High Cesarean Section Rates In Brazil Result in Large Part From Nonclinical Factors

D. Estrin

First published online:

The decision to perform a cesarean section is often strongly influenced by nonmedical factors, according to data from one city in southeastern Brazil. In Ribeirão Preto, in São Paulo State, the rate of cesarean section was 51% in 1994. The most important factors affecting mode of birth were related to health service delivery: The probability of delivery by cesarean section was elevated among women who had given birth between 7:00 a.m. and 7:00 p.m., had made four or more visits for prenatal care or had used the same physician for prenatal care and delivery.1

To examine risk factors associated with cesarean section, interviewers collected socioeconomic, demographic, reproductive and health service data from women having singleton births in maternity hospitals in Ribeirão Preto, São Paulo State, Brazil, in 1978-1979 and 1994. Information was obtained on 6,750 infants delivered in 1978-1979 and on 2,846 infants born in 1994.

Comparisons of the two samples indicated that women who delivered in 1994 were more likely than their counterparts in 1978-1979 to be employed outside the home, to be unmarried, to have had their first delivery before they were 20 or after they were 30, to be having their first birth or to have had up to two pregnancies, to have only one living child and to have made four or more prenatal visits. Between the two study periods, the rate of cesarean delivery rose from 30% to 51%, while the rate of preterm birth increased from 8% to 14% and the percentage of mothers whose baby weighed less than 3,000 g at birth rose from 28% to 37%. Infants born in 1994 were more likely than those born in 1978-1979 to have been delivered in private hospitals and between Monday and Wednesday.

In a multiple logistic regression analysis, the risk of a cesarean section in 1978-1979 was significantly higher among women who were aged 20 or older (odds ratios of 1.4-3.4 for various age-groups), had previously had a stillbirth (1.8) or had fewer than four living children (2.2-3.3); it was significantly lower among those having their third live birth than among those having their fourth (0.65). The risk of delivery by cesarean section was greater for infants with a birth weight of 3,500 g or more than for those weighing less than 3,000 g (1.4-2.0). Of the socioeconomic and demographic factors tested, four were associated with cesarean delivery--having had four or more years of education (odds ratios of 1.3-1.6), having delivered at a private rather than a public hospital (1.6-2.8), having delivered on any day other than Sunday (1.5) and having made four or more prenatal visits (1.7).

In 1994, the hour of delivery was a significant factor: Compared with infants born between 1:00 a.m. and 6:00 a.m., those born from 7:00 a.m. to 6:00 p.m. or from 7:00 p.m. to midnight were 3.4-4.7 times as likely to have been delivered by cesarean section. In addition, mothers who used the same physician for prenatal care and delivery were 2.5 times as likely to have had a cesarean section as were those who did not, and mothers who made four or more prenatal visits were 2.1 times as likely to have had a cesarean section as were those who made fewer visits. Women aged 30 or older were 2.7 times as likely to have had a cesarean as were those younger than 20, while women who had had two live births were 69% less likely than those with four or more births to have delivered by cesarean section. Having three living children was positively associated with operative delivery (odds ratios of 2.5-7.8). Cesarean deliveries were 50% more common for infants weighing 3,500-3,999 g at birth than for those weighing less than 3,000 g.

Finally, the investigators attempted to identify factors that played a major role in the increase in the cesarean section rate between the two study periods. They used adjusted population-attributable risks to compare the prevalence of each risk factor among women who had had an abdominal delivery in the two periods, as well as the variation of the odds ratio for each of those factors. Only three variables--having private insurance, being married and delivering in a private hospital--showed increases of 15% or more in population-attributable risk between the two periods.

The investigators speculate, based on the association of cesarean section with number of living children and maternal age, that many women may choose to have a cesarean section because in Brazil the procedure is an indication for tubal ligation. In addition, they say, the correlation of delivery timing with cesarean section suggests that physicians prefer cesarean to vaginal delivery because it allows them "to work a minimum of nonsocial hours." The investigators conclude that, in Brazil, nonclinical factors play a major role in the decision to perform a cesarean section. --D. Estrin


1. Gomes UA et al., Risk factors for the increasing cesarean section rate in southeast Brazil: a comparison of two birth cohorts, 1978-1979 and 1994, International Journal of Epidemiology, 1999, 28(4):687-694.