Blab AI
HCA Healthcare Uses AI to Streamline Nursing Handoffs and Scale Innovation Across 189 Hospitals
Imagine a nurse’s shift ending with a single click that guarantees no detail slips through. That’s the promise behind HCA Healthcare’s latest AI‑driven tool, a system designed to turn the high‑risk, high‑volume handoff process into a concise, reliable exchange.
The United States’ largest for‑profit health system is taking a systematic approach to digital transformation. Senior vice president and chief transformation officer Dr. Michael Schlosser explained that HCA’s digital‑transformation team is deploying an AI‑powered nurse‑handoff tool in eight hospitals today, with plans to roll it out network‑wide. The initiative is part of a broader strategy that leverages HCA’s scale, data volume, and governance framework to move from pilot projects to system‑wide adoption.
HCA’s size is a key asset for AI work. The system operates 189 hospitals and about 2,600 ambulatory sites, generating roughly 47 million patient encounters each year. According to Dr. Schlosser, that volume provides a rich data set for training models and for testing new solutions in a variety of care settings.
The organization also runs “innovation hubs” in three hospitals that serve as test beds for new ideas. When a pilot succeeds, HCA’s operational infrastructure—regional and hospital‑level teams, IT platforms, and a data office—supports rapid scaling.
The nurse‑handoff AI was born from frontline nurses who identified handoffs as a source of error and inefficiency. The team used large language models (LLMs) to generate concise summaries of a patient’s status, drawing on electronic health record (EHR) data and nurse notes. Over a year of iteration, the model’s performance metrics—cohesiveness, accuracy, conciseness and usefulness—reached the high 90s. The tool is now live in eight hospitals and is designed to be interactive, allowing nurses to flag inaccuracies and provide feedback.
Because LLMs can produce hallucinations, the system incorporates a human‑in‑the‑loop approach. Nurses are trained to review and correct the AI output, and the feedback is captured for continuous improvement. Dr. Schlosser noted that the organization is developing automated pipelines to feed this feedback back into the model, while a quality‑assurance team monitors performance over time.
Responsible AI is a core component of HCA’s approach. The company established a senior‑level Responsible AI Committee that includes legal, privacy, ethics and compliance leaders. The committee reviews each application against principles of transparency, privacy, security and robustness before it is tested or deployed. It also conducts ongoing governance, recognizing that models can drift and that patient populations evolve.
Scaling pilots to full deployment requires more than a working prototype. Dr. Schlosser emphasized that change management is embedded from day one. The implementation team collaborates with end users to adapt workflows, refine interfaces and provide training. The organization monitors adoption rates and, when uptake is lower than expected, investigates barriers and adjusts the rollout plan. The process is iterative; AI models are updated as new data arrive, and the infrastructure supports continuous learning.
Value tracking is integrated into every project. A small team of biostatisticians and data analysts evaluates potential impact during the discovery phase and measures outcomes after deployment. The team considers financial returns as well as patient safety, quality, experience and operational efficiency. Dr. Schlosser said that demonstrating measurable value is essential for maintaining momentum and securing resources.
Looking ahead, HCA Healthcare plans to expand its AI portfolio beyond nursing handoffs. The organization is exploring ambient documentation that records doctor‑patient encounters, sepsis‑detection models, and maternal‑fetal health monitoring. Dr. Schlosser expressed optimism that AI can reduce administrative burden for clinicians, allowing them to spend more time with patients. He also noted that patients will gain access to AI‑enabled tools that improve engagement and health literacy.
In sum, HCA Healthcare’s AI strategy combines data scale, rigorous governance, iterative development, and a focus on real‑world impact. The organization’s experience offers a roadmap for other large health systems seeking to deploy AI safely and effectively across complex, regulated environments.
The United States’ largest for‑profit health system is taking a systematic approach to digital transformation. Senior vice president and chief transformation officer Dr. Michael Schlosser explained that HCA’s digital‑transformation team is deploying an AI‑powered nurse‑handoff tool in eight hospitals today, with plans to roll it out network‑wide. The initiative is part of a broader strategy that leverages HCA’s scale, data volume, and governance framework to move from pilot projects to system‑wide adoption.
HCA’s size is a key asset for AI work. The system operates 189 hospitals and about 2,600 ambulatory sites, generating roughly 47 million patient encounters each year. According to Dr. Schlosser, that volume provides a rich data set for training models and for testing new solutions in a variety of care settings.
The organization also runs “innovation hubs” in three hospitals that serve as test beds for new ideas. When a pilot succeeds, HCA’s operational infrastructure—regional and hospital‑level teams, IT platforms, and a data office—supports rapid scaling.
The nurse‑handoff AI was born from frontline nurses who identified handoffs as a source of error and inefficiency. The team used large language models (LLMs) to generate concise summaries of a patient’s status, drawing on electronic health record (EHR) data and nurse notes. Over a year of iteration, the model’s performance metrics—cohesiveness, accuracy, conciseness and usefulness—reached the high 90s. The tool is now live in eight hospitals and is designed to be interactive, allowing nurses to flag inaccuracies and provide feedback.
Because LLMs can produce hallucinations, the system incorporates a human‑in‑the‑loop approach. Nurses are trained to review and correct the AI output, and the feedback is captured for continuous improvement. Dr. Schlosser noted that the organization is developing automated pipelines to feed this feedback back into the model, while a quality‑assurance team monitors performance over time.
Responsible AI is a core component of HCA’s approach. The company established a senior‑level Responsible AI Committee that includes legal, privacy, ethics and compliance leaders. The committee reviews each application against principles of transparency, privacy, security and robustness before it is tested or deployed. It also conducts ongoing governance, recognizing that models can drift and that patient populations evolve.
Scaling pilots to full deployment requires more than a working prototype. Dr. Schlosser emphasized that change management is embedded from day one. The implementation team collaborates with end users to adapt workflows, refine interfaces and provide training. The organization monitors adoption rates and, when uptake is lower than expected, investigates barriers and adjusts the rollout plan. The process is iterative; AI models are updated as new data arrive, and the infrastructure supports continuous learning.
Value tracking is integrated into every project. A small team of biostatisticians and data analysts evaluates potential impact during the discovery phase and measures outcomes after deployment. The team considers financial returns as well as patient safety, quality, experience and operational efficiency. Dr. Schlosser said that demonstrating measurable value is essential for maintaining momentum and securing resources.
Looking ahead, HCA Healthcare plans to expand its AI portfolio beyond nursing handoffs. The organization is exploring ambient documentation that records doctor‑patient encounters, sepsis‑detection models, and maternal‑fetal health monitoring. Dr. Schlosser expressed optimism that AI can reduce administrative burden for clinicians, allowing them to spend more time with patients. He also noted that patients will gain access to AI‑enabled tools that improve engagement and health literacy.
In sum, HCA Healthcare’s AI strategy combines data scale, rigorous governance, iterative development, and a focus on real‑world impact. The organization’s experience offers a roadmap for other large health systems seeking to deploy AI safely and effectively across complex, regulated environments.
