This Special Communication outlines a pragmatic, governance-aware schema for implementing digital twins (DTs) within a Learning Health System (LHS) and a companion Learning Research System (LRS), using multiple sclerosis (MS) as a translational model. Within the LHS, routine clinical data are transformed into actionable insights, while the LRS extends this process to discovery and iterative improvement. The NEurOinformatics Network (NEON) initiative initiated a DT framework for MS as a foundation for AI-enabled precision neurology. The DT integrates multimodal longitudinal data, including electronic health records, MRI, biomarkers, digital behavioral metrics, and registry datasets to model disease trajectories and guide individualized care. Synthetic data augment underrepresented cohorts, improving representativeness while safeguarding privacy. NEON’s multi-phase roadmap spans ideation through early clinical translation. Integrated within the Great Plains Institutional Development Award (GP IDeA) Clinical and Translational Research Network, the project advances data infrastructure, workforce development, and AI-driven translational science across IDeA states. Emerging insights highlight the importance of team science and institution-wide collaboration, while addressing barriers related to interoperability, digital biomarker validation, and limited AI workforce capacity. Collectively, this work positions DTs as scalable and generalizable assets for precision neurology within LHS and LRS environments.

Suicidal ideation and attempts are common in first-episode psychosis (FEP), but less is known about how these outcomes change in large, community-based, coordinated specialty care (CSC) settings. This longitudinal cohort study examined 2951 individuals with FEP enrolled in 87 CSC clinics across 15 states in the Early Psychosis Intervention Network between 2019 and 2024. Suicide-related outcomes were assessed at admission and every 6 months for up to 2 years. At admission, 34% reported suicidal ideation and 7% reported a suicide attempt in the previous 6 months. By 24 months, the odds of ideation and attempts had declined substantially. Younger age, female sex, White race, non-Hispanic ethnicity and longer duration of untreated psychosis (DUP) were associated with higher odds of ideation, with younger age, longer DUP and prior psychiatric hospitalisation being associated with higher odds of attempt. Six participants (0.2%) died by suicide. The findings support early detection and ongoing suicide risk monitoring during CSC.

Learning Health Systems (LHSs) enable data-driven improvement in clinical care. However, research processes that generate this evidence remain fragmented, limiting systematic learning across study portfolios and contributing to persistent translational inefficiencies. We introduce the Learning Research System (LRS) as a sociotechnical extension of the LHS that leverages LHS infrastructure to render research operations measurable, reusable, and governable across studies, positioning research as an object of continuous learning. Using targeted literature synthesis and supplementary searches, we traced the emergence of the LRS from foundational LHS scholarship and developed five core capabilities as proposed maturity indicators: cross-study measurement, reuse of cohorts and analytic assets, portfolio-level governance, feedback-enabled workflow optimization, and integration of research outputs into clinical and operational improvement. We illustrate the distinction between LHS and LRS roles across the NCATS Translational Science Spectrum. Finally, we applied the capabilities to a formative case at the University of Nebraska Medical Center through an institutional environmental scan, mapping them as indicators of emerging LRS maturity from an existing LHS. The LRS provides a framework for evaluating research-process learning, reducing fragmentation across translational stages, and addressing selected translational discontinuities, such as early-stage “valley of death” barriers between discovery and human studies.

Accelerating innovation translation is a priority for improving healthcare and health. Although dissemination and implementation (D&I) research has made significant advances over the past decade, it has attended primarily to the implementation of long-standing, well-established practices and policies. We present a conceptual architecture for speeding translation of promising innovations as candidates for iterative testing in practice. Our framework to Design for Accelerated Translation (DART) aims to clarify whether, when, and how to act on evolving evidence to improve healthcare. We view translation of evidence to practice as a dynamic process and argue that much evidence can be acted upon even when uncertainty is moderately high, recognizing that this evidence is evolving and subject to frequent reevaluation. The DART framework proposes that additional factors – demand, risk, and cost, in addition to the evolving evidence base – should influence the pace of translation over time. Attention to these underemphasized factors may lead to more dynamic decision-making about whether or not to adopt an emerging innovation or de-implement a suboptimal intervention. Finally, the DART framework outlines key actions that will speed movement from evidence to practice, including forming meaningful stakeholder partnerships, designing innovations for D&I, and engaging in a learning health system.

We review aspects of the recently released National Academies of Sciences, Engineering, and Medicine report A National Trauma Care System: Integrating Military and Civilian Trauma Systems to Achieve Zero Preventable Deaths After Injury most relevant to disaster health, particularly the concepts of focused empiricism and building a learning health system. The article references battlefield success utilizing these concepts and the emerging Disaster Research Response Program. We call upon disaster health researchers to apply the report’s recommendations to their work. (Disaster Med Public Health Preparedness. 2017;11:510–511)