Strengths and limitations
Limitations of our study need to be considered. First, inclusion of participants from our Kingston cohort required retrospective analysis of placentas that were originally submitted to Pathology for examination. This aspect of the study design may have resulted in selection bias, as although PE is an indication for pathology examination, not all placentas are sent.50 We did observe a significant difference in several important pregnancy parameters between our study sites, including blood pressure at delivery, anti-hypertensive use at delivery and gestational age at delivery. These differences may be due to placenta pathology referral practices, as it is likely that placentas from women with more severe PE disease are more frequently selected for submission to pathology (and thus available for inclusion in our study). As cardiovascular parameters were similar between our cohorts at the 6 month postpartum clinic visit, we do not feel these delivery parameters influenced our findings. Due to our small sample size, we may be underpowered to detect differences between low and high-risk groups for less common pathology, such as chronic inflammation. Confirming our results in an adequately powered prospective study may identify a predictive combination of placental lesions that robustly identifies high priority women for postpartum CVD screening at the time of delivery. Our study is strengthened by the standardised placental evaluations we conducted using our evidence-based synoptic data collection.33 Variability in placenta pathology examination is a known challenge and the use of this synoptic collection form ensures each placenta was evaluated in a reproducible manner.