Validation of Risk Stratification for Valvular Heart Disease in Pregnancy
Quick Takes
- Both DEVI and CARPREG-II risk prediction models are useful in pregnant patients with valvular heart disease.
- The DEVI score may better identify risk of adverse cardiac outcomes among pregnant patients with valvular heart disease.
Study Questions:
Among pregnant women with valvular heart disease, what is the accuracy of the DEVI and CARPREG-II risk stratification tools?
Methods:
A cohort of pregnant patients with valvular heart disease admitted to a tertiary care center in India was developed (2019–2022). Risk was calculated using the DEVI and CARPREG-II models. Performance was assessed with discrimination and calibration. Decision curve analysis was used to assess clinical utility.
Results:
Of 577 pregnancies with maternal valvular heart disease, 95% were related to rheumatic heart disease, and 48% had mitral regurgitation. Adverse cardiac events occurred in 12%. For DEVI, the area under the receiver-operating characteristic curve (AUC) was 0.884 (95% confidence interval [CI], 0.844-0.923). For CARPREG-II, the AUC was 0.808 (95% CI, 0.753-0.863). The DEVI score overestimated risk at higher probabilities and CARPREG-II overestimated risk at both extremes and underestimated risk at middle probabilities. Both models were useful across predicted probability thresholds between 10% and 50%.
Conclusions:
Both the DEVI and CARPREG-II scores showed good discriminative ability and clinical utility in pregnant patients with valvular heart disease. The DEVI score appeared superior in this cohort.
Perspective:
Risk stratification is an essential component of managing patients with cardiac disease during pregnancy. Unfortunately, few tools have been developed and validated. The CARPREG-II model was developed in Canada, a high-income country, in a cohort with a high percentage of patients with congenital heart disease. In contrast, the DEVI score was developed from a similar population to the one assessed in this study (i.e., maternal valvular heart disease in middle-income settings). While the DEVI score predicts in-hospital complications, CARPREG-II includes complications within 6 months of delivery. Overall, both models performed relatively well. Among patients with valvular heart disease, the DEVI score may be superior; however, this score still needs to be externally validated in other countries and diverse settings. Improving accurate risk stratification is an important step towards improving our care of high-risk pregnant patients globally.
Clinical Topics: Valvular Heart Disease
Keywords: Heart Valve Diseases, Pregnancy, Rheumatic Heart Disease, Risk Assessment
< Back to Listings