Strengths and limitations
This study included large numbers of women with severe disease which enabled us to assess different phenotypes of pre-eclampsia. To perform assessments in women with severe disease such as eclampsia is difficult in high income countries as the incidence is low, around 50/110,000.25 Performing this study in our setting, where eclampsia is more prevalent, enabled us to recruit this cohort in a short time period. Validated tests which included both subjective and objective cognitive function assessments were used. We also assessed cognition before and after delivery using different methods. To our knowledge, this study is the first to examine cognitive function in close proximity to onset of pre-eclampsia.
Our study does have limitations. Women with pre-eclampsia with severe features may have been more tired at the time of cognitive function testing. We tried to correct for this by adjusting for “time from delivery to discharge” in the analyses. In addition, no testing for anxiety or depression was performed. Though, very few women reported symptoms suggestive of depression during pregnancy and we adjusted for gestational age as a proxy for maternal and neonatal complications that could be related to anxiety and depression postpartum. The group of women with pulmonary oedema was small, introducing the risk of a type 2 error in the analyses where the differences in MoCA test were attenuated after adjustment for confounders. Lastly, there was no follow up after discharge. Women who deliver at Tygerberg hospital often live in informal settlements and it is challenging for them to return for follow up visits due to social circumstances. Finally, women rated their cognitive function prior to diagnosis or delivery retrospectively when doing the CFQ assessment that could introduce bias. However, the CFQ is the most common test used in the field and our results compare to previously published studies.