Strengths and limitations
The strengths of the study are: first, large sample size with prospectively collected data, second, focus on placental dysfunction related stillbirths, rather that treating all stillbirths as a homogeneous condition, and third, comparison of the predictive performance of two of our models that were previously internally validated1,7. We acknowledge the prerequisite for external validation to support generalization of our results and wide implementation of our model. Such external validation would require a large prospective multicenter study.
It is possible that in some cases the birthweight of the stillborn babies is lower than the weight at the time of death because there is a relationship between intrauterine retention interval and reduction in birthweight34. In our cases we did not have information on this interval and therefore the incidence of placental dysfunction related stillbirths maybe overestimated.
A Some of the risk factors included in the RCOG guideline for the prediction of SGA were not included in the competing risks model for SGA because we did not have such risk factors for any or some of our patients. For example, we did not have data on low fruit intake before pregnancy, paternal SGA, daily vigorous exercise, heavy bleeding similar to menses, or notching of the uterine artery Doppler waveforms, but these factors may well suffer from subjectivity or information bias. Similarly, we did not have available data on PAPP-A for all of our patients and did not use the criterion of <0.4 MoM for assessment of risk; in a previous study we reported that inclusion of PAPP-A as a binary variable (<0.4 MoMs) increases the screen positive rate without any significant improvement in the detection rate.33