Intraobserver and interobserver reproducibility of fetal biometry.
Academic Article
Overview
abstract
OBJECTIVE: To assess the intra- and interobserver reproducibility of ultrasound measurements of fetal biometric parameters. METHODS: We assessed the intraobserver and the interobserver agreement in measurements of fetal biparietal diameter (BPD), abdominal circumference (AC), head circumference (HC) and femur length (FL) on 122 singleton pregnancies. Patients were each examined twice by the first sonographer to determine the intraobserver reliability of measurements of fetal biometry. Subsequently, during the same ultrasound examination, a second blinded sonographer measured fetal biometric parameters to assess interobserver reliability. The consensus between and among observers was analyzed using the intraclass correlation coefficient (intra-CC) and interclass correlation coefficient (inter-CC) and the reliability coefficients (RC, alpha) for the four biometric measurements. A value > 0.75 was considered a reliable consensus for the intra-CC and inter-CC. A Bland and Altman plot was also created for the fetal biometric parameters to assess the repeatability of the measurements. RESULTS: Reliable consensus was observed for both the intra-CC and inter-CC and RC for all four biometric parameters. The intra-CC with the 95% CI and RC for the BPD, AC, HC and FL were as follows: 0.996 (0.995, 0.997), alpha 0.998; 0.994 (0.992, 0.996), alpha 0.997; 0.996 (0.994, 0.997), alpha 0.998; and 0.994 (0.992, 0.996), alpha 0.997, respectively. Similarly, the inter-CC with the 95% CI and RC for the same parameters were as follows: 0.995 (0.993, 0.997), alpha 0.998; 0.980 (0.971, 0.990), alpha 0.990; 0.994 (0.992, 0.996), alpha 0.997; and 0.990 (0.985,0.993), alpha 0.995, respectively. The Bland and Altman plots demonstrated a high degree of repeatability of BPD, AC, HC, and FL measurements. CONCLUSION: Our results demonstrate that the intra- and interobserver reproducibility of ultrasound measurements of fetal biometry are highly reliable.
