Developing HIV risk prediction tools in four African settings.

Area of Science:

• Public Health
• Epidemiology
• Infectious Disease Research

Background:

• HIV risk prediction tools are vital for pandemic control.
• Identifying individuals at high risk of prevalent and incident HIV is crucial, especially in African settings.

Purpose of the Study:

• To develop and validate tools for identifying individuals at highest risk of prevalent and incident HIV in an African context.
• To assess the performance of the VOICE risk score for predicting HIV incidence.

Main Methods:

• Logistic and Poisson regression were used to identify risk factors for HIV prevalence and incidence in a high-risk cohort.
• Identified factors were used to create and validate HIV risk prediction tools.
• The VOICE risk score's performance in predicting HIV incidence among women was evaluated.

Main Results:

• A prevalent HIV prediction tool was developed with good predictive ability (AUC=0.70).
• Key predictors for prevalent HIV included age, sex, drug use, unprotected anal sex, and partner's occupation.
• An HIV incidence prediction tool could not be validated; the VOICE risk score performed poorly (AUC=0.58).

Conclusions:

• A validated tool for prevalent HIV prediction can enhance HIV diagnosis and care linkage in sub-Saharan Africa.
• Existing incident HIV prediction tools may require adaptation with context-specific factors for different settings.