RESUMO
OBJECTIVES: To evaluate the agreement between multiplanar and Virtual Organ Computer-aided AnaLysis (VOCAL()) techniques for the measurement of total fetal thigh volume and to assess the repeatability and reproducibility of measurements performed using these methods; to derive birth weight-predicting models for both methods and to compare their accuracies. METHODS: This was a cross-sectional study of 150 singleton pregnancies at 30-42 weeks of gestation in which ultrasound volumes of the fetal thigh were obtained within 48 hours of delivery and measured using multiplanar and VOCAL techniques. Bland-Altman analyses were performed to determine the agreement between the two methods, and to evaluate intraobserver and interobserver variability in a subset of 40 patients. Birth weight-predicting models were derived using total fetal thigh volumes obtained using the VOCAL (ThiV) and multiplanar (ThiM) methods as independent variables. The accuracies of these formulas were compared. RESULTS: The mean percentage difference between measurements performed using the VOCAL technique and the multiplanar technique was -0.04 and the 95% limits of agreement were -8.17 and 8.09. The mean percentage difference and 95% limits of agreement between paired measurements in the assessment of intraobserver and interobserver variability were -1.10 (-7.67 to 5.47) and 0.61 (-7.68 to 8.91) for the VOCAL technique and 1.03 (-6.35 to 8.41) and -0.68 (-11.42 to 10.06) for the multiplanar method, respectively. The best-fit formulas for predicting birth weight (BW) were: BW = 1025.383 + 12.775x ThiV; and BW = 1033.286 + 12.733x ThiM. There was no significant difference between the accuracies of these formulas. CONCLUSIONS: There is good agreement between the VOCAL and multiplanar techniques for assessment of total fetal thigh volume. Measurements performed using both methods are repeatable and reproducible. For prediction of birth weight, the formulas generated in this study can be used interchangeably.
