Influence of Sonographic Fetal Weight Estimation Inaccuracies in Macrosomia on Perinatal Outcome

Fetal weight estimation based on deep neural network: a retrospective observational study

BACKGROUND

Improving the accuracy of estimated fetal weight (EFW) calculation can contribute to decision-making for obstetricians and decrease perinatal complications. This study aimed to develop a deep neural network (DNN) model for EFW based on obstetric electronic health records.

METHODS

This study retrospectively analyzed the electronic health records of pregnant women with live births delivery at the obstetrics department of International Peace Maternity & Child Health Hospital between January 2016 and December 2018. The DNN model was evaluated using Hadlock's formula and multiple linear regression.

RESULTS

A total of 34824 live births (23922 primiparas) from 49896 pregnant women were analyzed. The root-mean-square error of DNN model was 189.64 g (95% CI 187.95 g-191.16 g), and the mean absolute percentage error was 5.79% (95%CI: 5.70%-5.81%), significantly lower compared to Hadlock's formula (240.36 g and 6.46%, respectively). By combining with previously unreported factors, such as birth weight of prior pregnancies, a concise and effective DNN model was built based on only 10 parameters. Accuracy rate of a new model increased from 76.08% to 83.87%, with root-mean-square error of only 243.80 g.

CONCLUSIONS

Proposed DNN model for EFW calculation is more accurate than previous approaches in this area and be adopted for better decision making related to fetal monitoring.

Nomogram-based risk prediction of macrosomia: a case-control study

Background

Macrosomia is closely associated with poor maternal and fetal outcome. But there is short of studies on the risk of macrosomia in early pregnancy. The purpose of this study is to establish a nomogram for predicting macrosomia in the first trimester.

Methods

A case-control study involving 1549 pregnant women was performed. According to the birth weight of newborn, the subjects were divided into macrosomia group and non-macrosomia group. The risk factors for macrosomia in early pregnancy were analyzed by multivariate logistic regression. A nomogram was used to predict the risk of macrosomia.

Results

The prevalence of macrosomia was 6.13% (95/1549) in our hospital. Multivariate logistic regression analysis showed that prepregnancy overweight (OR: 2.13 95% CI: 1.18–3.83)/obesity (OR: 3.54, 95% CI: 1.56–8.04), multiparity (OR:1.88, 95% CI: 1.16–3.04), the history of macrosomia (OR: 36.97, 95% CI: 19.90–68.67), the history of GDM/DM (OR: 2.29, 95% CI: 1.31–3.98), the high levels of HbA1c (OR: 1.76, 95% CI: 1.00–3.10) and TC (OR: 1.36, 95% CI: 1.00–1.84) in the first trimester were the risk factors of macrosomia. The area under ROC (the receiver operating characteristic) curve of the nomogram model was 0.807 (95% CI: 0.755–0.859). The sensitivity and specificity of the model were 0.716 and 0.777, respectively.

Conclusion

The nomogram model provides an effective mothed for clinicians to predict macrosomia in the first trimester.

Percent error of ultrasound examination to estimate fetal weight at term in different categories of birth weight with focus on maternal diabetes and obesity

BACKGROUND

Sonography based estimate of fetal weight is a considerable issue for delivery planning. The study evaluated the influence of diabetes, obesity, excess weight gain, fetal and neonatal anthropometrics on accuracy of estimated fetal weight with respect to the extent of the percent error of estimated fetal weight to birth weight for different categories.

METHODS

Multicenter retrospective analysis from 11,049 term deliveries and fetal ultrasound biometry performed within 14 days to delivery. Estimated fetal weight was calculated by Hadlock IV. Percent error from birth weight was determined for categories in 250 g increments between 2500 g and 4500 g. Estimated fetal weight accuracy was categorized as accurate ≤ 10% of birth weight, under- and overestimated by >  ± 10% - ± 20% and > 20%.

RESULTS

Diabetes was diagnosed in 12.5%, obesity in 12.6% and weight gain exceeding IOM recommendation in 49.1% of the women. The percentage of accurate estimated fetal weight was not significantly different in the presence of maternal diabetes (70.0% vs. 71.8%, p = 0.17), obesity (69.6% vs. 71.9%, p = 0.08) or excess weight gain (71.2% vs. 72%, p = 0.352) but of preexisting diabetes (61.1% vs. 71.7%; p = 0.007) that was associated with the highest macrosomia rate (26.9%). Mean percent error of estimated fetal weight from birth weight was 2.39% ± 9.13%. The extent of percent error varied with birth weight with the lowest numbers for 3000 g-3249 g and increasing with the extent of birth weight variation: 5% ± 11% overestimation in the lowest and 12% ± 8% underestimation in the highest ranges.

CONCLUSION

Diabetes, obesity and excess weight gain are not necessarily confounders of estimated fetal weight accuracy. Percent error of estimated fetal weight is closely related to birth weight with clinically relevant over- and underestimation at both extremes. This work provides detailed data regarding the extent of percent error for different birth weight categories and may therefore improve delivery planning.