Key Population Size Estimation to Guide HIV Epidemic Responses in Nigeria: Bayesian Analysis of 3-Source...

Area of Science:

• Epidemiology
• Biostatistics
• Public Health

Background:

• Nigeria faces a significant HIV burden, disproportionately affecting key populations.
• Female sex workers, men who have sex with men, and people who inject drugs are highly vulnerable due to societal stigma and criminalization.
• Accurate population size estimates are crucial for effective HIV epidemic response and resource allocation.

Purpose of the Study:

• To improve the quality of population size estimates for key populations in Nigeria.
• To implement and demonstrate improved statistical estimation methods using empirical data.
• To provide reliable data for HIV program planning and intervention scale-up.

Main Methods:

• A 3-source capture-recapture methodology was employed across 7 Nigerian states from October to December 2018.
• Data collection involved 3 independent sampling rounds approximately one week apart at mapped hotspots.
• Bayesian nonparametric latent-class models were utilized to derive median population size estimates, accounting for factors influencing capture probabilities.

Main Results:

• Approximately 310,000 individuals were sampled across 9,015 hotspots.
• Estimated population sizes: female sex workers (14,500–64,300), men who have sex with men (3,200–41,400), and people who inject drugs (3,400–30,400).
• These estimates were consistently larger than previously documented figures for each key population in all studied states.

Conclusions:

• This study represents the first application of 3-source capture-recapture methods in Nigeria for these populations.
• The derived population size estimates are more reliable due to Bayesian models accounting for social visibility and heterogeneous capture probabilities.
• Larger population estimates underscore the urgent need to scale up HIV prevention and intervention programs for these high-risk groups.