Why AI matters at this scale

University of Utah Health is a major academic medical center and health system serving the Intermountain West. It operates a network of hospitals, clinics, and a renowned medical school, integrating cutting-edge research, education, and patient care. As a large, complex organization with over 10,000 employees, it generates vast amounts of clinical, operational, and financial data.

For an organization of this size and mission, AI is not a luxury but a strategic imperative to manage complexity and fulfill its tripartite mission. The scale creates both a challenge—data silos and bureaucratic inertia—and an opportunity: the volume and variety of data are perfect fuel for machine learning models. AI can help translate its academic research prowess into tangible improvements in patient outcomes, operational efficiency, and financial sustainability, creating a competitive edge in the healthcare market.

Concrete AI Opportunities with ROI

1. Predictive Analytics for Patient Deterioration: By applying AI to real-time EHR data (like vital signs and lab results), the system can predict adverse events like sepsis 6-12 hours earlier. The ROI is substantial: reduced ICU length of stay, lower mortality rates, and avoided penalties for hospital-acquired conditions. For a large hospital, this could prevent hundreds of critical incidents annually, saving millions in care costs and improving quality metrics.

2. Revenue Cycle Automation: AI can streamline the immensely complex billing and claims process. Machine learning models can predict claim denials before submission and automate prior authorization tasks. Given the system's billions in annual revenue, even a 1-2% improvement in collection efficiency and a reduction in administrative labor translates to tens of millions in recovered revenue and operational savings.

3. Operational Efficiency in Surgical Services: AI-powered tools can optimize OR scheduling by predicting case durations and turnover times with high accuracy. This maximizes the utilization of high-cost surgical suites and staff. For a health system performing tens of thousands of surgeries yearly, improving OR efficiency by even 5-10% can free up capacity for additional revenue-generating procedures and significantly reduce overtime costs.

Deployment Risks Specific to Large Health Systems

Deploying AI at this scale carries unique risks. First, integration complexity is high; any AI tool must interoperate seamlessly with core systems like the Epic EHR without disrupting critical clinical workflows. Second, change management is a massive undertaking; gaining buy-in from thousands of physicians, nurses, and staff requires careful communication and demonstrated value. Third, regulatory and compliance hurdles are significant, particularly for algorithms influencing clinical decisions, requiring rigorous validation and monitoring to meet FDA (if applicable) and institutional review board standards. Finally, data governance across a decentralized organization is challenging; ensuring clean, unified, and ethically sourced data for AI models requires substantial upfront investment in data infrastructure and policies.