A Multisource, Multilevel Contextual Database to Support Ending the HIV Epidemic Initiative: Protocol for Design, Construction, Harmonization, and Quality Assurance

The Ending the HIV Epidemic (EHE) initiative remains a national priority in the United States, aiming to reduce new HIV infections by 90% by 2030. As we cross the initiative's midpoint, there is renewed commitment to strengthening the HIV workforce's capacity to plan, implement, and sustain effective HIV prevention, treatment, and care interventions. Despite substantial improvement in HIV outcomes, uneven implementation of evidence-based interventions reflects persistent gaps between available evidence and its translation into locally actionable practice. Achieving EHE goals requires tailoring implementation to the diverse sociostructural conditions shaping HIV outcomes across jurisdictions. Research increasingly highlights the value of integrated, contextual data to strengthen public health decision-making. Linking indicators spanning multiple conceptual domains across various geographic levels can support a robust understanding of the distinct drivers of HIV outcomes, yet existing data systems remain fragmented. A harmonized, multisource, multilevel database is therefore essential to support targeted, needs-based, and data-driven implementation under the EHE initiative. We present a reproducible protocol detailing the transferable methodology of building and maintaining such a resource.

This project aims to (1) build a high-quality contextual database integrating multiple data sources using transparent, replicable, and updatable methods and (2) develop and document systematic workflows for ongoing updates and quality assurance (QA) and support future use aligned with open-science frameworks and standard data practices.

This project will follow best practices in data architecture, acquisition, standardization, and QA. For objective 1, we will integrate data across multiple geographic levels (eg, ZIP code, county) for 2020-2025, with measures categorized into conceptual domains (eg, epidemiologic and sociodemographic) guided by established theoretical frameworks. For objective 2, we will develop a tiered data structure using a GitHub repository to store documentation, processing scripts, and QA logs aligning with open science practices. Construction and QA methods were informed by targeted PubMed literature reviews. Data sources include existing data repositories, literature-identified datasets, and gray literature with consistent formatting and permissive terms of use suitable for web scraping. Stakeholder engagement will span all phases of database development, informing variable selection, usability, and validation to enable iterative refinement and revision.

Funding was obtained in August 2025. Literature reviews from October to November 2025 informed the database construction methods, source identification, and protocol development. Data acquisition, standardization, and QA began in May 2026, with acquisition remaining iterative. Cross-source integration, final QA checks, and dashboard development are expected to conclude in October 2026. Study results are anticipated for publication in summer 2027.

This protocol will provide a reproducible, transferable model for building and maintaining a contextual data resource that could support public health planning and advance implementation science through more context-responsive decision-making under the EHE initiative.