Methods of induction of labor and women's experience: a population-based cohort study with mediation analyses

International audienceBackground: Negative childbirth experience may affect mother wellbeing and health. However, it is rarely evaluated in studies comparing methods of induction of labor (IoL). Aim: To compare women's experience of IoL according to the method, considering the mediating role of interventions and complications of delivery. Methods: We used data from the MEDIP prospective population-based cohort, including all women with IoL during one month in seven French perinatal networks. The experience of IoL, assessed at 2 months postpartum, was first compared between cervical ripening and oxytocin, and secondarily between different cervical ripening methods. Mediation analyses were used to measure the direct and indirect effects of cervical ripening on maternal experience, through delivery with interventions or complications. Findings: The response rate was 47.8% (n = 1453/3042). Compared with oxytocin (n = 541), cervical ripening (n = 910) was associated less often with feelings that labor went 'as expected' (adjusted risk ratio for the direct effect 0.78, 95%CI [0.70-0.88]), length of labor was 'acceptable' (0.76[0.71-0.82]), 'vaginal discomfort' was absent (0.77[0.69-0.85]) and with lower global satisfaction (0.90[0.84-0.96]). Interventions and complications mediated between 6 and 35% of the total effect of cervical ripening on maternal experience. Compared to the dinoprostone insert, maternal experience was not significantly different with the other prostaglandins. The balloon catheter was associated with less pain. Discussion: Cervical ripening was associated with a less positive experience of childbirth, whatever the method, only partly explained by interventions and complications of delivery. Conclusion: Counselling and support of women requiring cervical ripening might be enhanced to improve the experience of IoL