Improving Sampling Efficiency for Determining Pediatric HIV Prevalence in National Surveys: Evidence From 8...

Area of Science:

• Epidemiology
• Biostatistics
• Public Health

Background:

• Measuring mother-to-child HIV transmission via population-based surveys necessitates large sample sizes due to low child HIV prevalence.
• Estimating HIV prevalence in children requires efficient survey designs.

Purpose of the Study:

• To evaluate the potential improvements in sampling efficiency offered by a targeted sample design for mother-to-child HIV transmission surveys.
• To compare the precision and sample size requirements of a targeted design versus a standard design.

Main Methods:

• Analysis of data from Population-based HIV Impact Assessment (PHIA) surveys in eight sub-Saharan African countries.
• Comparison of a standard geographically stratified 2-stage sample design with a 2-phase disproportionate sample design targeting strata based on maternal HIV status and mortality.

Main Results:

• Child HIV prevalence varied from 0.4% in Tanzania to 2.8% in Eswatini.
• The targeted sample design demonstrated improved precision across all analyzed countries.
• Reductions in mean squared error ranged from 27% to 61% with the targeted design, assuming equal sample size.

Conclusions:

• Population-based surveys measuring adult and child HIV prevalence should consider targeted child sampling strategies.
• Targeted sampling can reduce sample size, enhance precision, and increase the number of positive cases identified.
• Findings from PHIA surveys provide valuable baseline data for future survey design optimization.