Crossed Random Effect Models for Multiple Outcomes in a Study of Teratogenesis

• Published in: Journal of the American Statistical Association Vol. 96; no. 456; pp. 1194 - 1204
• Main Authors: Coull, Brent A, Hobert, James P, Ryan, Louise M, Holmes, Lewis B
• Format: Journal Article
• Language: English
• Published: Alexandria, VA Taylor & Francis 01.12.2001; American Statistical Association
• Subjects: Algorithms; Anticonvulsants; Applications; Applications and Case Studies; Exact sciences and technology; Foetus; Generalized linear mixed model; Health outcomes; Human development; Infants; Lasso; Linear inference, regression; Linear models; Logistic regression; Markov chain Monte Carlo; Markov processes; Mathematics; Medical sciences; Modeling; Monte Carlo EM algorithm; Monte Carlo Newton-Raphson; Monte Carlo simulation; Multilevel models; Numerical analysis; Numerical analysis. Scientific computation; Numerical methods in probability and statistics; Parametric models; Probability and statistics; Probability theory and stochastic processes; Sciences and techniques of general use; Standard error; Statistical methods; Statistical models; Statistics; Teratogenesis; Teratology; Monte Carlo method; Generalized linear model; Statistical analysis; Diagnostic program; In utero; Lognormal distribution; Newton Raphson method; Markov chain; Teratogenesis; Logistic regression; Fetal development; MCMC algorithm; Regression model; Mixed model; Random effect; Numerical simulation; MCEM algoritm
• ISSN: 0162-1459; 1537-274X
• DOI: 10.1198/016214501753381841

Human teratogens often manifest themselves through a broad spectrum of adverse effects. Although often not serious when considered individually, such outcomes taken together may represent a syndrome that can lead to serious developmental problems. Accordingly, studies that investigate the effect of human teratogens on fetal development typically record the presence or absence of a multitude of abnormalities, resulting in the data of multivariate binary form for each infant. Such studies typically have three objectives: (1) estimate an overall effect of exposure across outcomes, (2) identify subjects having the syndrome, and (3) identify those outcomes that constitute the syndrome so that doctors know what to look for when diagnosing the syndrome in other exposed newborns. This article proposes the use of a logistic regression model with crossed random effect structure to address all three questions simultaneously. We use the proposed models to analyze data from a study investigating the effects of in utero antiepileptic drug exposure on fetal development.