Objective: To compare a policy of elective caesarean delivery with a policy of selective caesarean delivery for women in spontaneous preterm labour.
Design: Systematic review of relevant randomised controlled trials.
Participants: One hundred and twenty-two women from six trials.
Interventions: Random allocation to "intention to deliver by caesarean section' (elective caesarean policy), compared with "intention to deliver vaginally' (selective caesarean policy).
Main outcome measures: Actual use of caesarean delivery, low Apgar score, neonatal intubation, intracranial pathology, perinatal death, and major maternal complications.
Results: One in six women in both groups were not delivered as allocated. Typical odds ratios for all adverse neonatal outcomes tended to favour the elective group, but for all outcomes studied the confidence intervals were wide and did not exclude clinically important hazards of the policy. The odds of serious maternal morbidity were increased in the caesarean section group (OR 6.2; 95% CI 1.3-30.1).
Conclusion: A policy of elective caesarean delivery increases the risks of maternal morbidity, but it remains unclear whether these are offset by benefits for the infant.