Personalized third-trimester fetal growth evaluation: comparisons of individualized growth assessment, percentile line and conditional probability methods.
This investigation has evaluated the ability of Rossavik growth models (derived from data obtained before 26 weeks) to predict birth characteristics in normal fetuses (assuming no growth after 38 weeks). Both weight and head circumference were predicted without systematic errors and the ranges of the random errors were approximately +/- 10% and +/- 5%, respectively. A range of approximately +/- 10% for the random error was also found for the abdominal and thigh circumferences, but both parameters were systematically overestimated (17.3% and 6.4%, respectively). No evidence of a relationship between the percent deviation and age at delivery (up to 41 weeks) was found for any of the parameters studied. As indicators of pregnancy outcome, Growth Potential Realization Index (GPRI) values were calculated for each parameter (predicted values corrected for systematic errors were used for the abdominal and thigh circumferences). Mean GPRI values were not significantly different from 100% and the ranges were approximately 90% to 110% except for the head circumference (range: 95% to 105%). These results indicate that GPRI values can be used to characterize the outcome of prenatal growth processes on an individual basis, with each fetus acting as its own control.