Health System Touts Huge GenAI Adoption
Parkview Health's CMIO says 88% of its providers called a recent generative AI rollout the "best EHR improvement they'd seen." In a recent podcast, Dr. Mark Mabus explained they scaled 18 new AI features by using physician builders and super-users for testing and training. The focus was on softer ROI, like reducing burnout, which proved more meaningful than initial financial metrics.
Parkview Health has been on Epic's Honor Roll for effective EHR implementation for five consecutive years, and was one of only 23 organizations to make the latest version of the honor roll out of 461 eligible organizations. The health system has been a leader in partnering with Epic to test and implement new features. This long-standing relationship provides a foundation for rolling out new technologies like generative AI. Epic's generative AI tools are integrated directly into the EHR to assist with tasks like drafting patient portal messages, formulating nurse handoffs, and creating patient summaries. For nurses, this can mean an 85% faster creation of end-of-shift notes. One early adopter, Mayo Clinic, reported that the AI tool saves nurses about 30 seconds per patient message while also generating more empathetic responses. However, AI is not a panacea for documentation issues. If existing nursing workflows are duplicative or inefficient, generative AI will expose those problems rather than fix them. Many nurses report that EHRs are designed around physicians, leading to click fatigue and redundant data entry. In fact, studies show that over two-thirds of nurses feel that poor EHR usability and the burden of digital documentation contribute to job dissatisfaction. For ICU nurses transitioning to informatics, understanding these end-user frustrations is key. A common path into the field is to serve as a subject matter expert on a technology team, where deep clinical knowledge is essential for designing systems that work for clinicians. This hands-on experience can be more valuable initially than a master's degree. To formalize informatics expertise, the American Nurses Credentialing Center (ANCC) offers the Nursing Informatics Certification (NI-BC). Eligibility generally requires a BSN, two years of RN experience, 30 hours of informatics continuing education, and a minimum of 1,000-2,000 hours of informatics practice. A core technical concept for informaticists is interoperability, driven by standards like HL7 FHIR (Fast Healthcare Interoperability Resources). FHIR uses modern web-based APIs to allow different health IT systems to exchange data securely and efficiently, which is a key mandate of the 21st Century Cures Act. This framework is critical for enabling everything from patient access to their own data to advanced AI-driven clinical decision support. Federal regulations from the ONC and CMS are major drivers of health IT priorities. These rules mandate the use of standardized APIs to prevent "information blocking" and ensure patients can access their health information from different providers and payers. This regulatory push forces healthcare organizations and EHR vendors to prioritize seamless data exchange. Parkview Health has also utilized predictive AI to improve patient outcomes and operational efficiency. One model they developed identifies patients likely to need post-acute care within 24 hours of admission. By proactively starting discharge planning for these high-risk patients, they reduced the average length of stay by over half a day and saved an estimated $7.5 million in one year.