Why AI matters at this scale

Intermountain Healthcare is a Utah-based, nonprofit integrated health system comprising 33 hospitals and hundreds of clinics. Founded in 1975, it operates on a unique model that combines insurance, physicians, and hospitals, with a longstanding reputation for data-driven practices and a focus on value-based care—improving patient outcomes while controlling costs. At its massive scale (10,001+ employees), Intermountain manages vast, structured datasets from electronic health records (EHRs), insurance claims, and operational systems. This creates a foundational asset for artificial intelligence. In the healthcare sector, where margins are thin and clinical precision is critical, AI is not merely an efficiency tool but a strategic lever for achieving the core mission: delivering high-quality, affordable care. For a system of Intermountain's size and complexity, AI can translate data into actionable insights at a pace and granularity impossible for human teams alone, impacting everything from individual patient prognoses to network-wide resource allocation.

Concrete AI Opportunities with ROI Framing

1. Operational Efficiency & Capacity Management: AI-driven predictive analytics can forecast patient admission rates, emergency department volume, and required ICU beds with high accuracy. By optimizing staff schedules and bed assignments in advance, Intermountain can reduce costly overtime, minimize patient wait times, and improve throughput. The ROI is direct: lower labor costs, higher revenue from increased capacity utilization, and better patient satisfaction scores.

2. Clinical Decision Support & Population Health: Deploying machine learning models on EHR data can identify patients at high risk for conditions like sepsis, heart failure readmissions, or diabetic complications. Early alerts enable proactive intervention, preventing adverse events and expensive hospitalizations. For a system deeply invested in value-based and bundled payment contracts, this directly protects revenue by improving quality metrics and reducing the cost of care for defined populations.

3. Administrative Automation: Natural Language Processing (NLP) can automate labor-intensive tasks such as clinical documentation, medical coding, and prior authorization. This reduces administrative burden on clinicians and staff, cutting processing time from days to minutes and reducing billing errors and claim denials. The ROI manifests in increased clinician productivity (more time for patients), lower administrative labor costs, and accelerated revenue cycles.

Deployment Risks Specific to Large Health Systems

Deploying AI at Intermountain's scale carries distinct risks. First, integration complexity is high. Any AI tool must interoperate seamlessly with core legacy systems, primarily its EHR (likely Epic or Cerner), without disrupting critical clinical workflows. Second, regulatory and compliance hurdles are stringent. Models must be rigorously validated, explainable to clinicians, and fully compliant with HIPAA and evolving FDA guidelines for software as a medical device. Third, change management is a monumental task. Gaining trust from thousands of physicians, nurses, and staff requires demonstrating clear utility, providing extensive training, and ensuring the technology augments rather than replaces professional judgment. Finally, data quality and bias are perennial concerns. Models trained on historical data may perpetuate existing disparities in care if not carefully audited and corrected, posing ethical and reputational risks. Successful deployment requires a centralized AI governance strategy that aligns clinical, operational, IT, and compliance leadership from the outset.