SUCCESS STORY

Accelerating Quality & Safety Case Reviews Through Scalable AI-Assisted Workflows

A health system sought to reduce the time and burden of reviewing complex clinical cases related to hospital-acquired conditions and infections. Impact Advisors leveraged deep healthcare expertise, a pragmatic AI implementation approach, and a platform-first strategy. Before investigating custom AI solutions, Impact Advisors first evaluates the capabilities within the client’s core platforms, then builds custom solutions only where true workflow gaps exist—ensuring clients receive the most effective, sustainable solution for their needs.
Depositphotos 395368276 web

Infection prevention teams at a health system were routinely spending hours manually reviewing charts spanning hundreds of pages, which limited time for education, rounding, and proactive quality improvement efforts.

After establishing the organization’s broader AI strategy, roadmap, and governance model, Impact Advisors designed and implemented custom scalable large language model (LLM) applications to streamline case review processes.

The health system engaged Impact Advisors to develop three custom LLM solutions to address Quality & Safety needs:

  • A CLABSI (central line-associated bloodstream infection) LLM tool to automatically compare chart data with NHSN criteria
  • A HAPI (hospital-acquired pressure injury) LLM solution to streamline pressure injury case reviews and documentation validation
  • A USNWR (U.S. News & World Report) LLM application to expedite case reviews supporting national performance rankings
 
Each solution synthesized structured and unstructured EHR data— including labs, flowsheets, and clinical notes—and applied relevant clinical logic to produce a concise summary for reviewers in under a minute.

A Capability-First Approach

Impact Advisors determined that the Quality & Safety domain was particularly well-positioned for a custom LLM application because it represented a niche workflow gap not addressed by the client’s core platforms, and it had dedicated stakeholders and reliable source data to support solution development and feedback. 

The Impact Advisors team built the tools using a capability-first approach. Rather than developing separate one-off solutions, the team designed a scalable application for AI-assisted chart abstraction. The CLABSI model served as the foundation, then expanded to HAPI and USNWR reviews, demonstrating how a single AI capability can accelerate multiple Quality & Safety workflows. 

The implementation followed a structured three-phase methodology:

  1. Lab Development Phase – Understanding clinical workflows, defining value measures, and building an MVP (minimally viable product) with clinician validation.
  2. User Validation Phase – Embedding the tool within real user workflows, refining logic, and validating outcomes.
  3. Factory/Adoption Phase – Establishing education, change management, workflow integration (e.g., within the EHR), and sustained governance to support long-term use.
 
A “two-in-the-box” model paired each Impact Advisors role (i.e., Delivery Lead, Technical Lead, and Data Engineer) with a corresponding client counterpart to ensure knowledge transfer and future autonomy. By the end of the engagement, the client could independently maintain and expand the solutions.

Positive Impact

The engagement resulted in measurable value in both efficiency and clinical workflow experience. By reducing the time required to review complex charts, the LLMs helped reinvest clinician time into infection prevention rounding and quality improvement activities.

Graphic for Lurie LLM Case Study

A strong value measurement and adoption strategy supported sustained success. The client and Impact Advisors collaborated to collect baseline and post-implementation data, including time savings, reinvestment of clinician time, user experience feedback, and longer-term outcomes such as reductions in HAC (hospital-acquired condition) and HAI (healthcare-associated infection) incidence. The Impact Advisors team also implemented tailored AI education and hands-on training programs to support user proficiency and confidence.

Strong clinical engagement, leadership advocacy, and a culture of co-development contributed to high adoption and trust in the new workflows. The scalable model built by Impact Advisors now serves as a blueprint for future AI-enabled clinical review processes across the health system.

Download PDF