#United States #Gilbert #Arizona State University #Drive Health #Nurse Assist

Drive Health and Arizona State University Collaborate for Safer Clinical AI

Drive Health has reached a significant milestone in its partnership with Arizona State University's Edson College of Nursing and Health Innovation. This collaboration has resulted in the Nurse Assist Initiative, a groundbreaking approach to the evaluation and credentialing of clinical artificial intelligence (AI) that is designed to augment, rather than replace, human nurses.

As healthcare institutions increasingly integrate AI, a pressing question arises: how can we ensure that these AI assistants are safe, trustworthy, and prepared to assist healthcare professionals in patient care? The Nurse Assist Initiative addresses this concern by establishing a new framework focused on evaluating AI using principles and assessments that closely mimic those used in nursing education.

Evaluation Beyond Tradition

Traditional AI evaluations often rely on standard metrics, focusing on knowledge and reasoning capabilities. However, the Nurse Assist Initiative introduces a holistic assessment model that evaluates AI performance in complex and realistic patient-care scenarios, similar to how nursing students are tested in their education programs.

AI assistants participating in this initiative must not only exhibit clinical knowledge but also demonstrate their ability to handle simulated clinical cases, effectively communicate with patients, and collaborate with healthcare teams while adhering to specific governance protocols.

The initiative is built on a fundamental principle: AI should bolster the nursing workforce rather than work against it. Kevin Longoria, the CEO and Co-Founder of Drive Health, emphasizes that the future of healthcare lies in a partnership between humans and AI, rather than a competitive relationship. He notes, 'The responsibility lies with technology companies to demonstrate that their systems can be integrated into patient care settings without compromising safety.'

Foundations of Clinical Safety

Central to the Nurse Assist Initiative is Drive Health’s proprietary Deterministic Clinical Engine (DCE), a governance-first architecture tailored for healthcare settings. The DCE operates on deterministic pathways and evidence-based guidelines, integrating human oversight to ensure safe patient interactions. This design negates the risks associated with generalized conversational AI, completely avoiding the possibility of hallucinations that could endanger patient safety.

Drive Health has already gathered substantial evidence highlighting the efficacy of this system, with achievements including:

• - Over 980,000 audited interactions.
• - 100% identification of high-risk (R3) events.
• - 99.7% accuracy for moderate-risk (R2) events.
• - Zero missed critical escalations and a false-positive escalation rate of only 0.7%.

This framework aims to enhance clinical judgment rather than replace it, serving as a supportive tool that identifies when human intervention is needed. The collaboration with ASU marks the evolution of this initiative from operational confirmation to academic evaluation and focused research.

Crafting New Standards for AI Competency

Over the past year, efforts have intensified to rigorously evaluate the Nurse Assist platform through a series of assessments mirroring nursing education methods. Activities included nursing examinations, clinical reasoning exercises, standardized patient simulations, and continuous research exploring clinician and patient trust in AI-assisted care.

Together, these efforts are laying the groundwork for what Drive Health hopes will be the new standard for responsible healthcare AI, anchored in evaluations by accredited institutions like Arizona State University. This innovative framework shifts focus from merely answering questions to ensuring safe and effective support for healthcare providers in real clinical situations.

Dr. Judith Karshmer, Dean of ASU's Edson College, notes the importance of involving nurses in technology development, stating, 'A nurse's expertise is vital to ensure that these tools are ethical, practical, and compatible with the realities of clinical care.' The goal is to make AI a powerful and reliable tool in nurses' arsenals.

Laying the Groundwork for Responsible AI

As the healthcare landscape continues to evolve with rapid advancements in artificial intelligence, Drive Health and Arizona State University recognize that merely achieving technical excellence is not enough. Future clinical AI must also prioritize safety, transparency, governance, and verifiable competency before making its way into patient-centered environments.

In their collaborative efforts, Drive Health and ASU aim to create a scalable framework that empowers health systems, academic institutions, regulators, and technology developers to consistently and responsibly evaluate AI. Dr. Sherine Gabriel, Executive Vice President of ASU Health, underscores the necessity of building trust in AI technologies, highlighting the shared aim of developing an evidence-based evaluation strategy to inform AI deployment in clinical settings.

Conclusion

As the Nurse Assist Initiative progresses, Drive Health and ASU are poised to refine the credentialing processes through ongoing research and validation efforts. While initially focused on nursing, the implications for this initiative are vast and extend beyond a singular institution or technology. A standardized evaluation framework for AI in healthcare can significantly influence regulations, insurance, accreditation, and reimbursement pathways.

In a world where AI continues to shape healthcare, initiatives like Nurse Assist represent crucial steps toward establishing the trustworthy framework necessary for responsibly integrating AI into clinical practice and beyond.