Diagnostic precision of Doppler flow echocardiography in fetuses at risk for atrioventricular block

Abstract

Objective: To investigate the diagnostic precision of three Doppler methods in their ability to predict postnatal first-degree atrioventricular (AV) block. Methods This was a prospective, observational study carried out from December 1999 to March 2008, including 95 fetuses of anti-SSA/Ro positive mothers undergoing weekly fetal echocardiograms at 18-24 weeks' gestation. Doppler-derived AV time intervals for left ventricular inflow (MV), inflow and outflow (MV-Ao) and superior vena cava a-wave to aortic flow (SVC-Ao) were compared with the PR interval on postnatal electrocardiography. Reference values for MV intervals were established from 102 healthy fetuses, with previously published reference ranges used for the two other methods. Bayesian and receiver-operating characteristics (ROC) curve analyses were performed. Results The prevalence of first-degree AV block at birth was 13.8%. Using a cut-off at the upper 95% confidence limit, MV-Ao and SVC-Ao time intervals had a sensitivity of 91.7%, and negative predictive value and negative likelihood ratio of 98.4% and 0.10, respectively. The corresponding positive predictive value/positive likelihood ratio for MV-Ao and SVC-Ao were 42.3%/4.5 and 47.8%/5.7, respectively. The areas under the ROC curve (AUC) for MV-Ao and SVC-Ao were 0.87 and 0.89, respectively (both P < 0.001), with generated cut-offs for abnormal AV time intervals at 134-138 and 132-138 ms. MV time intervals using a cut-off at the upper 95% confidence limit had a sensitivity of just 50% and an AUC of 0.74 (P < 0.01). Conclusion The MV-Ao and SVC-Ao Doppler methods make it possible to identify nearly all fetuses with first-degree AV block at birth and to exclude conduction disturbances in the case of a normal AV time measurement but at the cost of a positive predictive value of 50%. Copyright © 2010 ISUOG. Published by John Wiley & Sons, Ltd.