Discriminatory Capacity of Prenatal Ultrasound Measures for Large-for-Gestational-Age Birth: A Bayesian Approach to ROC Analysis Using Placement Values

Predicting large fetuses at birth is of great interest to obstetricians. Using an NICHD Scandinavian Study that collected longitudinal ultrasound examination data during pregnancy, we estimate diagnostic accuracy parameters of estimated fetal weight (EFW) at various times during pregnancy in predicting large for gestational age. We adopt a placement value-based Bayesian regression model with random effects to estimate ROC curves. The use of placement value allows us to model covariate effects directly on the ROC curves, and the adoption of a Bayesian approach accommodates the a priori constraint that an ROC curve of EFW near delivery should dominate another further away. The proposed methodology is shown to perform better than some alternative approaches in simulations and its application to the Scandinavian Study data suggest that diagnostic accuracy of EFW can improve about 65% from week 17 to 37 of gestation.

[1]  E. DeLong,et al.  Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. , 1988, Biometrics.

[2]  T. Moore,et al.  A comparison of single versus multiple growth ultrasonographic examinations in predicting birth weight. , 1994, American journal of obstetrics and gynecology.

[3]  Margaret Sullivan Pepe,et al.  Distribution-free ROC analysis using binary regression techniques. , 2002, Biostatistics.

[4]  Yvonne W. Cheng,et al.  The association between birthweight 4000 g or greater and perinatal outcomes in patients with and without gestational diabetes mellitus. , 2009, American journal of obstetrics and gynecology.

[5]  Paul S Albert,et al.  A linear mixed model for predicting a binary event from longitudinal data under random effects misspecification , 2012, Statistics in medicine.

[6]  G. Greisen,et al.  Prediction of birth weight by ultrasound-estimated fetal weight: a comparison between single and repeated estimates. , 1995, European journal of obstetrics, gynecology, and reproductive biology.

[7]  H. Hoffman,et al.  Pre‐pregnancy risk factors of small‐for‐gestational age births among parous women in Scandinavia , 1993, Acta obstetricia et gynecologica Scandinavica.

[8]  M. Hod,et al.  Accuracy of a single fetal weight estimation at 29-34 weeks in diabetic pregnancies: can it predict large-for-gestational-age infants at term? , 2007, American journal of obstetrics and gynecology.

[9]  H. Wolfe,et al.  Limited clinical utility of midtrimester fetal morphometric percentile rankings in screening for birth weight abnormalities. , 1997, American journal of obstetrics and gynecology.

[10]  D. Rubin,et al.  Inference from Iterative Simulation Using Multiple Sequences , 1992 .

[11]  Danping Liu,et al.  Identifying subgroups of enhanced predictive accuracy from longitudinal biomarker data by using tree‐based approaches: applications to fetal growth , 2017, Journal of the Royal Statistical Society. Series A,.

[12]  Lori E. Dodd,et al.  Partial AUC Estimation and Regression , 2003, Biometrics.

[13]  Beom Seuk Hwang,et al.  An Integrated Bayesian Nonparametric Approach for Stochastic and Variability Orders in ROC Curve Estimation: An Application to Endometriosis Diagnosis , 2015, Journal of the American Statistical Association.

[14]  M. Pepe.  An Interpretation for the ROC Curve and Inference Using GLM Procedures , 2000, Biometrics.

[15]  Margaret S. Pepe,et al.  Semiparametric Receiver Operating Characteristic Analysis to Evaluate Biomarkers for Disease , 2002 .

[16]  Jun Zhang,et al.  Predicting large fetuses at birth: do multiple ultrasound examinations and longitudinal statistical modelling improve prediction? , 2012, Paediatric and perinatal epidemiology.

[17]  S. Ghosal,et al.  Bayesian bootstrap estimation of ROC curve , 2008, Statistics in medicine.

[18]  Mitchell H Gail,et al.  On criteria for evaluating models of absolute risk. , 2005, Biostatistics.

[19]  Danping Liu,et al.  Combination of longitudinal biomarkers in predicting binary events. , 2014, Biostatistics.

[20]  Tianxi Cai,et al.  Semi-parametric ROC regression analysis with placement values. , 2004, Biostatistics.

[21]  S. Ghosal,et al.  A note on modeling placement values in the analysis of receiver operating characteristic curves , 2020, Biostatistics & Epidemiology.

[22]  Margaret S. Pepe,et al.  A regression modelling framework for receiver operating characteristic curves in medical diagnostic testing , 1997 .

[23]  T. Moore,et al.  A comparison of single versus multiple growth ultrasonographic examinations in predicting birth weight , 1994 .

[24]  D. Rubin.  The Bayesian Bootstrap , 1981 .

[25]  Zhen Chen,et al.  A Bayesian semiparametric approach to correlated ROC surfaces with stochastic order constraints , 2019, Biometrics.

[26]  Margaret Sullivan Pepe,et al.  The Analysis of Placement Values for Evaluating Discriminatory Measures , 2004, Biometrics.

[27]  M. Hod,et al.  Predictive value of a single early fetal weight estimate in normal pregnancies. , 2007, European journal of obstetrics, gynecology, and reproductive biology.

[28]  Gengsheng Qin,et al.  Empirical Likelihood Inference for the Area under the ROC Curve , 2006, Biometrics.

[29]  J. Tubbs,et al.  Beta Regression for Modeling a Covariate Adjusted ROC , 2018 .

[30]  T. Cai,et al.  Semi-parametric estimation of the binormal ROC curve for a continuous diagnostic test. , 2004, Biostatistics.

[31]  F. P. Hadlock,et al.  Estimation of fetal weight with the use of head, body, and femur measurements--a prospective study. , 1985, American journal of obstetrics and gynecology.

[32]  M. Pepe.  The Statistical Evaluation of Medical Tests for Classification and Prediction , 2003 .

[33]  Gang Li,et al.  A DIRECT SEMIPARAMETRIC RECEIVER OPERATING CHARACTERISTIC CURVE REGRESSION WITH UNKNOWN LINK AND BASELINE FUNCTIONS. , 2012, Statistica Sinica.

[34]  K. Blakemore,et al.  Prediction of Birth Weight by Ultrasound in the Third Trimester , 2000, Obstetrics and gynecology.

[35]  C. Metz.  ROC Methodology in Radiologic Imaging , 1986, Investigative radiology.

[36]  M. Rodríguez-Álvarez,et al.  Bayesian nonparametric inference for the covariate-adjusted ROC curve , 2018, 1806.00473.

[37]  J A Hanley,et al.  Sampling variability of nonparametric estimates of the areas under receiver operating characteristic curves: an update. , 1997, Academic radiology.