Risk of caesarean section after induced labour: Do hospitals make a difference?

Abstract

Background: There is a well-known relationship between induced labour and caesarean rates. However, it remains unknown whether this relationship reflects the impact of more complex obstetric conditions or the variability in obstetric practices. We sought to quantify the independent role of the hospital as a variable that can influence the occurrence of caesarean section after induced labour. Methods. As part of the Portuguese Generation XXI birth cohort, we evaluated 2041 consecutive women who underwent singleton pregnancies with labour induction, at five public level III obstetric units (April 2005-August 2006). The indications for induction were classified according to the guidelines of the American and the Royal Colleges of Obstetricians and Gynaecologists. Poisson regression models were adjusted to estimate the association between the hospital and surgical delivery after induction. Crude and adjusted prevalence ratios (PR) and a 95% confidence interval (95% CI) were computed. Results: The proportion of women who were induced without formal clinical indications varied among hospitals from 20.3% to 45.5% (p < 0.001). After adjusting for confounders, the risk of undergoing a caesarean section after induced labour remained significantly different between the hospitals, for the cases in which there was no evident indication for induction [the highest PR reaching 1.86 (95% CI, 1.23-2.82)] and also when at least one such indication was present [1.53 (95% CI, 1.12-2.10)]. This pattern was also observed among the primiparous cephalic term induced women [the highest PR reaching 2.06 (95% CI, 1.23-2.82) when there was no evident indication for induction and 1.61 (95% CI, 1.11-2.34) when at least one such indication was present]. Conclusions: Caesarean section after induced labour varied significantly across hospitals where similar outcomes were expected. The effect was more evident when the induction was not based on the unequivocal presence of commonly accepted indications. © 2013 Teixeira et al.; licensee BioMed Central Ltd.