Abstract
Background. Induction of labor (IOL) is among the most common obstetric interventions, but a meaningful minority of nulliparous women undergoing IOL ultimately deliver by cesarean. Locally relevant prediction tools, built from variables routinely available before induction, support patient counselling and protocol design. Objective. To identify independent clinical, anthropometric, and ultrasonographic predictors of failed IOL defined as cesarean delivery during an induction attempt among nulliparous women at a single tertiary center, and to compare a multivariable model with the modified Bishop score alone. Patients and methods. A retrospective cohort study was conducted at Bint al-Huda Hospital from January 2023 through December 2024 (24 months). Eligible records were those of nulliparous women with singleton, term (≥ 37 weeks), cephalic-presenting pregnancies undergoing IOL. Multiparous women, multiple pregnancies, prior cesarean delivery, intrauterine fetal death, and major fetal anomalies were excluded. The primary outcome was failed IOL. Univariable comparisons used the chi-squared, Fisher exact, or Mann–Whitney U test as appropriate. Multivariable logistic regression identified independent predictors. Discrimination was assessed by the area under the receiver operating characteristic curve (AUC). Results. Of 612 eligible inductions, 145 (23.7%) ended in cesarean delivery. Independent predictors of failed IOL were modified Bishop score ≤ 5 (adjusted odds ratio [aOR] 3.18, 95% confidence interval [CI] 2.18–4.62), pre-pregnancy body mass index (BMI) ≥ 30 kg/m² (aOR 2.55, 95% CI 1.66–3.92), sonographic cervical length ≥ 30 mm on transvaginal scan (aOR 2.11, 95% CI 1.34–3.32), maternal age ≥ 35 years (aOR 1.89, 95% CI 1.21–2.95), estimated fetal weight ≥ 3,800 g (aOR 1.96, 95% CI 1.27–3.02), hypertensive disorder of pregnancy (aOR 1.78, 95% CI 1.10–2.88), and maternal height < 155 cm (aOR 1.69, 95% CI 1.04–2.74). The combined model achieved an AUC of 0.82 (95% CI 0.78–0.86); the modified Bishop score alone achieved 0.76 (95% CI 0.71–0.80). Conclusions. Seven readily available pre-induction variables discriminated failed IOL with good performance. The combined model modestly improved on the bishop score alone and supports structured pre-induction counselling.
Keywords
bishop score
cervical length
cesarean delivery
failed induction
Induction of labor
nulliparous women
obstetric risk prediction