Abstract
Objectives: To evaluate the diagnostic performance of CVPs in predicting fetal Hb Bart's disease among pregnancies at risk and to study hemodynamic changes based on CVP components in response to fetal anemia. Methods: The database was assessed to retrieve the ultrasound records of fetuses at risk of Hb Bart's disease at 17-22 weeks and the relevant files including complete video sets of fetal echocardiography. The five components of CVPs of each case were blindly assigned. The definitive diagnosis of fetal Hb Bart's disease was based on cordocentesis or neonatal blood analysis. Results: Among 378 pregnancies at risk that were recruited into the study, there were 76 (20.1%) affected fetuses and 302 (79.9%) unaffected fetuses. Using a cut-off score of <9, CVPs had a sensitivity of 92.1% and specificity of 97.4% in predicting affected fetuses. However, the effectiveness was not much superior to cardio-thoracic area ratio (CTAR) alone (area under curve; AUC: 0.983 vs. 0.954). Of all parameters, CTAR provided the best diagnostic performance. The combination of CTAR and assessment of hydropic sign provided the best diagnostic values, comparable with full CVPs (AUC 0.982 vs. 0.983). The affected fetuses cope well with anemia by physically increasing in cardiac size and functionally increasing in Tei index with minimally reduced shortening fraction, without compromising arterial and venous Doppler indices. Conclusions: CVPs are highly effective in predicting affected fetuses among pregnancies at risk of fetal Hb Bart's disease. Nevertheless, only two components (CTAR and hydropic sign) are adequate to yield the best diagnostic performance.