Reply

The majority of parents report that investigation of stillbirth is important to obtain information about why their baby died and to inform future care1. Despite this, investigation is frequently incomplete; data from 2014 show that autopsy was performed in around 40% and histopathological placental examination in around 90% of perinatal deaths2. Concerns have also been raised about the variable quality of placental examination and recommendations have been made for this to be carried out by specialist pathologists3. The intention of the recent series of papers on the investigation of stillbirth as part of a special issue focusing on pregnancy loss was to encourage discussion of various components of investigation after stillbirth and to highlight areas of difficulty or controversy. Therefore, we are grateful to Cox et al. for their contribution, which raises several issues. With regard to the role of ascending infection, Cox et al. are of course correct that most cases with ascending infection were second-trimester losses; data presented in the original manuscript were categorized by gestational age (Table 1 of Man et al.4). These findings are similar to those of other groups investigating stillbirth and we agree that the significance of such findings in the third trimester in the absence of fetal involvement/response remains uncertain5. The study data were indeed focused on cause of death. Whilst we agree that documenting changes of uncertain significance may be of value for future research, we would argue that reporting associations of unknown significance to parents without due caveats may be unhelpful for counseling. Cox et al. cite the ReCoDe classification system6, which aims to identify factors potentially relevant or contributory to stillbirth. Conceptual issues around classification of stillbirths remain challenging and beyond the scope of these manuscripts, but a recent consensus study proposed that classification systems should focus on the main factor leading to death as recorded, with associated factors noted but clearly distinguished from the cause of death7. Interestingly, however, almost all of the classification categories of the ReCoDe system are derived from either clinical review or placental examination; indeed, in the main ReCoDe publication6, it states that the proportion of stillbirths that were unexplained was not significantly different regardless of whether a postmortem examination had been carried out (P = 0.3). Regarding histological examination of tissue samples, data were presented regarding determination of cause of death in addition to data on all histological ‘abnormalities’, which were much more common than those directly contributory to the cause of death8. Interpretation of the value of such findings is subjective. Whilst concepts such as ‘exclusion of disease’ are indeed philosophically appealing, in practice, data are required regarding the practical utility and opportunity cost of this approach; the relevance of such findings appears