What is it about?

We compared statistical Methods that can be used to compute HIV prevalence estimates in not targeted geographical areas where survey based sample size of existing surveys yield estimates with wide confidence intervals, which are not reliable for decision making.

Featured Image

Why is it important?

Our findings supports use of existing datasets from surveys and routine service delivery to compute more precise prevalence estimates for non targeted geographical areas by national and regional surveys. National population surveys are designed and powered to generate estimates for decision making at national and regional levels only. Lower administration areas and regions may not have resources to conduct surveys to inform decision making at their levels. The methods presented in our work make use of all available data to obtain estimates with no additional cost of collecting data.


This research contributes to the body of work on survey methods which are not widely researched and used. Most researchers seek to design surveys, collect data and make inferences based on the data collected. Use of approaches such as the one in our work presents new perspectives of tackling public health questions/problems when they arise without incurring additional costs of planning for a survey, data collection and analysis. Our approaches also computes and presents the errors associated with each of the methods.

Joseph Ouma
University of the Witwatersrand

Read the Original

This page is a summary of: Model-based small area estimation methods and precise district-level HIV prevalence estimates in Uganda, PLoS ONE, August 2021, PLOS, DOI: 10.1371/journal.pone.0253375.
You can read the full text:



The following have contributed to this page