Umbilical artery (UA) Doppler is integral in the management of fetal growth restriction (FGR). Doppler sampling technique recommendations sampling the UA in a free loop of cord (Archaya 2005). Yet, when both UAs are sampled, Doppler index discrepancies >20% may occur between the two UA in up to 1/3 of normal pregnancies (Predanic 1998). This has led to variation in sampling methodology across institutions which may influence clinical management. Objectives: 1) to determine the frequency of discrepant Doppler pulsatility indices (PI) between paired UAs, and 2) to establish how varying UA analysis approaches affect the diagnosis of abnormal UA PI. In this prospective cohort of 121 fetuses with an EFW < 10th%ile, the final ultrasound prior to delivery was used to obtain UA PI from both UAs in a mid-segment of cord. We first determined the percent discrepancy between the two UA PI. Using an ANOVA, we compared differences between commonly used UA PI analysis methods: 1) mean UA PI calculated as the average PI from both UA vessels, 2) UA PI obtained from one UA chosen at random (by random number generator), and 3) UA PI using the highest of the two UA PIs. We then determined whether UA PI evaluation method altered the detection rate of abnormal PI within our cohort. Fig. 1 depicts the percent discrepancies in UA PIs between patients. The ANOVA revealed that all three analysis methods yielded significantly different absolute UA PI values (Table 1). When the PIs from each approach were converted to PI %ile, the difference in the number of patients diagnosed with abnormal UA PI (>95th %ile) trended towards significance (Table 1). While different UA Doppler analysis approaches result in significantly different absolute UA PI values, they do not impact the percentage of patients diagnosed with clinically abnormal UA PI %ile. Though this cohort was under-powered to evaluate for such a difference, we speculate that in a larger cohort of FGR fetuses, using the higher of two UA PIs may increase detection of abnormal UA PI %ile and affect clinical management.View Large Image Figure ViewerDownload Hi-res image Download (PPT)
Read full abstract