A new formula for calculating weight in the fetus of < or = 1600 g

Evaluating the accuracy and precision of sonographic fetal weight estimation models in extremely early-onset fetal growth restriction

INTRODUCTION

Birthweight is a critical predictor of survival in extremely early-onset fetal growth restriction (diagnosed pre-28 weeks' gestation, with abnormal umbilical/uterine artery Doppler waveforms), therefore accurate fetal weight estimation is a crucial component of antenatal management. Currently available sonographic fetal weight estimation models were predominantly developed in populations of mixed gestational age and varying fetal weights, but not specifically tested within the context of extremely early-onset fetal growth restriction. This study aimed to determine the accuracy and precision of fetal weight estimation in this population and investigate whether model performance is affected by other factors.

MATERIAL AND METHODS

Cases where a growth scan was performed within 48 hours of delivery (n = 65) were identified from a cohort of extremely early-onset fetal growth-restricted pregnancies at a single tertiary maternity center (n = 159). Fetal biometry measurements were used to calculate estimated fetal weight using 21 previously published models. Systematic and random errors were calculated for each model and used to identify the best performing model, which in turn was used to explore the relationship between error and gestation, estimated fetal weight, fetal presentation, fetal asymmetry and amniotic fluid volume.

RESULTS

Both systematic (median 8.2%; range -44.1 to 49.5%) and random error (median 11.6%; range 9.7-23.8%) varied widely across models. The best performing model was Hadlock head circumference-abdominal circumference-femur length (HC-AC-FL), regardless of gestational age, fetal size, fetal presentation or asymmetry, with an overall systematic error of 1.5% and random error of 9.7%. Despite this, it only calculated the estimated fetal weight within 10% of birthweight in 64.6% of cases. There was a weak negative relation between mean percentage error with Hadlock HC-AC-FL and amniotic fluid volume, suggesting fetal weight is overestimated at lower liquor volumes and underestimated at higher liquor volumes (P = 0.002, adjusted R²  = 0.08).

CONCLUSIONS

Hadlock HC-AC-FL is the most accurate model currently available to estimate fetal weight in extremely early-onset fetal growth restriction independent of gestation or fetal size, asymmetry or presentation. However, for 35.4% of cases in this study, estimated fetal weight calculated using this model deviates by more than 10% from birthweight, highlighting a need for an improved model.

Ultrasonographic estimation of fetal weight: development of new model and assessment of performance of previous models

OBJECTIVES

To develop a new formula for ultrasonographic estimation of fetal weight and evaluate the accuracy of this and all previous formulae in the prediction of birth weight.

METHODS

The study population consisted of 5163 singleton pregnancies with fetal biometry at 22-43 weeks' gestation and live birth of a phenotypically normal neonate within 2 days of the ultrasound examination. Multivariable fractional polynomial analysis was used to determine the combination of variables that provided the best-fitting models for estimated fetal weight (EFW). A systematic review was also carried out of articles reporting formulae for EFW and comparing EFW to actual birth weight. The accuracy of each model for EFW was assessed by comparing mean percentage error, absolute mean error (AE), proportion of pregnancies with AE ≤ 10% and Euclidean distance.

RESULTS

The most accurate models, with the lowest Euclidean distance and highest proportion of AE ≤ 10%, were provided by the formulae incorporating ≥ 3 rather than < 3 biometrical measurements. The systematic review identified 45 studies describing a total of 70 models for EFW by various combinations of measurements of fetal head circumference (HC), biparietal diameter, femur length (FL) and abdominal circumference (AC). The most accurate model with the lowest Euclidean distance and highest proportion of AE ≤ 10% was provided by the formula of Hadlock et al., published in 1985, which incorporated measurements of HC, AC and FL; there was a highly significant linear association between EFW and birth weight (r = 0.959; P < 0.0001), and EFW was within 10% of birth weight in 80% of cases. The performance of the best model developed in this study, utilizing HC, AC and FL, was very similar to that of Hadlock et al. CONCLUSION: Despite many efforts to develop new models for EFW, the one reported in 1985 by Hadlock et al., from measurements of HC, AC and FL, provides the most accurate prediction of birth weight and can be used for assessment of all babies, including those suspected to be either small or large. Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd.

Comparison of Errors of 35 Weight Estimation Formulae in a Standard Collective

Issue: The estimation of foetal weight is an integral part of prenatal care and obstetric routine. In spite of its known susceptibility to errors in cases of underweight or overweight babies, important obstetric decisions depend on it. In the present contribution we have examined the accuracy and error distribution of 35 weight estimation formulae within the normal weight range of 2500-4000 g. The aim of the study was to identify the weight estimation formulae with the best possible correspondence to the requirements of clinical routine. Materials and Methods: 35 clinically established weight estimation formulae were analysed in 3416 foetuses with weights between 2500 and 4000 g. For this we determined and compared the mean percentage error (MPE), the mean absolute percentage error (MAPE), and the proportions of estimates within the error ranges of 5, 10, 20 and 30 %. In addition, separate regression lines were calculated for the relationship between estimated and actual birth weights for the weight range 2500-4000 g. The formulae were thus examined for possible inhomogeneities. Results: The lowest MPE were achieved with the Hadlock III and V formulae (0.8 %, STW 9.2 % or, respectively, -0.8 %, STW 10.0 %). The lowest absolute error (6.6 %) as well as the most favourable frequency distribution in cases below 5 % and 10 % error (43.9 and 77.5) were seen for the Halaska formula. In graphic representations of the regression lines, 16 formulae revealed a weight overestimation in the lower weight range and an underestimation in the upper range. 14 formulae gave underestimations and merely 5 gave overestimations over the entire tested weight range. Conclusion: The majority of the tested formulae gave underestimations of the actual birth weight over the entire weight range or at least in the upper weight range. This result supports the current strategy of a two-stage weight estimation in which a formula is first chosen after a pre-estimation of the weight range.

A New Sonographic Weight Estimation Formula for Small-for-Gestational-Age Fetuses

OBJECTIVES

The purpose of this study was to develop a new specific weight estimation formula for small-for-gestational-age (SGA) fetuses that differentiated between symmetric and asymmetric growth patterns.

METHODS

A statistical estimation technique known as component-wise gradient boosting was applied to a group of 898 SGA fetuses (symmetric, n = 750; asymmetric, n = 148). A new formula was derived from the data obtained and was then compared to other commonly used equations.

RESULTS

The new formula derived is as follows: estimated fetal weight = e^[1.3734627 + 0.0057133 × biparietal diameter + 0.0011282 × head circumference + 0.0201147 × abdominal circumference + 0.0183081 × femur length - 0.0000177 × biparietal diameter(2) - 0.0000018 × head circumference(2) - 0.0000297 × abdominal circumference(2) -0.0001007 × femur length(2) + 0.0397563 × I(sex = male) + 0.0064505 × gestational age (days) + 0.0096528 × I(SGA = asymmetric)], where the function I denotes an indicator function, which is 1 if the expression is fulfilled (sex = male; SGA type = asymmetric) and otherwise 0. In the whole study group and the 2 subgroups, the new formula showed the lowest median absolute percentage error, mean percentage error, and random error and the best distribution of absolute percentage errors within prespecified error bounds.

CONCLUSIONS

The new formula substantially improves weight estimation in SGA fetuses.

Prediction of Small for Gestational Age: Accuracy of Different Sonographic Fetal Weight Estimation Formulas

Objective: To compare the accuracy of various sonographic estimated fetal weight (sEFW) formulas for the prediction of small for gestational age (SGA) neonates. Methods: A retrospective analysis of 6,126 fetal biometrical measurements performed within 3 days of delivery. SGA prediction was evaluated for various sEFW formulas by calculating the sensitivity, specificity, positive/negative predictive value (PPV/NPV), likelihood ratio (+LR/-LR), overall accuracy and area under the receiver operating characteristic curve (AUC). Systematic error, random error, proportion of estimates >10% of birth weights, actual and absolute weight differences were compared between SGA and non-SGA neonates. Results: Overall, 638 (10.4%) neonates were SGA. There was considerable variation among formulas in sensitivity (mean ± SD, 62 ± 14.4%; range, 32.4-91.2), PPV (72.5 ± 10.7%; 45.8-95.6) and +LR (24.2 ± 10.9; 7.2-57.3), mild variation in specificity (96.6 ± 2.7%; 87.4-99.4), NPV (94.6 ± 5.3%; 72.2-98.9) and -LR (0.4 ± 0.1; 0.1-0.7) and minimal variation in AUC (mean, 0.93; range, 0.91-0.93). The majority of formulas had a lower accuracy for the SGA neonates, with systematic error and random error ranging from -4.2 to 14.3% and from 8.4 to 12.9% for SGA, and from -8.7 to 16.1% and from 7.2 to 10.5% for non-SGA, respectively. Conclusion: sEFW formulas differ in their accuracy for SGA prediction. In our population, the most accurate formula for SGA prediction was Hadlock's formula utilizing femur length, abdominal and head circumference.

Choice of Formula and Accuracy of Fetal Weight Estimation in Small-for-Gestational-Age Fetuses

OBJECTIVES

The purpose of this study was to identify the most accurate sonographic models for fetal weight estimation in specific subgroups of small-for-gestational-age (SGA) fetuses.

METHODS

We conducted a retrospective study of women who delivered an SGA neonate and underwent a sonographic estimation of fetal weight within 7 days of delivery in a single tertiary center (n = 370). The accuracy of fetal weight estimation was compared for 33 sonographic models (27 nontargeted and 6 targeted SGA- or low-birth-weight-specific models) in specific subgroups of SGA fetuses: early versus late SGA, asymmetric versus symmetric, and presence of Doppler abnormalities.

RESULTS

A wide variation in the accuracy of the different models was found (systematic error, -12.5% to 15.1%; random error, 7.8% to 15.5%). Most nontargeted models tended to systematically overestimate the weight of SGA fetuses. The best performing model in the overall SGA group was the targeted model of Scott et al (J Ultrasound Med 1996; 15:669-672; systematic error ± random error, -2.8% ± 8.3%). However, the optimal models varied for different subgroups of SGA fetuses, and in most cases the targeted models were the most accurate. An approach that used the optimal model for each subgroup of SGA fetuses compared with the uniform use of the model of Scott et al for all SGA fetuses was associated with a lower systematic error (-0.38% versus -2.8%; P < .001) and a higher proportion of weight estimations within 5%, 10%, and 15% of birth weight (48.4% versus 40.8%; P= .038; 78.6% versus 71.4%; P= .022; 95.1% versus 89.2%; P = .003, respectively).

CONCLUSIONS

Sonographic models in current use for fetal weight estimation in SGA fetuses have significant errors, and their performance varies for specific subgroups of SGA fetuses. An approach that uses subgroup-specific models may improve the accuracy of weight estimation among SGA fetuses.

The accuracy of ultrasound-estimated fetal weight in extremely preterm infants: a comparison of small for gestational age and appropriate for gestational age

OBJECTIVES

To compare the accuracy of estimated fetal weight (EFW) in extremely preterm small for gestational age (SGA) and appropriate for gestational age (AGA) infants and report other significant factors influencing the accuracy of EFW.

METHODS

A retrospective cohort study of singleton pregnancies 22(+0) -27(+6)  weeks. Women were included in the study if an ultrasound scan had been performed within seven days of delivery, with no major fetal anomaly and data available to calculate customised birthweight (BW) centiles. Mean error of EFW and actual BW and mean % error of EFW and actual birthweight were compared for SGA and AGA infants. A stepwise backward elimination linear regression model was used to determine the significant factors influencing the accuracy of EFW.

RESULTS

A total of 134 cases (51 SGA and 83 AGA) were analysed. The mean gestational age at delivery was 25(+2)  weeks (SD 11.5 days) and mean BW 711 g (SD 227 g). Overall mean percentage error of EFW and actual BW was 8.8% (range 0-34.6%). There was a significant difference in mean error of EFW and actual BW for SGA and AGA deliveries (mean +16 g versus -23 g, respectively, P = 0.01) and in mean % error of EFW (11.2%, 95%CI 9.1-13.3 versus 7.4%, 95% CI 6.2-8.6 P = 0.009). Factors that significantly influenced the accuracy of EFW included SGA (P = 0.001, coeff. = -3.73, 95% CI -5.94/-1.52), scan to delivery interval (P = 0.02, coeff. = 0.66, 95% CI 0.12/1.21) and reduced amniotic fluid (P = 0.008, coeff = 3.61, 95% CI -5.47/-0.85).

CONCLUSIONS

Ultrasonographic EFW for extreme preterm SGA fetuses is less accurate than AGA fetuses and is more likely to overestimate EFW. This should be considered when counselling women with growth restricted fetuses at the limits of viability.

Specific formulas improve the estimation of fetal weight by ultrasound scan

OBJECTIVE

To develop and evaluate local, sex specific, small for gestational age (SGA) specific, large for gestational age (LGA) specific and combined (biometry, sex and Doppler indices) formulas for ultrasound estimated fetal weight (EFW).

METHOD

Low-risk singleton pregnancies that delivered within 7 days from ultrasound examination were assessed. A formula-generating group (1407 pregnancies) and a validation group (469 pregnancies) were created. Fractional regression analysis was used to develop the formulas. Systematic error, random error, fraction within the 10% of actual birth weight and Bland-Altman analysis were used.

RESULTS

The local formula and the Hadlock formula with local co-efficients performed better than the Hadlock formula. The SGA-specific formula, the LGA-specific formula and the combined formula had the lower systematic error (MSE: +0.0022291, -0.4226888, +0.8386222, respectively) and the narrower 95% LOA (-292.8 to +292.23, -485.6 to +461.5, -425.7 to +450.46, respectively). The SGA- and the LGA-specific formulas had higher fraction within the 10% of actual birth weight (81.5% and 84%, respectively).

CONCLUSIONS

Local formulas improve the EFW calculation. The combined formula can further optimize the accuracy and precision. Application of specific formulas for the small and the large fetus had the most pronounced effect in improving fetal weight estimation.

Does use of a sex-specific model improve the accuracy of sonographic weight estimation?

OBJECTIVE

To determine whether the use of a sex-specific sonographic model improves the accuracy of fetal weight estimation.

METHODS

New regression models (sex-independent and sex-specific) were developed, based on 1708 sonographic weight estimations performed within 3 days prior to delivery. The accuracy of these models was compared to that of several published models including two of the original Hadlock models (which incorporate the biometric indices abdominal circumference (AC), biparietal diameter (BPD), femur diaphysis length (FL) and head circumference (HC) as follows: AC-FL-BPD and AC-FL-HC, designated here as Hadlock I and Hadlock II, respectively), modified versions of the Hadlock I and II models for which coefficients were adjusted to our local cohort, sex-specific versions of the Hadlock I and II models and Schild's model (a previously published sex-specific model).

RESULTS

The unadjusted models of Hadlock and Schild were associated with the highest systematic error (1.6-4.9%; P < 0.001) which was significantly higher for females (2.3-4.9%) compared to males (1.6-2.0%; P < 0.001). Adjustment of model coefficients to the local population decreased the systematic error (-1.4% to 1.5%) and resulted in a systematic error that was of similar magnitude (P = 0.3) but opposite in direction for male and female fetuses. The sex-specific models (adjusted or newly developed) were associated with the lowest systematic error (-0.4 to 0.5%) and were the only models for which the systematic error was similar for male and female fetuses. There were no differences in the systematic error between adjusted sex-specific versions of the Hadlock I and II models and the newly developed sex-specific models (0.0% to 0.4% vs. - 0.4% to 0.5%; P = 0.4). The random error was similar for all models and, for most of the models, was unrelated to fetal sex.

CONCLUSIONS

The use of sex-specific models appears to improve the accuracy of fetal weight estimation, principally because the optimal set of model coefficients differs for male and female fetuses. The improved accuracy is mainly the result of a decrease in systematic error, as the random error was not affected by the use of such sex-specific models.

Sonographic fetal weight estimation: which model should be used?

OBJECTIVE

The purpose of this study was to compare the accuracy of different sonographic models for fetal weight estimation.

METHODS

We evaluated 26 different models using 3705 sonographic weight estimations performed less than 3 days before delivery. Models were ranked on the basis of systematic and random errors and were grouped according to the combination of biometric indices in each model. Cluster analysis was used to compare the accuracy of the different model groups.

RESULTS

A considerable variation in the accuracy of the different models was found. For birth weights (BWs) in the range of 1000 to 4500 g, models based on 3 or 4 fetal biometric indices were significantly more accurate than models that incorporated only 1 or 2 indices. The accuracy of weight estimation decreased at the extremes of BWs, leading to overestimation in low-BW categories as opposed to underestimation when the BW exceeded 4000 g. The precision of most models was lowest in the low-BW groups.

CONCLUSIONS

To improve the accuracy of fetal weight estimation, sonographic models that are based on 3 or 4 fetal biometric indices should be preferred. Recognizing the accuracy and the tendency for underestimation or overestimation of each of the available models is important for the judicious interpretation of fetal weight estimations, especially at the extremes of fetal weight.

New regression formulas for sonographic weight estimation within 10, 7, and 3 days of delivery

OBJECTIVES

The purpose of this study was to develop new regression formulas based on large numbers of sonographic examinations performed within 10, 7, and 3 days of delivery.

METHODS

Sonographic fetal biometric measurements and delivery ward data for an unselected population were analyzed. Multivariate linear regression models were fitted to the sonographic data to predict the actual birth weight (BW) within 10, 7, and 3 days.

RESULTS

The analyses included 6289, 5449, and 4007 patients who underwent sonographic examinations within 10, 7, and 3 days of delivery, respectively. All models yielded very high correlation coefficients (r = 0.927-0.958; R(2) = 0.859-0.918), low mean deviations between the calculated and actual BWs (6.4%-6.6% +/- 1 SD of 5.5%-5.9%), and high percentages of the calculated BW within 10% of the actual BW (78.5%-80.4%). Estimated fetal weight analyses made within 3 days of delivery yielded slightly better results than within 7 and 10 days.

CONCLUSIONS

The new regression formulas yielded overall similar results, with a small advantage for estimates calculated within 3 days of delivery. Further prospective studies are needed to compare the accuracy of these formulas with those used to date.

Weight estimation by three-dimensional ultrasound imaging in the small fetus

OBJECTIVES

To improve birth weight estimation in fetuses weighing <or= 1600 g at birth by deriving a new formula including measurements obtained using three-dimensional (3D) sonography.

METHODS

In a prospective cohort study, biometric data of 150 singleton fetuses weighing <or= 1600 g at birth were obtained by sonographic examination within 1 week before delivery. Exclusion criteria were multiple pregnancy, intrauterine death as well as major structural or chromosomal anomalies. A new formula was derived using our data, and was then compared with currently available equations for estimating weight in the preterm fetus.

RESULTS

Different statistical estimation strategies were pursued. Gradient boosting with component- wise smoothing splines achieved the best results. The resulting new formula (estimated fetal weight = 656.41 + 1.8321 x volABDO + 31.1981 x HC + 5.7787 x volFEM + 73.5214 x FL + 8.3009 x AC - 449.8863 x BPD + 32.5340 x BPD(2), where volABDO is abdominal volume determined by 3D volumetry, HC is head circumference, volFEM is thigh volume determined by 3D volumetry, FL is femur length and BPD is biparietal diameter) proved to be superior to established equations in terms of mean squared prediction errors, signed percentage errors and absolute percentage errors.

CONCLUSIONS

Our new formula is relatively easy to use and needs no adjustment to weight percentiles or to fetal lie. In fetuses weighing <or= 1600 g at birth it is superior to weight estimation by traditional formulae using two-dimensional measurements.

Ultrasonic estimation of fetal weight at term: An evaluation of eight formulae

AIM

To compare the accuracy of eight sonographic formulae for predicting fetal birth weight at term in a multiethnic population.

METHODS

Pregnant women at term who were booked for induction of labor or elective cesarean section were included in the study. Eight ultrasonic fetal biometric formulae were used to predict fetal birth weight.

RESULTS

A total of 173 patients were included in the study; 53 (30.6%) patients were from the Indian subcontinent, 44 (25.4%) patients were from Africa, 33 (19.1%) patients were from the Arabian Peninsula and 43 (24.9%) were from other ethnic groups. The mean absolute error ranged from a minimum of 0.3% (+/-11.3) for Hadlock (biparietal diameter [BPD], head circumference [HC], abdominal circumference [AC], femur length [FL]) to a maximum of 37.5% (+/-10.0) for Warsof (FL). The correlation of estimated fetal weight with actual birth weight ranged from a minimum of 0.09 with Warsof (FL) to a maximum of 0.77 with Shepard and Warsof (BPD, AC) and Hadlock (BPD, HC, AC, FL). The combination of AC with BPD measurements rather than FL achieves a high level of accuracy.

CONCLUSIONS

Shepard (BPD, AC) provides a simple and accurate logarithm for the prediction of fetal weight at term in the studied multiethnic population.