Accuracy of Risk Prediction Scores in Pregnant Women

Since the end of the 20th century, surgical and medical care progresses have significantly improved prognosis and quality of life of patients with congenital heart diseases (CHD). Currently, women with CHD expect to experience a "normal" life and express a strong maternity desire.

During the last decade, several scores have been used to predict outcome of pregnancy in women with cardiovascular heart diseases. It has been suggested that the mWHO classification was more appropriate than the previous ones (e.g. Harris, Zahara and CAPREG classifications) in predicting cardiac complications in CHD. More recently, a new risk score stratification (CARPREG II) have been published to improve maternal risk stratification.

This retrospective study will include all pregnant women with CHD who were follow up in our institution from January 2007 to June 2018 until 6 months postpartum. The investigators will collect demographic and clinical data, type and complexity of CHD (Houyel and Bethesda classification), and pregnancy outcomes. Pregnancy scores will be based on information obtained from the relevant clinical letter and clinical notes. The discriminatory power of each risk scores was assessed by the area-under-the receiver-operating characteristic (ROC) curve (AUC). In order to assess the reproducibility of scores classification assigned to CHD patients, all pregnancy scores from this cohort were assessed by 3 independent observers and, thereafter, estimated Cronbach's coefficient alpha.

Inclusion criteria

• All patients with congenital heart diseases who deliver in our institution during the study period

Exclusion criteria

• Women with acquired cardiovascular diseases