The assessment of placental sharing using X-ray angiogram versus digital photograph: A prospective study

We aim to compare two different methods for the assessment of placental sharing in monochorionic diamniotic twins: X-ray-angiogram and digital photograph of the placenta.

We included the placentas of a prospective series of twins that were followed from the first trimester onward and resulted in a double live birth or double stillbirth between April 2016 and February 2019. Injection was performed after delivery and an X-ray angiogram was made, as well as a digital photograph. On both of these, the territory of each twin was measured two investigators (IC and LL). Placental sharing discordance was determined using the following formula: (larger territory - smaller territory)/larger territory. We calculated the intra-class correlation coefficients for intra-observer and inter-observer reliability and used Bland-Altman analysis to compare both methods.

77 placentas were included in the analysis. For both methods, there was an excellent intra- and inter-observer reliability. The mean difference in sharing (bias) on the X-ray and digital photograph using Bland-Altman analysis was 3,7% (95% CI 1,1% - 6,3%), where the digital photograph tends to overestimate the discordance. Limits of agreement were between -19% and 26%.

Delineation of the placental sharing on a digital photograph slightly overestimates the discordance. Since the venous territory on X-ray angiogram physically determines where each twin gets its oxygenated blood, X-ray angiogram may be a better method to determine placental sharing, although the digital photograph constitutes a valid alternative.