Improving Abstraction Quality Through Registry Metric Deconstruction

Despite regular and complete data submission to the CathPCI Registry, a participating hospital continued to struggle to improve metric 4462 ("Elective PCI with Stress Imaging"), indicating that data capture alone was not translating into measurable quality improvement.

This discrepancy suggested a gap between data collection and meaningful metric analysis, prompting a structured review of data abstraction and documentation processes.

We deconstructed metric 4462 using raw data, a custom spreadsheet, and logic-based filtering to identify abstraction errors and missed documentation. Of 66 elective percutaneous coronary intervention cases performed in 2024, 22 were initially flagged as metric fallouts. After review, only 7 cases were confirmed to be true fallouts, demonstrating a 68% reduction. This led to specific educational interventions and adjustments to data abstraction protocols.

Structured metric analysis reveals actionable abstraction errors, improves clinical documentation, and enhances data integrity. Integrating artificial intelligence-assisted tools in the future could further optimize and scale this quality improvement approach.