Major Health System Tests AI That Tells Doctors What to Do Next During Patient Visits
Sentara Health's 90-day pilot moves beyond documentation AI to real-time clinical decision support that recommends specific treatments and actions for each patient.
A $14.2 billion health system is testing whether artificial intelligence can do more than just take notes—it wants AI to guide doctors through increasingly complex patient visits by recommending specific next steps in real time.
Sentara Health has launched a 90-day pilot with GW RhythmX's AI precision care platform, deploying the technology with 30 physicians and advanced practice providers across family and internal medicine practices in two markets. Unlike ambient AI tools that primarily handle documentation, this platform analyzes patient histories, evidence-based guidelines, care gaps, social determinants, payer requirements, and formulary considerations to deliver patient-specific clinical recommendations during the visit itself.
From Documentation to Decision Support
The pilot represents a strategic shift in how health systems are deploying AI in clinical settings. While ambient documentation tools have proliferated across healthcare, Sentara is testing whether AI can address what executives describe as the cognitive burden of managing complex patient populations with limited clinician capacity.
GW RhythmX—formed through a 2025 merger of RhythmX AI and Get Well—positions its platform as "precision care" technology that goes beyond transcription. The system integrates multiple data sources to provide what the company calls "next-best-action recommendations" with transparent data provenance, allowing clinicians to see exactly how the AI arrived at its suggestions.
Presbyterian Healthcare Services deployed a similar approach in December, generating 17,000 hyper-personalized treatment responses and reducing the hours clinicians spent manually hunting for data across fragmented systems, according to the company.
Structured Evaluation with Control Groups
Sentara's pilot design reflects the caution health systems are applying to clinical AI adoption. The 90-day phase one includes a control group using current Sentara tools, allowing direct comparison of outcomes. The health system is also running a parallel pilot of team-care model redesign, enabling it to analyze the impacts of AI and care model changes separately before integrating them at scale.
Evaluation metrics include clinician interaction patterns, adoption rates, trust levels, recommendation utility, workflow integration, clinical action rates, and specialist referral patterns. Prasanna Mohanty, executive vice president and president for ambulatory care at Sentara Health, indicated the pilot is part of a broader ambulatory care redesign strategy.
Sentara operates 12 hospitals and more than 400 care sites across Virginia and Northeast North Carolina, with 473 primary care physicians and advanced practice providers at Sentara Medical Group. The health system reported $14.2 billion in operating revenue in 2025.
Market Positioning in Crowded AI Landscape
GW RhythmX CEO and founder Deepthi Bathina emphasized the company's differentiation in a market increasingly saturated with ambient AI documentation tools. The platform's integration of clinical guidelines, payer requirements, formulary data, and social determinants into actionable recommendations represents positioning toward higher-value clinical applications rather than pure administrative efficiency.
The pilot's structured approach—with control groups, phased deployment, and parallel testing of care model changes—may provide a framework for other health systems considering clinical decision support AI while managing implementation risk. The focus on reducing cognitive load and enabling primary care teams to manage more complex patients could address a key bottleneck in healthcare delivery as patient complexity increases and clinician capacity remains constrained.
Whether AI-generated clinical recommendations will achieve meaningful adoption depends on factors the pilot is specifically designed to measure: trust, utility, workflow integration, and ultimately whether clinicians act on the system's suggestions.
Sources
Companies in this article
Google Search
Follow Health AI Central on Google
Choose Health AI Central as a preferred source, and Google is more likely to show our reporting in Top Stories.
Reporting & accountability
How we maintain accuracy
Health AI Central separates reported facts, analysis, and public-source data. Readers can review our standards, request a correction, or consult the relevant research methodology before relying on a claim.