Fairview Health Services has selected Layer Health to automate clinical registry abstraction across its Minneapolis-based hospital and clinic network, targeting a labor-intensive process that consumes significant clinical staff time. The partnership will begin with the American Heart Association's Get With The Guidelines®–Stroke registry before expanding to cancer and cardiology quality programs.

The deployment addresses a persistent challenge in healthcare quality measurement: extracting structured data from complex patient charts for mandatory registry reporting. Traditional abstraction requires clinical staff to manually review entire medical records, a time-consuming process that diverts resources from direct patient care and quality improvement initiatives.

AI-Powered Chart Analysis with Human Validation

Layer Health's platform uses clinical AI to analyze longitudinal patient charts, including narrative physician notes, diagnostic imaging reports, pathology reports, and discrete telemetry data. According to the company, the technology delivers more than 70 percent time savings per abstracted chart while achieving greater than 98 percent accuracy against audited registry benchmarks, based on data from peer health systems.

The system's distinguishing feature is its source-linked evidence approach. Rather than simply extracting data fields, the platform provides citations to specific chart sections supporting each extracted value, enabling human abstractors to validate findings without re-reading entire records. This human-in-the-loop design preserves clinical oversight while accelerating the abstraction workflow.

"Fairview Health Services is a leader in quality measurement and improvement, and we're honored to support their mission with AI that augments—not replaces—the expertise of their clinical teams," said David Sontag, Chief Executive Officer and Co-Founder of Layer Health.

Phased Rollout Across Multiple Registry Programs

Fairview's implementation strategy begins with internal validation against the health system's historical clinical data before scaling to high-volume registries. Following the initial stroke registry deployment, the platform will expand to the Commission on Cancer's National Cancer Database (NCDB) and the American College of Cardiology's NCDR registries.

The phased approach reflects the complexity of registry abstraction, where different clinical domains require distinct data elements and validation criteria. Stroke registries track acute care metrics and treatment timelines, while cancer registries document staging, treatment protocols, and long-term outcomes. Cardiology registries focus on procedural data and device tracking.

Implications for Quality Measurement Workflows

The enterprise agreement represents a test case for clinical AI in quality measurement, an area where transparency and validation requirements have limited automation adoption. By providing traceable evidence for extracted data, the technology addresses concerns about "black box" AI systems in clinical documentation review.

For Fairview, the potential 70 percent time reduction could free clinical staff to focus on quality improvement initiatives rather than data extraction. The technology may also enable broader registry participation by reducing the administrative burden of compliance with mandatory reporting requirements.

The deployment's success will likely influence other health systems evaluating AI-powered solutions for clinical documentation workflows, particularly as registry reporting requirements continue to expand across specialty areas.

Sources

  1. Fairview Health Services Selects Layer Health to Automate Enterprise Clinical Registry Abstraction — HIT Consultant