Development and validation of an infant morbidity index using latent variable models.

Birth defect, abnormal condition of the newborn, developmental delay or disability and low birth weight are four major infant morbidity outcomes. Most studies have focused on assessment of the effects of risk factors on each of these outcomes or of the relationship among these outcomes or both. Little attention has been paid to the development of a composite index, which is a summary construct of infant morbidity outcomes. In this paper, we develop extended latent variable (LV) models and modified Gauss-Newton algorithms for multiple multinomial morbidity outcomes with complete responses. By assuming the marginal distribution of the LV to be log-normal, we model the conditional probability of each outcome as a nonlinear function of the LV, which has properties similar to the logistic function. The estimated generalized nonlinear least-square method is used to solve equations for parameters of interest. The models are applied to an infant morbidity data set. A new single variable, called infant morbidity index (IMI) that functions as a summary of four infant morbidity outcomes and represents propensity for infant morbidity, is developed. The validity of this index is then assessed in detail. It is shown that the IMI is correlated with each of the individual outcomes, with infant mortality and with a face-valid index of morbidity outcomes, and can be used in future research as a measure of propensity for infant morbidity.