis limited to those who were most likely to represent spontaneous preterm birth (excluding those with medical indications for delivery such as preeclampsia, fetal growth restriction, and labor inductions) and results were not substantially different (Table 4). Although there is clearly a strong link between cesarean delivery and maternal morbidity, we found even higher risks of maternal complications among women who delivered vaginally within the periviable period compared with term vaginal deliveries.
In this study, we identified higher rates of maternal complications among women who underwent a periviable delivery. Women who delivered within the 20- to 22-week group had the highest composite adverse outcome rate and more than a sixfold increased risk in morbidity compared with term pregnancies. Likewise, the rate and risk of maternal complications were inversely related to gestational age when analyses were stratified by weeks of gestational age at birth. Previous studies on this topic are limited in scope, often excluding nonviable gestational ages (less than 23 weeks of gestation) and focusing primarily on women undergoing cesarean delivery at early gestational ages. Our results are nonetheless similar to these previous smaller retrospective studies of maternal complications associated with premature birth, but provide more detailed analysis of observed rates and risks of maternal complications when birth occurs at the threshold of viability.
Reddy et al7 evaluated a cohort of women delivering between 23 and 33 weeks of gestation and found composite complication rates of nearly 9%, particularly in births occurring between 23 and 27 weeks of gestation. The complications included hemorrhage, infection, ICU admission, and death. Kawakita et al6 studied (2005–2014) maternal outcomes associated with cesarean delivery after stratifying by gestational age and found ICU admission rates of 1.8–3.2% between 23 and 27 weeks of gestation with transfusion rates of 10.0–11.9%, but again this study was limited to women who underwent cesarean delivery only. Recently, Berholdt et al16 compared cesarean deliveries between 24 and 25 weeks of gestation and 26–27 weeks of gestation and found higher composite intraoperative complications (63.5 vs 30.8%, adjusted odds ratio [OR] 5.04) including classic incision, transplacental incision, difficulty in delivering the fetus, postpartum hemorrhage, and injury to internal organs in the 24- to 25-week group. Luthra et al17 (2002–2009) found transfusion rates of 4.4–8.6% and high rates of postpartum antibiotic utilization (20–28%) for presumed infection morbidity in women delivering by cesarean between 23 and 34 weeks of gestation at their medical center.
Not only do periviable births place the mother at greater risk in the index pregnancy, but they also increase her risk in future pregnancies.18 Lannon et al demonstrated an increased risk of uterine rupture (1.8%) in subsequent pregnancy regardless of uterine incision type for those who underwent a periviable cesarean delivery and even higher (2.5%) in those women with a classical incision.4,5 These risks are concerning because there is only limited retrospective evidence suggesting short-term benefit in performing cesarean deliveries for breech periviable pregnancies. Tucker Edmonds et al19 performed a retrospective cohort of periviable (23–24 weeks of gestation) pregnancies complicated by breech-presenting neonates and found cesarean-born neonates had overall improved survival (adjusted OR 2.91), but among intubated neonates, despite a short-term survival advantage, there was no difference in survival to greater than 6 months corrected age. In addition, the morbidity associated with repeat cesarean delivery is a growing concern in the United States as highlighted in the Safe Prevention of the Primary Cesarean Delivery initiative.20,21 In our study, 46% of the population was primiparous. More than 90% of women who undergo a primary cesarean delivery will have a subsequent cesarean delivery, and this rate is likely higher in this population of women, who are inherently at greater risk of having vertical incisions.21,22
Limitations to a study using vital statistics data include potential inaccuracies in reporting of data variables. Pregnancy complications and medical comorbidities have been shown to be underreported23; however, gestational age appears to be a highly reliable variable.24 In this study, we found lower rates of transfusion and ICU admission in our periviable cohort compared with previous studies in an early preterm cohort (23–27 weeks of gestation), which may be a result of underreporting. Cesarean delivery and hysterectomy rates were similar, however. There are ongoing efforts by the Ohio Perinatal Quality Collaborative to improve data accuracy of the birth certificate in Ohio and multiple publications regarding these efforts.25–28 Although the complications we studied are all plausibly associated with cesarean delivery, we performed a separate analysis excluding women who underwent cesarean delivery and still found strong associations with chorioamnionitis, transfusion, ICU admission, hysterectomy, and need for additional operations among women who delivered a periviable neonate. Similar to the observational studies discussed previously, the retrospective nature of our study precludes the determination of causality for periviable birth on the outcome of maternal morbidity.
Nearly one in five women in this cohort had a serious morbidity associated with their periviable delivery. This association remained regardless of mode of delivery. Obstetric and neonatal care providers as well as their patients should be cognizant of these risks when outlining a treatment plan for the management of threatened periviable delivery.
1. Lemyre B, Daboval T, Dunn S, Kekewich M, Jones G, Wang D, et al. Shared decision making for extremely preterm birth. J Perinatol 2016;36:503–9.
2. Periviable birth. Obstetric Care Consensus No. 6. American College of Obstetricians and Gynecologists. Obstet Gynecol 2017;130:e187–99.
3. Raju TN, Mercer BM, Burchfield DJ, Joseph GF Jr. Periviable birth: executive summary of a joint workshop by the Eunice Kennedy Shriver National Institute of Child Health and Human Development, Society for Maternal-Fetal Medicine, American Academy of Pediatrics, and American College of Obstetricians and Gynecologists. Am J Obstet Gynecol 2014;210:406–17.
4. Lannon S, Guthrie KA, Reed SD, Gammill HS. Mode of delivery at periviable gestational ages: impact on subsequent reproductive outcomes. J Perinat Med 2013;41:691–7.
5. Lannon SM, Guthrie KA, Vanderhoeven JP, Gammill HS. Uterine rupture after periviable cesarean delivery. Obstet Gynecol 2015;125:1095–100.
6. Kawakita T, Reddy UM, Grantz KL, Landy HJ, Desale S, Iqbal SN. Maternal outcomes associated with early preterm cesarean delivery. Am J Obstet Gynecol 2016;216:312.e1–9.
7. Reddy UM, Rice MM, Grobman WA, Bailit JL, Wapner RJ, Varner MW, et al. Serious maternal complications after early preterm delivery (24-33 weeks' gestation). Am J Obstet Gynecol 2015;213:538.e1–9.
8. Crane JG, Magee LA, Lee T, Synnes A, von Dadelszen P, Dahlgren L, et al. Maternal and perinatal outcomes of pregnancies delivered at 23 weeks' gestation. J Obstet Gynaecol Can 2015;37:214–24.
11. Martin JA, Osterman MH, Kimeyer SE, Gregory EC. Measuring gestational age in vital statistics data: transitioning to the obstetric estimate. Natl Vital Stat Rep 2015;64:1–20.
12. Obesity: preventing and managing the global epidemic: report of a WHO consultation. World Health Organ Tech Rep Ser 2000;894:i–xii, 1–253.
13. Peduzzi P, Concato J, Feinstein AR, Holford TR. The importance of events per independent variable in multivariable analysis. II. Accuracy and precision of regression estimates. J Clin Epidemiol 1995;48:1503–10.
14. Peduzzi PN, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol 1996;49:1373–9.
15. Szklo M, Nieto F. Identifying noncausal associations: confounding. In: Epidemiology: beyond the basics. Burlington (MA): Jones & Bartlett Learning; 2014. p.159–60.
16. Berholdt C, Menard S, Delorme P, Lamau MC, Goffinet F, Le Ray C. Intraoperative adverse events associated with extremely preterm cesarean deliveries. Acta Obstet Gynecol Scand 2018;97:608–14.
17. Luthra G, Gawade P, Starikov R, Markenson G. Uterine incision-to-delivery interval and perinatal outcomes in transverse versus vertical incisions in preterm cesarean deliveries. J Matern Fetal Neonatal Med 2013;26:1788–91.
18. Patterson LS, O'Connell CM, Baskett TF. Maternal and perinatal morbidity associated with classic and inverted T cesarean incisions. Obstet Gynecol 2002;100:633–7.
19. Tucker Edmonds B, McKenzie F, Macheras M, Srinivas SK, Lorch SA. Morbidity and mortality associated with mode of delivery for breech periviable deliveries. Am J Obstet Gynecol 2015;213:70.e1–12.
20. Safe prevention of the primary cesarean delivery. Obstetric Care Consensus No. 1. American College of Obstetricians and Gynecologists. Obstet Gynecol 2014;123:693–711.
21. Spong CY, Berghella V, Wenstrom KD, Mercer BM, Saade GR. Preventing the first cesarean delivery: summary of a joint Eunice Kennedy Shriver National Institute of Child Health and Human Development, Society for Maternal-Fetal Medicine, and American College of Obstetricians and Gynecologists Workshop. Obstet Gynecol 2012;120:1181–93.
22. Martin JA, Hamilton BE, Ventura SJ, Osterman MJ, Kirmeyer S, Mathews TJ, et al. Births: final data for 2009. Natl Vital Stat Rep 2011;60:1–70.
23. Reichman NE, Schwartz-Soicher O. Accuracy of birth certificate data by risk factors and outcomes: analysis of data from New Jersey. Am J Obstet Gynecol 2007;197:32.e1–8.
24. Dietz PM, Bombard JM, Hutchings YI, Gauthier JP, Gambatese MA, Ko JY, et al. Validation of obstetric estimate of gestational age on US birth certificates. Am J Obstet Gynecol 2014;210:335.e1–5.
27. Lannon C, Kaplan HC, Friar K, Fuller S, Ford S, White B, et al. Using a state birth registry as a quality improvement tool. Am J Perinatol 2017;34:958–65.
© 2018 by the American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
28. Kaplan HC, King E, White BE, Ford SE, Fuller S, Krew MA, et al. Statewide quality improvement initiative to reduce early elective deliveries and improve birth registry accuracy. Obstet Gynecol 2018;131:688–95.