Atrium Health, formerly Carolinas HealthCare System, is a not for profit hospital network which operates nearly 40 hospitals, six freestanding emergency departments, more than 30 urgent care centers, and medical practices in the American states of North Carolina, South Carolina and Georgia. About the majority of the hospitals affiliated with the system are located within 75 miles of Carolinas Medical Center, the system's flagship hospital and headquarters, in Charlotte, North Carolina. The system has over 55,000 employees. Legally, Atrium Health is The Charlotte-Mecklenburg Hospital Authority, a municipal hospital authority established under North Carolina's Hospital Authorities Act (North Carolina General Statutes chapter 131E, part 2). The authority is governed by a self-perpetuating board of commissioners which nominates new commissioners to fill its own vacancies; the chair of the Mecklenburg Board of County Commissioners can approve or veto those nominations but not make nominations of her own.
Objectives: Artificial intelligence (AI) offers health systems opportunities to enhance care delivery, improve efficiency, and expand patient access. However, rapid innovation introduces new risks requiring careful oversight. This study examines how diverse stakeholders shaped the design and early evaluation of the Framework for the Appropriate Implementation and Review of AI (FAIR-AI), a system-wide AI governance framework implemented within a large, multi-state health system. Methods: We conducted two rounds of semi-structured interviews-before FAIR-AI development and shortly after FAIR-AI was approved-with executive leaders (N = 5), risk/compliance/legal leaders (N = 11), and data developers (N = 8) to identify initial design needs and evaluate the approved framework. Pre-development interviews also included patients (N = 5) and clinicians (N = 5) to capture AI end-user expectations. Data were analyzed using thematic analysis and inductive and deductive coding methodologies. Results: Pre-development interviews highlighted three central priorities: balancing risk tolerance with potential benefits, ensuring direct human oversight, and streamlining review for low-risk solutions. Patients and clinicians emphasized the need for clinician control over care decisions, with AI serving as supplemental support. Post-approval interviews identified seven elements critical to success: (1) transparent and consistent reviews; (2) timely evaluations; (3) ongoing solution monitoring; (4) iterative framework refinement; (5) alignment with institutional priorities and regulatory standards; (6) multi-modal teammate education; and (7) diverse patient dissemination efforts. Conclusions: Our findings highlight the importance of AI governance frameworks integrating both pre-deployment risk assessment and post-implementation solution monitoring, while remaining adaptable through feedback loops and in response to changing regulatory and technological contexts. This stakeholder-informed approach provides practical guidance for responsible AI at enterprise scale.
BACKGROUND AND OBJECTIVES:Health-related social needs (HRSN) affect a wide range of short- and long-term outcomes, health care use, functioning, and quality of life. Although HRSN screening is valuable, it is likely ineffective unless coupled with interventions. This project integrated HRSN screening and intervention(s) across 9 inpatient and outpatient divisions, varied in composition, patient populations, process flows, and resource availability. METHODS:A quality improvement team with a standardized change package facilitated a multimodal initiative for divisions spanning hospital medicine, newborn nurseries, and specialty outpatient clinics at an urban, quaternary pediatric hospital system. We aimed for 80% of patient encounters to be screened for HRSN and that each positive screen receive an agreed upon resource intervention using a centralized resource bank. RESULTS:From January 2021 to October 2023, 31 834 screenings were conducted across 9 divisions. Performance increased to a mean 92%, with positive screens receiving interventions. The last 7 months of the project were sustained at 92% or higher. Food insecurity was identified in 17.6% of encounters (n = 10 007, 1765 positive), with a 56% decrease in prevalence on repeat screening after identification/intervention. CONCLUSIONS:A centralized quality team and change package can facilitate successful implementation of HRSN screening and connection to resources across multiple disciplines and sites. These interventions may lead to a decrease in subsequent HRSN positivity. Rescreening patients over time is important to capture the full spectrum of HRSN needs of a family.
Background: Code Stroke protocols are recommended to facilitate efficient evaluation and management of patients with suspected acute ischemic stroke (AIS) and expedite stroke treatment, namely thrombolysis and/or thrombectomy in eligible patients. However, the diagnostic and therapeutic performance of these pathways is under-reported. We evaluated the rates of stroke diagnosis and treatment in patients evaluated via Emergency Department (ED) Code Stroke protocols across four large hub and spoke stroke networks within one large health system. Methods: Anonymized Code Stroke data from four stroke networks including a total of 51 EDs under a single healthcare system were prospectively collected into databases. Data from adult (≥18 years old) ED stroke activations occurring between January 2024-March 2025 were analyzed. The primary response variables were the proportions of patients with stroke as their primary discharge diagnosis, as identified via ICD codes, and proportions of activated patients receiving intravenous thrombolysis and/or mechanical thrombectomy. Secondary analyses evaluated differences in diagnosis and treatment between hospital stroke capabilities, stroke network, and patient demographics. Results: 24,974 consecutive Code Stroke activations were evaluated. As described in the accompanying table, 23.1% of patients evaluated for suspected stroke were diagnosed with AIS. 5.6% of activated patients were treated for AIS via thrombolysis (4.3-9.6% across networks), 2.2% underwent mechanical thrombectomy (1.6-3.3%), and 1.3% received both therapies (0.9-2.2%). Further secondary analyses, including differences in diagnosis and treatment rates between hospitals by stroke-treatment capability and patient demographics, will be reported as part of the abstract presentation. Conclusions: Acute ischemic strokes are diagnosed in a minority of ED Code Stroke activations and rates of stroke treatment with thrombolysis and/or thrombectomy among activated patients remain low. Further research is needed to optimize stroke identification and treatment in patients presenting to the ED and better understand variation across a large healthcare system.