Clinical risk factors for pre-eclampsia early pregnancy: problems with systematic review

We read Bartsch and colleagues’ review on clinical risk factors for pre-eclampsia with interest.1 This timely topic is of great clinical interest because it reworks the established institutional guidelines for using maternal characteristics to screen for pre-eclampsia risk in pregnancy. Although we applaud their diligent and meticulous approach, we have some reservations about this study.

The authors attributed specific risk thresholds to individual maternal risk factors in an attempt to allow clinicians to identify women who may benefit from aspirin therapy. However, they failed to establish the risk lowering effect when these risk factors are absent. For instance, maternal age >40 years justifies aspirin treatment according to their analysis, but would this still be true if the woman was multiparous with no previous pre-eclampsia and a normal body mass index? Because this is a systematic review, the data cannot assess the significance of interactions between maternal risk factors. A large targeted population based study or an individual participant data meta-analysis would be most appropriate for developing a reliable prediction model.2 3

Secondly, the authors produce aggregate data on pre-eclampsia as a single disease entity rather than acknowledge that early (<34 weeks) and late (>34 weeks) pre-eclampsia have different risk profiles, recurrence rates, and response to aspirin therapy.4 5 Such an approach to classification of pre-eclampsia—independent of gestation—would probably lead to the unjustified use of aspirin in a large proportion of women. In addition, the population attributable risks for pre-eclampsia are modest at best, so a risk factor based screening programme will not identify the majority of at-risk pregnancies.5 6 The authors propose a maternal history based approach to screening because of the lack of availability of uterine artery Doppler assessment in early pregnancy. However, the influence of ethnicity—a factor that affects the risk of pre-eclampsia and would probably be of clinical significance if their algorithm is to be used in low and middle income countries—is not available in their review.7

Interestingly, the most effective risk factors identified are all useful for predicting adult cardiovascular disease as well—antiphospholipid antibody syndrome, chronic hypertension, diabetes, high BMI, chronic kidney disease, and previous pre-eclampsia. It is possible that this is a coincidence and that these risk factors have as yet unspecified mechanisms for deleteriously affecting placental development—the widely accepted cause of pre-eclampsia. Alternatively, it is possible that this observation supports the emerging hypothesis that the pre-eclampsia is cardiovascular in origin.8 9