# Factors affecting the umbilical artery Doppler reference values in the second and third trimesters

## TL;DR

This study shows that the selection of fetal weight distribution in study populations affects umbilical artery Doppler reference charts more than other factors.

## Contribution

The study identifies that fetal weight inclusion criteria significantly impact Doppler reference chart variability.

## Key findings

- Higher exclusion of low fetal weight fetuses leads to higher UA‐PI values in reference charts.
- Charts based on fetuses above the 10th EFW centile show fewer abnormal UA‐PI readings.
- Parity and statistical methods have minimal effect on reference chart variability.

## Abstract

To explore the contribution of selected methodological factors to the heterogeneity in published umbilical artery pulsatility index (UA‐PI) reference charts.

Cross‐sectional study of uncomplicated singleton pregnancies that underwent assessment of UA Doppler at a single center. We explored the effects of the cohort characteristics (parity and estimated fetal weight [EFW] centile cut‐off) and the statistical modeling approach used to construct the UA‐PI centile reference charts.

25 069 UA‐PI measurements from 12 394 patients were analyzed. UA‐PI centile values decreased as the minimal EFW centile inclusion cut‐off of the study cohort increased. Interpretation of UA‐PI using charts constructed from fetuses with EFW > 25th or 50th centiles resulted in a higher proportion of examinations with UA‐PI > 95th centile compared with the chart based on fetuses with EFW⟩10th centile (11.5% and 13.2% versus 10.6% of the subgroup of small‐for‐gestational‐age fetuses, respectively, P < 0.001). In contrast, parity and the statistical method used to construct the chart had minimal impact on the UA‐PI centile reference charts. Considerable heterogeneity was identified among published UA‐PI reference charts.

Variation in the distribution of EFW centiles across the populations used to construct the UA‐PI reference chart may contribute to the heterogeneity observed in published charts.