OBJECTIVE: To develop a nomogram to predict macrosomia with a combination of clinical and ultrasound variables. METHODS: Data from 194 women who underwent sonographic fetal weight estimation were used to develop and calibrate a nomogram to predict fetal macrosomia. The nomogram was subjected to 200 bootstrap resamples for internal validation and to reduce overfit bias. An Internet-based tool was developed to facilitate use of the nomogram. RESULTS: The macrosomia prediction nomogram, based on parity, ethnicity, body mass index and fetal weight estimated macrosomia, had good discrimination and calibration before and after bootstrapping (area under curve (AUC), 0.860 and 0.850, respectively). The predictive accuracy of our nomogram was significantly better than was sonographically estimated fetal weight using Hadlock's formula (AUC, 0.740; P<0.001). We have provided a web-based interface to predict the individual probability of macrosomia. CONCLUSION: We have developed a nomogram to predict the individual probability of macrosomia based on clinical and ultrasound findings. Our web-based interface should help to guide patients and physicians in decision-making. Copyright (c) 2007 ISUOG.
OBJECTIVE: To develop a nomogram to predict macrosomia with a combination of clinical and ultrasound variables. METHODS: Data from 194 women who underwent sonographic fetal weight estimation were used to develop and calibrate a nomogram to predict fetal macrosomia. The nomogram was subjected to 200 bootstrap resamples for internal validation and to reduce overfit bias. An Internet-based tool was developed to facilitate use of the nomogram. RESULTS: The macrosomia prediction nomogram, based on parity, ethnicity, body mass index and fetal weight estimated macrosomia, had good discrimination and calibration before and after bootstrapping (area under curve (AUC), 0.860 and 0.850, respectively). The predictive accuracy of our nomogram was significantly better than was sonographically estimated fetal weight using Hadlock's formula (AUC, 0.740; P<0.001). We have provided a web-based interface to predict the individual probability of macrosomia. CONCLUSION: We have developed a nomogram to predict the individual probability of macrosomia based on clinical and ultrasound findings. Our web-based interface should help to guide patients and physicians in decision-making. Copyright (c) 2007 ISUOG.