Human-Centered Design of a Multistakeholder Reporting Dashboard for Disseminating HIV Implementation Research in the...

Interactive dashboards have expanded public access to HIV surveillance and programmatic data, but existing tools do not synthesize implementation research evidence into formats accessible to the diverse stakeholders who could use it. Designing such tools requires deliberate alignment with users' needs, language, and decision-making contexts.

We describe the human-centered design (HCD) process used to develop and conduct a formative evaluation of the Ending the HIV Epidemic (EHE) Implementation Science (IS) Reporting Dashboard, a public-facing tool intended to synthesize HIV implementation research conducted in EHE priority jurisdictions, and to identify generalizable lessons on translating IS evidence for diverse academic and nonacademic audiences.

Guided by the Design Council's Double Diamond framework, the Johns Hopkins HIV IS Hub used iterative Discover, Define, Develop, and Deliver phases to conceptualize, prototype, and refine the dashboard. Initial consultations with representatives from the National Institutes of Health and a national HIV IS coordination team identified 5 primary user groups: funders, academics, community partners, implementing partners, and domestic HIV IS experts. The Define phase used a purposeful working corpus of 15 EHE IS publications to develop a theory-informed, inductively refined extraction structure. From June 2024 to April 2025, we conducted 14 virtual user-testing sessions across 3 rounds, using purposeful recruitment through established professional networks. Round 1 included 4 academics; rounds 2 and 3 each included 5 participants, with 1 representative from each user group. Sessions used a semistructured dashboard walkthrough, were recorded and transcribed, and were analyzed by 3 team members using rapid qualitative analysis. Notes and transcript observations were organized into structured insight-translation tables and synthesized through team consensus meetings.

The HCD process produced a dashboard concept, a user-group and use-case framework, a data extraction architecture, wireframes, and a prototype organized around study design, populations, outcomes, and community engagement. User testing generated 5 cross-cutting themes: IS terminology created translation barriers; users needed clearer orientation to the dashboard's purpose and audience; nonacademic users wanted plain-language, action-oriented summaries of research findings; representing community organizations and population affiliations required attention to consent and political sensitivity; and sustained use would require active dissemination beyond launch. These findings informed revisions to navigation, terminology, glossary features, plain-language study profiles, consent-based organization profiles, and dissemination planning.

This study illustrates how HCD can inform both interface design and upstream evidence synthesis decisions when developing dissemination tools for implementation research. We offer practical lessons on translating IS evidence for diverse academic, public health, and community audiences.