Abstract:
|
Health disparities are differences in health status across different racial, ethnic and socioeconomic groups. Analytic methods, such as the delta method, are commonly used to estimate the standard errors of estimated measures of health disparities, and to construct confidence intervals for these measures. However, the confidence intervals constructed using these analytic methods do not have good coverage properties for situations involving sparse data. In this presentation we introduce a new method to construct confidence intervals for measures of health disparities based on approximate fiducial quantities. We use an extensive simulation study to compare the coverage properties of the confidence intervals constructed using the two types of methods, and we also illustrate the use of these methods on several real data sets.
|