Developing a dynamic HIV transmission model for 6 U.S. cities: An evidence synthesis

Area of Science:

• Epidemiology
• Mathematical Modeling
• Public Health

Background:

• Dynamic HIV transmission models are vital for guiding HIV/AIDS treatment and prevention strategies.
• Data availability for regional HIV microepidemics varies significantly, posing challenges for model calibration.
• This study focuses on a transparent evidence synthesis process for a compartmental HIV transmission model in six US cities.

Purpose of the Study:

• To describe a comprehensive evidence synthesis and reporting framework for populating and calibrating a dynamic HIV transmission model.
• To address data intensity challenges in HIV modeling by synthesizing existing evidence.
• To provide city-specific model parameters for six US cities.

Main Methods:

• A mixed-method evidence synthesis strategy was employed, covering population, transmission probability, treatment, prevention, costs, and utility parameters.
• Parameters were identified and stratified by city, gender, risk group, and race/ethnicity a priori.
• Parameter quality was ranked, and sources were verified with a scientific advisory committee.

Main Results:

• 1,667 parameters were synthesized from 59 publications and 24 reports, with primary analyses from 11 datasets.
• 91% of parameters were city-specific, with significant variation found across and within cities for different demographic groups.
• 39% of common parameters and 56%-60% of city-specific parameters were of best- to moderate-quality evidence.

Conclusions:

• Systematic reporting of evidence synthesis processes enhances the integration of modeling in decision-making.
• Explicit assessment of data quality strengthens model conclusions and reduces decision uncertainty.
• Prioritizing data collection for informative components of local HIV services is essential for robust model outcomes.