Improving Estimation of the Infant Mortality Rate

说明（由申请人提供）：婴儿死亡率（IMR）是衡量人口健康的一个基本指标，已得到广泛研究。该项目将研究一个关键的测量问题，许多研究，研究婴儿死亡率：如何在活产分类的变化影响测量婴儿死亡率和比较婴儿死亡率跨时间段，组，和地方？具体而言，除非首先将死亡归类为活产，否则死亡不会对IMR产生影响，而这种分类可能存在一些模糊性。此外，随着医疗技术提高了存活率，医生可能会更加积极地管理边缘存活的胎儿，从而越来越多地将这些病例归类为活产。使用生命统计（VS）队列链接的出生/婴儿死亡文件，出生文件和胎儿eath文件，我们将追求两个具体的目标，第一个评估活产内分泌对IMR趋势的重要性，第二个评估IMR差异。我们建议使用最先进的统计方法和几种直接和间接的方法来研究这些问题。

DESCRIPTION (provided by applicant): The infant mortality rate (IMR) is a fundamental and widely studied measure of a population's health. This project will examine a key measurement issue for many studies that examine the IMR: how does variation in the classification of live births influence the measured IMR and comparisons of IMR across time periods, group, and place? Specifically, a death cannot contribute to the IMR unless it is first categorized as a live birth, and this classification may involve some ambiguity. Moreover, as medical technology improves survival at the margins of viability, doctors may become more aggressive in the management of marginally viable fetuses, thereby increasingly classifying these cases as live births. Using Vital Statistics (VS) cohort linked birth/infant death files, birth files, and fetal eath files, we will pursue two specific aims, the first assessing the importance of live birth endogeneity to IMR trends and the second to IMR disparities. We propose to use state-of-the-art statistical methods and several direct and indirect methods to examine these issues.