Progress in the management of adverse pregnancy outcomes demands increased understanding of the relevant pathology and pathophysiology. Taking pre-eclampsia as an example, the study of this disorder has been challenging. It is a multisystem disease and, for the woman with the disorder who is ill, it is very difficult to identify the proximate pathophysiology. Pre-eclampsia occurs in 3-7% of pregnancies and, thus, the ideal approach — obtaining information in women prior to disease — requires a large number of subjects. Predictive tests would be quite useful to enrich populations for studies and trials but the low prevalence again demands large numbers of subjects for studies and a test with high specificity. In addition, the varied clinical presentation and very different outcomes in women with early vs. late or single occurrence vs. recurrent pre-eclampsia suggest that pre-eclampsia may not be one disease.
These observations also contribute to the difficulty of treatments to attempt to prevent the disorder in which (i) many women must be treated in order to prevent disease in one woman and (ii) a single treatment may not be effective in all women with the disorder. Furthermore, the diagnosis is not easily obtainable from all databases because the findings of hypertension and proteinuria, which are the diagnostic criteria, are not always rigorously adhered to in published studies or clinical care. Ideally, what is needed are large numbers of subjects with excellent demographic data, and reliable clinical data across pregnancy with accurate outcomes data.
In oncology and cardiovascular disease similar challenges have been met by bringing together data and biosamples from several sources to address important questions. It seems a similar strategy would be useful in the study of adverse pregnancy outcomes.