Agreement between umbilical vein volume blood flow measurements obtained at the intra-abdominal portion and free loop of the umbilical cord

Objective The umbilical vein (UV) is a single vessel and theoretically the volume of blood flowing through it should be equal whether it is measured at the intra-abdominal portion (IA) or a free loop of the umbilical cord (FL). However, the reported values vary considerably depending on the technique and site of measurement. Our objective was to investigate the correlation and agreement between UV volume blood flows (Q(uv)) measured at the IA and FL. Methods Blood flow velocities and inner diameter of the UV were measured cross-sectionally at the IA and FL in 13.1 pregnant women at 22-24 weeks' gestation, and in 53 of them longitudinally at 4-weekly intervals until delivery. For each sampling site, the Q(uv) was calculated as: 0.5 x tune-averaged maximum velocity x pi x (UV diameter/2)(2). Results The mean Q(uv) measured at the IA (61.5 +/- 23.3 mL/min) and FL (60.7 +/- 17.6 mL/min) cross-sectionally in 131 fetuses at 22-24 weeks of gestation were similar, but the bivariate correlation between them was not strong (r = 0.38; P < 0.0001) and the intraclass correlation coefficient (ICC) was 0.37 (95% CI, 0.21-0.5.1). When the agreement between 131 pairs of Q(uv) measurements was tested with Bland-Altman analysis, the mean of the ratio IA-Q(uv)/FL-Q(uv) was found to be 1.05 (i.e. IA-Q(uv) exceeded FL-Q(uv) on average by 5%) with 95% limits of agreement of 0.31-1.78. In the longitudinal analysis of 232 pairs of Q(uv) measurements from 53 fetuses at 22-40 weeks, the mean Q(uv) obtained at the IA and FL were similar, i.e. 164 +/- 87 (range, 25-484) mL/min vs. 159 82 (range, 30-470) mL/min, and the correlation between them, assessed after grouping the observations in five different gestational age groups of 4-week intervals, was slightly better (r = 0.45-0.62; P < 0.0001). The ICC for the Q(uv) measurements obtained at the two sites ranged front 0.38 to 0.54. The mean of the ratio IA-Q(uv)/FL-Q(uv) was 1.07 (i.e. IA-Q(uv) exceeded FL-Q(uv) on average by 7%), with 95% limits of agreement of 0.39-1.75. Conclusion Average Q(uv) measured at the IA and FL teas similar, but the agreement between individual pairs of measurements was not good enough to be able to use theta interchangeably. Therefore, any clinical application of Q(uv) measurement would require standardization of the technique, strict adherence to methodology and use of appropriate reference ranges for it to he useful. Copyright (C) 2009 ISUOG. Published by John Wiley & Sons, Ltd.