Introduction
In 2012, national guidelines were published in the United States with the goal of reducing the first or “primary” cesarean birth.1,2 These guidelines considered a broad range of opportunities to reduce cesarean birth but focused predominantly on strategies to reduce the rate of primary cesarean in labor (“intrapartum cesarean”). The contribution of pre-labor cesarean, cesarean without an attempt at labor or vaginal birth, has received much less attention.
Recently Hehir et al.3 used US vital statistics birth certificate data and the Robson 10-Group Classification System4 to examine cesarean birth rates in the US between 2005 and 2014. Cesareans identified in the birth certificate as being performed without a trial of labor were assigned as “pre-labor”. The study suggested pre-labor cesareans in singleton term cephalic pregnancies with no history of cesarean constituted 27% of all cesareans in the US. Additionally, studies in Ireland5and Australia6 identified increases in both repeat cesarean and pre-labor cesarean in singleton term cephalic nulliparas as contributing to an increase in the total cesarean rate between 2005 and 2014.
The goal of our study was to evaluate the contribution of cesarean performed without attempting vaginal birth to all cesareans, and to cesareans within subgroups of the pregnant population. Additionally, stimulated by the studies by Hehir et al.3, we wanted to evaluate the rate and indications for cesarean without attempting vaginal birth in singleton term cephalic pregnancies with no history of cesarean.
Methods :
This retrospective cohort study included births of 23+0-42+6 weeks’ gestation at 15 hospitals (5 Level I, 4 Level II, and 6 Level III/IV) in the Pacific Northwest of the US between January 1, 2017 and March 31, 2018. Intra-partum transfers from other hospitals or planned community-based birth settings were excluded.
The data for the study came from the Foundation for Health Care Quality’s Obstetrical Care Outcomes Assessment Program (OB COAP), an ongoing multicenter, clinician-led, perinatal quality collaborative that has been previously described in detail.7 The program collects data from the medical records of consecutive births> 20 weeks’ gestation at hospitals in urban, suburban, and rural settings. Trained abstractors at each site enter the data into a cloud-based database. Subsets of variables are also uploaded electronically from electronic medical records at some sites. Data undergo real-time data quality and validation checks, both at the site and aggregate level. Ad hoc quality checks are also conducted on a routine basis by OB COAP staff.   Patients were not involved in the study.
Cesareans performed without attempting vaginal birth in the OB COAP database are defined as cesareans where “surgical delivery has been previously planned (planned primary or repeat cesarean) or if [the] patient presents with an immediate need to deliver surgically (unexpected maternal or fetal compromise upon presentation)”. These cesareans were assigned as “pre-labor”, although it is likely that some of these women were in labor at the time of initial presentation to hospital.
Characteristics of the study population were explored including demographic, pre-pregnancy, and pregnancy characteristics, and hospital level of neonatal care (I – III/IV). Fisher’s exact test was used to test for association between demographic characteristics and whether vaginal birth was attempted.
Births in the study population were then grouped using the Robson 10-Group Classification System,4 described in detail on the World Health Organization website.8 The Robson classification categorizes all births in the population into ten mutually exclusive groups based on parity, type of labor (spontaneous or induced), previous cesarean birth, presentation, plurality, and gestational age at birth (term or preterm) (Table 1).
The contribution of each of the Robson classification groups to all births in the study population, the cesarean birth rate in each group, the contribution of cesareans in the group to all births (“absolute contribution”), and the contribution of cesareans in the group to all cesarean births (“relative contribution”), were calculated. In addition, we calculated the percentage of cesareans in each Robson group that were pre-labor cesareans. For multiple pregnancies, cesarean rates were based on the birth of the presenting baby (Baby A). Births that could not be assigned to a Robson group because of missing data were categorized as “non-classifiable”. The rate of pre-labor cesarean in the nulliparous term singleton vertex (NTSV) population (Robson groups 1, 2a, and 2b), the standardized metric for cesarean birth in nulliparas in the US, was additionally examined.
The primary cesarean (first cesarean irrespective of parity) rate cannot be calculated directly from the Robson classification because some of the Robson groups (7, 8, 9 and 10) include patients with and without previous cesarean births. As a result, we separately calculated the rate of primary cesarean, the percentage of all births that were pre-labor cesareans, and the percentage of all cesareans that were pre-labor in women with no history of cesarean birth.
Indications for pre-labor cesareans in singleton term cephalic pregnancies with no history of cesarean (Robson groups 2b and 4b) were also examined. The “indication for cesarean” field in the OB COAP database includes the primary indication for cesarean stated in the surgeon’s operative note. If a patient had a cesarean planned but presented in labor with or without ruptured membranes, data abstractors were directed to enter the original indication for the planned cesarean in the “indication for cesarean” field. The “indication for cesarean” field includes 20 specific indications for cesarean birth and the option to choose “Other” if the indication does not fall into one of the specified indications. For cesareans that fall into the “Other” category, the indication is entered into a separate free text field. Indications (including those in the free text field) were assigned by one of the study authors (VS) into groups including maternal medical condition, concern for fetal status, placental or cord complication, fetal abnormality, pre-eclampsia/gestational hypertension, maternal request, suspected macrosomia, suspected cephalopelvic disproportion, complication in previous pregnancy, previous uterine surgery, active genital herpes, other, and not known.
Descriptive statistics were calculated using R version 3.4.2.