Pregnancy: Prognostic models for the probability of achieving an ongoing pregnancy after in-vitro fertilization and the importance of testing their predictive value

The aim of this study was to create reliable models to predict the probability of achieving an ongoing pregnancy during in-vitro fertilization (IVF) treatment: model A, at the start of the first treatment, model B at the time of embryo transfer, and model C, during the second treatment at the end of the first IVF treatment. Prognostic models were created using data from the University Hospital Nijmegen (n = 757) and applied to the data from the Catharina Hospital Eindhoven (n = 432), The Netherlands, to test their predictive performance. The predictions of model B (made at time of embryo transfer) were fairly good (c = 0.672 in the test population). For instance, 93% of the patients who had a predicted probability of achieving an ongoing pregnancy of < 10% did not achieve an ongoing pregnancy. However, the predictions of the other two models (A and C) for Eindhoven were less reliable. The predictive value of model C was fairly high in Nijmegen (c = 0.673). Its poor performance in the test population may be explained partly by differences in effectiveness of the ovulation stimulation protocols and the decision about when to discontinue the cycle. Thus, before using prognostic models at an IVF centre, their reliability at that specific centre should be tested.