About Emory

Emory is a not-for profit organization based in Atlanta, Georgia. The system includes 10 hospitals and the Emory Clinic (a physician group with 490+ provider location), all of which are non-unionized.

Emory’s AI Governance

In 2023, Emory nursing leadership created their Center for Care Delivery and Innovation, a designated department solely focused on driving four critical outcomes: attracting talent, retaining staff, caring for employees and patients, and making it easy to practice. The Center is led by the system’s Vice President Center of Care Delivery and Innovation (a former CNO) and supported by a nurse scientist dedicated to innovation and implementation nursing science. The Center collaborates with Emory Digital teams to implement new innovations into the nursing workflow seamlessly.

Among other care transformation work, Center leaders lead (or co-lead) AI-enabled care delivery innovation with the Emory digital team. For example, the Center’s VP also serves as co-chair on several system councils that play a critical role in deciding AI strategy. Those councils include:

• Digital Experience Council: a subgroup of the system practice council, which prioritizes Emory’s digital and IT initiatives and provides recommendations on practice implications.

• Interprofessional Predictive Model Group: comprised of nursing and physician leaders, this group trains and reviews AI models with a focus on safety, equity, and accuracy. They also manage model approvals and deployment readiness.

• Adoption Workgroup: engineers and representatives from the school of medicine provide workflow insights and design deployment processes and protocols.

In addition to their in-house resourcing, Emory partners with vendors including Impact Advisors, VirtuSense, and Abridge to co-design system AI solutions.

Emory’s (Current) AI Initiatives

Emory is currently working on AI-powered models for the following use cases:

• Sepsis prediction: This decision support model provides a score that identifies patients at risk of sepsis and alerts the bedside team of high-risk patients

• Deterioration index: AI leverages EHR data to detect (and alert clinicians to) emerging signs of clinical deterioration

• Readmission risk: AI leverages EHR data to identify patient readmission risk, especially for recently discharged, chronically ill patients

• Operating Room efficiency: AI leverages EHR data to identify who have a high probability of not showing up for their scheduled appointments. This data is used to drive reminder outreach and/or find alternate transportation options for patients who have a higher no-show probability.

EmChat Pilot Summary

Emory’s nursing and Emory Digital teams partnered with Impact Advisors in 2024 to create EmChat, an AI-augmented LLM chatbot. EmChat is designed for two purposes:

1. To support bedside staff to quickly find organizational protocols and policies

2. To allow infection preventionists to rapidly verify CLABSI cases

Leaders conducted proof-of-value pilots with several service lines cross the organization. Nine months later, Emory not only established proof of value, but fully integrated EmChat within the system. CLABSI cases have rolled out the clinical-reference development to one of their hospitals.

Internal Policy Look-Up

Emory discovered that it could take nurses up to 45 minutes to locate internal policies and protocols. To not delay care for their patients, nurses typically relied on their peers or memory for answers rather than system guidelines. To provide an easier and more accurate alternative, nurses worked directly with Impact Advisors to create, test, and deploy the EmChat, a clinical reference AI-augmented LLM chatbot. When nurses ask EmChat a question regarding their organization’s internal policies, EmChat leverages AI to locate the correct information from Emory’s internal policy manager system and Epic learning home dashboard.

EmChat is designed to streamline workflows by saving nurses time and ensuring greater consistency in care delivery. Emory leaders also plan to use EmChat’s software policy look-up to proactively identify differences in policies between sites. With this information, Emory plans to create greater policy consistency across their system.

CLABSI Verification

In addition to its policy look-up function, EmChat also helps the system’s infection preventionists quickly find the data they need to verify infection. This process begins when a pre-existing alert mechanism within the EHR signals to IPs that a patient is at high risk of CLABSI. IPs ask EmChat “Does this patient have a CLABSI?”, which triggers the chatbot to pull the relevant clinical indicators IPs need to detect and verify the presence of a CLABSI (informed by the NHSN CLABSI criteria). Rather than having to search through the patient chart to identify CLABSI criteria, EmChat makes the information readily available.

Outcomes

Emory has completed the pilot proof of value and has now moved into the industrialization phase, with plans to scale both models widely across the health system.

Their outcomes to date include:

• Reduction in RN and clinician time spent searching for internal clinical reference

• Reduction in time spent to verify the presence of a CLABSI

• 15+ min estimation of clinician time saved per EmChat augmented task

Acknowledgements and Implementation Rolodex

Nursing Catalyst would like to thank the following health system leaders for their time, insight, and generosity in sharing the outcomes of their impactful work.

For more information on these pilots, please reach out to them at:

• Nancye Feistritzer, DNP, RN, NEA-BC, Vice President Center for Care Delivery and Innovation, Emory Healthcare, EmChat, nancye.feistritzer@emoryhealthcare.org

• Monique Bouvier, PhD, RN, PNP-BC, Nurse Scientist, Emory Healthcare, Emchat, monique.bouvier@emoryhealthcare.org

• Jason Atkins, Vice President, Chief Clinical Informatics Officer, Emory Healthcare, Emchat, wjatkin@emory.edu