Human-Centered Innovation Driving Better Nurse Experiences - with Umesh Rustogi of Microsoft

65% of nurses report high stress and burnout, with the US facing a shortage of 78,000 registered nurses projected to grow to 500,000 by 2030. Nurses spend 25-40% of their shift time on documentation and administrative tasks, particularly manual flow sheet documentation that requires row-by-row data entry into EHR systems, taking time away from direct patient care.

• •Current shortage of 78,000 RNs in US, projected to reach 500,000 by 2030
• •25-40% of nursing shift time consumed by documentation and administrative tasks
• •Flow sheet documentation requires manual entry of structured observations (vitals, assessments) into EHR systems
• •Documentation burden is primary driver of cognitive load and time away from patients

Dragon Copilot integrates directly into Epic Rover mobile app, enabling ambient flow sheet capture through conversational recording at patient bedside. AI automatically extracts relevant observations and populates EHR systems. Additional capabilities include access to trusted clinical sources (CDC, FDA, MORT Manuals) with proper citations, automated note generation (narrative notes, physician notifications, incident reports), and self-service AI customization tools.

• •Mobile-first integration with Epic Rover allows bedside recording without tethering to workstations
• •AI extracts structured observations from nurse-patient conversations and auto-populates EHR flow sheets
• •Provides access to trusted clinical sources (CDC, FDA, MedlinePlus) with citations and hallucination safeguards
• •One-click generation of narrative notes, physician notifications, and incident reports from recorded conversations
• •Self-service AI tuning and customization for different health system schemas and workflows

Microsoft spent 12-15 months co-creating with leading healthcare organizations including Mercy, Advocate, and Stanford Healthcare. The process involved product management, engineering, design, and research teams working directly with frontline nurses, nursing informatics teams, and nursing leadership. This iterative approach was essential because nursing workflows are fundamentally different from physician workflows—more mobile, faster-paced, and focused on structured observations rather than narrative notes.

• •12-15 month co-innovation period with Mercy, Advocate, Stanford Healthcare and others
• •Cross-functional engagement: product, engineering, design, research, and customer teams
• •Gathered feedback from frontline nurses, nursing informatics, and nursing leadership
• •Purpose-built for nursing rather than repurposing physician-focused technology
• •Nursing workflows are uniquely mobile, fast-paced, and structured-observation focused

Early deployments demonstrate significant quantifiable outcomes: Mercy reported 21% reduction in documentation latency (one customer saw 70% reduction), 8-24 minutes saved per shift, and 29% reduction in incremental overtime. Patient satisfaction increased 4.5% for patients served by nurses using the technology. Documentation quality improved through capture of 'invisible care'—observations nurses previously didn't document due to time constraints.

• •21-70% reduction in documentation latency across different health systems
• •8-24 minutes saved per nursing shift, reducing cognitive burden and overtime
• •29% reduction in incremental overtime costs
• •4.5% increase in patient satisfaction scores compared to nurses not using technology
• •Improved documentation quality and thoroughness by capturing previously undocumented 'invisible care'
• •Near real-time documentation enables better care team coordination and patient outcomes

The next evolution will involve agentic AI handling administrative tasks like booking patient transportation and managing nursing schedules (which currently takes nurse managers 4-6 hours weekly). Hands-free workflows will integrate with smart hospital room technologies, Vocera badges, and Meta AI glasses, allowing ambient observation without nurses needing to interact with devices. The goal is friction-free documentation where audio, video, and sensors automatically capture and process information.

• •Agentic AI will automate administrative tasks: patient transportation booking, schedule management
• •Nurse managers currently spend 4-6 hours per week on scheduling—prime target for AI agents
• •Integration with wearables (Vocera badges, Meta AI glasses) and smart hospital room sensors
• •Hands-free ambient capture eliminates need to open apps or start recordings manually
• •Vision: nurses focus entirely on patient care while systems observe and document automatically
• •Timeline: significant advances expected within 2-3 years, not 5-10 years