Objective: To determine whether using color flow Doppler to identify the umbilical cord affects amniotic fluid index (AFI) measurements.
Methods: A total of 2236 AFI measurements between 24 and 42 weeks in singleton gestations with no known or suspected fetal anomalies and < 14 days' discrepancy between menstrual and ultrasonographic dating were included. Color flow Doppler was used to identify loops of umbilical cord; these were excluded from the measurement. Polynomial regression was used to generate means and centiles. Data were grouped according to completed weeks of gestation and descriptive statistics were calculated. At each week of gestation, the number and percentage of pregnancies diagnosed as < or = 2.5th, < or = 5th, > or = 95th, and > or = 97.5th centile according to a 'standard' nomogram derived without using color flow Doppler were calculated.
Results: The AFI decreased significantly over gestational age, starting at 31 weeks (p < 0.05 by ANOVA). The relationship between AFI and gestational age was best modeled by a second-degree polynomial (p < 0.001). The median and range of the proportion of AFIs that fell outside the ranges of the standard nomogram at each completed gestational age was: 6.0 (2.3-35.4)% for the < or = 2.5th centile, 9.9 (3.1-37.5)% for the < or = 5th centile, 3.8 (0-30)% for the > or = 95th centile, and 1.8 (0-20)% for the > or = 97.5th centile. The 2.5th and 5th centiles using the current data were lower than those of the 'standard', and the difference increased with advancing gestation. Upper centiles were also different.
Conclusion: The AFI measured using color flow Doppler overestimates oligohydramnios and may underestimate polyhydramnios when a standard AFI table obtained without color flow Doppler is used. Normal values specific for measurement method should be used for reference.