Study Design, Methods, and Modeling in Networks to Inform HIV Interventions and Policy in Marginalized Populations

Area of Science:

• Public Health Research
• Network Science
• Causal Inference

Background:

• The Networks and Causal Inference for Public Health Research (NCIPHER) Lab was established to develop methods for evaluating interventions in connected populations.
• Interventions in real-world settings must account for social and healthcare network structures that influence health outcomes.

Purpose of the Study:

• To develop and apply methodological and computational approaches for estimating causal intervention effects in networked populations.
• To evaluate the impact of interventions considering spillover effects through social and healthcare networks.

Main Methods:

• Integration of empirical and simulation methods using bioinformatics and high-performance computing.
• Estimation of causal intervention effects, including indirect effects (spillover) on individuals not directly treated.
• Analysis of intervention effectiveness in HIV and opioid use disorder (OUD) contexts.

Main Results:

• Increased contact exposure to HIV risk alerts correlated with reduced unsafe injection behaviors.
• Widespread medication for opioid use disorder treatment in networks significantly reduced reports compared to limited treatment.
• Accounting for network spillover enhances the understanding of intervention effectiveness.

Conclusions:

• Network-based approaches are essential for accurately assessing public health intervention effectiveness.
• Understanding spillover effects improves the translation of interventions to population-level health impacts.
• This research supports evidence-based policy for public health initiatives in networked environments.