0624. Report of the 2005 WHO global survey on caesarean section rates and perinatal outcome

The questions around indications for and safety of caesarean section (c/s), and in particular the rising c/s rates around the world, are about to surface yet again as results of the WHO survey are published. The first report covering 24 regions in 8 Latin American countries appeared recently, and results for Africa are due. Surveys for Canada and Asia are under way. These large scale, statistically-sophisticated surveys hope to minimise the usual confounders viz. indication bias (reasons for the c/s are themselves associated with higher maternal and/or neonatal risk) and residual confounding by socio-economic status (c/s is commonest in affluent groups in which baseline risk is intrinsically low). The South American study included >100 000 deliveries, mostly in public and ‘social security’ hospitals, and with a smattering of private facilities. The median c/s rate was 33%, with the highest rates in private hospitals (53%). Rates in primigravidae were around 66% in both private and public facilities. The data suggest that indications for c/s were similar in private and public/social security facilities, but one should perhaps question this because while it is acknowledged that many private c/sections are performed in response to maternal choice (i.e. no specific medical necessity), few practitioners will record this as the indication (particularly if a third party insurer is involved). Overall the analyses show associations between c/s rates and maternal morbidity and mortality, and between c/s rates and fetal mortality and neonatal admission to ICU, and the authors conclude that high c/s rates do not necessarily indicate better perinatal care. In fact their data suggest that very low c/s rates are associated with fewer preterm deliveries and perinatal deaths. This is unlikely given that optimal obstetric care requires a baseline c/s rate to deal with fetal distress, disproportion, fetal or maternal disease etc, and failure to perform a c/s will have negative consequences for mother and infant. One must therefore ask whether facilities with such low c/s rates were low risk units that referred problems out. The authors suggest that a randomised trial would answer the question of whether c/s on demand is an acceptable alternative to c/s only when clinically indicated, but one wonders how researchers would possibly find subjects for such a trial.

Read more:
Lancet 2006; 367: 1819-29
BMJ 1999; 319: 1397-402
S Afr Med J 2005; 95: 257-60