Interpretation
The rates of pre-labor and laboring cesarean in our data are very similar to those in a 2016 hospital birth population in Sydney, Australia6. However, comparable studies on the contribution of primary cesarean without attempting vaginal birth to the total cesarean birth rate in the US are limited and are based on births that pre-date the 2014 ACOG guidance aimed at reducing cesarean birth.2
Our observation thatprimary pre-labor cesareans in singleton term cephalic pregnancies, constitute 6.6% of all cesareans births, is similar to the 7.9% rate reported by Zhang et al.9 using chart abstracted data on births between 2002 and 2008 from the Consortium on Safe Labor cohort. Joesch et al.10, using administrative (procedural code) data, reported that between 1979 and 2004 the rate of primary pre-labor cesarean as a proportion ofall live births varied between 4.5% and 6.6% which is similar to the 6.7% observed in our study.
Huesch et al.11, reported variation in pre-labor cesarean rates among hospitals in their study and highlighted the pre-labor cesarean rate as a potentially important metric. However, the heavy reliance on administrative data for quality improvement in maternity care in the US is a barrier to using pre-labor cesarean as a metric. Pre-labor and laboring cesarean are not easily distinguishable in administrative data and rely on the assumption that if there is an ICD procedural code for cesarean in the absence of a code consistent with labor, pre-labor cesarean was the mode of delivery. This involves searching for multiple procedural codes and may be subject to reliability issues.
National birth certificates provide another data source for evaluating cesarean birth in the US. However, the finding by Hehir et al3 that pre-labor cesareans in Robson groups 2b and 4b (singleton term cephalic pregnancies with no history of cesarean), constitute 27% of all cesarean births when using US Birth Certificate data compared to 6.6% in our data raises the possibility of misclassification in the US national birth certificate data.
On examining indications for cesarean in Robson groups 2b and 4b in our clinical dataset, the rate of pre-labor cesarean without a medical indication in the nulliparous term cephalic singleton population was low 0.6%, suggesting that cesarean for maternal request is uncommon in first time mothers. This is consistent with other authors.12 The finding that suspected macrosomia was the most common indication for primary cesarean without attempting vaginal birth in singleton term cephalic pregnancies was unexpected and may be an increasingly common indication.13 A study from France reporting on potentially avoidable planned cesarean births also reported suspected macrosomia to be one of the more common stated indications.14 ACOG does not consider suspected macrosomia to be an indication for cesarean birth unless the estimated fetal weight is >4500g in a diabetic woman or >5000g in a non-diabetic woman, when “prophylactic cesarean delivery may be considered”.15 It is therefore of concern that we found birth weight was <4500g in more than 3/4 of cesareans performed without attempting vaginal birth for suspected macrosomia in singleton term cephalic nulliparous pregnancies. Our findings are consistent those of Boyle et al.16 who reported 80.3% of primary cesareans performed for fetal macrosomia were associated with an actual birth weight of <4500g.
The basis for “suspected macrosomia” could not be determined from the OB COAP database but it seems likely that ultrasound-derived estimated fetal weights contributed in many cases. A “large for dates” diagnosis based on ultrasound measurements has been associated with an increased risk for both pre-labor and intrapartum cesarean independent of the actual birth weight of the baby.17,18