A mathematical model to predict mean time to delivery following cervical ripening with dinoprostone vaginal insert

The main objective of our study was to analyze the mean time to delivery following cervical ripening with a 10 mg dinoprostone vaginal insert. We performed a retrospective observational study at the level III maternity ward of Angers university hospital. We included all women who had cervical ripening with dinoprostone between January 1^st, 2015 and September 30^th, 2016. Overall, 405 patients were included, and 59.3% (240/405) were nulliparous. The mean time to delivery was 20h39 min ± 10h49 min. 21% of deliveries (86/405) occurred between midnight and 6 h a.m., and the cesarean section rate was 33% (132/405). Multiple regression analysis showed that nulliparity, overweight (BMI ≥ 25), a closed cervix on initial examination and the absence of premature rupture of membranes (PRM) all significantly increased the mean time to delivery. We developed a mathematical model integrating the aforementioned factors and their impact to help predict the mean time to delivery following cervical ripening with dinoprostone vaginal insert: Y = 961.188-80.346 × parity + 21.437 × BMI-165.263 × cervical dilation-241.759 × PRM. This equation allows obstetricians to calculate a personalized time to delivery for each patient, allowing a precise scheduling of dinoprostone insert placement, and thus improving the organization in busy maternity wards.