Ultrasonic fetal weight estimation and tolerance to measurement error: A comparative analysis

Statistical limits in sonographic estimation of birth weight

PURPOSE

The accuracy of sonographic estimation of birth weight (EBW) is compromised by the precision of the biometrical measurements and the quality of the algorithms. This prospective study was to evaluate technical aspects to derive new equations for the EBW.

METHODS

Three consecutive phases were carried out (1) to recruit a homogenous population, (2) to derive eight new algorithms using a multiple stepwise mathematical/statistical method, and (3) to test the accuracy of the developed equations. Only those patients with a singleton pregnancy who delivered within 48 h from the scan were considered for the analysis.

RESULTS

The study population was made of 473 women. Four polynomial, two square root and two logarithmic algorithms were derived from a balanced study group of 200 women selected from the original study population. These formulas were subsequently applied and compared between them and showed a significant correlation with birth weight (p < 0.0001) and satisfactory statistical performances (r > 0.9), nevertheless they performed similarly to other equations previously published.

CONCLUSIONS

The present findings define better the limitations associated with the intrinsic properties of algorithms and highlight that the possibility to improve the precision of sonographic measurements remains the only point at issue to increase the accuracy in the prediction of birth weight.

Estimation of fetal weight by measurement of fetal thigh soft-tissue thickness in the late third trimester

OBJECTIVE

The accuracy of current formulae for the sonographic estimation of fetal weight (EFW) is compromised by significant intra- and interobserver variability of biometrical measurements, particularly circumferences. The aim of this study was to assess the reliability of the linear measurement of mid-thigh soft-tissue thickness (STT) and to derive a novel formula for EFW.

METHODS

This was a prospective study involving 388 singleton uncomplicated pregnancies. There were three consecutive phases: (1) to verify the relationship between STT and birth weight, (2) to derive a novel formula for EFW using femur length and STT only, and (3) to test the accuracy of the new equation. Only the 290 patients who delivered within 48 h of measurement were considered for the analysis. A comparison with other formulae was performed.

RESULTS

STT was significantly correlated with both abdominal circumference and birth weight (r(2) = 0.36 and 0.46, respectively; P < 0.001). Both intra- and interobserver variability were satisfactory (0.44 +/- 0.27 and 0.57 +/- 0.35 mm, respectively). The equation for EFW was developed using multiple stepwise regression analysis (EFW = - 1687.47 + (54.1 x femur length) + (76.68 x STT)) and tested prospectively on 69 patients. The new formula yielded results (r = 0.79) that were slightly better in accuracy than two other published equations, and had an absolute mean error of < 15% in 97% of cases.

CONCLUSIONS

Our findings confirm the potential of the linear measurement of mid-thigh STT as a valuable parameter for the sonographic assessment of fetal growth and EFW. Our new equation is apparently at least as reliable as the most widely used formulae for EFW.

Estimation of birth weight by two-dimensional ultrasonography: a critical appraisal of its accuracy

OBJECTIVE

To assess the accuracy and characterize two-dimensional ultrasonographic formulas for the estimation of birth weight according to the type of fetal biometric parameters these formulas rely on to make fetal weight predictions.

METHODS

A prospective recruitment of 589 pregnant women was carried out for this cross-sectional study. Different biometric parameters were taken ultrasonographically to estimate birth weight using 35 different formulas. Only those patients who delivered within 48 hours were considered for the analysis (n=441). Differences between the estimated and actual birth weight were assessed by percentage error, accuracy in predictions within +/-10% and +/-15% of error, and use of the Bland-Altman method. All formulas were assessed individually and clustered on the basis of the type of fetal biometric information that they incorporate.

RESULTS

Twenty-nine formulas provided an overall mean absolute percentage error less than or equal to 10%, with overall predictions within +/-10% and +/-15% of the actual birth weight (69.2% and 86.5%, respectively). Twenty formulas showed a good accuracy (bias 0.50 or less) and low variability (mean standard deviation 1.2). Among the categorized algorithms, formulas based on head-abdomen-femur measurements showed the lowest mean absolute percentage error. Upon stratification for birth weight, the group of formulas that rely on abdomen and femur measurements performed best for fetuses weighing more than 3,500 g (P<.01).

CONCLUSION

Our findings show that most formulas are relatively accurate at predicting birth weight up to 3,500 g, and all algorithms tend to underestimate large fetuses.

LEVEL OF EVIDENCE

III.

Prediction of birth weight by ultrasound in Turkish population. Which formula should be used in Turkey to estimate fetal weight?

To determine optimal sonographic fetal weight estimation formula for male and female infants, a prospective study has been performed. Uncomplicated pregnancies and 465 newborns were evaluated. Measurements included birth weight, length and head circumference in addition to fetal head circumference, femur length, abdominal circumference and biparietal diameter. Actual weights were compared with estimated weights calculated by ten different formula. Estimated fetal weight obtained from all formula, except those of Merz, Warsof and Ferrero, tended to be lower than the measured birth weight. The smallest mean differences were obtained with Hadlock 1, Hadlock 2, Hadlock 4 and Shepard formula (19 g-85 g), whereas Merz and Woo produced largest mean differences (110 g-364 g). Intraclass correlation coefficients (ICCs) generated with Hadlock 1 and Hadlock 2 formula were identical (0.90). ICCs obtained with Hadlock 3 and Hadlock 4, Shepard, Merz, Warsof and Campbell formula varied between 0.84 and 0.88. Hadlock 1 and 2 formula gave the closest approximation of birth weight in Turkish population.