Why AI matters at this scale

Avera PACE, part of the large Avera Health system, operates a Program of All-inclusive Care for the Elderly (PACE). This model delivers integrated medical, therapeutic, and social services to frail seniors who qualify for nursing home-level care but wish to remain in their communities. As a large-scale provider (10,001+ employees), Avera PACE manages vast amounts of complex, longitudinal data across clinical, operational, and financial domains. At this scale, even marginal improvements in care coordination, preventive intervention, and operational efficiency can yield substantial clinical benefits and financial returns, particularly under the capitated payment structure of PACE, where the organization bears the full risk for participant costs.

Concrete AI Opportunities with ROI Framing

1. Predictive Analytics for Acute Care Utilization: The most significant financial risk in PACE is unplanned hospitalizations and emergency department visits. Machine learning models can synthesize EHR data, medication adherence, recent fall incidents, and social determinants of health to generate a daily 'risk score' for each participant. By alerting care teams to individuals with rapidly escalating risk, interventions like urgent home visits, medication reconciliation, or specialist consults can be deployed preemptively. For an organization of this size, preventing just a few dozen hospitalizations per month can translate to millions in annual savings and improved quality metrics.

2. AI-Optimized Interdisciplinary Team (IDT) Workflow: PACE relies on daily collaboration between nurses, doctors, therapists, social workers, and drivers. An AI-powered workflow engine can analyze pending tasks, participant needs, and staff credentials to intelligently assign and prioritize actions. For example, it could automatically route a participant reporting new dizziness to a physical therapist for a fall risk assessment and schedule a pharmacy review, ensuring a coordinated response. This reduces administrative overhead, improves team efficiency, and enhances the participant experience.

3. Intelligent Transportation and Logistics: Transportation is a critical, costly service. AI-driven routing algorithms can dynamically optimize schedules and routes for a fleet of vehicles based on real-time appointment changes, traffic conditions, weather, and individual participant needs (e.g., wheelchair requirements). This maximizes vehicle utilization, reduces fuel costs and driver overtime, and minimizes participant wait times—a key factor in satisfaction and adherence. The ROI is direct operational savings and improved service quality.

Deployment Risks Specific to Large Healthcare Organizations

Deploying AI at this scale within a large, regulated health system introduces specific risks. Integration Complexity is paramount; new AI tools must interface seamlessly with entrenched legacy systems like Epic or Cerner EHRs, which can be costly and time-consuming. Change Management across thousands of clinical and administrative staff requires extensive training and clear communication of benefits to avoid resistance. Regulatory and Compliance Risk is ever-present, as AI models handling PHI must be rigorously validated, explainable, and fully compliant with HIPAA, while also meeting evolving CMS guidelines for PACE programs. Finally, Data Quality and Silos can undermine model performance; clinical data may be fragmented across departments, requiring significant upfront investment in data governance and engineering to create a unified, reliable analytics foundation.