We read with interest the article of Eller et al.1 which provides a detailed description of the characteristics and outcomes of patients with placenta accreta in a series where the main policy was caesarean hysterectomy. We have some concerns regarding the title of the article and the conclusion of the authors, which may lead the reader to believe that these results constitute evidence for defining management strategies for placenta accreta, and, more specifically, that preoperative ureteric stent placement is recommended or that hypogastric ligation is useless. In fact, the study design – an observational study without any adjustment for potential confounding factors – does not allow any conclusions to be drawn regarding the effectiveness of the management procedures used. The long study period and the inclusion of patients from two different centres make us doubt the consistency of the medical and surgical policies for management of placenta accreta used. In addition, the issue of a ‘selection by indication’ bias, a major concern in observational studies on the association between components of care and health outcomes, is not discussed although we consider it likely. The outcomes differences found between women who received different types of management may actually result from pre-existing differences in the characteristics of the patients.
Regarding ureteric stent placement, it is possible that only one centre, or only a few practitioners in one centre, used it during only a part of the 13-year study period. This strategy implies specific attention to ureteric damage by the surgical/obstetrical team. It is also possible that the characteristics of the women who had stents placed were different from those who did not. Therefore, the reported difference in early maternal morbidity may be related not only to this procedure but also to differences in the clinical context, the centre and/or the period.
We have similar methodological concerns related to the suggested ineffectiveness of hypogastric ligation. Hypogastric ligation is usually performed in cases of established postpartum haemorrhage. Therefore, using hypogastric ligation ‘prophylactically’, as stated in the discussion, is difficult to understand. We can speculate that ‘prophylactic’ ligation was performed after birth of the baby but before the delivery of the placenta. How can this treatment be prophylactic after delivery in cases of anterior placenta praevia associated with a caesarean scar, which is the most frequent reason for placenta accreta? Why did the practitioner choose to perform a hypogastric artery ligation? It seems likely that hypogastric artery ligation was only possible in particular clinical and technical contexts. Therefore, the two groups (with or without hypogastric ligation) should not be compared.
There are few data on the effectiveness of hypogastric arteries ligation in cases of postpartum haemorrhage,2,3 and those that exist are controversial. Eller et al. suggest that its efficiency may be lower in cases of placenta accreta. Other techniques such as selective embolisation may be as, or even more, efficient.4 Nonetheless, we consider that hypogastric ligation remains an option when trying to avoid hysterectomy, even in cases of post partum haemorrhage caused by placenta accreta.