A1403
Title: Statistics integration of health survey data for estimating disease spatial patterns
Authors: Samuel Manda - University of Pretoria (South Africa) [presenting]
Abstract: In a two-phasing health survey sampling, disease data may only be observed in a subsample (nested data). The reduced sample could only be used for robust estimates of the disease spatial patterns at high levels of administrative aggregation. However, disease spatial patterns are increasingly needed at local levels for public health decision-making. Statistical methods are considered for imputation of the disease data for the remaining sample or independent samples that collect auxiliary information (non-nested structure). Combining information from different sources to obtain an improved official health estimate and association could be desired by health policymakers to reduce public health costs. A spatial analysis of adult HIV in sub-Saharan Africa exemplifies the methodology.
