Induction of labor with prostaglandins for medical reasons: Determining explanatory variables of the induction to delivery time interval for vaginal deliveries and caesarean section

Abstract

Objective Aim of the study was to examine the influence of the various medical indications on a successful induction of labor by prostaglandins and to determine the explanatory variables. Study design Retrospective analysis of 729 pregnancies, which fulfilled following criteria: only one medical indication for induction, 37 completed weeks of gestation or more, alive singleton pregnancy with cephalic presentation and induction with prostaglandins locally. The chosen endpoint was delivery. Results The Kaplan-Meier curves demonstrating the percentage of deliveries in function of the time interval from induction to delivery showed a significant longer time interval for those without prelabor rupture of membranes (PROM) than for those with PROM as indicated by log rank testing (vaginal deliveries and caesarean sections: hazard ratio = 0.67, 95% CI = 0.48–0.82, P < 0.001; vaginal deliveries only: hazard ratio = 0.65; 95% CI = 0.45–0.80; P < 0.001). In a similar way, those without diabetes had a shorter induction to delivery time interval than those with diabetes when all deliveries, including caesarean sections, were considered (hazard ratio = 1.59; 95% CI = 1.05–2.06; P = 0.02), but there was a “borderline missed” statistical difference when only vaginal deliveries were considered (hazard ratio = 1.48; 95% CI = 0.96–2.03; P = 0.08). Subsequently, univariate analysis in a Cox proportional hazards regression model was used to identify possible explanatory variables of the outcome, followed by multivariate analysis using the Cox proportional hazards regression model again in order to determine the independent contribution of each of these variables to the outcome. Uni- and multivariate analysis showed the cervix score, parity, the number of applied vaginal prostaglandin tablets, the gestational age and PROM to be significant explanatory variables of the induction to delivery time interval. Diabetes, body mass index and body weight were related to the outcome in a univariate analysis, but after adjustment in a multivariate analysis these variables were not significant. Conclusion PROM was the only medical reason to influence the outcome of an induction with prostaglandins locally, other significant explanatory variables were the cervix score, parity, the number of applied prostaglandin tablets and the gestational age, whereas other factors, such as diabetes, body mass index and body weight had no significant influence.