Estimating incidence from prevalence in generalised HIV epidemics: methods and validation

Area of Science:

• Epidemiology
• Biostatistics
• Public Health

Background:

• Traditional HIV surveillance in Africa relies on antenatal clinic data, which may not reflect general population trends.
• Estimating HIV incidence in the general population is crucial for understanding and managing generalized epidemics.
• Increasing availability of repeated cross-sectional HIV prevalence data offers new opportunities for incidence estimation.

Purpose of the Study:

• To develop and validate methods for estimating HIV incidence using repeated cross-sectional prevalence data.
• To provide tools for more accurate monitoring of HIV epidemics in general populations.

Main Methods:

• Developed two novel methods to decompose changes in HIV prevalence into incidence and mortality components.
• Method 1 utilizes cohort mortality rates; Method 2 incorporates survival information post-infection.
• Validated methods using simulated data from mathematical models and real-world data from African cohort studies.

Main Results:

• Both developed methods accurately estimate HIV incidence rates and their changes across various epidemic conditions.
• Method 1 requires local mortality data, while Method 2 offers broader applicability.
• Estimates from both methods align well with actual HIV incidence measurements in adults and young people, capturing age-specific patterns.

Conclusions:

• Estimating HIV incidence from cross-sectional prevalence data is feasible and accurate enough for epidemic monitoring.
• These methods are applicable in contexts with mature, generalized HIV epidemics, such as southern and eastern Africa.
• The selection of the appropriate method depends on the availability of local HIV mortality data.