Constructing Normalcy and Discrepancy Indexes for Births Using Threshold Regression

• Activity Number: 660
• Type: Contributed
• Date/Time: Thursday, August 4, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Section on Government Statistics
• Author(s): George Alexander Whitmore and Guangyu Zhang*+ and Mei-Ling Ting Lee
• Companies: McGill University and Centers for Disease Control and Prevention and University of Maryland
• Keywords: latent measurements ; boundary crossing ; first hitting time ; subordinated process ; mixture model ; threshold regression

Abstract

Birth weight and gestational age are important measures of a newborn's intrinsic health. We propose two summary indexes for birth weight and gestational age that we call the birth normalcy index (BNI) and the birth discrepancy index (BDI). BNI summarizes the relative balance of birth weight and gestational age and BDI measures the discrepancy of these two measurements. The BNI distribution is derived from a threshold regression model and takes the form of a mixture of inverse Gaussian distributions, with mixture components representing normal and abnormal births. We use a non-central t-distribution to model the BDI. The BNI and BDI distributions can be estimated for births in any reference population of interest. Covariates can be taken into account using threshold regression methods. We use the 2002 US birth data reported by the National Vital Statistics System to demonstrate the indexes. BNI and BDI are also well suited for making comparisons of birth outcomes in other reference populations. Standardized scores are proposed for such comparisons.