Predictive models for long term survival after premature rupture of membranes

Premature rupture of the membranes complicates 30 to 40% of the preterm deliveries. The purpose of this study is to assess the outcome of preterm prelabor rupture of the membranes (PPROM) before 26 weeks of gestation and to develop a predictive model for survival. We apply four statistical methods in order to construct a predictive model: Cox's proportional hazards model, logistic regression, least squares - support vector machines and generalized estimating equations. The obtained predictive models are tested on a prospective set and have a very good performance of > 95% for the area under the ROC-curve.