Performance of Risk Models Predicting Cardiac Complications in Pregnant Women With Congenital Heart Disease: A Meta-Analysis

Abstract

Background: With advances in medical and surgical treatment, there is a growing population of women with congenital heart disease (CHD). Cardiac and obstetric complications are prevalent in pregnant woman with moderate- and high-complexity CHD. Accurate risk stratification, although it has not received great attention, can be important in guiding optimal management and counselling. A meta-analysis was undertaken to assess the performance of risk models for predicting cardiac complications in pregnant women with CHD. Methods: MEDLINE, Embase, and Cochrane databases from 1 January 1980 to 30 June 2018 were searched for original studies. To be included, studies needed to have externally assessed the prognostic utility of risk models at predicting cardiac complications in pregnant women with CHD and reported their respective areas under the receiver-operative characteristics curve (AUC). Data were then extracted for pooled meta-analysis. Results: The search obtained 1,013 articles for screening: 32 full-text articles were reviewed and six studies were included for analysis. There was a total of 3,426 pregnancies, and cardiac complications for pregnant women with CHD were reported at 6.7%-20.6% in these studies. Only the mWHO, CARPREG and ZAHARA risk models have been externally assessed for pregnant women with congenital heart disease. The pooled AUCs (95% CI) for cardiac complications were: mWHO, 0.71 (0.65–0.76); CARPREG, 0.66 (0.61-0.71); and ZAHARA, 0.71 (0.65–0.76). Conclusion: Cardiac complications are common in pregnant women with CHD. The mWHO and ZAHARA risk models moderately discriminated cardiac complications in these women, while the CARPREG risk model had modest performance.