Summary

It was the objective of this study to analyse the influence of confounders, such as ethnicity,severity of illness and method of testing, in articles concerning the still moot relationship of thrombophilias to adverse pregnancy outcome (APO). Relevant casecontrol studies were identified using Medline and EMBASE databases between 1966 and 2006. Search terms were recurrent fetal loss, intrauterine fetal death, preeclampsia, HELLP-syndrome, eclampsia, fetal growth restriction, abruptio placentae, combined with maternal thrombophilias. Data was extracted from the articles per subgroup ofAPO regardless of confounder. These subgroups were tested if they fulfilled the heterogeneity testing criterion (I2 > 35%) to weigh the influence of the confounder. Confounders were selected and examined with Mantel- Haenszel method. Increased thrombophilia prevalence was confirmed in most adverse pregnancy outcomes. Ethnicity, genetic testing only and severity of illness were confounders in the various forms of APO. Stronger relationships between factor V Leiden and severity of disease were found in 2nd and 3rd trimester than 1st trimester recurrent fetal loss, in preeclampsia with: blood pressure ≥160/110 mmHg than ≥140/90 mmHg; proteinuria ≥5 grams per day than < 5 grams; onset before than after 28 weeks, in fetal growth restriction <3rd percentile than <5th, than <10th, and in earlier occurrence of abruptio placentae than 3rd trimester. In conclusion, reports on the prevalence of maternal thrombophilias and APO are influenced by various confounders, which are not always appropriately analysed.The differences we have identified reflect the differential impact of these confounders. These data emphasise the importance of more uniform research.