Predictors of perinatal outcomes and economic costs for late-term induction of labour.

Objective: We aimed to predict the perinatal outcomes and costs of health services following labour induction for late-term pregnancies. Materials and methods: We conducted a cohort study of 245 women who underwent labour induction during their 41st week of gestation. The cervical condition was assessed upon admission using the Bishop score and ultrasound cervical length measurements. We estimated the direct costs of labour induction, and a predictive model for perinatal outcomes was constructed using the decision tree analysis algorithm and a logit model. Results: A very unfavourable Bishop score at admission (Bishop score <2) (OR, 3.43 [95% Cl, 1.77-6.59]), and a history of previous caesarean section (OR, 7.72 [95% CI, 2.43-24.43]) or previous vaginal delivery (OR, 0.24 [95% CI, 0.09-0.58]) were the only variables with predictive capacity for caesarean section in our model. The mean cost of labour induction was (sic)3465.56 (95% confidence interval [CI], 3339.53 -3591.58). Unfavourable Bishop scores upon admission and no history of previous deliveries significantly increased the cost of labour induction. Both of these criteria significantly predicted the likelihood of a caesarean section in the decision tree analysis. Conclusion: The cost of labour induction mostly depends on the likelihood of successful trial of labour. Combined use of the Bishop score and previous vaginal or caesarean deliveries improves the ability to predict the likelihood of a caesarean section and the economic costs associated with labour induction for late-term pregnancies. This information is useful for patient counselling. (C) 2017 Taiwan Association of Obstetrics & Gynecology. Publishing services by Elsevier B.V.