Large Country – Lot Quality Assurance Sampling: Health surveys for local and regional program management

Sampling theory provides a means of economically, effectively and rapidly measuring a population. The focus of the inference is sometimes a large body, such as a province or a nation. However, inference about smaller units is more beneficial if the outcome of the survey is action at the local level. It is often not possible to disaggregate the national survey to obtain local information since that was not the intent of the original survey design. We present Large Country - Lot Quality Assurance Sampling, a method for obtaining local level information, with a desired level of accuracy, for policy decisions--we use Lot Quality Assurance Sampling--and subsequently, combining these local samples to provide a national, or aggregated survey. Attention is paid to the statistical laws of surveys to ensure that the resulting inference is valid. The proposed methodology is depicted by an example. We show how the HIV/AIDS, Malaria, Sexually Transmitted Diseases and Tuberculosis project in Eritrea provides both classifications of local health departments into performance categories and overall provincial estimates of coverage, with associated confidence intervals, for various health indicators of interest.