We have read with interest the article by Glavind et al. published in your last edition. The study concluded that elective caesarean section scheduled before 39 weeks of gestation carried a similar risk of neonatal intensive care unit admission, compared with those scheduled after 39 weeks of gestation.
Some points, however, need to be addressed. The authors discussed only briefly the high observed neonatal intensive care unit (NICU) admission rate, and did not compare this with previously published data where NICU admission rates were 8.1, 5.9, and 4.8% at 38, 39, and 40 weeks of gestation, respectively. A 13.9% rate of NICU admissions and an 8.7% rate of continuous positive airway pressure (CPAP) treatment among low-risk neonates seems unacceptable.
With the available evidence in favour of vaginal birth for women without a clear indication for caesarean section, and with observational data indicating higher neonatal risks for early elective delivery (before 39 weeks of gestation), including higher mortality rates, it is surprising that a randomised cotrolled trial allocating women to elective caesarean section at 38 versus 39 weeks of gestation was deemed ethically appropriate. Reasons stated by the authors include a potential selection bias in previous cohort studies, which is likely to have led to an over-representation of high-risk neonates in the early-delivery group; however, a proper statistical analysis of observational data could address this issue without conducting a study that raises ethical concerns, by exposing newborns and women to significant risks. It is already known that caesarean section at term for non-medical reasons affects overall maternal morbidity rates, and increases the mortality rate by three to five times, when compared with vaginal birth. A total of 1274 women (243 nulliparous) had caesarean sections without precise indication in the study.
The authors affirmed their intention to ‘fully investigate neonatal and maternal benefits or adverse events associated with elective caesarean section timing’, which cannot be achieved without a comparison group (i.e. spontaneous labour and vaginal birth, a mode of delivery with evidence showing better neonatal and maternal outcomes). If the authors insist on merely making comparisons between different timings of caesarean section, current evidence would point towards a comparison between elective caesarean at 39 and 40 weeks of gestation.
The sample size was not sufficient for most of the outcomes examined, and the authors reported that the expected difference employed to calculate the sample was abstracted from a cohort study; however, these data were not available in the published version.
Figures 2 and 3 of the original paper indicate a slightly greater number of births before 38 weeks of gestation in the group assigned to elective caesarean at 39 weeks of gestation, and it is reasonable to question if this adversely interfered with the neonatal outcomes for this group.
It is also important to examine the potential impact of these conclusions on public health, particularly in settings where caesarean section rates are rising. Despite a statistically insignificant difference (2%) in the primary outcome, we could consider this magnitude as the incremental risk of elective caesarean at 38 weeks of gestation to project possible scenarios following the adoption of this practice. In Brazil, considering that 35% of caesareans performed are elective, we could estimate about 500 000 elective procedures each year. If 200 000 of these women have no clear medical indication for surgical birth, and were scheduled at 38 weeks of gestation, we could project 4000 neonates being admitted to NICUs unnecessarily.