Medical interventions during pregnancy and childbirth may not be the best way to improve newborn survival

Medical interventions during pregnancy and childbirth may not be the best way to improve newborn survival in middle-income countries, suggests a study published online today by The Lancet.

In Brazil neonatal mortality declined in the 1980’s but rates have virtually remained unchanged since the 1990’s. Fernando Barros (Hospital de Clinicas, Montevideo, Uruguay) and colleagues studied births in 1992, 1993, and 2004 in the Brazilian city of Pelotas. They found that important changes in maternal characteristics occurred over the 22 years covered by the study. Mothers were on average 4·5 cm taller, 5·1kg heavier, smoked less and were more educated than those who gave birth in 1982. The women’s mean number of antenatal visits increased as did the proportion whose initial visit was in their first trimester. 97% of all women had at least one ultrasound scan and 31% had three or more scans during pregnancy. Other antenatal procedures were less frequent: 23% (543 of 2373) of women had no vaginal examination and 68% (1189 of 1748) of women, not previously immunised against tetanus, reported receiving one or more doses of this vaccine. The percentage of induced deliveries increased from 2.5% (147 of 5914) in 1982 to 43% (1026 of 2386) in 2004. Rates of caesarean section increased greatly from 28% (1632 of 5914) in 1982 to 43% (1039 of 2403) in 2004, reaching 82% (374 of 456) of all private deliveries in 2004.

The mean birthweight of babies decreased and the proportion of children lighter than 2500 g rose, primarily due to an increase in pre-term births. The prevalence of pre-term births increased from 8·5% (437 of 5139) of babies in 1993 to 13·5% (316 of 2340) in 2004. The authors suggest that the increase in preterm birth rates might have resulted from either acceleration of delivery thorough induction or caesarean section of foetuses whose gestational age has been overestimated by inaccurate ultrasound.

Dr Barros comments: “The existence of three successive cohorts has enabled us to understand the relatively small change in neonatal mortality since 1990 was due to a combination of increased rates of low birthweight and preterm deliveries with improved care of neonates. This information presents new challenges for the local and national policymakers, because in order to reach the Millennium Development Goal for child survival, strategies will have to be implemented to improve birthweight and reduce preterm deliveries.”

In an accompanying commentary Anthony Costello (Institute of Child Health, UK) and colleagues conclude: “In middle income countries with high institutional usage, a focus on quality of perinatal care is a priority, and health outcomes must be monitored rigorously. Doing the simple things better is probably the most cost-effective policy: increasing coverage of syphilis screening, making sure that unimmunised women receive tetanus toxoid, and careful monitoring in labour. As Barros and colleagues show, the risks of medicalisation should not be ignored because they might offset the gains resulting from improved maternal health and survival of newborns.”